From xen-changelog-bounces@lists.xenproject.org Tue Jun 01 10:55:13 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 01 Jun 2021 10:55:13 +0000 Received: from list by lists.xenproject.org with outflank-mailman.134790.250722 (Exim 4.92) (envelope-from ) id 1lo236-00041G-6T; Tue, 01 Jun 2021 10:55:04 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 134790.250722; Tue, 01 Jun 2021 10:55:04 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lo236-000418-3W; Tue, 01 Jun 2021 10:55:04 +0000 Received: by outflank-mailman (input) for mailman id 134790; Tue, 01 Jun 2021 10:55:02 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lo234-000411-Ox for xen-changelog@lists.xenproject.org; Tue, 01 Jun 2021 10:55:02 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lo234-0002i2-O2 for xen-changelog@lists.xenproject.org; Tue, 01 Jun 2021 10:55:02 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lo234-0002Mr-N2 for xen-changelog@lists.xenproject.org; Tue, 01 Jun 2021 10:55:02 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=yRH5Ijbl15pmTINWfZ77Fnz1VR00rzuVituBoCyUHRo=; b=oDl7PTE1FnXQCWbrkG3WZTnckz 1DUCxh5HREZv5ueEgtY+GAjaXvOKqYX4S0/+6s/PSsMwPfgH3L/xOpPQ2saL1QC+zq7PbgxjPAsGN rVzXdpvWNVffhrpoDFBU/7VxYkz0X8EzajaY4Qz/o+t6WkYky9mx+oBcYllYFvZ18U3o=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] x86/tboot: include all valid frame table entries in S3 integrity check Message-Id: Date: Tue, 01 Jun 2021 10:55:02 +0000 commit 8701f68d26afc641527c775d2e56a8709f535ffd Author: Jan Beulich AuthorDate: Mon May 31 12:45:51 2021 +0200 Commit: Jan Beulich CommitDate: Mon May 31 12:45:51 2021 +0200 x86/tboot: include all valid frame table entries in S3 integrity check The difference of two pdx_to_page() return values is a number of pages, not the number of bytes covered by the corresponding frame table entries. Fixes: 3cb68d2b59ab ("tboot: fix S3 issue for Intel Trusted Execution Technology.") Signed-off-by: Jan Beulich Reviewed-by: Lukasz Hawrylko --- xen/arch/x86/tboot.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/xen/arch/x86/tboot.c b/xen/arch/x86/tboot.c index aadcce591f..f394bd33b1 100644 --- a/xen/arch/x86/tboot.c +++ b/xen/arch/x86/tboot.c @@ -323,12 +323,12 @@ static void tboot_gen_frametable_integrity(const uint8_t key[TB_KEY_SIZE], if ( nidx >= max_idx ) break; vmac_update((uint8_t *)pdx_to_page(sidx * PDX_GROUP_COUNT), - pdx_to_page(eidx * PDX_GROUP_COUNT) - - pdx_to_page(sidx * PDX_GROUP_COUNT), &ctx); + (eidx - sidx) * PDX_GROUP_COUNT * sizeof(*frame_table), + &ctx); } vmac_update((uint8_t *)pdx_to_page(sidx * PDX_GROUP_COUNT), - pdx_to_page(max_pdx - 1) + 1 - - pdx_to_page(sidx * PDX_GROUP_COUNT), &ctx); + (max_pdx - sidx * PDX_GROUP_COUNT) * sizeof(*frame_table), + &ctx); *mac = vmac(NULL, 0, nonce, NULL, &ctx); -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Tue Jun 01 10:55:14 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 01 Jun 2021 10:55:14 +0000 Received: from list by lists.xenproject.org with outflank-mailman.134791.250726 (Exim 4.92) (envelope-from ) id 1lo23G-00043L-85; Tue, 01 Jun 2021 10:55:14 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 134791.250726; Tue, 01 Jun 2021 10:55:14 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lo23G-00043D-55; Tue, 01 Jun 2021 10:55:14 +0000 Received: by outflank-mailman (input) for mailman id 134791; Tue, 01 Jun 2021 10:55:12 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lo23E-00042x-SD for xen-changelog@lists.xenproject.org; Tue, 01 Jun 2021 10:55:12 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lo23E-0002i8-RL for xen-changelog@lists.xenproject.org; Tue, 01 Jun 2021 10:55:12 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lo23E-0002O5-QD for xen-changelog@lists.xenproject.org; Tue, 01 Jun 2021 10:55:12 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=ol0iDPLzPpR5XozAYOO/ZbVoai+LclS6qmmLQ4vqLz0=; b=BCQ9xq+I0dr2UIt2pq/kqb1al5 83Tv1pMoyMG5ecmg3SstiMz2bClEJc8GJSXITY1P3avxbmVLDoHUj5XH0ijMXkWLTKKEteFaMrGCz JMDqmynsudco0KasqJm3V5Xvj5F5cBXwGgk4BrYa2XWa1cwAUMiOvWBDgrMrZt1RG310=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] x86/tboot: adjust UUID check Message-Id: Date: Tue, 01 Jun 2021 10:55:12 +0000 commit e95c243f67a95bca8b4be62b4e024c64ab082e56 Author: Jan Beulich AuthorDate: Mon May 31 12:46:48 2021 +0200 Commit: Jan Beulich CommitDate: Mon May 31 12:46:48 2021 +0200 x86/tboot: adjust UUID check Replace a bogus cast, move the static variable into the only function using it, and add __initconst. While there, also remove a pointless NULL check. Signed-off-by: Jan Beulich Reviewed-by: Lukasz Hawrylko --- xen/arch/x86/tboot.c | 7 ++----- 1 file changed, 2 insertions(+), 5 deletions(-) diff --git a/xen/arch/x86/tboot.c b/xen/arch/x86/tboot.c index f394bd33b1..529367ed81 100644 --- a/xen/arch/x86/tboot.c +++ b/xen/arch/x86/tboot.c @@ -27,8 +27,6 @@ static vmac_t domain_mac; /* MAC for all domains during S3 */ static vmac_t xenheap_mac; /* MAC for xen heap during S3 */ static vmac_t frametable_mac; /* MAC for frame table during S3 */ -static const uuid_t tboot_shared_uuid = TBOOT_SHARED_UUID; - /* used by tboot_protect_mem_regions() and/or tboot_parse_dmar_table() */ static uint64_t __initdata txt_heap_base, __initdata txt_heap_size; static uint64_t __initdata sinit_base, __initdata sinit_size; @@ -93,6 +91,7 @@ static void __init tboot_copy_memory(unsigned char *va, uint32_t size, void __init tboot_probe(void) { tboot_shared_t *tboot_shared; + static const uuid_t __initconst tboot_shared_uuid = TBOOT_SHARED_UUID; /* Look for valid page-aligned address for shared page. */ if ( !opt_tboot_pa || (opt_tboot_pa & ~PAGE_MASK) ) @@ -101,9 +100,7 @@ void __init tboot_probe(void) /* Map and check for tboot UUID. */ set_fixmap(FIX_TBOOT_SHARED_BASE, opt_tboot_pa); tboot_shared = fix_to_virt(FIX_TBOOT_SHARED_BASE); - if ( tboot_shared == NULL ) - return; - if ( memcmp(&tboot_shared_uuid, (uuid_t *)tboot_shared, sizeof(uuid_t)) ) + if ( memcmp(&tboot_shared_uuid, &tboot_shared->uuid, sizeof(uuid_t)) ) return; /* new tboot_shared (w/ GAS support, integrity, etc.) is not backwards -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Tue Jun 01 10:55:24 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 01 Jun 2021 10:55:24 +0000 Received: from list by lists.xenproject.org with outflank-mailman.134792.250730 (Exim 4.92) (envelope-from ) id 1lo23Q-00046X-9Z; Tue, 01 Jun 2021 10:55:24 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 134792.250730; Tue, 01 Jun 2021 10:55:24 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lo23Q-00046P-6Y; Tue, 01 Jun 2021 10:55:24 +0000 Received: by outflank-mailman (input) for mailman id 134792; Tue, 01 Jun 2021 10:55:23 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lo23O-00046G-VF for xen-changelog@lists.xenproject.org; Tue, 01 Jun 2021 10:55:22 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lo23O-0002iK-US for xen-changelog@lists.xenproject.org; Tue, 01 Jun 2021 10:55:22 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lo23O-0002Ox-TW for xen-changelog@lists.xenproject.org; Tue, 01 Jun 2021 10:55:22 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=VawYpoiGUc25Z5piGWI0xzRprwQlUoo6gbJwWHz2tUg=; b=lKqtPDFy2qJWLwRv1K3nQ3Fdns c6kO6kNYePF2UIGz/bHBxYd6zjdkhwTyEN57uo6kLNMsbYBk73e9cEXM6G8vilEyMEKVdt0e6qtx+ fhB/4hoh7ATwKW2FTT4jO6EYa4ZQY1Q7W+c11gTiWreAkfxRJxpN76NKnsm2y4MZZXeo=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] x86/mtrr: remove stale function prototype Message-Id: Date: Tue, 01 Jun 2021 10:55:22 +0000 commit 57f68dfd2d111a2ad381df740543c901b41f2299 Author: Roger Pau Monné AuthorDate: Mon May 31 12:47:12 2021 +0200 Commit: Jan Beulich CommitDate: Mon May 31 12:47:12 2021 +0200 x86/mtrr: remove stale function prototype Fixes: 1c84d04673 ('VMX: remove the problematic set_uc_mode logic') Signed-off-by: Roger Pau Monné Acked-by: Jan Beulich --- xen/include/asm-x86/mtrr.h | 2 -- 1 file changed, 2 deletions(-) diff --git a/xen/include/asm-x86/mtrr.h b/xen/include/asm-x86/mtrr.h index 4be704cb6a..24e5de5c22 100644 --- a/xen/include/asm-x86/mtrr.h +++ b/xen/include/asm-x86/mtrr.h @@ -78,8 +78,6 @@ extern u32 get_pat_flags(struct vcpu *v, u32 gl1e_flags, paddr_t gpaddr, extern int epte_get_entry_emt(struct domain *, unsigned long gfn, mfn_t mfn, unsigned int order, uint8_t *ipat, bool_t direct_mmio); -extern void ept_change_entry_emt_with_range( - struct domain *d, unsigned long start_gfn, unsigned long end_gfn); extern unsigned char pat_type_2_pte_flags(unsigned char pat_type); extern int hold_mtrr_updates_on_aps; extern void mtrr_aps_sync_begin(void); -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Tue Jun 01 15:44:10 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 01 Jun 2021 15:44:10 +0000 Received: from list by lists.xenproject.org with outflank-mailman.134938.250929 (Exim 4.92) (envelope-from ) id 1lo6Yn-0007wn-5V; Tue, 01 Jun 2021 15:44:05 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 134938.250929; Tue, 01 Jun 2021 15:44:05 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lo6Yn-0007wf-2M; Tue, 01 Jun 2021 15:44:05 +0000 Received: by outflank-mailman (input) for mailman id 134938; Tue, 01 Jun 2021 15:44:04 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lo6Ym-0007wX-Gc for xen-changelog@lists.xenproject.org; Tue, 01 Jun 2021 15:44:04 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lo6Ym-0008Rk-Dl for xen-changelog@lists.xenproject.org; Tue, 01 Jun 2021 15:44:04 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lo6Ym-0007Al-Cr for xen-changelog@lists.xenproject.org; Tue, 01 Jun 2021 15:44:04 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=m9JnmFDsaIwb0NMIAuFE01uo2zPh8dHbokV6ZC0I37Q=; b=3qZ8VbD82/V5bUzJef2R5OG/SE +48PSJoYbS26kGsuuiwlJAnWhGsijVxuUrW5e10dGHbbu2sUVxiZRi9Uk2exL8+PrTISkmKg+cUh/ 5YpfQXvWo0HZK0tFoPJiQSwJoiBARPfdmHHuhXE5TITftnhpa/D8PHge/FNN4zXDTVjE=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] xen/page_alloc: Remove dead code in alloc_domheap_pages() Message-Id: Date: Tue, 01 Jun 2021 15:44:04 +0000 commit 5268b2dcf7e5342c8a51ceb4bed3e7740c69f5c1 Author: Julien Grall AuthorDate: Mon Feb 22 12:01:18 2021 +0000 Commit: Julien Grall CommitDate: Tue Jun 1 16:34:50 2021 +0100 xen/page_alloc: Remove dead code in alloc_domheap_pages() Since commit 1aac966e24e9 "xen: support RAM at addresses 0 and 4096", bits_to_zone() will never return 0 and it is expected that we have minimum 2 zones. Therefore the check in alloc_domheap_pages() is unnecessary and can be removed. However, for sanity, it is replaced with an ASSERT(). Also take the opportunity to switch from min_t() to min() as bits_to_zone() cannot return a negative value. The macro is tweaked to make it clearer. This bug was discovered and resolved using Coverity Static Analysis Security Testing (SAST) by Synopsys, Inc. Signed-off-by: Julien Grall Reviewed-by: Jan Beulich --- xen/common/page_alloc.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/xen/common/page_alloc.c b/xen/common/page_alloc.c index ace6333c18..958ba0cd92 100644 --- a/xen/common/page_alloc.c +++ b/xen/common/page_alloc.c @@ -441,7 +441,7 @@ mfn_t __init alloc_boot_pages(unsigned long nr_pfns, unsigned long pfn_align) #define MEMZONE_XEN 0 #define NR_ZONES (PADDR_BITS - PAGE_SHIFT + 1) -#define bits_to_zone(b) (((b) < (PAGE_SHIFT + 1)) ? 1 : ((b) - PAGE_SHIFT)) +#define bits_to_zone(b) (((b) < (PAGE_SHIFT + 1)) ? 1U : ((b) - PAGE_SHIFT)) #define page_to_zone(pg) (is_xen_heap_page(pg) ? MEMZONE_XEN : \ (flsl(mfn_x(page_to_mfn(pg))) ? : 1)) @@ -2336,8 +2336,9 @@ struct page_info *alloc_domheap_pages( bits = domain_clamp_alloc_bitsize(memflags & MEMF_no_owner ? NULL : d, bits ? : (BITS_PER_LONG+PAGE_SHIFT)); - if ( (zone_hi = min_t(unsigned int, bits_to_zone(bits), zone_hi)) == 0 ) - return NULL; + + zone_hi = min(bits_to_zone(bits), zone_hi); + ASSERT(zone_hi != 0); if ( memflags & MEMF_no_owner ) memflags |= MEMF_no_refcount; -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Wed Jun 02 10:44:07 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Wed, 02 Jun 2021 10:44:07 +0000 Received: from list by lists.xenproject.org with outflank-mailman.135939.252293 (Exim 4.92) (envelope-from ) id 1loOLz-0007gZ-36; Wed, 02 Jun 2021 10:44:03 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 135939.252293; Wed, 02 Jun 2021 10:44:03 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1loOLz-0007gR-0A; Wed, 02 Jun 2021 10:44:03 +0000 Received: by outflank-mailman (input) for mailman id 135939; Wed, 02 Jun 2021 10:44:01 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1loOLx-0007g0-PX for xen-changelog@lists.xenproject.org; Wed, 02 Jun 2021 10:44:01 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1loOLx-0004nK-Om for xen-changelog@lists.xenproject.org; Wed, 02 Jun 2021 10:44:01 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1loOLx-0008Tm-Nl for xen-changelog@lists.xenproject.org; Wed, 02 Jun 2021 10:44:01 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=JyG3bqxts/QdHE8XmFo3RUJF2y/WPmtPQn03fmhbaE8=; b=Te2dsn6gIDnD70sBLJ6Wu4ES22 sK+Qa+YJk7rzcZOCnhKO6ndlF0pfjZQP8S6CVySKKNIO0+CTkyBLEUgcpIutV+z4g+uI2j7YXNxF6 ne8A333ldK1jK7Ggy6flp1fQi2ifT9CfoKv/qkqJiJ79oxBa0Wssl+Q9LST4CIwCIPdE=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] xen/page_alloc: Remove dead code in alloc_domheap_pages() Message-Id: Date: Wed, 02 Jun 2021 10:44:01 +0000 commit 5268b2dcf7e5342c8a51ceb4bed3e7740c69f5c1 Author: Julien Grall AuthorDate: Mon Feb 22 12:01:18 2021 +0000 Commit: Julien Grall CommitDate: Tue Jun 1 16:34:50 2021 +0100 xen/page_alloc: Remove dead code in alloc_domheap_pages() Since commit 1aac966e24e9 "xen: support RAM at addresses 0 and 4096", bits_to_zone() will never return 0 and it is expected that we have minimum 2 zones. Therefore the check in alloc_domheap_pages() is unnecessary and can be removed. However, for sanity, it is replaced with an ASSERT(). Also take the opportunity to switch from min_t() to min() as bits_to_zone() cannot return a negative value. The macro is tweaked to make it clearer. This bug was discovered and resolved using Coverity Static Analysis Security Testing (SAST) by Synopsys, Inc. Signed-off-by: Julien Grall Reviewed-by: Jan Beulich --- xen/common/page_alloc.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/xen/common/page_alloc.c b/xen/common/page_alloc.c index ace6333c18..958ba0cd92 100644 --- a/xen/common/page_alloc.c +++ b/xen/common/page_alloc.c @@ -441,7 +441,7 @@ mfn_t __init alloc_boot_pages(unsigned long nr_pfns, unsigned long pfn_align) #define MEMZONE_XEN 0 #define NR_ZONES (PADDR_BITS - PAGE_SHIFT + 1) -#define bits_to_zone(b) (((b) < (PAGE_SHIFT + 1)) ? 1 : ((b) - PAGE_SHIFT)) +#define bits_to_zone(b) (((b) < (PAGE_SHIFT + 1)) ? 1U : ((b) - PAGE_SHIFT)) #define page_to_zone(pg) (is_xen_heap_page(pg) ? MEMZONE_XEN : \ (flsl(mfn_x(page_to_mfn(pg))) ? : 1)) @@ -2336,8 +2336,9 @@ struct page_info *alloc_domheap_pages( bits = domain_clamp_alloc_bitsize(memflags & MEMF_no_owner ? NULL : d, bits ? : (BITS_PER_LONG+PAGE_SHIFT)); - if ( (zone_hi = min_t(unsigned int, bits_to_zone(bits), zone_hi)) == 0 ) - return NULL; + + zone_hi = min(bits_to_zone(bits), zone_hi); + ASSERT(zone_hi != 0); if ( memflags & MEMF_no_owner ) memflags |= MEMF_no_refcount; -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Fri Jun 04 12:55:13 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Fri, 04 Jun 2021 12:55:13 +0000 Received: from list by lists.xenproject.org with outflank-mailman.136812.253488 (Exim 4.92) (envelope-from ) id 1lp9Lu-0004R4-Ou; Fri, 04 Jun 2021 12:55:06 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 136812.253488; Fri, 04 Jun 2021 12:55:06 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lp9Lu-0004Qw-Lw; Fri, 04 Jun 2021 12:55:06 +0000 Received: by outflank-mailman (input) for mailman id 136812; Fri, 04 Jun 2021 12:55:05 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lp9Lt-0004Qq-FP for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 12:55:05 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lp9Lt-0004HG-Dj for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 12:55:05 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lp9Lt-00035e-CN for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 12:55:05 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=89GKVvX0/7E8zJcwH7Bna+pRTon5dSwwxCOCcwEDm8w=; b=OHJ6K2+h2diNiCpFnzL5DanHfX mQ44z6bdtNlW4k5IW5XerJ+sIDkNO+cZEsMF1Pf2FIx6jIRUjHS3bWkEESUJLTfKIV24YaAVAwzXM 3W8m/Qhy4Au3Q7rmkHLr1NiXtIXytCuuJ2JVAJ2Eg7LB/t94SjbhFCMzHqNzIolj7hKQ=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.15] VT-d: Don't assume register-based invalidation is always supported Message-Id: Date: Fri, 04 Jun 2021 12:55:05 +0000 commit 78a7c3b1f2e80be44cfcb27c102edfc8e7741555 Author: Chao Gao AuthorDate: Fri Jun 4 14:43:42 2021 +0200 Commit: Jan Beulich CommitDate: Fri Jun 4 14:43:42 2021 +0200 VT-d: Don't assume register-based invalidation is always supported According to Intel VT-d SPEC rev3.3 Section 6.5, Register-based Invalidation isn't supported by Intel VT-d version 6 and beyond. This hardware change impacts following two scenarios: admin can disable queued invalidation via 'qinval' cmdline and use register-based interface; VT-d switches to register-based invalidation when queued invalidation needs to be disabled, for example, during disabling x2apic or during system suspension or after enabling queued invalidation fails. To deal with this hardware change, if register-based invalidation isn't supported, queued invalidation cannot be disabled through Xen cmdline; and if queued invalidation has to be disabled temporarily in some scenarios, VT-d won't switch to register-based interface but use some dummy functions to catch errors in case there is any invalidation request issued when queued invalidation is disabled. Signed-off-by: Chao Gao Reviewed-by: Jan Beulich Reviewed-by: Kevin Tian master commit: 6773b1a7584a75a486e9774541ad5bd84c9aa5ee master date: 2021-04-26 10:16:50 +0200 --- docs/misc/xen-command-line.pandoc | 4 +++- xen/drivers/passthrough/vtd/iommu.c | 27 +++++++++++++++++++++++++-- xen/drivers/passthrough/vtd/iommu.h | 7 +++++++ xen/drivers/passthrough/vtd/qinval.c | 33 +++++++++++++++++++++++++++++++-- 4 files changed, 66 insertions(+), 5 deletions(-) diff --git a/docs/misc/xen-command-line.pandoc b/docs/misc/xen-command-line.pandoc index a4bd3f12c5..c32a397a12 100644 --- a/docs/misc/xen-command-line.pandoc +++ b/docs/misc/xen-command-line.pandoc @@ -1458,7 +1458,9 @@ The following options are specific to Intel VT-d hardware: * The `qinval` boolean controls the Queued Invalidation sub-feature, and is active by default on compatible hardware. Queued Invalidation is a feature in second-generation IOMMUs and is a functional prerequisite for - Interrupt Remapping. + Interrupt Remapping. Note that Xen disregards this setting for Intel VT-d + version 6 and greater as Registered-Based Invalidation isn't supported + by them. * The `igfx` boolean is active by default, and controls whether the IOMMU in front of an Intel Graphics Device is enabled or not. diff --git a/xen/drivers/passthrough/vtd/iommu.c b/xen/drivers/passthrough/vtd/iommu.c index 6428c8fe3e..d14fb43182 100644 --- a/xen/drivers/passthrough/vtd/iommu.c +++ b/xen/drivers/passthrough/vtd/iommu.c @@ -1193,6 +1193,14 @@ int __init iommu_alloc(struct acpi_drhd_unit *drhd) iommu->cap = dmar_readq(iommu->reg, DMAR_CAP_REG); iommu->ecap = dmar_readq(iommu->reg, DMAR_ECAP_REG); + iommu->version = dmar_readl(iommu->reg, DMAR_VER_REG); + + if ( !iommu_qinval && !has_register_based_invalidation(iommu) ) + { + printk(XENLOG_WARNING VTDPREFIX "IOMMU %d: cannot disable Queued Invalidation\n", + iommu->index); + iommu_qinval = true; + } if ( iommu_verbose ) { @@ -2141,6 +2149,10 @@ static int __must_check init_vtd_hw(bool resume) */ if ( enable_qinval(iommu) != 0 ) { + /* Ensure register-based invalidation is available */ + if ( !has_register_based_invalidation(iommu) ) + return -EIO; + iommu->flush.context = vtd_flush_context_reg; iommu->flush.iotlb = vtd_flush_iotlb_reg; } @@ -2231,6 +2243,7 @@ static int __init vtd_setup(void) struct acpi_drhd_unit *drhd; struct vtd_iommu *iommu; int ret; + bool reg_inval_supported = true; if ( list_empty(&acpi_drhd_units) ) { @@ -2252,8 +2265,8 @@ static int __init vtd_setup(void) } /* We enable the following features only if they are supported by all VT-d - * engines: Snoop Control, DMA passthrough, Queued Invalidation, Interrupt - * Remapping, and Posted Interrupt + * engines: Snoop Control, DMA passthrough, Register-based Invalidation, + * Queued Invalidation, Interrupt Remapping, and Posted Interrupt. */ for_each_drhd_unit ( drhd ) { @@ -2275,6 +2288,9 @@ static int __init vtd_setup(void) if ( iommu_qinval && !ecap_queued_inval(iommu->ecap) ) iommu_qinval = 0; + if ( !has_register_based_invalidation(iommu) ) + reg_inval_supported = false; + if ( iommu_intremap && !ecap_intr_remap(iommu->ecap) ) iommu_intremap = iommu_intremap_off; @@ -2301,6 +2317,13 @@ static int __init vtd_setup(void) softirq_tasklet_init(&vtd_fault_tasklet, do_iommu_page_fault, NULL); + if ( !iommu_qinval && !reg_inval_supported ) + { + dprintk(XENLOG_ERR VTDPREFIX, "No available invalidation interface\n"); + ret = -ENODEV; + goto error; + } + if ( !iommu_qinval && iommu_intremap ) { iommu_intremap = iommu_intremap_off; diff --git a/xen/drivers/passthrough/vtd/iommu.h b/xen/drivers/passthrough/vtd/iommu.h index 216791b3d6..57737b1a4a 100644 --- a/xen/drivers/passthrough/vtd/iommu.h +++ b/xen/drivers/passthrough/vtd/iommu.h @@ -540,6 +540,7 @@ struct vtd_iommu { struct list_head ats_devices; unsigned long *domid_bitmap; /* domain id bitmap */ u16 *domid_map; /* domain id mapping array */ + uint32_t version; }; #define INTEL_IOMMU_DEBUG(fmt, args...) \ @@ -549,4 +550,10 @@ struct vtd_iommu { dprintk(XENLOG_WARNING VTDPREFIX, fmt, ## args); \ } while(0) +/* Register-based invalidation isn't supported by VT-d version 6 and beyond. */ +static inline bool has_register_based_invalidation(const struct vtd_iommu *vtd) +{ + return VER_MAJOR(vtd->version) < 6; +} + #endif diff --git a/xen/drivers/passthrough/vtd/qinval.c b/xen/drivers/passthrough/vtd/qinval.c index 764ef9f0fc..dfcbb66c8a 100644 --- a/xen/drivers/passthrough/vtd/qinval.c +++ b/xen/drivers/passthrough/vtd/qinval.c @@ -442,6 +442,23 @@ int enable_qinval(struct vtd_iommu *iommu) return 0; } +static int vtd_flush_context_noop(struct vtd_iommu *iommu, uint16_t did, + uint16_t source_id, uint8_t function_mask, + uint64_t type, bool flush_non_present_entry) +{ + WARN(); + return -EIO; +} + +static int vtd_flush_iotlb_noop(struct vtd_iommu *iommu, uint16_t did, + uint64_t addr, unsigned int size_order, + uint64_t type, bool flush_non_present_entry, + bool flush_dev_iotlb) +{ + WARN(); + return -EIO; +} + void disable_qinval(struct vtd_iommu *iommu) { u32 sts; @@ -463,6 +480,18 @@ void disable_qinval(struct vtd_iommu *iommu) out: spin_unlock_irqrestore(&iommu->register_lock, flags); - iommu->flush.context = vtd_flush_context_reg; - iommu->flush.iotlb = vtd_flush_iotlb_reg; + /* + * Assign callbacks to noop to catch errors if register-based invalidation + * isn't supported. + */ + if ( has_register_based_invalidation(iommu) ) + { + iommu->flush.context = vtd_flush_context_reg; + iommu->flush.iotlb = vtd_flush_iotlb_reg; + } + else + { + iommu->flush.context = vtd_flush_context_noop; + iommu->flush.iotlb = vtd_flush_iotlb_noop; + } } -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.15 From xen-changelog-bounces@lists.xenproject.org Fri Jun 04 12:55:16 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Fri, 04 Jun 2021 12:55:16 +0000 Received: from list by lists.xenproject.org with outflank-mailman.136813.253492 (Exim 4.92) (envelope-from ) id 1lp9M4-0004Sr-QX; Fri, 04 Jun 2021 12:55:16 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 136813.253492; Fri, 04 Jun 2021 12:55:16 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lp9M4-0004Sj-Nb; Fri, 04 Jun 2021 12:55:16 +0000 Received: by outflank-mailman (input) for mailman id 136813; Fri, 04 Jun 2021 12:55:15 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lp9M3-0004SX-I8 for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 12:55:15 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lp9M3-0004HK-HH for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 12:55:15 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lp9M3-00036b-GC for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 12:55:15 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=AS5hpwfjKD8nJwa4kSR8VgYkh5zNQOe78W540OPqoeY=; b=dPU00yug1Kz50Y4nnh8T+15wgm Vdhej2RSpv7yp3WYolN7DaKT14NvjkM7eyzcDKDG1ZAVWy2AZm0g7/J3z6TL78iPhzEjWYa67kUIy Yg/LJ8zpCyRjUBfcZMqYYSNXTgb6mxa5vtcII5Ik3S1wfnUNA9H0vJKbiQcQzHN9Hj3g=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.15] x86/vtx: add LBR_SELECT to the list of LBR MSRs Message-Id: Date: Fri, 04 Jun 2021 12:55:15 +0000 commit 3e6c1b6cccd8c7bc63871836c14ac104d2a6e86a Author: Igor Druzhinin AuthorDate: Fri Jun 4 14:45:00 2021 +0200 Commit: Jan Beulich CommitDate: Fri Jun 4 14:45:00 2021 +0200 x86/vtx: add LBR_SELECT to the list of LBR MSRs This MSR exists since Nehalem / Silvermont and is actively used by Linux, for instance, to improve sampling efficiency. Signed-off-by: Igor Druzhinin Reviewed-by: Jan Beulich Reviewed-by: Kevin Tian master commit: 08693c03e00ea3448adc4406c891e707f0068eb6 master date: 2021-04-26 10:22:04 +0200 --- xen/arch/x86/hvm/vmx/vmx.c | 20 ++++++++++++++++---- xen/include/asm-x86/msr-index.h | 10 ++++++++-- 2 files changed, 24 insertions(+), 6 deletions(-) diff --git a/xen/arch/x86/hvm/vmx/vmx.c b/xen/arch/x86/hvm/vmx/vmx.c index b52824677e..8e0aa976c2 100644 --- a/xen/arch/x86/hvm/vmx/vmx.c +++ b/xen/arch/x86/hvm/vmx/vmx.c @@ -2898,14 +2898,16 @@ static const struct lbr_info { }, nh_lbr[] = { { MSR_IA32_LASTINTFROMIP, 1 }, { MSR_IA32_LASTINTTOIP, 1 }, - { MSR_C2_LASTBRANCH_TOS, 1 }, + { MSR_NHL_LBR_SELECT, 1 }, + { MSR_NHL_LASTBRANCH_TOS, 1 }, { MSR_P4_LASTBRANCH_0_FROM_LIP, NUM_MSR_P4_LASTBRANCH_FROM_TO }, { MSR_P4_LASTBRANCH_0_TO_LIP, NUM_MSR_P4_LASTBRANCH_FROM_TO }, { 0, 0 } }, sk_lbr[] = { { MSR_IA32_LASTINTFROMIP, 1 }, { MSR_IA32_LASTINTTOIP, 1 }, - { MSR_SKL_LASTBRANCH_TOS, 1 }, + { MSR_NHL_LBR_SELECT, 1 }, + { MSR_NHL_LASTBRANCH_TOS, 1 }, { MSR_SKL_LASTBRANCH_0_FROM_IP, NUM_MSR_SKL_LASTBRANCH }, { MSR_SKL_LASTBRANCH_0_TO_IP, NUM_MSR_SKL_LASTBRANCH }, { MSR_SKL_LASTBRANCH_0_INFO, NUM_MSR_SKL_LASTBRANCH }, @@ -2917,10 +2919,19 @@ static const struct lbr_info { { MSR_C2_LASTBRANCH_0_FROM_IP, NUM_MSR_ATOM_LASTBRANCH_FROM_TO }, { MSR_C2_LASTBRANCH_0_TO_IP, NUM_MSR_ATOM_LASTBRANCH_FROM_TO }, { 0, 0 } +}, sm_lbr[] = { + { MSR_IA32_LASTINTFROMIP, 1 }, + { MSR_IA32_LASTINTTOIP, 1 }, + { MSR_SM_LBR_SELECT, 1 }, + { MSR_SM_LASTBRANCH_TOS, 1 }, + { MSR_C2_LASTBRANCH_0_FROM_IP, NUM_MSR_ATOM_LASTBRANCH_FROM_TO }, + { MSR_C2_LASTBRANCH_0_TO_IP, NUM_MSR_ATOM_LASTBRANCH_FROM_TO }, + { 0, 0 } }, gm_lbr[] = { { MSR_IA32_LASTINTFROMIP, 1 }, { MSR_IA32_LASTINTTOIP, 1 }, - { MSR_GM_LASTBRANCH_TOS, 1 }, + { MSR_SM_LBR_SELECT, 1 }, + { MSR_SM_LASTBRANCH_TOS, 1 }, { MSR_GM_LASTBRANCH_0_FROM_IP, NUM_MSR_GM_LASTBRANCH_FROM_TO }, { MSR_GM_LASTBRANCH_0_TO_IP, NUM_MSR_GM_LASTBRANCH_FROM_TO }, { 0, 0 } @@ -2974,6 +2985,7 @@ static const struct lbr_info *last_branch_msr_get(void) return sk_lbr; /* Atom */ case 0x1c: case 0x26: case 0x27: case 0x35: case 0x36: + return at_lbr; /* Silvermont */ case 0x37: case 0x4a: case 0x4d: case 0x5a: case 0x5d: /* Xeon Phi Knights Landing */ @@ -2982,7 +2994,7 @@ static const struct lbr_info *last_branch_msr_get(void) case 0x85: /* Airmont */ case 0x4c: - return at_lbr; + return sm_lbr; /* Goldmont */ case 0x5c: case 0x5f: return gm_lbr; diff --git a/xen/include/asm-x86/msr-index.h b/xen/include/asm-x86/msr-index.h index f2e34dd22b..bd3a3a1e7f 100644 --- a/xen/include/asm-x86/msr-index.h +++ b/xen/include/asm-x86/msr-index.h @@ -603,15 +603,21 @@ #define NUM_MSR_C2_LASTBRANCH_FROM_TO 4 #define NUM_MSR_ATOM_LASTBRANCH_FROM_TO 8 +/* Nehalem (and newer) last-branch recording */ +#define MSR_NHL_LBR_SELECT 0x000001c8 +#define MSR_NHL_LASTBRANCH_TOS 0x000001c9 + /* Skylake (and newer) last-branch recording */ -#define MSR_SKL_LASTBRANCH_TOS 0x000001c9 #define MSR_SKL_LASTBRANCH_0_FROM_IP 0x00000680 #define MSR_SKL_LASTBRANCH_0_TO_IP 0x000006c0 #define MSR_SKL_LASTBRANCH_0_INFO 0x00000dc0 #define NUM_MSR_SKL_LASTBRANCH 32 +/* Silvermont (and newer) last-branch recording */ +#define MSR_SM_LBR_SELECT 0x000001c8 +#define MSR_SM_LASTBRANCH_TOS 0x000001c9 + /* Goldmont last-branch recording */ -#define MSR_GM_LASTBRANCH_TOS 0x000001c9 #define MSR_GM_LASTBRANCH_0_FROM_IP 0x00000680 #define MSR_GM_LASTBRANCH_0_TO_IP 0x000006c0 #define NUM_MSR_GM_LASTBRANCH_FROM_TO 32 -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.15 From xen-changelog-bounces@lists.xenproject.org Fri Jun 04 12:55:26 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Fri, 04 Jun 2021 12:55:26 +0000 Received: from list by lists.xenproject.org with outflank-mailman.136814.253496 (Exim 4.92) (envelope-from ) id 1lp9ME-0004VH-S4; Fri, 04 Jun 2021 12:55:26 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 136814.253496; Fri, 04 Jun 2021 12:55:26 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lp9ME-0004V9-P9; Fri, 04 Jun 2021 12:55:26 +0000 Received: by outflank-mailman (input) for mailman id 136814; Fri, 04 Jun 2021 12:55:25 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lp9MD-0004Ux-Le for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 12:55:25 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lp9MD-0004HV-Kp for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 12:55:25 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lp9MD-00037D-Jo for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 12:55:25 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=K3FHZtkjJLW53QYL3xyihYL9guqDDNojw/1Sv1LxCjE=; b=Hubjv6UH3kQzBOcx0EMt0+QElf Bga75MKw55RsET2iNIZfChJYiKn+eKJYmPTeMkEIYzamsBMcHaZryXq1CjrV2LP05m3OcVB5Mrxqy vamIwAojw2ytLCnxL8YJ1E7TzQNAxUIVFpjwAEfLCYt/HZ6vFUY5SQ/WVgqRaKgBt2bU=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.15] x86/intel: insert Ice Lake-SP and Ice Lake-D model numbers Message-Id: Date: Fri, 04 Jun 2021 12:55:25 +0000 commit cacad0cdfcf2763e3377316e779532a4a115ae8c Author: Igor Druzhinin AuthorDate: Fri Jun 4 14:45:39 2021 +0200 Commit: Jan Beulich CommitDate: Fri Jun 4 14:45:39 2021 +0200 x86/intel: insert Ice Lake-SP and Ice Lake-D model numbers LBR, C-state MSRs should correspond to Ice Lake desktop according to SDM rev. 74 for both models. Ice Lake-SP is known to expose IF_PSCHANGE_MC_NO in IA32_ARCH_CAPABILITIES MSR (as advisory tells and Whitley SDP confirms) which means the erratum is fixed in hardware for that model and therefore it shouldn't be present in has_if_pschange_mc list. Provisionally assume the same to be the case for Ice Lake-D. Signed-off-by: Igor Druzhinin Reviewed-by: Jan Beulich Reviewed-by: Kevin Tian master commit: 95419adfd4b275cffe24b96edcc2f15bc4db8907 master date: 2021-04-26 10:22:48 +0200 --- xen/arch/x86/acpi/cpu_idle.c | 2 ++ xen/arch/x86/hvm/vmx/vmx.c | 2 +- 2 files changed, 3 insertions(+), 1 deletion(-) diff --git a/xen/arch/x86/acpi/cpu_idle.c b/xen/arch/x86/acpi/cpu_idle.c index c092086b33..d788c8bffc 100644 --- a/xen/arch/x86/acpi/cpu_idle.c +++ b/xen/arch/x86/acpi/cpu_idle.c @@ -181,6 +181,8 @@ static void do_get_hw_residencies(void *arg) case 0x55: case 0x5E: /* Ice Lake */ + case 0x6A: + case 0x6C: case 0x7D: case 0x7E: /* Tiger Lake */ diff --git a/xen/arch/x86/hvm/vmx/vmx.c b/xen/arch/x86/hvm/vmx/vmx.c index 8e0aa976c2..dde4f3b70d 100644 --- a/xen/arch/x86/hvm/vmx/vmx.c +++ b/xen/arch/x86/hvm/vmx/vmx.c @@ -2973,7 +2973,7 @@ static const struct lbr_info *last_branch_msr_get(void) /* Goldmont Plus */ case 0x7a: /* Ice Lake */ - case 0x7d: case 0x7e: + case 0x6a: case 0x6c: case 0x7d: case 0x7e: /* Tiger Lake */ case 0x8c: case 0x8d: /* Tremont */ -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.15 From xen-changelog-bounces@lists.xenproject.org Fri Jun 04 12:55:36 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Fri, 04 Jun 2021 12:55:36 +0000 Received: from list by lists.xenproject.org with outflank-mailman.136815.253500 (Exim 4.92) (envelope-from ) id 1lp9MO-0004Z9-Tn; Fri, 04 Jun 2021 12:55:36 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 136815.253500; Fri, 04 Jun 2021 12:55:36 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lp9MO-0004Yz-Qm; Fri, 04 Jun 2021 12:55:36 +0000 Received: by outflank-mailman (input) for mailman id 136815; Fri, 04 Jun 2021 12:55:35 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lp9MN-0004Ya-Os for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 12:55:35 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lp9MN-0004Hf-O6 for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 12:55:35 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lp9MN-00037q-N0 for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 12:55:35 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=m2ZkIIjKr/cF24qsa/LxHg0DRqVGAHadnbYbvPBpMAU=; b=q9U2jEHHiOVcmEaF8RgBgMTEPj t0YxEOi8byIRonp67xGcK66pHfae8Rm1ezlj/Qda5qyjeBaf36yweRFAxtBRVWypDjH95c5tRpCXI ER/dD4AhhfJR6g6FpKXymL0JJmy+FlC8VBtjVQtaGMfrgvZvbJSa0ohuuNNkEhPpy93I=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.15] x86/vhpet: fix RTC special casing Message-Id: Date: Fri, 04 Jun 2021 12:55:35 +0000 commit edeaa04ab895a7283a57532ee349df19fe34c56b Author: Roger Pau Monné AuthorDate: Fri Jun 4 14:46:09 2021 +0200 Commit: Jan Beulich CommitDate: Fri Jun 4 14:46:09 2021 +0200 x86/vhpet: fix RTC special casing Restore setting the virtual timer callback private data to NULL if the timer is not level triggered. This fixes the special casing done in pt_update_irq so that the RTC interrupt when originating from the HPET is suspended if the interrupt source is masked. Note the RTC special casing done in pt_update_irq should only apply to the RTC interrupt originating from the emulated RTC device (which does set the callback private data), as in that case the callback itself will destroy the virtual timer if the interrupt is ignored. While there also use RTC_IRQ instead of 8 when the HPET is configured in LegacyReplacement Mode. Fixes: be07023be115 ("x86/vhpet: add support for level triggered interrupts") Reported-by: Jan Beulich Signed-off-by: Roger Pau Monné Reviewed-by: Jan Beulich master commit: 472a13988a051e5ae84b95815c6caf4378062abe master date: 2021-05-07 10:43:29 +0200 --- xen/arch/x86/hvm/hpet.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/xen/arch/x86/hvm/hpet.c b/xen/arch/x86/hvm/hpet.c index ca94e8b453..ee756abb82 100644 --- a/xen/arch/x86/hvm/hpet.c +++ b/xen/arch/x86/hvm/hpet.c @@ -22,6 +22,7 @@ #include #include #include +#include #include #include #include @@ -290,7 +291,7 @@ static void hpet_set_timer(HPETState *h, unsigned int tn, /* if LegacyReplacementRoute bit is set, HPET specification requires timer0 be routed to IRQ0 in NON-APIC or IRQ2 in the I/O APIC, timer1 be routed to IRQ8 in NON-APIC or IRQ8 in the I/O APIC. */ - irq = (tn == 0) ? 0 : 8; + irq = (tn == 0) ? 0 : RTC_IRQ; h->pt[tn].source = PTSRC_isa; } else @@ -318,7 +319,8 @@ static void hpet_set_timer(HPETState *h, unsigned int tn, hpet_tick_to_ns(h, diff), oneshot ? 0 : hpet_tick_to_ns(h, h->hpet.period[tn]), irq, timer_level(h, tn) ? hpet_timer_fired : NULL, - (void *)(unsigned long)tn, timer_level(h, tn)); + timer_level(h, tn) ? (void *)(unsigned long)tn : NULL, + timer_level(h, tn)); } static inline uint64_t hpet_fixup_reg( -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.15 From xen-changelog-bounces@lists.xenproject.org Fri Jun 04 12:55:47 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Fri, 04 Jun 2021 12:55:47 +0000 Received: from list by lists.xenproject.org with outflank-mailman.136816.253504 (Exim 4.92) (envelope-from ) id 1lp9MY-0004cZ-Vd; Fri, 04 Jun 2021 12:55:46 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 136816.253504; Fri, 04 Jun 2021 12:55:46 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lp9MY-0004cQ-SJ; Fri, 04 Jun 2021 12:55:46 +0000 Received: by outflank-mailman (input) for mailman id 136816; Fri, 04 Jun 2021 12:55:45 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lp9MX-0004cG-Rm for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 12:55:45 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lp9MX-0004Jn-R5 for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 12:55:45 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lp9MX-00038P-Q8 for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 12:55:45 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=9B1eV9LS+GIw/eODc/4/Vnz/1KyFJyaDpIZngxb3Vfo=; b=Dxx5ZNrDiri5qzn/cnqfxP+rYn Sr1rabIQzvNgjHMvNaWfHR2HCBSqIK3SAROQiHdCDIJHEAEg0f+0pF0Uv670HVYJMNis3lE7Zr3l9 PSIf7SfG4s1fiRguDRd5uOMqy1ulhpBnavRWfY5BYnrCSO2EeTYFaj4u8drF843TV5IA=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.15] x86emul: fix test harness build for gas 2.36 Message-Id: Date: Fri, 04 Jun 2021 12:55:45 +0000 commit ee2b1d616e8dda8edf1946c403c781983d1888b7 Author: Jan Beulich AuthorDate: Fri Jun 4 14:46:42 2021 +0200 Commit: Jan Beulich CommitDate: Fri Jun 4 14:46:42 2021 +0200 x86emul: fix test harness build for gas 2.36 All of the sudden, besides .text and .rodata and alike, an always present .note.gnu.property section has appeared. This section, when converting to binary format output, gets placed according to its linked address, causing the resulting blobs to be about 128Mb in size. The resulting headers with a C representation of the binary blobs then are, of course all a multiple of that size (and take accordingly long to create). I didn't bother waiting to see what size the final test_x86_emulator binary then would have had. See also https://sourceware.org/bugzilla/show_bug.cgi?id=27753. Rather than figuring out whether gas supports -mx86-used-note=, simply remove the section while creating *.bin. Signed-off-by: Jan Beulich master commit: aa803ba38a867551917d11059eaa044955556e05 master date: 2021-05-17 15:41:28 +0200 --- tools/tests/x86_emulator/testcase.mk | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/tools/tests/x86_emulator/testcase.mk b/tools/tests/x86_emulator/testcase.mk index dafeb6caf7..d4aa82e3f3 100644 --- a/tools/tests/x86_emulator/testcase.mk +++ b/tools/tests/x86_emulator/testcase.mk @@ -12,11 +12,11 @@ all: $(TESTCASE).bin %.bin: %.c $(CC) $(filter-out -M% .%,$(CFLAGS)) -c $< $(LD) $(LDFLAGS_DIRECT) -N -Ttext 0x100000 -o $*.tmp $*.o - $(OBJCOPY) -O binary $*.tmp $@ + $(OBJCOPY) -O binary -R .note.gnu.property $*.tmp $@ rm -f $*.tmp %-opmask.bin: opmask.S $(CC) $(filter-out -M% .%,$(CFLAGS)) -c $< -o $(basename $@).o $(LD) $(LDFLAGS_DIRECT) -N -Ttext 0x100000 -o $(basename $@).tmp $(basename $@).o - $(OBJCOPY) -O binary $(basename $@).tmp $@ + $(OBJCOPY) -O binary -R .note.gnu.property $(basename $@).tmp $@ rm -f $(basename $@).tmp -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.15 From xen-changelog-bounces@lists.xenproject.org Fri Jun 04 12:55:57 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Fri, 04 Jun 2021 12:55:57 +0000 Received: from list by lists.xenproject.org with outflank-mailman.136817.253507 (Exim 4.92) (envelope-from ) id 1lp9Mj-0004fR-1A; Fri, 04 Jun 2021 12:55:57 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 136817.253507; Fri, 04 Jun 2021 12:55:56 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lp9Mi-0004fJ-Tq; Fri, 04 Jun 2021 12:55:56 +0000 Received: by outflank-mailman (input) for mailman id 136817; Fri, 04 Jun 2021 12:55:56 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lp9Mh-0004f1-Uh for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 12:55:55 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lp9Mh-0004Jx-Tz for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 12:55:55 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lp9Mh-00039W-TB for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 12:55:55 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=HDCcLwu17zh/HcHbmfhJEY8FGMdItzR9wHFoRXVIhek=; b=mWpaKTGvwo9A2ChBJR0zQNB8Vw wZ2BTMia18RpzkM+1miISgu1JYPE9xJgYugb0Cq7Dqp1yfNGbVXjsfmqPHZDKGmnn7SC1zFwU77Sj +6zVYnK3zh+pRCdqIpQcyXrFOW89yUYgT2DcBo9P0dL2sDQoimsARs0Dsy6xph9mqnc4=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.15] firmware/shim: update linkfarm exclusions Message-Id: Date: Fri, 04 Jun 2021 12:55:55 +0000 commit 6a7e21a135148ef5bf6a62177dfbd5c104e3db13 Author: Jan Beulich AuthorDate: Fri Jun 4 14:47:14 2021 +0200 Commit: Jan Beulich CommitDate: Fri Jun 4 14:47:14 2021 +0200 firmware/shim: update linkfarm exclusions Some intermediate files weren't considered at all at the time. Also after its introduction, various changes to the build environment have rendered the exclusion sets stale. For example, we now have some .*.cmd files in the build tree. Combine all respective patterns into a single .* one, seeing that we don't have any actual source files matching this pattern in the tree. Add other patterns as well as individual files. Also introduce LINK_EXCLUDE_PATHS to deal with entire directories full of generated headers as well as a few specific files the names of which are too generic to list under LINK_EXCLUDES. Signed-off-by: Jan Beulich Acked-by: Roger Pau Monné master commit: 022da003747a7f4336f8c581fdf898a7d5cd47e4 master date: 2021-05-26 09:34:07 +0200 --- tools/firmware/xen-dir/Makefile | 26 +++++++++++++++++++------- 1 file changed, 19 insertions(+), 7 deletions(-) diff --git a/tools/firmware/xen-dir/Makefile b/tools/firmware/xen-dir/Makefile index 5413f86790..afc651f861 100644 --- a/tools/firmware/xen-dir/Makefile +++ b/tools/firmware/xen-dir/Makefile @@ -15,9 +15,19 @@ DEP_DIRS=$(foreach i, $(LINK_DIRS), $(XEN_ROOT)/$(i)) DEP_FILES=$(foreach i, $(LINK_FILES), $(XEN_ROOT)/$(i)) # Exclude some intermediate files and final build products -LINK_EXCLUDES := '*.[isoa]' '.*.d' '.*.d2' '.config' -LINK_EXCLUDES += '*.map' 'xen' 'xen.gz' 'xen.efi' 'xen-syms' -LINK_EXCLUDES += '.*.tmp' +LINK_EXCLUDES := '*.[isoa]' '*.bin' '*.chk' '*.lnk' '*.gz' '.*' +LINK_EXCLUDES += lexer.lex.? parser.tab.? conf +LINK_EXCLUDES += asm-offsets.h asm-macros.h compile.h '*-autogen.h' +LINK_EXCLUDES += mkelf32 mkreloc symbols config_data.S xen.lds efi.lds +LINK_EXCLUDES += '*.map' xen xen.gz xen.efi xen-syms check.efi + +# To exclude full subtrees or individual files of not sufficiently specific +# names, regular expressions are used: +LINK_EXCLUDE_PATHS := xen/include/compat/.* +LINK_EXCLUDE_PATHS += xen/include/config/.* +LINK_EXCLUDE_PATHS += xen/include/generated/.* +LINK_EXCLUDE_PATHS += xen/arch/x86/boot/reloc[.]S +LINK_EXCLUDE_PATHS += xen/arch/x86/boot/cmdline[.]S # This is all a giant mess and doesn't really work. # @@ -32,9 +42,10 @@ LINK_EXCLUDES += '.*.tmp' # support easy development of the shim, but has a side effect of clobbering # the already-built shim. # -# $(LINK_EXCLUDES) should be set such that a parallel build of shim and xen/ -# doesn't cause a subsequent `make install` to decide to regenerate the -# linkfarm. This means that all final build artefacts must be excluded. +# $(LINK_EXCLUDES) and $(LINK_EXCLUDE_DIRS) should be set such that a parallel +# build of shim and xen/ doesn't cause a subsequent `make install` to decide to +# regenerate the linkfarm. This means that all intermediate and final build +# artefacts must be excluded. linkfarm.stamp: $(DEP_DIRS) $(DEP_FILES) FORCE mkdir -p $(D) rm -f linkfarm.stamp.tmp @@ -46,7 +57,8 @@ linkfarm.stamp: $(DEP_DIRS) $(DEP_FILES) FORCE sed 's,^$(XEN_ROOT)/$(d)/,,g' | xargs mkdir -p .);) \ $(foreach d, $(LINK_DIRS), \ (cd $(XEN_ROOT); \ - find $(d) ! -type l -type f $(addprefix ! -name ,$(LINK_EXCLUDES))) \ + find $(d) ! -type l -type f $(addprefix ! -name ,$(LINK_EXCLUDES)) \ + | grep -v $(patsubst %,-e '^%$$',$(LINK_EXCLUDE_PATHS))) \ >> linkfarm.stamp.tmp ; ) \ $(foreach f, $(LINK_FILES), \ echo $(f) >> linkfarm.stamp.tmp ;) -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.15 From xen-changelog-bounces@lists.xenproject.org Fri Jun 04 12:56:07 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Fri, 04 Jun 2021 12:56:07 +0000 Received: from list by lists.xenproject.org with outflank-mailman.136818.253512 (Exim 4.92) (envelope-from ) id 1lp9Mt-0004im-4o; Fri, 04 Jun 2021 12:56:07 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 136818.253512; Fri, 04 Jun 2021 12:56:07 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lp9Mt-0004ie-1N; Fri, 04 Jun 2021 12:56:07 +0000 Received: by outflank-mailman (input) for mailman id 136818; Fri, 04 Jun 2021 12:56:06 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lp9Ms-0004iR-27 for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 12:56:06 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lp9Ms-0004KL-1G for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 12:56:06 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lp9Ms-0003AR-0S for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 12:56:06 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=vpnnI8a9lYlhFsW2tCPxUdMb02ej2Yuz17bw9P2jxR8=; b=OepWtl1JYbHH9OkoPD2ayWslNo +btR0Uz7cCK7pe9xVrSK/Q3Z0YPVCXXtzcP16fz+5Mxu2a2GUxePFQaKmx8zkf6WHq1Fe3aOSEdkQ Kcho6U6bB27jFsASDlYrOX/ZuTwgZ6jeFRywqInFysDWlo4+Op9FaR4+1O9Da0PG+fOE=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.15] x86: make hypervisor build with gcc11 Message-Id: Date: Fri, 04 Jun 2021 12:56:06 +0000 commit 7c3c98497f8b5e413ca1858f71f369170ead8d5b Author: Jan Beulich AuthorDate: Fri Jun 4 14:47:36 2021 +0200 Commit: Jan Beulich CommitDate: Fri Jun 4 14:47:36 2021 +0200 x86: make hypervisor build with gcc11 Gcc 11 looks to make incorrect assumptions about valid ranges that pointers may be used for addressing when they are derived from e.g. a plain constant. See https://gcc.gnu.org/bugzilla/show_bug.cgi?id=100680. Utilize RELOC_HIDE() to work around the issue, which for x86 manifests in at least - mpparse.c:efi_check_config(), - tboot.c:tboot_probe(), - tboot.c:tboot_gen_frametable_integrity(), - x86_emulate.c:x86_emulate() (at -O2 only). The last case is particularly odd not just because it only triggers at higher optimization levels, but also because it only affects one of at least three similar constructs. Various "note" diagnostics claim the valid index range to be [0, 2⁶³-1]. Signed-off-by: Jan Beulich Tested-by: Jason Andryuk Acked-by: Roger Pau Monné master commit: 722f59d38c710a940ab05e542a83020eb5546dea master date: 2021-05-27 14:40:29 +0200 --- tools/tests/x86_emulator/x86-emulate.c | 7 +++++++ xen/arch/x86/x86_emulate/x86_emulate.c | 2 +- xen/include/asm-x86/fixmap.h | 2 +- xen/include/xen/compiler.h | 6 ++++++ xen/include/xen/pdx.h | 2 +- 5 files changed, 16 insertions(+), 3 deletions(-) diff --git a/tools/tests/x86_emulator/x86-emulate.c b/tools/tests/x86_emulator/x86-emulate.c index fb34933e82..adb4f7aca4 100644 --- a/tools/tests/x86_emulator/x86-emulate.c +++ b/tools/tests/x86_emulator/x86-emulate.c @@ -8,6 +8,13 @@ #define ERR_PTR(val) NULL +/* See gcc bug 100680, but here don't bother making this version dependent. */ +#define gcc11_wrap(x) ({ \ + unsigned long x_; \ + __asm__ ( "" : "=g" (x_) : "0" (x) ); \ + (typeof(x))x_; \ +}) + #define cpu_has_amd_erratum(nr) 0 #define cpu_has_mpx false #define read_bndcfgu() 0 diff --git a/xen/arch/x86/x86_emulate/x86_emulate.c b/xen/arch/x86/x86_emulate/x86_emulate.c index 4a33fe9613..a8000cbcbd 100644 --- a/xen/arch/x86/x86_emulate/x86_emulate.c +++ b/xen/arch/x86/x86_emulate/x86_emulate.c @@ -726,7 +726,7 @@ union vex { #define copy_VEX(ptr, vex) ({ \ if ( !mode_64bit() ) \ (vex).reg |= 8; \ - (ptr)[0 - PFX_BYTES] = ext < ext_8f08 ? 0xc4 : 0x8f; \ + gcc11_wrap(ptr)[0 - PFX_BYTES] = ext < ext_8f08 ? 0xc4 : 0x8f; \ (ptr)[1 - PFX_BYTES] = (vex).raw[0]; \ (ptr)[2 - PFX_BYTES] = (vex).raw[1]; \ container_of((ptr) + 1 - PFX_BYTES, typeof(vex), raw[0]); \ diff --git a/xen/include/asm-x86/fixmap.h b/xen/include/asm-x86/fixmap.h index 0db314baeb..20746afd0a 100644 --- a/xen/include/asm-x86/fixmap.h +++ b/xen/include/asm-x86/fixmap.h @@ -78,7 +78,7 @@ extern void __set_fixmap( #define clear_fixmap(idx) __set_fixmap(idx, 0, 0) -#define __fix_to_virt(x) (FIXADDR_TOP - ((x) << PAGE_SHIFT)) +#define __fix_to_virt(x) gcc11_wrap(FIXADDR_TOP - ((x) << PAGE_SHIFT)) #define __virt_to_fix(x) ((FIXADDR_TOP - ((x)&PAGE_MASK)) >> PAGE_SHIFT) #define fix_to_virt(x) ((void *)__fix_to_virt(x)) diff --git a/xen/include/xen/compiler.h b/xen/include/xen/compiler.h index 17cf00e1ec..696c7eb89e 100644 --- a/xen/include/xen/compiler.h +++ b/xen/include/xen/compiler.h @@ -140,6 +140,12 @@ __asm__ ("" : "=r"(__ptr) : "0"(ptr)); \ (typeof(ptr)) (__ptr + (off)); }) +#if CONFIG_GCC_VERSION >= 110000 /* See gcc bug 100680. */ +# define gcc11_wrap(x) RELOC_HIDE(x, 0) +#else +# define gcc11_wrap(x) (x) +#endif + #ifdef __GCC_ASM_FLAG_OUTPUTS__ # define ASM_FLAG_OUT(yes, no) yes #else diff --git a/xen/include/xen/pdx.h b/xen/include/xen/pdx.h index 770fadc06c..9fcfb0ce52 100644 --- a/xen/include/xen/pdx.h +++ b/xen/include/xen/pdx.h @@ -19,7 +19,7 @@ extern u64 pdx_region_mask(u64 base, u64 len); extern void set_pdx_range(unsigned long smfn, unsigned long emfn); #define page_to_pdx(pg) ((pg) - frame_table) -#define pdx_to_page(pdx) (frame_table + (pdx)) +#define pdx_to_page(pdx) gcc11_wrap(frame_table + (pdx)) bool __mfn_valid(unsigned long mfn); -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.15 From xen-changelog-bounces@lists.xenproject.org Fri Jun 04 12:56:17 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Fri, 04 Jun 2021 12:56:17 +0000 Received: from list by lists.xenproject.org with outflank-mailman.136819.253516 (Exim 4.92) (envelope-from ) id 1lp9N3-0004m1-5z; Fri, 04 Jun 2021 12:56:17 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 136819.253516; Fri, 04 Jun 2021 12:56:17 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lp9N3-0004lt-2y; Fri, 04 Jun 2021 12:56:17 +0000 Received: by outflank-mailman (input) for mailman id 136819; Fri, 04 Jun 2021 12:56:16 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lp9N2-0004ln-5D for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 12:56:16 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lp9N2-0004KY-4U for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 12:56:16 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lp9N2-0003BY-3k for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 12:56:16 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=jFV0e3bGDl8YGDMVh/Ab+JC3PHnAovUc+amna35XZwA=; b=VolnB6nCx1RsaqY4pm1fLjUM63 adYt5KgC8ngpHQyhUoybTY/jN/A6xzW7Dq99mLpfAfBcnDqVpOVvUR+2Y4ndCRk+MKM9lQ4reTkp3 nOLle/XSy1JOo6VjBvnykk81ujt1xd+xIdEcpBR/6mG/H2WQ6aAlUtFRzNgLX6jSr+YU=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.15] x86/cpuid: Rework HLE and RTM handling Message-Id: Date: Fri, 04 Jun 2021 12:56:16 +0000 commit 89c6e84f6dd31a73ed9037194c9a61081e000c38 Author: Andrew Cooper AuthorDate: Fri Jun 4 14:49:05 2021 +0200 Commit: Jan Beulich CommitDate: Fri Jun 4 14:49:05 2021 +0200 x86/cpuid: Rework HLE and RTM handling The TAA mitigation offered the option to hide the HLE and RTM CPUID bits, which has caused some migration compatibility problems. These two bits are special. Annotate them with ! to emphasise this point. Hardware Lock Elision (HLE) may or may not be visible in CPUID, but is disabled in microcode on all CPUs, and has been removed from the architecture. Do not advertise it to VMs by default. Restricted Transactional Memory (RTM) may or may not be visible in CPUID, and may or may not be configured in force-abort mode. Have tsx_init() note whether RTM has been configured into force-abort mode, so guest_common_feature_adjustments() can conditionally hide it from VMs by default. The host policy values for HLE/RTM may or may not be set, depending on any previous running kernel's choice of visibility, and Xen's choice. TSX is available on any CPU which enumerates a TSX-hiding mechanism, so instead of doing a two-step to clobber any hiding, scan CPUID, then set the visibility, just force visibility of the bits in the first place. With the HLE/RTM bits now unilaterally visible in the host policy, xc_cpuid_apply_policy() can construct a more appropriate policy out of thin air for pre-4.13 VMs with no CPUID data in their migration stream, and specifically one where HLE/RTM doesn't potentially disappear behind the back of a running VM. Fixes: 8c4330818f6 ("x86/spec-ctrl: Mitigate the TSX Asynchronous Abort sidechannel") Signed-off-by: Andrew Cooper Reviewed-by: Jan Beulich Reviewed-by: Roger Pau Monné master commit: 8fe24090d940d760145ccd5e234290be7418b175 master date: 2021-05-27 19:34:00 +0100 --- tools/libs/guest/xg_cpuid_x86.c | 2 ++ xen/arch/x86/cpuid.c | 24 ++++++++++------------ xen/arch/x86/spec_ctrl.c | 3 --- xen/arch/x86/tsx.c | 31 +++++++++++++++++++++++++++-- xen/include/asm-x86/processor.h | 1 + xen/include/public/arch-x86/cpufeatureset.h | 4 ++-- 6 files changed, 44 insertions(+), 21 deletions(-) diff --git a/tools/libs/guest/xg_cpuid_x86.c b/tools/libs/guest/xg_cpuid_x86.c index aae6931a11..c6c806884a 100644 --- a/tools/libs/guest/xg_cpuid_x86.c +++ b/tools/libs/guest/xg_cpuid_x86.c @@ -504,6 +504,8 @@ int xc_cpuid_apply_policy(xc_interface *xch, uint32_t domid, bool restore, if ( restore ) { p->basic.rdrand = test_bit(X86_FEATURE_RDRAND, host_featureset); + p->feat.hle = test_bit(X86_FEATURE_HLE, host_featureset); + p->feat.rtm = test_bit(X86_FEATURE_RTM, host_featureset); if ( di.hvm ) { diff --git a/xen/arch/x86/cpuid.c b/xen/arch/x86/cpuid.c index dde3e95284..7ac6636efc 100644 --- a/xen/arch/x86/cpuid.c +++ b/xen/arch/x86/cpuid.c @@ -357,6 +357,16 @@ static void __init guest_common_default_feature_adjustments(uint32_t *fs) boot_cpu_data.x86 == 6 && boot_cpu_data.x86_model == 0x3a && cpu_has_rdrand && !is_forced_cpu_cap(X86_FEATURE_RDRAND) ) __clear_bit(X86_FEATURE_RDRAND, fs); + + /* + * On certain hardware, speculative or errata workarounds can result in + * TSX being placed in "force-abort" mode, where it doesn't actually + * function as expected, but is technically compatible with the ISA. + * + * Do not advertise RTM to guests by default if it won't actually work. + */ + if ( rtm_disabled ) + __clear_bit(X86_FEATURE_RTM, fs); } static void __init guest_common_feature_adjustments(uint32_t *fs) @@ -620,20 +630,6 @@ void recalculate_cpuid_policy(struct domain *d) __clear_bit(X86_FEATURE_SYSCALL, max_fs); } - /* - * On hardware with MSR_TSX_CTRL, the admin may have elected to disable - * TSX and hide the feature bits. Migrating-in VMs may have been booted - * pre-mitigation when the TSX features were visbile. - * - * This situation is compatible (albeit with a perf hit to any TSX code in - * the guest), so allow the feature bits to remain set. - */ - if ( cpu_has_tsx_ctrl ) - { - __set_bit(X86_FEATURE_HLE, max_fs); - __set_bit(X86_FEATURE_RTM, max_fs); - } - /* Clamp the toolstacks choices to reality. */ for ( i = 0; i < ARRAY_SIZE(fs); i++ ) fs[i] &= max_fs[i]; diff --git a/xen/arch/x86/spec_ctrl.c b/xen/arch/x86/spec_ctrl.c index c9f78ead62..bea4c3e8b4 100644 --- a/xen/arch/x86/spec_ctrl.c +++ b/xen/arch/x86/spec_ctrl.c @@ -1158,9 +1158,6 @@ void __init init_speculation_mitigations(void) ((hw_smt_enabled && opt_smt) || !boot_cpu_has(X86_FEATURE_SC_VERW_IDLE)) ) { - setup_clear_cpu_cap(X86_FEATURE_HLE); - setup_clear_cpu_cap(X86_FEATURE_RTM); - opt_tsx = 0; tsx_init(); } diff --git a/xen/arch/x86/tsx.c b/xen/arch/x86/tsx.c index 39e483640a..e09e819dce 100644 --- a/xen/arch/x86/tsx.c +++ b/xen/arch/x86/tsx.c @@ -15,6 +15,7 @@ */ int8_t __read_mostly opt_tsx = -1; int8_t __read_mostly cpu_has_tsx_ctrl = -1; +bool __read_mostly rtm_disabled; static int __init parse_tsx(const char *s) { @@ -45,6 +46,30 @@ void tsx_init(void) rdmsrl(MSR_ARCH_CAPABILITIES, caps); cpu_has_tsx_ctrl = !!(caps & ARCH_CAPS_TSX_CTRL); + + /* + * The TSX features (HLE/RTM) are handled specially. They both + * enumerate features but, on certain parts, have mechanisms to be + * hidden without disrupting running software. + * + * At the moment, we're running in an unknown context (WRT hiding - + * particularly if another fully fledged kernel ran before us) and + * depending on user settings, may elect to continue hiding them from + * native CPUID instructions. + * + * Xen doesn't use TSX itself, but use cpu_has_{hle,rtm} for various + * system reasons, mostly errata detection, so the meaning is more + * useful as "TSX infrastructure available", as opposed to "features + * advertised and working". + * + * Force the features to be visible in Xen's view if we see any of the + * infrastructure capable of hiding them. + */ + if ( cpu_has_tsx_ctrl ) + { + setup_force_cpu_cap(X86_FEATURE_HLE); + setup_force_cpu_cap(X86_FEATURE_RTM); + } } if ( cpu_has_tsx_ctrl ) @@ -53,9 +78,11 @@ void tsx_init(void) rdmsrl(MSR_TSX_CTRL, val); + /* Check bottom bit only. Higher bits are various sentinels. */ + rtm_disabled = !(opt_tsx & 1); + val &= ~(TSX_CTRL_RTM_DISABLE | TSX_CTRL_CPUID_CLEAR); - /* Check bottom bit only. Higher bits are various sentinals. */ - if ( !(opt_tsx & 1) ) + if ( rtm_disabled ) val |= TSX_CTRL_RTM_DISABLE | TSX_CTRL_CPUID_CLEAR; wrmsrl(MSR_TSX_CTRL, val); diff --git a/xen/include/asm-x86/processor.h b/xen/include/asm-x86/processor.h index d5f467d245..44752bfdf6 100644 --- a/xen/include/asm-x86/processor.h +++ b/xen/include/asm-x86/processor.h @@ -629,6 +629,7 @@ static inline uint8_t get_cpu_family(uint32_t raw, uint8_t *model, } extern int8_t opt_tsx, cpu_has_tsx_ctrl; +extern bool rtm_disabled; void tsx_init(void); enum ap_boot_method { diff --git a/xen/include/public/arch-x86/cpufeatureset.h b/xen/include/public/arch-x86/cpufeatureset.h index a501479820..6477f75505 100644 --- a/xen/include/public/arch-x86/cpufeatureset.h +++ b/xen/include/public/arch-x86/cpufeatureset.h @@ -197,14 +197,14 @@ XEN_CPUFEATURE(FSGSBASE, 5*32+ 0) /*A {RD,WR}{FS,GS}BASE instructions */ XEN_CPUFEATURE(TSC_ADJUST, 5*32+ 1) /*S TSC_ADJUST MSR available */ XEN_CPUFEATURE(SGX, 5*32+ 2) /* Software Guard extensions */ XEN_CPUFEATURE(BMI1, 5*32+ 3) /*A 1st bit manipulation extensions */ -XEN_CPUFEATURE(HLE, 5*32+ 4) /*A Hardware Lock Elision */ +XEN_CPUFEATURE(HLE, 5*32+ 4) /*!a Hardware Lock Elision */ XEN_CPUFEATURE(AVX2, 5*32+ 5) /*A AVX2 instructions */ XEN_CPUFEATURE(FDP_EXCP_ONLY, 5*32+ 6) /*! x87 FDP only updated on exception. */ XEN_CPUFEATURE(SMEP, 5*32+ 7) /*S Supervisor Mode Execution Protection */ XEN_CPUFEATURE(BMI2, 5*32+ 8) /*A 2nd bit manipulation extensions */ XEN_CPUFEATURE(ERMS, 5*32+ 9) /*A Enhanced REP MOVSB/STOSB */ XEN_CPUFEATURE(INVPCID, 5*32+10) /*H Invalidate Process Context ID */ -XEN_CPUFEATURE(RTM, 5*32+11) /*A Restricted Transactional Memory */ +XEN_CPUFEATURE(RTM, 5*32+11) /*!A Restricted Transactional Memory */ XEN_CPUFEATURE(PQM, 5*32+12) /* Platform QoS Monitoring */ XEN_CPUFEATURE(NO_FPU_SEL, 5*32+13) /*! FPU CS/DS stored as zero */ XEN_CPUFEATURE(MPX, 5*32+14) /*s Memory Protection Extensions */ -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.15 From xen-changelog-bounces@lists.xenproject.org Fri Jun 04 12:56:27 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Fri, 04 Jun 2021 12:56:27 +0000 Received: from list by lists.xenproject.org with outflank-mailman.136820.253520 (Exim 4.92) (envelope-from ) id 1lp9ND-0004or-7a; Fri, 04 Jun 2021 12:56:27 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 136820.253520; Fri, 04 Jun 2021 12:56:27 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lp9ND-0004oi-4c; Fri, 04 Jun 2021 12:56:27 +0000 Received: by outflank-mailman (input) for mailman id 136820; Fri, 04 Jun 2021 12:56:26 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lp9NC-0004oY-87 for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 12:56:26 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lp9NC-0004Kl-7R for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 12:56:26 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lp9NC-0003CC-6W for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 12:56:26 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=kuX1bTVUdTGenWtkmQSfCBiiAnlIJNoCcbAkvTzEhgQ=; b=D3thZ8bn3hPLslhSoHx7Nb6hqL VO8amnHV2omTyd3Asf+t33gdtrWU4+C6Kjj0d7GLjynceyO1pMNOh7hBAUr9jHSge+qdRVH5bRH1F n88UHVrEBMUR3/34pVrTcAT7frxr3wHnqViFME5AT+D72jo0h5J+932gNHtQmNrjEXdI=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.15] x86: fix build race when generating temporary object files (take 2) Message-Id: Date: Fri, 04 Jun 2021 12:56:26 +0000 commit eae0dfac891f521ceb6c4733e22a0cd718f336c0 Author: Jan Beulich AuthorDate: Fri Jun 4 14:49:40 2021 +0200 Commit: Jan Beulich CommitDate: Fri Jun 4 14:49:40 2021 +0200 x86: fix build race when generating temporary object files (take 2) The original commit wasn't quite sufficient: Emptying DEPS is helpful only when nothing will get added to it subsequently. xen/Rules.mk will, after including the local Makefile, amend DEPS by dependencies for objects living in sub-directories though. For the purpose of suppressing dependencies of the makefiles on the .*.d2 files (and thus to avoid their re-generation) it is, however, not necessary at all to play with DEPS. Instead we can override DEPS_INCLUDE (which generally is a late- expansion variable). Fixes: 761bb575ce97 ("x86: fix build race when generating temporary object files") Signed-off-by: Jan Beulich master commit: 8c90dbb99907f3b471d558775777a84daec7c3f6 master date: 2021-05-28 09:12:24 +0200 --- xen/arch/x86/Makefile | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/xen/arch/x86/Makefile b/xen/arch/x86/Makefile index b6567c4127..b560c12a58 100644 --- a/xen/arch/x86/Makefile +++ b/xen/arch/x86/Makefile @@ -286,5 +286,5 @@ clean:: # Suppress loading of DEPS files for internal, temporary target files. This # then also suppresses re-generation of the respective .*.d2 files. ifeq ($(filter-out .xen%.o,$(notdir $(MAKECMDGOALS))),) -DEPS:= +DEPS_INCLUDE:= endif -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.15 From xen-changelog-bounces@lists.xenproject.org Fri Jun 04 13:00:07 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Fri, 04 Jun 2021 13:00:07 +0000 Received: from list by lists.xenproject.org with outflank-mailman.136821.253524 (Exim 4.92) (envelope-from ) id 1lp9Ql-0005eg-Ha; Fri, 04 Jun 2021 13:00:07 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 136821.253524; Fri, 04 Jun 2021 13:00:07 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lp9Ql-0005eY-EO; Fri, 04 Jun 2021 13:00:07 +0000 Received: by outflank-mailman (input) for mailman id 136821; Fri, 04 Jun 2021 13:00:05 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lp9Qj-0005WF-Gi for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 13:00:05 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lp9Qj-0004PX-G1 for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 13:00:05 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lp9Qj-0003bf-FD for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 13:00:05 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=F3nrbUImyomK37UI6CPMRnmKRD9aLYM4eRd9e91KBH8=; b=vEES9ZI3JZ/DhiIYbdXjXE7ZHK 4TcvYDY/DlF/VS1Iy1FqiBJHbB6lKh1N90BtCLbAz1JeyJEtoHCUFqZeLehMiussfKqBjaw9qcZ/l 7crxobNF5WvzpoaQwCbaV9A5pBWx8X1EQ/WJznKT3Hzj69ljMMEfBDABkapDlysOKI6s=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.14] x86/Intel: insert Tiger Lake model numbers Message-Id: Date: Fri, 04 Jun 2021 13:00:05 +0000 commit 02f9760498ce1c4fb320045f3f3b8f515d124ad4 Author: Jan Beulich AuthorDate: Fri Jun 4 14:51:25 2021 +0200 Commit: Jan Beulich CommitDate: Fri Jun 4 14:51:25 2021 +0200 x86/Intel: insert Tiger Lake model numbers Both match prior generation processors as far as LBR and C-state MSRs go (SDM rev 073). The if_pschange_mc erratum, according to the spec update, is not applicable. Signed-off-by: Jan Beulich Acked-by: Andrew Cooper master commit: e93c3712d67098453760fd61c338cbf62dd08da1 master date: 2020-12-22 09:00:03 +0100 --- xen/arch/x86/acpi/cpu_idle.c | 3 +++ xen/arch/x86/hvm/vmx/vmx.c | 2 ++ 2 files changed, 5 insertions(+) diff --git a/xen/arch/x86/acpi/cpu_idle.c b/xen/arch/x86/acpi/cpu_idle.c index 27e0b52621..c092086b33 100644 --- a/xen/arch/x86/acpi/cpu_idle.c +++ b/xen/arch/x86/acpi/cpu_idle.c @@ -183,6 +183,9 @@ static void do_get_hw_residencies(void *arg) /* Ice Lake */ case 0x7D: case 0x7E: + /* Tiger Lake */ + case 0x8C: + case 0x8D: /* Kaby Lake */ case 0x8E: case 0x9E: diff --git a/xen/arch/x86/hvm/vmx/vmx.c b/xen/arch/x86/hvm/vmx/vmx.c index cc6d4ece22..ca47f83cd4 100644 --- a/xen/arch/x86/hvm/vmx/vmx.c +++ b/xen/arch/x86/hvm/vmx/vmx.c @@ -2787,6 +2787,8 @@ static const struct lbr_info *last_branch_msr_get(void) case 0x7a: /* Ice Lake */ case 0x7d: case 0x7e: + /* Tiger Lake */ + case 0x8c: case 0x8d: /* Tremont */ case 0x86: /* Kaby Lake */ -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.14 From xen-changelog-bounces@lists.xenproject.org Fri Jun 04 13:00:19 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Fri, 04 Jun 2021 13:00:19 +0000 Received: from list by lists.xenproject.org with outflank-mailman.136822.253528 (Exim 4.92) (envelope-from ) id 1lp9Qv-0005gp-Iv; Fri, 04 Jun 2021 13:00:17 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 136822.253528; Fri, 04 Jun 2021 13:00:17 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lp9Qv-0005gh-Fy; Fri, 04 Jun 2021 13:00:17 +0000 Received: by outflank-mailman (input) for mailman id 136822; Fri, 04 Jun 2021 13:00:15 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lp9Qt-0005gb-NF for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 13:00:15 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lp9Qt-0004Ph-JT for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 13:00:15 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lp9Qt-0003di-IT for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 13:00:15 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=K10dEZOWOBZlfvZgFz7O4AcyO3o528Qr20vObbqP+Sc=; b=4LmkOh1QTa9w40mhoav5L0MnyR XS7zwEIokg56Vty5CV/yRRtWk1zM298b5K5RnWg79k7uoKmheTwADtEdgeyU8RXPdjWJ3J3zj6eHH l6Y/tvRq3B4nOccWTQ0MmkxOkjr1aEKxorh3/3EHBQIHqs3Jd0VLqeUn9qhNVxy79/Ss=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.14] update Xen version to 4.14.3-pre Message-Id: Date: Fri, 04 Jun 2021 13:00:15 +0000 commit a351751d88973da710a3a902a7ced2ede9df78b0 Author: Jan Beulich AuthorDate: Fri Jun 4 14:52:04 2021 +0200 Commit: Jan Beulich CommitDate: Fri Jun 4 14:52:04 2021 +0200 update Xen version to 4.14.3-pre --- xen/Makefile | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/xen/Makefile b/xen/Makefile index 4ffa2ed7d5..22362d7348 100644 --- a/xen/Makefile +++ b/xen/Makefile @@ -2,7 +2,7 @@ # All other places this is stored (eg. compile.h) should be autogenerated. export XEN_VERSION = 4 export XEN_SUBVERSION = 14 -export XEN_EXTRAVERSION ?= .2$(XEN_VENDORVERSION) +export XEN_EXTRAVERSION ?= .3-pre$(XEN_VENDORVERSION) export XEN_FULLVERSION = $(XEN_VERSION).$(XEN_SUBVERSION)$(XEN_EXTRAVERSION) -include xen-version -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.14 From xen-changelog-bounces@lists.xenproject.org Fri Jun 04 13:00:27 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Fri, 04 Jun 2021 13:00:27 +0000 Received: from list by lists.xenproject.org with outflank-mailman.136823.253532 (Exim 4.92) (envelope-from ) id 1lp9R5-0005jq-Kg; Fri, 04 Jun 2021 13:00:27 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 136823.253532; Fri, 04 Jun 2021 13:00:27 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lp9R5-0005ji-HU; Fri, 04 Jun 2021 13:00:27 +0000 Received: by outflank-mailman (input) for mailman id 136823; Fri, 04 Jun 2021 13:00:25 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lp9R3-0005jD-Nq for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 13:00:25 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lp9R3-0004Pq-N2 for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 13:00:25 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lp9R3-0003ee-M2 for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 13:00:25 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=b53RIwY2o08/wtbo8zSpGFO5EZKPRSQgGEJnVekWhRw=; b=fJH6olUxoZMKoMutnF6YwZ1eoo RpqbxZXBFHc7ZRSIsI07kCUQvH5rsRndXzjyIiv2ZgHgHqlgqWF3TfvGiZbbQst2LD99s92zVt04G 6cak5ldEQOuIm+3kbKpj/AZU4V4zrtkByrY0SaKkmRAHBXxyuNZs7uY8q30JmCfKycG0=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.14] VT-d: Don't assume register-based invalidation is always supported Message-Id: Date: Fri, 04 Jun 2021 13:00:25 +0000 commit c2f78b49050335f9a0871263e622d7a10fdccbc4 Author: Chao Gao AuthorDate: Fri Jun 4 14:53:25 2021 +0200 Commit: Jan Beulich CommitDate: Fri Jun 4 14:53:25 2021 +0200 VT-d: Don't assume register-based invalidation is always supported According to Intel VT-d SPEC rev3.3 Section 6.5, Register-based Invalidation isn't supported by Intel VT-d version 6 and beyond. This hardware change impacts following two scenarios: admin can disable queued invalidation via 'qinval' cmdline and use register-based interface; VT-d switches to register-based invalidation when queued invalidation needs to be disabled, for example, during disabling x2apic or during system suspension or after enabling queued invalidation fails. To deal with this hardware change, if register-based invalidation isn't supported, queued invalidation cannot be disabled through Xen cmdline; and if queued invalidation has to be disabled temporarily in some scenarios, VT-d won't switch to register-based interface but use some dummy functions to catch errors in case there is any invalidation request issued when queued invalidation is disabled. Signed-off-by: Chao Gao Reviewed-by: Jan Beulich Reviewed-by: Kevin Tian master commit: 6773b1a7584a75a486e9774541ad5bd84c9aa5ee master date: 2021-04-26 10:16:50 +0200 --- docs/misc/xen-command-line.pandoc | 4 +++- xen/drivers/passthrough/vtd/iommu.c | 27 +++++++++++++++++++++++++-- xen/drivers/passthrough/vtd/iommu.h | 7 +++++++ xen/drivers/passthrough/vtd/qinval.c | 33 +++++++++++++++++++++++++++++++-- 4 files changed, 66 insertions(+), 5 deletions(-) diff --git a/docs/misc/xen-command-line.pandoc b/docs/misc/xen-command-line.pandoc index f63fe9680a..b9824d443f 100644 --- a/docs/misc/xen-command-line.pandoc +++ b/docs/misc/xen-command-line.pandoc @@ -1405,7 +1405,9 @@ The following options are specific to Intel VT-d hardware: * The `qinval` boolean controls the Queued Invalidation sub-feature, and is active by default on compatible hardware. Queued Invalidation is a feature in second-generation IOMMUs and is a functional prerequisite for - Interrupt Remapping. + Interrupt Remapping. Note that Xen disregards this setting for Intel VT-d + version 6 and greater as Registered-Based Invalidation isn't supported + by them. * The `igfx` boolean is active by default, and controls whether the IOMMU in front of an Intel Graphics Device is enabled or not. diff --git a/xen/drivers/passthrough/vtd/iommu.c b/xen/drivers/passthrough/vtd/iommu.c index 74c7aca24b..32ff60bfe3 100644 --- a/xen/drivers/passthrough/vtd/iommu.c +++ b/xen/drivers/passthrough/vtd/iommu.c @@ -1188,6 +1188,14 @@ int __init iommu_alloc(struct acpi_drhd_unit *drhd) iommu->cap = dmar_readq(iommu->reg, DMAR_CAP_REG); iommu->ecap = dmar_readq(iommu->reg, DMAR_ECAP_REG); + iommu->version = dmar_readl(iommu->reg, DMAR_VER_REG); + + if ( !iommu_qinval && !has_register_based_invalidation(iommu) ) + { + printk(XENLOG_WARNING VTDPREFIX "IOMMU %d: cannot disable Queued Invalidation\n", + iommu->index); + iommu_qinval = true; + } if ( iommu_verbose ) { @@ -2161,6 +2169,10 @@ static int __must_check init_vtd_hw(bool resume) */ if ( enable_qinval(iommu) != 0 ) { + /* Ensure register-based invalidation is available */ + if ( !has_register_based_invalidation(iommu) ) + return -EIO; + iommu->flush.context = vtd_flush_context_reg; iommu->flush.iotlb = vtd_flush_iotlb_reg; } @@ -2251,6 +2263,7 @@ static int __init vtd_setup(void) struct acpi_drhd_unit *drhd; struct vtd_iommu *iommu; int ret; + bool reg_inval_supported = true; if ( list_empty(&acpi_drhd_units) ) { @@ -2272,8 +2285,8 @@ static int __init vtd_setup(void) } /* We enable the following features only if they are supported by all VT-d - * engines: Snoop Control, DMA passthrough, Queued Invalidation, Interrupt - * Remapping, and Posted Interrupt + * engines: Snoop Control, DMA passthrough, Register-based Invalidation, + * Queued Invalidation, Interrupt Remapping, and Posted Interrupt. */ for_each_drhd_unit ( drhd ) { @@ -2295,6 +2308,9 @@ static int __init vtd_setup(void) if ( iommu_qinval && !ecap_queued_inval(iommu->ecap) ) iommu_qinval = 0; + if ( !has_register_based_invalidation(iommu) ) + reg_inval_supported = false; + if ( iommu_intremap && !ecap_intr_remap(iommu->ecap) ) iommu_intremap = iommu_intremap_off; @@ -2321,6 +2337,13 @@ static int __init vtd_setup(void) softirq_tasklet_init(&vtd_fault_tasklet, do_iommu_page_fault, NULL); + if ( !iommu_qinval && !reg_inval_supported ) + { + dprintk(XENLOG_ERR VTDPREFIX, "No available invalidation interface\n"); + ret = -ENODEV; + goto error; + } + if ( !iommu_qinval && iommu_intremap ) { iommu_intremap = iommu_intremap_off; diff --git a/xen/drivers/passthrough/vtd/iommu.h b/xen/drivers/passthrough/vtd/iommu.h index 216791b3d6..57737b1a4a 100644 --- a/xen/drivers/passthrough/vtd/iommu.h +++ b/xen/drivers/passthrough/vtd/iommu.h @@ -540,6 +540,7 @@ struct vtd_iommu { struct list_head ats_devices; unsigned long *domid_bitmap; /* domain id bitmap */ u16 *domid_map; /* domain id mapping array */ + uint32_t version; }; #define INTEL_IOMMU_DEBUG(fmt, args...) \ @@ -549,4 +550,10 @@ struct vtd_iommu { dprintk(XENLOG_WARNING VTDPREFIX, fmt, ## args); \ } while(0) +/* Register-based invalidation isn't supported by VT-d version 6 and beyond. */ +static inline bool has_register_based_invalidation(const struct vtd_iommu *vtd) +{ + return VER_MAJOR(vtd->version) < 6; +} + #endif diff --git a/xen/drivers/passthrough/vtd/qinval.c b/xen/drivers/passthrough/vtd/qinval.c index 764ef9f0fc..dfcbb66c8a 100644 --- a/xen/drivers/passthrough/vtd/qinval.c +++ b/xen/drivers/passthrough/vtd/qinval.c @@ -442,6 +442,23 @@ int enable_qinval(struct vtd_iommu *iommu) return 0; } +static int vtd_flush_context_noop(struct vtd_iommu *iommu, uint16_t did, + uint16_t source_id, uint8_t function_mask, + uint64_t type, bool flush_non_present_entry) +{ + WARN(); + return -EIO; +} + +static int vtd_flush_iotlb_noop(struct vtd_iommu *iommu, uint16_t did, + uint64_t addr, unsigned int size_order, + uint64_t type, bool flush_non_present_entry, + bool flush_dev_iotlb) +{ + WARN(); + return -EIO; +} + void disable_qinval(struct vtd_iommu *iommu) { u32 sts; @@ -463,6 +480,18 @@ void disable_qinval(struct vtd_iommu *iommu) out: spin_unlock_irqrestore(&iommu->register_lock, flags); - iommu->flush.context = vtd_flush_context_reg; - iommu->flush.iotlb = vtd_flush_iotlb_reg; + /* + * Assign callbacks to noop to catch errors if register-based invalidation + * isn't supported. + */ + if ( has_register_based_invalidation(iommu) ) + { + iommu->flush.context = vtd_flush_context_reg; + iommu->flush.iotlb = vtd_flush_iotlb_reg; + } + else + { + iommu->flush.context = vtd_flush_context_noop; + iommu->flush.iotlb = vtd_flush_iotlb_noop; + } } -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.14 From xen-changelog-bounces@lists.xenproject.org Fri Jun 04 13:00:37 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Fri, 04 Jun 2021 13:00:37 +0000 Received: from list by lists.xenproject.org with outflank-mailman.136824.253536 (Exim 4.92) (envelope-from ) id 1lp9RF-0005nb-Ox; Fri, 04 Jun 2021 13:00:37 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 136824.253536; Fri, 04 Jun 2021 13:00:37 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lp9RF-0005nT-Ls; Fri, 04 Jun 2021 13:00:37 +0000 Received: by outflank-mailman (input) for mailman id 136824; Fri, 04 Jun 2021 13:00:35 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lp9RD-0005n4-R1 for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 13:00:35 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lp9RD-0004Q1-QC for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 13:00:35 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lp9RD-0003fO-PS for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 13:00:35 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=4EMWKdMzoi6ZW/Yk3RSTKvWtbsDqnwIDxn2+7lLDXLM=; b=CVjLwOsfV4hT1R1MEsvOWe+qOf 2xWnuD3o6m/T1zJ7AInqPgU0wUQU2byiMb3D8TVGtmC9M53PSMAVEw5heX6bD+OD91lQNlJtolEGb 9Wjrq3YgI1wp18w+WAsezV9et6tIgTRh8x4k/4YwkDtDi50bgcyVCKV2FF4Ua90eEImc=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.14] x86/vtx: add LBR_SELECT to the list of LBR MSRs Message-Id: Date: Fri, 04 Jun 2021 13:00:35 +0000 commit f23cb474e8183f1e3e27c5d20342173d8869b234 Author: Igor Druzhinin AuthorDate: Fri Jun 4 14:54:10 2021 +0200 Commit: Jan Beulich CommitDate: Fri Jun 4 14:54:10 2021 +0200 x86/vtx: add LBR_SELECT to the list of LBR MSRs This MSR exists since Nehalem / Silvermont and is actively used by Linux, for instance, to improve sampling efficiency. Signed-off-by: Igor Druzhinin Reviewed-by: Jan Beulich Reviewed-by: Kevin Tian master commit: 08693c03e00ea3448adc4406c891e707f0068eb6 master date: 2021-04-26 10:22:04 +0200 --- xen/arch/x86/hvm/vmx/vmx.c | 20 ++++++++++++++++---- xen/include/asm-x86/msr-index.h | 10 ++++++++-- 2 files changed, 24 insertions(+), 6 deletions(-) diff --git a/xen/arch/x86/hvm/vmx/vmx.c b/xen/arch/x86/hvm/vmx/vmx.c index ca47f83cd4..09221c4811 100644 --- a/xen/arch/x86/hvm/vmx/vmx.c +++ b/xen/arch/x86/hvm/vmx/vmx.c @@ -2722,14 +2722,16 @@ static const struct lbr_info { }, nh_lbr[] = { { MSR_IA32_LASTINTFROMIP, 1 }, { MSR_IA32_LASTINTTOIP, 1 }, - { MSR_C2_LASTBRANCH_TOS, 1 }, + { MSR_NHL_LBR_SELECT, 1 }, + { MSR_NHL_LASTBRANCH_TOS, 1 }, { MSR_P4_LASTBRANCH_0_FROM_LIP, NUM_MSR_P4_LASTBRANCH_FROM_TO }, { MSR_P4_LASTBRANCH_0_TO_LIP, NUM_MSR_P4_LASTBRANCH_FROM_TO }, { 0, 0 } }, sk_lbr[] = { { MSR_IA32_LASTINTFROMIP, 1 }, { MSR_IA32_LASTINTTOIP, 1 }, - { MSR_SKL_LASTBRANCH_TOS, 1 }, + { MSR_NHL_LBR_SELECT, 1 }, + { MSR_NHL_LASTBRANCH_TOS, 1 }, { MSR_SKL_LASTBRANCH_0_FROM_IP, NUM_MSR_SKL_LASTBRANCH }, { MSR_SKL_LASTBRANCH_0_TO_IP, NUM_MSR_SKL_LASTBRANCH }, { MSR_SKL_LASTBRANCH_0_INFO, NUM_MSR_SKL_LASTBRANCH }, @@ -2741,10 +2743,19 @@ static const struct lbr_info { { MSR_C2_LASTBRANCH_0_FROM_IP, NUM_MSR_ATOM_LASTBRANCH_FROM_TO }, { MSR_C2_LASTBRANCH_0_TO_IP, NUM_MSR_ATOM_LASTBRANCH_FROM_TO }, { 0, 0 } +}, sm_lbr[] = { + { MSR_IA32_LASTINTFROMIP, 1 }, + { MSR_IA32_LASTINTTOIP, 1 }, + { MSR_SM_LBR_SELECT, 1 }, + { MSR_SM_LASTBRANCH_TOS, 1 }, + { MSR_C2_LASTBRANCH_0_FROM_IP, NUM_MSR_ATOM_LASTBRANCH_FROM_TO }, + { MSR_C2_LASTBRANCH_0_TO_IP, NUM_MSR_ATOM_LASTBRANCH_FROM_TO }, + { 0, 0 } }, gm_lbr[] = { { MSR_IA32_LASTINTFROMIP, 1 }, { MSR_IA32_LASTINTTOIP, 1 }, - { MSR_GM_LASTBRANCH_TOS, 1 }, + { MSR_SM_LBR_SELECT, 1 }, + { MSR_SM_LASTBRANCH_TOS, 1 }, { MSR_GM_LASTBRANCH_0_FROM_IP, NUM_MSR_GM_LASTBRANCH_FROM_TO }, { MSR_GM_LASTBRANCH_0_TO_IP, NUM_MSR_GM_LASTBRANCH_FROM_TO }, { 0, 0 } @@ -2798,6 +2809,7 @@ static const struct lbr_info *last_branch_msr_get(void) return sk_lbr; /* Atom */ case 0x1c: case 0x26: case 0x27: case 0x35: case 0x36: + return at_lbr; /* Silvermont */ case 0x37: case 0x4a: case 0x4d: case 0x5a: case 0x5d: /* Xeon Phi Knights Landing */ @@ -2806,7 +2818,7 @@ static const struct lbr_info *last_branch_msr_get(void) case 0x85: /* Airmont */ case 0x4c: - return at_lbr; + return sm_lbr; /* Goldmont */ case 0x5c: case 0x5f: return gm_lbr; diff --git a/xen/include/asm-x86/msr-index.h b/xen/include/asm-x86/msr-index.h index 5e64ecff91..d66c92b946 100644 --- a/xen/include/asm-x86/msr-index.h +++ b/xen/include/asm-x86/msr-index.h @@ -616,15 +616,21 @@ #define NUM_MSR_C2_LASTBRANCH_FROM_TO 4 #define NUM_MSR_ATOM_LASTBRANCH_FROM_TO 8 +/* Nehalem (and newer) last-branch recording */ +#define MSR_NHL_LBR_SELECT 0x000001c8 +#define MSR_NHL_LASTBRANCH_TOS 0x000001c9 + /* Skylake (and newer) last-branch recording */ -#define MSR_SKL_LASTBRANCH_TOS 0x000001c9 #define MSR_SKL_LASTBRANCH_0_FROM_IP 0x00000680 #define MSR_SKL_LASTBRANCH_0_TO_IP 0x000006c0 #define MSR_SKL_LASTBRANCH_0_INFO 0x00000dc0 #define NUM_MSR_SKL_LASTBRANCH 32 +/* Silvermont (and newer) last-branch recording */ +#define MSR_SM_LBR_SELECT 0x000001c8 +#define MSR_SM_LASTBRANCH_TOS 0x000001c9 + /* Goldmont last-branch recording */ -#define MSR_GM_LASTBRANCH_TOS 0x000001c9 #define MSR_GM_LASTBRANCH_0_FROM_IP 0x00000680 #define MSR_GM_LASTBRANCH_0_TO_IP 0x000006c0 #define NUM_MSR_GM_LASTBRANCH_FROM_TO 32 -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.14 From xen-changelog-bounces@lists.xenproject.org Fri Jun 04 13:00:47 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Fri, 04 Jun 2021 13:00:47 +0000 Received: from list by lists.xenproject.org with outflank-mailman.136825.253539 (Exim 4.92) (envelope-from ) id 1lp9RP-0005qD-QD; Fri, 04 Jun 2021 13:00:47 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 136825.253539; Fri, 04 Jun 2021 13:00:47 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lp9RP-0005q7-NL; Fri, 04 Jun 2021 13:00:47 +0000 Received: by outflank-mailman (input) for mailman id 136825; Fri, 04 Jun 2021 13:00:45 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lp9RN-0005po-UR for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 13:00:45 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lp9RN-0004Ry-Tc for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 13:00:45 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lp9RN-0003gc-Se for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 13:00:45 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=zyFry2PnvULDjWjvUhno3K08QMO2Y7E+ZWEpup0+ND0=; b=HYZJLsUPh6Sx7KUIBZ3lhCtpR+ gOu48a8hLYDJsma+M2vlF4jIYqCm6BxWKl1GeltpNtlJPhVRQ4uHJIkVq3cGjkiSDaRvjlDvvR+T8 rce/igei8AnR1MufulPacqYBuYkeV8WqjyxjZKa7t60ih4sFBJwLRGgiWolcF+5H9uz0=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.14] x86/intel: insert Ice Lake-SP and Ice Lake-D model numbers Message-Id: Date: Fri, 04 Jun 2021 13:00:45 +0000 commit b15c24a70c8be8b32f61b1962a6dc9df3d65ce78 Author: Igor Druzhinin AuthorDate: Fri Jun 4 14:54:43 2021 +0200 Commit: Jan Beulich CommitDate: Fri Jun 4 14:54:43 2021 +0200 x86/intel: insert Ice Lake-SP and Ice Lake-D model numbers LBR, C-state MSRs should correspond to Ice Lake desktop according to SDM rev. 74 for both models. Ice Lake-SP is known to expose IF_PSCHANGE_MC_NO in IA32_ARCH_CAPABILITIES MSR (as advisory tells and Whitley SDP confirms) which means the erratum is fixed in hardware for that model and therefore it shouldn't be present in has_if_pschange_mc list. Provisionally assume the same to be the case for Ice Lake-D. Signed-off-by: Igor Druzhinin Reviewed-by: Jan Beulich Reviewed-by: Kevin Tian master commit: 95419adfd4b275cffe24b96edcc2f15bc4db8907 master date: 2021-04-26 10:22:48 +0200 --- xen/arch/x86/acpi/cpu_idle.c | 2 ++ xen/arch/x86/hvm/vmx/vmx.c | 2 +- 2 files changed, 3 insertions(+), 1 deletion(-) diff --git a/xen/arch/x86/acpi/cpu_idle.c b/xen/arch/x86/acpi/cpu_idle.c index c092086b33..d788c8bffc 100644 --- a/xen/arch/x86/acpi/cpu_idle.c +++ b/xen/arch/x86/acpi/cpu_idle.c @@ -181,6 +181,8 @@ static void do_get_hw_residencies(void *arg) case 0x55: case 0x5E: /* Ice Lake */ + case 0x6A: + case 0x6C: case 0x7D: case 0x7E: /* Tiger Lake */ diff --git a/xen/arch/x86/hvm/vmx/vmx.c b/xen/arch/x86/hvm/vmx/vmx.c index 09221c4811..6972d4ab23 100644 --- a/xen/arch/x86/hvm/vmx/vmx.c +++ b/xen/arch/x86/hvm/vmx/vmx.c @@ -2797,7 +2797,7 @@ static const struct lbr_info *last_branch_msr_get(void) /* Goldmont Plus */ case 0x7a: /* Ice Lake */ - case 0x7d: case 0x7e: + case 0x6a: case 0x6c: case 0x7d: case 0x7e: /* Tiger Lake */ case 0x8c: case 0x8d: /* Tremont */ -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.14 From xen-changelog-bounces@lists.xenproject.org Fri Jun 04 13:00:57 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Fri, 04 Jun 2021 13:00:57 +0000 Received: from list by lists.xenproject.org with outflank-mailman.136826.253544 (Exim 4.92) (envelope-from ) id 1lp9RZ-0005te-Rm; Fri, 04 Jun 2021 13:00:57 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 136826.253544; Fri, 04 Jun 2021 13:00:57 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lp9RZ-0005tW-Oo; Fri, 04 Jun 2021 13:00:57 +0000 Received: by outflank-mailman (input) for mailman id 136826; Fri, 04 Jun 2021 13:00:56 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lp9RY-0005tB-0x for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 13:00:56 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lp9RY-0004S8-0H for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 13:00:56 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lp9RX-0003hq-Vf for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 13:00:55 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=tWb6p801fqrTe9yyBq5E+im/E4q0X8HViVQwRZyfu3E=; b=6SHcgBaELpbltxIUNtmsmMUADB zhpifoazRJt6PHk0hWVs8ffSjL9wsPE5VdwZ3FtIwWK84iCHXO0LdKasOrepK644iGA3X/KmrrbWc l73upQzPMIgod4TZMGIR1+QFQgjPV42Zt8feRHKQXikQWWsAYkb9BN7U8l4zktOXoNjE=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.14] x86/vhpet: fix RTC special casing Message-Id: Date: Fri, 04 Jun 2021 13:00:55 +0000 commit 9d963a7169a3b6e2e0fb9c8abc31c48064b2906d Author: Roger Pau Monné AuthorDate: Fri Jun 4 14:55:12 2021 +0200 Commit: Jan Beulich CommitDate: Fri Jun 4 14:55:12 2021 +0200 x86/vhpet: fix RTC special casing Restore setting the virtual timer callback private data to NULL if the timer is not level triggered. This fixes the special casing done in pt_update_irq so that the RTC interrupt when originating from the HPET is suspended if the interrupt source is masked. Note the RTC special casing done in pt_update_irq should only apply to the RTC interrupt originating from the emulated RTC device (which does set the callback private data), as in that case the callback itself will destroy the virtual timer if the interrupt is ignored. While there also use RTC_IRQ instead of 8 when the HPET is configured in LegacyReplacement Mode. Fixes: be07023be115 ("x86/vhpet: add support for level triggered interrupts") Reported-by: Jan Beulich Signed-off-by: Roger Pau Monné Reviewed-by: Jan Beulich master commit: 472a13988a051e5ae84b95815c6caf4378062abe master date: 2021-05-07 10:43:29 +0200 --- xen/arch/x86/hvm/hpet.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/xen/arch/x86/hvm/hpet.c b/xen/arch/x86/hvm/hpet.c index ca94e8b453..ee756abb82 100644 --- a/xen/arch/x86/hvm/hpet.c +++ b/xen/arch/x86/hvm/hpet.c @@ -22,6 +22,7 @@ #include #include #include +#include #include #include #include @@ -290,7 +291,7 @@ static void hpet_set_timer(HPETState *h, unsigned int tn, /* if LegacyReplacementRoute bit is set, HPET specification requires timer0 be routed to IRQ0 in NON-APIC or IRQ2 in the I/O APIC, timer1 be routed to IRQ8 in NON-APIC or IRQ8 in the I/O APIC. */ - irq = (tn == 0) ? 0 : 8; + irq = (tn == 0) ? 0 : RTC_IRQ; h->pt[tn].source = PTSRC_isa; } else @@ -318,7 +319,8 @@ static void hpet_set_timer(HPETState *h, unsigned int tn, hpet_tick_to_ns(h, diff), oneshot ? 0 : hpet_tick_to_ns(h, h->hpet.period[tn]), irq, timer_level(h, tn) ? hpet_timer_fired : NULL, - (void *)(unsigned long)tn, timer_level(h, tn)); + timer_level(h, tn) ? (void *)(unsigned long)tn : NULL, + timer_level(h, tn)); } static inline uint64_t hpet_fixup_reg( -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.14 From xen-changelog-bounces@lists.xenproject.org Fri Jun 04 13:01:07 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Fri, 04 Jun 2021 13:01:07 +0000 Received: from list by lists.xenproject.org with outflank-mailman.136827.253548 (Exim 4.92) (envelope-from ) id 1lp9Rj-0005wc-Td; Fri, 04 Jun 2021 13:01:07 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 136827.253548; Fri, 04 Jun 2021 13:01:07 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lp9Rj-0005wS-QP; Fri, 04 Jun 2021 13:01:07 +0000 Received: by outflank-mailman (input) for mailman id 136827; Fri, 04 Jun 2021 13:01:06 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lp9Ri-0005wB-4b for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 13:01:06 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lp9Ri-0004SV-3m for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 13:01:06 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lp9Ri-0003jf-2U for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 13:01:06 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=sdgeHs2ATK/pSjah8nG6fDJwji6t65QBM9R9I1oqvlk=; b=DKnCoJeIsFgC/fakiLeK/5CjJx 5nMLGg+5rnE8vMlcMXwuMn12BBiWWXdlKH32/1v32MHIjGgKyMQt3fQhRShYY4MMJ8mNigeHrFCqs 7F70oio6830r/RdYDPTp1MaDJeBxldoqbUOF//W1C9rHbrDxFbbYiXsAyLJASWeDRLMs=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.14] x86emul: fix test harness build for gas 2.36 Message-Id: Date: Fri, 04 Jun 2021 13:01:06 +0000 commit ebfdf0c7eb0ba26a9e83f3f39bdf51452fa11b03 Author: Jan Beulich AuthorDate: Fri Jun 4 14:55:55 2021 +0200 Commit: Jan Beulich CommitDate: Fri Jun 4 14:55:55 2021 +0200 x86emul: fix test harness build for gas 2.36 All of the sudden, besides .text and .rodata and alike, an always present .note.gnu.property section has appeared. This section, when converting to binary format output, gets placed according to its linked address, causing the resulting blobs to be about 128Mb in size. The resulting headers with a C representation of the binary blobs then are, of course all a multiple of that size (and take accordingly long to create). I didn't bother waiting to see what size the final test_x86_emulator binary then would have had. See also https://sourceware.org/bugzilla/show_bug.cgi?id=27753. Rather than figuring out whether gas supports -mx86-used-note=, simply remove the section while creating *.bin. Signed-off-by: Jan Beulich master commit: aa803ba38a867551917d11059eaa044955556e05 master date: 2021-05-17 15:41:28 +0200 --- tools/tests/x86_emulator/testcase.mk | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/tools/tests/x86_emulator/testcase.mk b/tools/tests/x86_emulator/testcase.mk index dafeb6caf7..d4aa82e3f3 100644 --- a/tools/tests/x86_emulator/testcase.mk +++ b/tools/tests/x86_emulator/testcase.mk @@ -12,11 +12,11 @@ all: $(TESTCASE).bin %.bin: %.c $(CC) $(filter-out -M% .%,$(CFLAGS)) -c $< $(LD) $(LDFLAGS_DIRECT) -N -Ttext 0x100000 -o $*.tmp $*.o - $(OBJCOPY) -O binary $*.tmp $@ + $(OBJCOPY) -O binary -R .note.gnu.property $*.tmp $@ rm -f $*.tmp %-opmask.bin: opmask.S $(CC) $(filter-out -M% .%,$(CFLAGS)) -c $< -o $(basename $@).o $(LD) $(LDFLAGS_DIRECT) -N -Ttext 0x100000 -o $(basename $@).tmp $(basename $@).o - $(OBJCOPY) -O binary $(basename $@).tmp $@ + $(OBJCOPY) -O binary -R .note.gnu.property $(basename $@).tmp $@ rm -f $(basename $@).tmp -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.14 From xen-changelog-bounces@lists.xenproject.org Fri Jun 04 13:01:17 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Fri, 04 Jun 2021 13:01:17 +0000 Received: from list by lists.xenproject.org with outflank-mailman.136828.253552 (Exim 4.92) (envelope-from ) id 1lp9Rs-0005z1-VA; Fri, 04 Jun 2021 13:01:16 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 136828.253552; Fri, 04 Jun 2021 13:01:16 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lp9Rs-0005yt-S1; Fri, 04 Jun 2021 13:01:16 +0000 Received: by outflank-mailman (input) for mailman id 136828; Fri, 04 Jun 2021 13:01:16 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lp9Rs-0005yl-8C for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 13:01:16 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lp9Rs-0004Sl-7S for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 13:01:16 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lp9Rs-0003lg-6g for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 13:01:16 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=J6o9rPjtpth5VsPX3EUYo/B9FCVf1eAc6/31Gf96AT0=; b=5CNIdYxJX+fxnn5PP+Hta1MId5 ehE7MMcN3lRBUE2UDnQOEEyVlIvqfJp+4Dfb5g3y3ew+CSwZqCvTFbBGsCVg0LE+XIIfJxG4/rRYK k5hxcHWPWNw07aYPAl/GT1FsiOfXSJcWtkBFOdUErJyuJlanrk61gawOg/tT29xDxhbw=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.14] x86: make hypervisor build with gcc11 Message-Id: Date: Fri, 04 Jun 2021 13:01:16 +0000 commit ac507e078edcc18215d86c436d6c26dd76ffa15c Author: Jan Beulich AuthorDate: Fri Jun 4 14:56:19 2021 +0200 Commit: Jan Beulich CommitDate: Fri Jun 4 14:56:19 2021 +0200 x86: make hypervisor build with gcc11 Gcc 11 looks to make incorrect assumptions about valid ranges that pointers may be used for addressing when they are derived from e.g. a plain constant. See https://gcc.gnu.org/bugzilla/show_bug.cgi?id=100680. Utilize RELOC_HIDE() to work around the issue, which for x86 manifests in at least - mpparse.c:efi_check_config(), - tboot.c:tboot_probe(), - tboot.c:tboot_gen_frametable_integrity(), - x86_emulate.c:x86_emulate() (at -O2 only). The last case is particularly odd not just because it only triggers at higher optimization levels, but also because it only affects one of at least three similar constructs. Various "note" diagnostics claim the valid index range to be [0, 2⁶³-1]. Signed-off-by: Jan Beulich Tested-by: Jason Andryuk Acked-by: Roger Pau Monné master commit: 722f59d38c710a940ab05e542a83020eb5546dea master date: 2021-05-27 14:40:29 +0200 --- tools/tests/x86_emulator/x86-emulate.c | 7 +++++++ xen/arch/x86/x86_emulate/x86_emulate.c | 2 +- xen/include/asm-x86/fixmap.h | 2 +- xen/include/xen/compiler.h | 6 ++++++ xen/include/xen/pdx.h | 2 +- 5 files changed, 16 insertions(+), 3 deletions(-) diff --git a/tools/tests/x86_emulator/x86-emulate.c b/tools/tests/x86_emulator/x86-emulate.c index 572989078a..c32d9a55b4 100644 --- a/tools/tests/x86_emulator/x86-emulate.c +++ b/tools/tests/x86_emulator/x86-emulate.c @@ -8,6 +8,13 @@ #define ERR_PTR(val) NULL +/* See gcc bug 100680, but here don't bother making this version dependent. */ +#define gcc11_wrap(x) ({ \ + unsigned long x_; \ + __asm__ ( "" : "=g" (x_) : "0" (x) ); \ + (typeof(x))x_; \ +}) + #define cpu_has_amd_erratum(nr) 0 #define cpu_has_mpx false #define read_bndcfgu() 0 diff --git a/xen/arch/x86/x86_emulate/x86_emulate.c b/xen/arch/x86/x86_emulate/x86_emulate.c index 8ef66d824d..16d164904c 100644 --- a/xen/arch/x86/x86_emulate/x86_emulate.c +++ b/xen/arch/x86/x86_emulate/x86_emulate.c @@ -725,7 +725,7 @@ union vex { #define copy_VEX(ptr, vex) ({ \ if ( !mode_64bit() ) \ (vex).reg |= 8; \ - (ptr)[0 - PFX_BYTES] = ext < ext_8f08 ? 0xc4 : 0x8f; \ + gcc11_wrap(ptr)[0 - PFX_BYTES] = ext < ext_8f08 ? 0xc4 : 0x8f; \ (ptr)[1 - PFX_BYTES] = (vex).raw[0]; \ (ptr)[2 - PFX_BYTES] = (vex).raw[1]; \ container_of((ptr) + 1 - PFX_BYTES, typeof(vex), raw[0]); \ diff --git a/xen/include/asm-x86/fixmap.h b/xen/include/asm-x86/fixmap.h index 8330097a74..22afb92ff0 100644 --- a/xen/include/asm-x86/fixmap.h +++ b/xen/include/asm-x86/fixmap.h @@ -79,7 +79,7 @@ extern void __set_fixmap( #define clear_fixmap(idx) __set_fixmap(idx, 0, 0) -#define __fix_to_virt(x) (FIXADDR_TOP - ((x) << PAGE_SHIFT)) +#define __fix_to_virt(x) gcc11_wrap(FIXADDR_TOP - ((x) << PAGE_SHIFT)) #define __virt_to_fix(x) ((FIXADDR_TOP - ((x)&PAGE_MASK)) >> PAGE_SHIFT) #define fix_to_virt(x) ((void *)__fix_to_virt(x)) diff --git a/xen/include/xen/compiler.h b/xen/include/xen/compiler.h index c0e0ee9f27..eb819fab46 100644 --- a/xen/include/xen/compiler.h +++ b/xen/include/xen/compiler.h @@ -105,6 +105,12 @@ __asm__ ("" : "=r"(__ptr) : "0"(ptr)); \ (typeof(ptr)) (__ptr + (off)); }) +#if CONFIG_GCC_VERSION >= 110000 /* See gcc bug 100680. */ +# define gcc11_wrap(x) RELOC_HIDE(x, 0) +#else +# define gcc11_wrap(x) (x) +#endif + #ifdef __GCC_ASM_FLAG_OUTPUTS__ # define ASM_FLAG_OUT(yes, no) yes #else diff --git a/xen/include/xen/pdx.h b/xen/include/xen/pdx.h index 770fadc06c..9fcfb0ce52 100644 --- a/xen/include/xen/pdx.h +++ b/xen/include/xen/pdx.h @@ -19,7 +19,7 @@ extern u64 pdx_region_mask(u64 base, u64 len); extern void set_pdx_range(unsigned long smfn, unsigned long emfn); #define page_to_pdx(pg) ((pg) - frame_table) -#define pdx_to_page(pdx) (frame_table + (pdx)) +#define pdx_to_page(pdx) gcc11_wrap(frame_table + (pdx)) bool __mfn_valid(unsigned long mfn); -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.14 From xen-changelog-bounces@lists.xenproject.org Fri Jun 04 13:01:27 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Fri, 04 Jun 2021 13:01:27 +0000 Received: from list by lists.xenproject.org with outflank-mailman.136829.253556 (Exim 4.92) (envelope-from ) id 1lp9S3-000624-0N; Fri, 04 Jun 2021 13:01:27 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 136829.253556; Fri, 04 Jun 2021 13:01:26 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lp9S2-00061w-Tk; Fri, 04 Jun 2021 13:01:26 +0000 Received: by outflank-mailman (input) for mailman id 136829; Fri, 04 Jun 2021 13:01:26 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lp9S2-00061q-BO for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 13:01:26 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lp9S2-0004T2-Ad for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 13:01:26 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lp9S2-0003n5-9v for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 13:01:26 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=cXc/egKuB0XSpPZd6Uk27N0El7InG7vxOhNeDnePKm0=; b=RG8omF26FduFKIuul+/7HdLKbV hCU4u4YooTlLN9YSgbl/wPIBCN9i4D/0c01diW6nyYynoQHQDquuHncbB5zq4Z5KDYYDkTtfRKIpU hIxj6R4p/cQQHG/oTp6O8YCNe6omKeAbuquUKsRc9I0yJxgkgkzniPyjM/VdXDlbrDfE=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.14] x86/cpuid: Rework HLE and RTM handling Message-Id: Date: Fri, 04 Jun 2021 13:01:26 +0000 commit 3f8549386bc72b0ddc7020eec97f0b497785193e Author: Andrew Cooper AuthorDate: Fri Jun 4 14:56:56 2021 +0200 Commit: Jan Beulich CommitDate: Fri Jun 4 14:56:56 2021 +0200 x86/cpuid: Rework HLE and RTM handling The TAA mitigation offered the option to hide the HLE and RTM CPUID bits, which has caused some migration compatibility problems. These two bits are special. Annotate them with ! to emphasise this point. Hardware Lock Elision (HLE) may or may not be visible in CPUID, but is disabled in microcode on all CPUs, and has been removed from the architecture. Do not advertise it to VMs by default. Restricted Transactional Memory (RTM) may or may not be visible in CPUID, and may or may not be configured in force-abort mode. Have tsx_init() note whether RTM has been configured into force-abort mode, so guest_common_feature_adjustments() can conditionally hide it from VMs by default. The host policy values for HLE/RTM may or may not be set, depending on any previous running kernel's choice of visibility, and Xen's choice. TSX is available on any CPU which enumerates a TSX-hiding mechanism, so instead of doing a two-step to clobber any hiding, scan CPUID, then set the visibility, just force visibility of the bits in the first place. With the HLE/RTM bits now unilaterally visible in the host policy, xc_cpuid_apply_policy() can construct a more appropriate policy out of thin air for pre-4.13 VMs with no CPUID data in their migration stream, and specifically one where HLE/RTM doesn't potentially disappear behind the back of a running VM. Fixes: 8c4330818f6 ("x86/spec-ctrl: Mitigate the TSX Asynchronous Abort sidechannel") Signed-off-by: Andrew Cooper Reviewed-by: Jan Beulich Reviewed-by: Roger Pau Monné master commit: 8fe24090d940d760145ccd5e234290be7418b175 master date: 2021-05-27 19:34:00 +0100 --- tools/libxc/xc_cpuid_x86.c | 2 ++ xen/arch/x86/cpuid.c | 24 ++++++++++------------ xen/arch/x86/spec_ctrl.c | 3 --- xen/arch/x86/tsx.c | 31 +++++++++++++++++++++++++++-- xen/include/asm-x86/processor.h | 1 + xen/include/public/arch-x86/cpufeatureset.h | 4 ++-- 6 files changed, 44 insertions(+), 21 deletions(-) diff --git a/tools/libxc/xc_cpuid_x86.c b/tools/libxc/xc_cpuid_x86.c index 0f24d6dd08..512eb1f78f 100644 --- a/tools/libxc/xc_cpuid_x86.c +++ b/tools/libxc/xc_cpuid_x86.c @@ -504,6 +504,8 @@ int xc_cpuid_apply_policy(xc_interface *xch, uint32_t domid, bool restore, if ( restore ) { p->basic.rdrand = test_bit(X86_FEATURE_RDRAND, host_featureset); + p->feat.hle = test_bit(X86_FEATURE_HLE, host_featureset); + p->feat.rtm = test_bit(X86_FEATURE_RTM, host_featureset); if ( di.hvm ) { diff --git a/xen/arch/x86/cpuid.c b/xen/arch/x86/cpuid.c index 4b424fac95..425dce7bc7 100644 --- a/xen/arch/x86/cpuid.c +++ b/xen/arch/x86/cpuid.c @@ -357,6 +357,16 @@ static void __init guest_common_default_feature_adjustments(uint32_t *fs) boot_cpu_data.x86 == 6 && boot_cpu_data.x86_model == 0x3a && cpu_has_rdrand && !is_forced_cpu_cap(X86_FEATURE_RDRAND) ) __clear_bit(X86_FEATURE_RDRAND, fs); + + /* + * On certain hardware, speculative or errata workarounds can result in + * TSX being placed in "force-abort" mode, where it doesn't actually + * function as expected, but is technically compatible with the ISA. + * + * Do not advertise RTM to guests by default if it won't actually work. + */ + if ( rtm_disabled ) + __clear_bit(X86_FEATURE_RTM, fs); } static void __init guest_common_feature_adjustments(uint32_t *fs) @@ -628,20 +638,6 @@ void recalculate_cpuid_policy(struct domain *d) if ( cpu_has_itsc && (d->disable_migrate || d->arch.vtsc) ) __set_bit(X86_FEATURE_ITSC, max_fs); - /* - * On hardware with MSR_TSX_CTRL, the admin may have elected to disable - * TSX and hide the feature bits. Migrating-in VMs may have been booted - * pre-mitigation when the TSX features were visbile. - * - * This situation is compatible (albeit with a perf hit to any TSX code in - * the guest), so allow the feature bits to remain set. - */ - if ( cpu_has_tsx_ctrl ) - { - __set_bit(X86_FEATURE_HLE, max_fs); - __set_bit(X86_FEATURE_RTM, max_fs); - } - /* Clamp the toolstacks choices to reality. */ for ( i = 0; i < ARRAY_SIZE(fs); i++ ) fs[i] &= max_fs[i]; diff --git a/xen/arch/x86/spec_ctrl.c b/xen/arch/x86/spec_ctrl.c index c9f78ead62..bea4c3e8b4 100644 --- a/xen/arch/x86/spec_ctrl.c +++ b/xen/arch/x86/spec_ctrl.c @@ -1158,9 +1158,6 @@ void __init init_speculation_mitigations(void) ((hw_smt_enabled && opt_smt) || !boot_cpu_has(X86_FEATURE_SC_VERW_IDLE)) ) { - setup_clear_cpu_cap(X86_FEATURE_HLE); - setup_clear_cpu_cap(X86_FEATURE_RTM); - opt_tsx = 0; tsx_init(); } diff --git a/xen/arch/x86/tsx.c b/xen/arch/x86/tsx.c index 39e483640a..e09e819dce 100644 --- a/xen/arch/x86/tsx.c +++ b/xen/arch/x86/tsx.c @@ -15,6 +15,7 @@ */ int8_t __read_mostly opt_tsx = -1; int8_t __read_mostly cpu_has_tsx_ctrl = -1; +bool __read_mostly rtm_disabled; static int __init parse_tsx(const char *s) { @@ -45,6 +46,30 @@ void tsx_init(void) rdmsrl(MSR_ARCH_CAPABILITIES, caps); cpu_has_tsx_ctrl = !!(caps & ARCH_CAPS_TSX_CTRL); + + /* + * The TSX features (HLE/RTM) are handled specially. They both + * enumerate features but, on certain parts, have mechanisms to be + * hidden without disrupting running software. + * + * At the moment, we're running in an unknown context (WRT hiding - + * particularly if another fully fledged kernel ran before us) and + * depending on user settings, may elect to continue hiding them from + * native CPUID instructions. + * + * Xen doesn't use TSX itself, but use cpu_has_{hle,rtm} for various + * system reasons, mostly errata detection, so the meaning is more + * useful as "TSX infrastructure available", as opposed to "features + * advertised and working". + * + * Force the features to be visible in Xen's view if we see any of the + * infrastructure capable of hiding them. + */ + if ( cpu_has_tsx_ctrl ) + { + setup_force_cpu_cap(X86_FEATURE_HLE); + setup_force_cpu_cap(X86_FEATURE_RTM); + } } if ( cpu_has_tsx_ctrl ) @@ -53,9 +78,11 @@ void tsx_init(void) rdmsrl(MSR_TSX_CTRL, val); + /* Check bottom bit only. Higher bits are various sentinels. */ + rtm_disabled = !(opt_tsx & 1); + val &= ~(TSX_CTRL_RTM_DISABLE | TSX_CTRL_CPUID_CLEAR); - /* Check bottom bit only. Higher bits are various sentinals. */ - if ( !(opt_tsx & 1) ) + if ( rtm_disabled ) val |= TSX_CTRL_RTM_DISABLE | TSX_CTRL_CPUID_CLEAR; wrmsrl(MSR_TSX_CTRL, val); diff --git a/xen/include/asm-x86/processor.h b/xen/include/asm-x86/processor.h index 9acb80fdcd..f8e1e4d523 100644 --- a/xen/include/asm-x86/processor.h +++ b/xen/include/asm-x86/processor.h @@ -629,6 +629,7 @@ static inline uint8_t get_cpu_family(uint32_t raw, uint8_t *model, } extern int8_t opt_tsx, cpu_has_tsx_ctrl; +extern bool rtm_disabled; void tsx_init(void); #endif /* !__ASSEMBLY__ */ diff --git a/xen/include/public/arch-x86/cpufeatureset.h b/xen/include/public/arch-x86/cpufeatureset.h index d06a2fd4c8..02b3a2a265 100644 --- a/xen/include/public/arch-x86/cpufeatureset.h +++ b/xen/include/public/arch-x86/cpufeatureset.h @@ -197,14 +197,14 @@ XEN_CPUFEATURE(FSGSBASE, 5*32+ 0) /*A {RD,WR}{FS,GS}BASE instructions */ XEN_CPUFEATURE(TSC_ADJUST, 5*32+ 1) /*S TSC_ADJUST MSR available */ XEN_CPUFEATURE(SGX, 5*32+ 2) /* Software Guard extensions */ XEN_CPUFEATURE(BMI1, 5*32+ 3) /*A 1st bit manipulation extensions */ -XEN_CPUFEATURE(HLE, 5*32+ 4) /*A Hardware Lock Elision */ +XEN_CPUFEATURE(HLE, 5*32+ 4) /*!a Hardware Lock Elision */ XEN_CPUFEATURE(AVX2, 5*32+ 5) /*A AVX2 instructions */ XEN_CPUFEATURE(FDP_EXCP_ONLY, 5*32+ 6) /*! x87 FDP only updated on exception. */ XEN_CPUFEATURE(SMEP, 5*32+ 7) /*S Supervisor Mode Execution Protection */ XEN_CPUFEATURE(BMI2, 5*32+ 8) /*A 2nd bit manipulation extensions */ XEN_CPUFEATURE(ERMS, 5*32+ 9) /*A Enhanced REP MOVSB/STOSB */ XEN_CPUFEATURE(INVPCID, 5*32+10) /*H Invalidate Process Context ID */ -XEN_CPUFEATURE(RTM, 5*32+11) /*A Restricted Transactional Memory */ +XEN_CPUFEATURE(RTM, 5*32+11) /*!A Restricted Transactional Memory */ XEN_CPUFEATURE(PQM, 5*32+12) /* Platform QoS Monitoring */ XEN_CPUFEATURE(NO_FPU_SEL, 5*32+13) /*! FPU CS/DS stored as zero */ XEN_CPUFEATURE(MPX, 5*32+14) /*s Memory Protection Extensions */ -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.14 From xen-changelog-bounces@lists.xenproject.org Fri Jun 04 13:01:37 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Fri, 04 Jun 2021 13:01:37 +0000 Received: from list by lists.xenproject.org with outflank-mailman.136830.253560 (Exim 4.92) (envelope-from ) id 1lp9SD-00065c-3z; Fri, 04 Jun 2021 13:01:37 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 136830.253560; Fri, 04 Jun 2021 13:01:37 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lp9SD-00065U-0k; Fri, 04 Jun 2021 13:01:37 +0000 Received: by outflank-mailman (input) for mailman id 136830; Fri, 04 Jun 2021 13:01:36 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lp9SC-00065K-Eb for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 13:01:36 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lp9SC-0004Tt-Dh for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 13:01:36 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lp9SC-0003oE-Co for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 13:01:36 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=knNY7a8kV1DGc/Eemj2gXhQAITzdojCTxYPGT0NbD18=; b=BCEmM5NQlvkJjMQHcE4hV3vclL W0GNWRmXr0LyYJ1+42Od+SiiOzDYdAbKo4FaVYgJbnww73bBftBwASCQl/MKf5wE4OS1Q1xpNt2YE cl2aSFIm2bGbtXHdhjxjJT/cV9lFASKEhMEFmWARQy139o+KLBKAYtMcdwlYiwzIjeXo=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.14] x86: fix build race when generating temporary object files (take 2) Message-Id: Date: Fri, 04 Jun 2021 13:01:36 +0000 commit b046e05736deecbd8254540c5e45444115fb1c98 Author: Jan Beulich AuthorDate: Fri Jun 4 14:57:42 2021 +0200 Commit: Jan Beulich CommitDate: Fri Jun 4 14:57:42 2021 +0200 x86: fix build race when generating temporary object files (take 2) The original commit wasn't quite sufficient: Emptying DEPS is helpful only when nothing will get added to it subsequently. xen/Rules.mk will, after including the local Makefile, amend DEPS by dependencies for objects living in sub-directories though. For the purpose of suppressing dependencies of the makefiles on the .*.d2 files (and thus to avoid their re-generation) it is, however, not necessary at all to play with DEPS. Instead we can override DEPS_INCLUDE (which generally is a late- expansion variable). Fixes: 761bb575ce97 ("x86: fix build race when generating temporary object files") Signed-off-by: Jan Beulich master commit: 8c90dbb99907f3b471d558775777a84daec7c3f6 master date: 2021-05-28 09:12:24 +0200 --- xen/arch/x86/Makefile | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/xen/arch/x86/Makefile b/xen/arch/x86/Makefile index d45b213ab6..bbd11a4726 100644 --- a/xen/arch/x86/Makefile +++ b/xen/arch/x86/Makefile @@ -271,5 +271,5 @@ clean:: # Suppress loading of DEPS files for internal, temporary target files. This # then also suppresses re-generation of the respective .*.d2 files. ifeq ($(filter-out .xen%.o,$(notdir $(MAKECMDGOALS))),) -DEPS:= +DEPS_INCLUDE:= endif -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.14 From xen-changelog-bounces@lists.xenproject.org Fri Jun 04 13:11:08 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Fri, 04 Jun 2021 13:11:08 +0000 Received: from list by lists.xenproject.org with outflank-mailman.136831.253564 (Exim 4.92) (envelope-from ) id 1lp9bN-0006zY-PH; Fri, 04 Jun 2021 13:11:05 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 136831.253564; Fri, 04 Jun 2021 13:11:05 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lp9bN-0006zO-M8; Fri, 04 Jun 2021 13:11:05 +0000 Received: by outflank-mailman (input) for mailman id 136831; Fri, 04 Jun 2021 13:11:05 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lp9bN-0006zI-7g for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 13:11:05 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lp9bN-0004dj-6Z for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 13:11:05 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lp9bN-0004eS-5Y for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 13:11:05 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=/AdON+NOEGHnc7BJHabH08pqXIxWaS0nGbBZtdPNdww=; b=ii+FUlXp9I+U0RiFp5NMzEErAv wh3ArKFeAi5jiJOtEVMDa3JIgmVTZ3ZRO18OisIpWjN37AajduHZ+YZR8XcbVKjDyniA845sbNYBz TAJ/gNBmZkaErne6oipLvWnqjLi5UPbwJyGs2zeTpQ7xSKPFQr/+Fdl89uc034xgJQ6c=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.13] VT-d: Don't assume register-based invalidation is always supported Message-Id: Date: Fri, 04 Jun 2021 13:11:05 +0000 commit e289ed670712782f119d18a9f7b56c93680b4b43 Author: Chao Gao AuthorDate: Fri Jun 4 15:00:37 2021 +0200 Commit: Jan Beulich CommitDate: Fri Jun 4 15:00:37 2021 +0200 VT-d: Don't assume register-based invalidation is always supported According to Intel VT-d SPEC rev3.3 Section 6.5, Register-based Invalidation isn't supported by Intel VT-d version 6 and beyond. This hardware change impacts following two scenarios: admin can disable queued invalidation via 'qinval' cmdline and use register-based interface; VT-d switches to register-based invalidation when queued invalidation needs to be disabled, for example, during disabling x2apic or during system suspension or after enabling queued invalidation fails. To deal with this hardware change, if register-based invalidation isn't supported, queued invalidation cannot be disabled through Xen cmdline; and if queued invalidation has to be disabled temporarily in some scenarios, VT-d won't switch to register-based interface but use some dummy functions to catch errors in case there is any invalidation request issued when queued invalidation is disabled. Signed-off-by: Chao Gao Reviewed-by: Jan Beulich Reviewed-by: Kevin Tian master commit: 6773b1a7584a75a486e9774541ad5bd84c9aa5ee master date: 2021-04-26 10:16:50 +0200 --- docs/misc/xen-command-line.pandoc | 4 +++- xen/drivers/passthrough/vtd/iommu.c | 27 +++++++++++++++++++++++++-- xen/drivers/passthrough/vtd/iommu.h | 7 +++++++ xen/drivers/passthrough/vtd/qinval.c | 33 +++++++++++++++++++++++++++++++-- 4 files changed, 66 insertions(+), 5 deletions(-) diff --git a/docs/misc/xen-command-line.pandoc b/docs/misc/xen-command-line.pandoc index 0b6ec8b6a1..cf9dea62db 100644 --- a/docs/misc/xen-command-line.pandoc +++ b/docs/misc/xen-command-line.pandoc @@ -1370,7 +1370,9 @@ The following options are specific to Intel VT-d hardware: * The `qinval` boolean controls the Queued Invalidation sub-feature, and is active by default on compatible hardware. Queued Invalidation is a feature in second-generation IOMMUs and is a functional prerequisite for - Interrupt Remapping. + Interrupt Remapping. Note that Xen disregards this setting for Intel VT-d + version 6 and greater as Registered-Based Invalidation isn't supported + by them. * The `igfx` boolean is active by default, and controls whether the IOMMU in front of an Intel Graphics Device is enabled or not. diff --git a/xen/drivers/passthrough/vtd/iommu.c b/xen/drivers/passthrough/vtd/iommu.c index 01a2a461ba..34ee5d22f1 100644 --- a/xen/drivers/passthrough/vtd/iommu.c +++ b/xen/drivers/passthrough/vtd/iommu.c @@ -1185,6 +1185,14 @@ int __init iommu_alloc(struct acpi_drhd_unit *drhd) iommu->cap = dmar_readq(iommu->reg, DMAR_CAP_REG); iommu->ecap = dmar_readq(iommu->reg, DMAR_ECAP_REG); + iommu->version = dmar_readl(iommu->reg, DMAR_VER_REG); + + if ( !iommu_qinval && !has_register_based_invalidation(iommu) ) + { + printk(XENLOG_WARNING VTDPREFIX "IOMMU %d: cannot disable Queued Invalidation\n", + iommu->index); + iommu_qinval = true; + } if ( iommu_verbose ) { @@ -2158,6 +2166,10 @@ static int __must_check init_vtd_hw(bool resume) */ if ( enable_qinval(iommu) != 0 ) { + /* Ensure register-based invalidation is available */ + if ( !has_register_based_invalidation(iommu) ) + return -EIO; + iommu->flush.context = vtd_flush_context_reg; iommu->flush.iotlb = vtd_flush_iotlb_reg; } @@ -2248,6 +2260,7 @@ static int __init vtd_setup(void) struct acpi_drhd_unit *drhd; struct vtd_iommu *iommu; int ret; + bool reg_inval_supported = true; if ( list_empty(&acpi_drhd_units) ) { @@ -2269,8 +2282,8 @@ static int __init vtd_setup(void) } /* We enable the following features only if they are supported by all VT-d - * engines: Snoop Control, DMA passthrough, Queued Invalidation, Interrupt - * Remapping, and Posted Interrupt + * engines: Snoop Control, DMA passthrough, Register-based Invalidation, + * Queued Invalidation, Interrupt Remapping, and Posted Interrupt. */ for_each_drhd_unit ( drhd ) { @@ -2290,6 +2303,9 @@ static int __init vtd_setup(void) if ( iommu_qinval && !ecap_queued_inval(iommu->ecap) ) iommu_qinval = 0; + if ( !has_register_based_invalidation(iommu) ) + reg_inval_supported = false; + if ( iommu_intremap && !ecap_intr_remap(iommu->ecap) ) iommu_intremap = 0; @@ -2314,6 +2330,13 @@ static int __init vtd_setup(void) softirq_tasklet_init(&vtd_fault_tasklet, do_iommu_page_fault, 0); + if ( !iommu_qinval && !reg_inval_supported ) + { + dprintk(XENLOG_ERR VTDPREFIX, "No available invalidation interface\n"); + ret = -ENODEV; + goto error; + } + if ( !iommu_qinval && iommu_intremap ) { iommu_intremap = 0; diff --git a/xen/drivers/passthrough/vtd/iommu.h b/xen/drivers/passthrough/vtd/iommu.h index 216791b3d6..57737b1a4a 100644 --- a/xen/drivers/passthrough/vtd/iommu.h +++ b/xen/drivers/passthrough/vtd/iommu.h @@ -540,6 +540,7 @@ struct vtd_iommu { struct list_head ats_devices; unsigned long *domid_bitmap; /* domain id bitmap */ u16 *domid_map; /* domain id mapping array */ + uint32_t version; }; #define INTEL_IOMMU_DEBUG(fmt, args...) \ @@ -549,4 +550,10 @@ struct vtd_iommu { dprintk(XENLOG_WARNING VTDPREFIX, fmt, ## args); \ } while(0) +/* Register-based invalidation isn't supported by VT-d version 6 and beyond. */ +static inline bool has_register_based_invalidation(const struct vtd_iommu *vtd) +{ + return VER_MAJOR(vtd->version) < 6; +} + #endif diff --git a/xen/drivers/passthrough/vtd/qinval.c b/xen/drivers/passthrough/vtd/qinval.c index 764ef9f0fc..dfcbb66c8a 100644 --- a/xen/drivers/passthrough/vtd/qinval.c +++ b/xen/drivers/passthrough/vtd/qinval.c @@ -442,6 +442,23 @@ int enable_qinval(struct vtd_iommu *iommu) return 0; } +static int vtd_flush_context_noop(struct vtd_iommu *iommu, uint16_t did, + uint16_t source_id, uint8_t function_mask, + uint64_t type, bool flush_non_present_entry) +{ + WARN(); + return -EIO; +} + +static int vtd_flush_iotlb_noop(struct vtd_iommu *iommu, uint16_t did, + uint64_t addr, unsigned int size_order, + uint64_t type, bool flush_non_present_entry, + bool flush_dev_iotlb) +{ + WARN(); + return -EIO; +} + void disable_qinval(struct vtd_iommu *iommu) { u32 sts; @@ -463,6 +480,18 @@ void disable_qinval(struct vtd_iommu *iommu) out: spin_unlock_irqrestore(&iommu->register_lock, flags); - iommu->flush.context = vtd_flush_context_reg; - iommu->flush.iotlb = vtd_flush_iotlb_reg; + /* + * Assign callbacks to noop to catch errors if register-based invalidation + * isn't supported. + */ + if ( has_register_based_invalidation(iommu) ) + { + iommu->flush.context = vtd_flush_context_reg; + iommu->flush.iotlb = vtd_flush_iotlb_reg; + } + else + { + iommu->flush.context = vtd_flush_context_noop; + iommu->flush.iotlb = vtd_flush_iotlb_noop; + } } -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.13 From xen-changelog-bounces@lists.xenproject.org Fri Jun 04 13:11:16 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Fri, 04 Jun 2021 13:11:16 +0000 Received: from list by lists.xenproject.org with outflank-mailman.136832.253567 (Exim 4.92) (envelope-from ) id 1lp9bY-00071M-Qe; Fri, 04 Jun 2021 13:11:16 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 136832.253567; Fri, 04 Jun 2021 13:11:16 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lp9bY-00071E-Nj; Fri, 04 Jun 2021 13:11:16 +0000 Received: by outflank-mailman (input) for mailman id 136832; Fri, 04 Jun 2021 13:11:15 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lp9bX-000718-Ah for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 13:11:15 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lp9bX-0004dy-9q for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 13:11:15 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lp9bX-0004f0-8j for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 13:11:15 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=+2715KM1PjLYGW/nzRbSE3CEg+rHAhOXWsomWl7vhyA=; b=gIzQr4MLfoXLHrlI44r+1XvR24 Q7DsnxFeQ2wdguzdTKm2wuYHIEbiZAJAKuusCziTELh3fYjjki10GbtWtoJq+mRxlhy7EJTTJmyRs iAytmlB+JVG+Kf5xDxgA2cZUmvVCe+lnD75mKB42r5mF5AY0VAc0Ofas5p1WeY2m736I=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.13] x86/vhpet: fix RTC special casing Message-Id: Date: Fri, 04 Jun 2021 13:11:15 +0000 commit 287f2292c38575dcf059358ce3d9a70294017554 Author: Roger Pau Monné AuthorDate: Fri Jun 4 15:01:33 2021 +0200 Commit: Jan Beulich CommitDate: Fri Jun 4 15:01:33 2021 +0200 x86/vhpet: fix RTC special casing Restore setting the virtual timer callback private data to NULL if the timer is not level triggered. This fixes the special casing done in pt_update_irq so that the RTC interrupt when originating from the HPET is suspended if the interrupt source is masked. Note the RTC special casing done in pt_update_irq should only apply to the RTC interrupt originating from the emulated RTC device (which does set the callback private data), as in that case the callback itself will destroy the virtual timer if the interrupt is ignored. While there also use RTC_IRQ instead of 8 when the HPET is configured in LegacyReplacement Mode. Fixes: be07023be115 ("x86/vhpet: add support for level triggered interrupts") Reported-by: Jan Beulich Signed-off-by: Roger Pau Monné Reviewed-by: Jan Beulich master commit: 472a13988a051e5ae84b95815c6caf4378062abe master date: 2021-05-07 10:43:29 +0200 --- xen/arch/x86/hvm/hpet.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/xen/arch/x86/hvm/hpet.c b/xen/arch/x86/hvm/hpet.c index 9fbdc5302b..1dcd38cebc 100644 --- a/xen/arch/x86/hvm/hpet.c +++ b/xen/arch/x86/hvm/hpet.c @@ -22,6 +22,7 @@ #include #include #include +#include #include #include #include @@ -289,7 +290,7 @@ static void hpet_set_timer(HPETState *h, unsigned int tn, /* if LegacyReplacementRoute bit is set, HPET specification requires timer0 be routed to IRQ0 in NON-APIC or IRQ2 in the I/O APIC, timer1 be routed to IRQ8 in NON-APIC or IRQ8 in the I/O APIC. */ - irq = (tn == 0) ? 0 : 8; + irq = (tn == 0) ? 0 : RTC_IRQ; h->pt[tn].source = PTSRC_isa; } else @@ -317,7 +318,8 @@ static void hpet_set_timer(HPETState *h, unsigned int tn, hpet_tick_to_ns(h, diff), oneshot ? 0 : hpet_tick_to_ns(h, h->hpet.period[tn]), irq, timer_level(h, tn) ? hpet_timer_fired : NULL, - (void *)(unsigned long)tn, timer_level(h, tn)); + timer_level(h, tn) ? (void *)(unsigned long)tn : NULL, + timer_level(h, tn)); } static inline uint64_t hpet_fixup_reg( -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.13 From xen-changelog-bounces@lists.xenproject.org Fri Jun 04 13:11:26 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Fri, 04 Jun 2021 13:11:26 +0000 Received: from list by lists.xenproject.org with outflank-mailman.136833.253572 (Exim 4.92) (envelope-from ) id 1lp9bi-00074S-SH; Fri, 04 Jun 2021 13:11:26 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 136833.253572; Fri, 04 Jun 2021 13:11:26 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lp9bi-00074K-PH; Fri, 04 Jun 2021 13:11:26 +0000 Received: by outflank-mailman (input) for mailman id 136833; Fri, 04 Jun 2021 13:11:25 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lp9bh-000742-DY for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 13:11:25 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lp9bh-0004eB-Cl for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 13:11:25 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lp9bh-0004fR-Bo for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 13:11:25 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=JyKSC7koOdsOR0JHXguV7COIZCIj2MJVpx9NTZbfRq8=; b=bEMHjtenS0XD2b07WtGzhTgTOg Nzx5rhKKfyyRYcSZxXjKmVRVptxXa+MwOf3hTu/gW3g/Rb9Jcqg1HpUl38PiFVCP86B1lvyesqqmE YhbHv3hK3pCfLD5aSjQLeKlxpzW7w17+zkpi2TUaQhqp4+w3uoqfKGTKiEFG6qtWOvb0=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.13] x86emul: fix test harness build for gas 2.36 Message-Id: Date: Fri, 04 Jun 2021 13:11:25 +0000 commit 4d544141ebab31dabb542c7dbacad658c5d3757c Author: Jan Beulich AuthorDate: Fri Jun 4 15:02:13 2021 +0200 Commit: Jan Beulich CommitDate: Fri Jun 4 15:02:13 2021 +0200 x86emul: fix test harness build for gas 2.36 All of the sudden, besides .text and .rodata and alike, an always present .note.gnu.property section has appeared. This section, when converting to binary format output, gets placed according to its linked address, causing the resulting blobs to be about 128Mb in size. The resulting headers with a C representation of the binary blobs then are, of course all a multiple of that size (and take accordingly long to create). I didn't bother waiting to see what size the final test_x86_emulator binary then would have had. See also https://sourceware.org/bugzilla/show_bug.cgi?id=27753. Rather than figuring out whether gas supports -mx86-used-note=, simply remove the section while creating *.bin. Signed-off-by: Jan Beulich master commit: aa803ba38a867551917d11059eaa044955556e05 master date: 2021-05-17 15:41:28 +0200 --- tools/tests/x86_emulator/testcase.mk | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/tools/tests/x86_emulator/testcase.mk b/tools/tests/x86_emulator/testcase.mk index dafeb6caf7..d4aa82e3f3 100644 --- a/tools/tests/x86_emulator/testcase.mk +++ b/tools/tests/x86_emulator/testcase.mk @@ -12,11 +12,11 @@ all: $(TESTCASE).bin %.bin: %.c $(CC) $(filter-out -M% .%,$(CFLAGS)) -c $< $(LD) $(LDFLAGS_DIRECT) -N -Ttext 0x100000 -o $*.tmp $*.o - $(OBJCOPY) -O binary $*.tmp $@ + $(OBJCOPY) -O binary -R .note.gnu.property $*.tmp $@ rm -f $*.tmp %-opmask.bin: opmask.S $(CC) $(filter-out -M% .%,$(CFLAGS)) -c $< -o $(basename $@).o $(LD) $(LDFLAGS_DIRECT) -N -Ttext 0x100000 -o $(basename $@).tmp $(basename $@).o - $(OBJCOPY) -O binary $(basename $@).tmp $@ + $(OBJCOPY) -O binary -R .note.gnu.property $(basename $@).tmp $@ rm -f $(basename $@).tmp -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.13 From xen-changelog-bounces@lists.xenproject.org Fri Jun 04 13:11:37 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Fri, 04 Jun 2021 13:11:37 +0000 Received: from list by lists.xenproject.org with outflank-mailman.136834.253577 (Exim 4.92) (envelope-from ) id 1lp9bs-00077l-VI; Fri, 04 Jun 2021 13:11:36 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 136834.253577; Fri, 04 Jun 2021 13:11:36 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lp9bs-00077d-Qo; Fri, 04 Jun 2021 13:11:36 +0000 Received: by outflank-mailman (input) for mailman id 136834; Fri, 04 Jun 2021 13:11:35 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lp9br-00077Q-Hn for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 13:11:35 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lp9br-0004eR-Gv for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 13:11:35 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lp9br-0004g9-FX for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 13:11:35 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=HLxWD3bUJNA+zr3IUatYSL8InL/+jEdQY21xhTtM9CI=; b=afL8/wPkIYyu8JT+3xoQ27mz8R zXd1/h+sp2vkaskZfXdJ/65mG0IbOIm+NCNVCsaNctFA4fYm/26HLzEcuqxt6DjXxe7oNk+2jaXwx 0PS3GoZlt+ACOaKuEcZcsVYnmiA3cI8x6UojJNDoxOO6co76zNmxezHvly803U6JaI2s=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.13] x86: make hypervisor build with gcc11 Message-Id: Date: Fri, 04 Jun 2021 13:11:35 +0000 commit fa5afbbc20ef3577c5338f9d0b24dad45cef59cd Author: Jan Beulich AuthorDate: Fri Jun 4 15:02:33 2021 +0200 Commit: Jan Beulich CommitDate: Fri Jun 4 15:02:33 2021 +0200 x86: make hypervisor build with gcc11 Gcc 11 looks to make incorrect assumptions about valid ranges that pointers may be used for addressing when they are derived from e.g. a plain constant. See https://gcc.gnu.org/bugzilla/show_bug.cgi?id=100680. Utilize RELOC_HIDE() to work around the issue, which for x86 manifests in at least - mpparse.c:efi_check_config(), - tboot.c:tboot_probe(), - tboot.c:tboot_gen_frametable_integrity(), - x86_emulate.c:x86_emulate() (at -O2 only). The last case is particularly odd not just because it only triggers at higher optimization levels, but also because it only affects one of at least three similar constructs. Various "note" diagnostics claim the valid index range to be [0, 2⁶³-1]. Signed-off-by: Jan Beulich Tested-by: Jason Andryuk Acked-by: Roger Pau Monné master commit: 722f59d38c710a940ab05e542a83020eb5546dea master date: 2021-05-27 14:40:29 +0200 --- tools/tests/x86_emulator/x86-emulate.c | 7 +++++++ xen/arch/x86/x86_emulate/x86_emulate.c | 2 +- xen/include/asm-x86/fixmap.h | 2 +- xen/include/xen/compiler.h | 6 ++++++ xen/include/xen/pdx.h | 2 +- 5 files changed, 16 insertions(+), 3 deletions(-) diff --git a/tools/tests/x86_emulator/x86-emulate.c b/tools/tests/x86_emulator/x86-emulate.c index 23dcbe4247..ecb4b6245e 100644 --- a/tools/tests/x86_emulator/x86-emulate.c +++ b/tools/tests/x86_emulator/x86-emulate.c @@ -2,6 +2,13 @@ #include +/* See gcc bug 100680, but here don't bother making this version dependent. */ +#define gcc11_wrap(x) ({ \ + unsigned long x_; \ + __asm__ ( "" : "=g" (x_) : "0" (x) ); \ + (typeof(x))x_; \ +}) + #define cpu_has_amd_erratum(nr) 0 #define cpu_has_mpx false #define read_bndcfgu() 0 diff --git a/xen/arch/x86/x86_emulate/x86_emulate.c b/xen/arch/x86/x86_emulate/x86_emulate.c index bba6dd0187..2dee940adb 100644 --- a/xen/arch/x86/x86_emulate/x86_emulate.c +++ b/xen/arch/x86/x86_emulate/x86_emulate.c @@ -723,7 +723,7 @@ union vex { #define copy_VEX(ptr, vex) ({ \ if ( !mode_64bit() ) \ (vex).reg |= 8; \ - (ptr)[0 - PFX_BYTES] = ext < ext_8f08 ? 0xc4 : 0x8f; \ + gcc11_wrap(ptr)[0 - PFX_BYTES] = ext < ext_8f08 ? 0xc4 : 0x8f; \ (ptr)[1 - PFX_BYTES] = (vex).raw[0]; \ (ptr)[2 - PFX_BYTES] = (vex).raw[1]; \ container_of((ptr) + 1 - PFX_BYTES, typeof(vex), raw[0]); \ diff --git a/xen/include/asm-x86/fixmap.h b/xen/include/asm-x86/fixmap.h index 9fb2f47946..b6b981b621 100644 --- a/xen/include/asm-x86/fixmap.h +++ b/xen/include/asm-x86/fixmap.h @@ -78,7 +78,7 @@ extern void __set_fixmap( #define clear_fixmap(idx) __set_fixmap(idx, 0, 0) -#define __fix_to_virt(x) (FIXADDR_TOP - ((x) << PAGE_SHIFT)) +#define __fix_to_virt(x) gcc11_wrap(FIXADDR_TOP - ((x) << PAGE_SHIFT)) #define __virt_to_fix(x) ((FIXADDR_TOP - ((x)&PAGE_MASK)) >> PAGE_SHIFT) #define fix_to_virt(x) ((void *)__fix_to_virt(x)) diff --git a/xen/include/xen/compiler.h b/xen/include/xen/compiler.h index ff6c0f5cdd..85cbd1ab00 100644 --- a/xen/include/xen/compiler.h +++ b/xen/include/xen/compiler.h @@ -99,6 +99,12 @@ __asm__ ("" : "=r"(__ptr) : "0"(ptr)); \ (typeof(ptr)) (__ptr + (off)); }) +#if CONFIG_GCC_VERSION >= 110000 /* See gcc bug 100680. */ +# define gcc11_wrap(x) RELOC_HIDE(x, 0) +#else +# define gcc11_wrap(x) (x) +#endif + #ifdef __GCC_ASM_FLAG_OUTPUTS__ # define ASM_FLAG_OUT(yes, no) yes #else diff --git a/xen/include/xen/pdx.h b/xen/include/xen/pdx.h index 770fadc06c..9fcfb0ce52 100644 --- a/xen/include/xen/pdx.h +++ b/xen/include/xen/pdx.h @@ -19,7 +19,7 @@ extern u64 pdx_region_mask(u64 base, u64 len); extern void set_pdx_range(unsigned long smfn, unsigned long emfn); #define page_to_pdx(pg) ((pg) - frame_table) -#define pdx_to_page(pdx) (frame_table + (pdx)) +#define pdx_to_page(pdx) gcc11_wrap(frame_table + (pdx)) bool __mfn_valid(unsigned long mfn); -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.13 From xen-changelog-bounces@lists.xenproject.org Fri Jun 04 13:11:47 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Fri, 04 Jun 2021 13:11:47 +0000 Received: from list by lists.xenproject.org with outflank-mailman.136835.253580 (Exim 4.92) (envelope-from ) id 1lp9c2-0007Aq-Vg; Fri, 04 Jun 2021 13:11:46 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 136835.253580; Fri, 04 Jun 2021 13:11:46 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lp9c2-0007Ag-Sg; Fri, 04 Jun 2021 13:11:46 +0000 Received: by outflank-mailman (input) for mailman id 136835; Fri, 04 Jun 2021 13:11:45 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lp9c1-0007AU-KX for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 13:11:45 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lp9c1-0004ev-Jl for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 13:11:45 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lp9c1-0004gj-J1 for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 13:11:45 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=Z1nSy5oWxoXpAgadgbm3C+9O59DcBSIf4Dc3vfyW2EM=; b=UzG7Zh3TTreMPWsVhEFWGs6UE4 nk16Hje1NUdT04XM7Qj8jQ+ZG8Fkydp1xPONYe6GaVhfUcA/bL9XLH8GI6kJjtkKcM3i88NDtEva0 k3VQAw4Wenm8bmkFF+LtJ3E84br9MuTZTom/+KW9/knMN55Avu5yuedSdj5/zKn4NpUI=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.13] x86/cpuid: Rework HLE and RTM handling Message-Id: Date: Fri, 04 Jun 2021 13:11:45 +0000 commit f17d848c4caa4fa562f3121c430667b6df838700 Author: Andrew Cooper AuthorDate: Fri Jun 4 15:04:56 2021 +0200 Commit: Jan Beulich CommitDate: Fri Jun 4 15:04:56 2021 +0200 x86/cpuid: Rework HLE and RTM handling The TAA mitigation offered the option to hide the HLE and RTM CPUID bits, which has caused some migration compatibility problems. These two bits are special. Annotate them with ! to emphasise this point. Hardware Lock Elision (HLE) may or may not be visible in CPUID, but is disabled in microcode on all CPUs, and has been removed from the architecture. Do not advertise it to VMs by default. Restricted Transactional Memory (RTM) may or may not be visible in CPUID, and may or may not be configured in force-abort mode. Have tsx_init() note whether RTM has been configured into force-abort mode, so guest_common_feature_adjustments() can conditionally hide it from VMs by default. The host policy values for HLE/RTM may or may not be set, depending on any previous running kernel's choice of visibility, and Xen's choice. TSX is available on any CPU which enumerates a TSX-hiding mechanism, so instead of doing a two-step to clobber any hiding, scan CPUID, then set the visibility, just force visibility of the bits in the first place. With the HLE/RTM bits now unilaterally visible in the host policy, xc_cpuid_apply_policy() can construct a more appropriate policy out of thin air for pre-4.13 VMs with no CPUID data in their migration stream, and specifically one where HLE/RTM doesn't potentially disappear behind the back of a running VM. Fixes: 8c4330818f6 ("x86/spec-ctrl: Mitigate the TSX Asynchronous Abort sidechannel") Signed-off-by: Andrew Cooper Reviewed-by: Jan Beulich Reviewed-by: Roger Pau Monné master commit: 8fe24090d940d760145ccd5e234290be7418b175 master date: 2021-05-27 19:34:00 +0100 --- xen/arch/x86/tsx.c | 24 ++++++++++++++++++++++++ xen/include/public/arch-x86/cpufeatureset.h | 4 ++-- 2 files changed, 26 insertions(+), 2 deletions(-) diff --git a/xen/arch/x86/tsx.c b/xen/arch/x86/tsx.c index 2d202a0d4e..a5d5944fe8 100644 --- a/xen/arch/x86/tsx.c +++ b/xen/arch/x86/tsx.c @@ -44,6 +44,30 @@ void tsx_init(void) rdmsrl(MSR_ARCH_CAPABILITIES, caps); cpu_has_tsx_ctrl = !!(caps & ARCH_CAPS_TSX_CTRL); + + /* + * The TSX features (HLE/RTM) are handled specially. They both + * enumerate features but, on certain parts, have mechanisms to be + * hidden without disrupting running software. + * + * At the moment, we're running in an unknown context (WRT hiding - + * particularly if another fully fledged kernel ran before us) and + * depending on user settings, may elect to continue hiding them from + * native CPUID instructions. + * + * Xen doesn't use TSX itself, but use cpu_has_{hle,rtm} for various + * system reasons, mostly errata detection, so the meaning is more + * useful as "TSX infrastructure available", as opposed to "features + * advertised and working". + * + * Force the features to be visible in Xen's view if we see any of the + * infrastructure capable of hiding them. + */ + if ( cpu_has_tsx_ctrl ) + { + setup_force_cpu_cap(X86_FEATURE_HLE); + setup_force_cpu_cap(X86_FEATURE_RTM); + } } if ( cpu_has_tsx_ctrl ) diff --git a/xen/include/public/arch-x86/cpufeatureset.h b/xen/include/public/arch-x86/cpufeatureset.h index 3162ee0378..a1d1619643 100644 --- a/xen/include/public/arch-x86/cpufeatureset.h +++ b/xen/include/public/arch-x86/cpufeatureset.h @@ -195,14 +195,14 @@ XEN_CPUFEATURE(FSGSBASE, 5*32+ 0) /*A {RD,WR}{FS,GS}BASE instructions */ XEN_CPUFEATURE(TSC_ADJUST, 5*32+ 1) /*S TSC_ADJUST MSR available */ XEN_CPUFEATURE(SGX, 5*32+ 2) /* Software Guard extensions */ XEN_CPUFEATURE(BMI1, 5*32+ 3) /*A 1st bit manipulation extensions */ -XEN_CPUFEATURE(HLE, 5*32+ 4) /*A Hardware Lock Elision */ +XEN_CPUFEATURE(HLE, 5*32+ 4) /*!A Hardware Lock Elision */ XEN_CPUFEATURE(AVX2, 5*32+ 5) /*A AVX2 instructions */ XEN_CPUFEATURE(FDP_EXCP_ONLY, 5*32+ 6) /*! x87 FDP only updated on exception. */ XEN_CPUFEATURE(SMEP, 5*32+ 7) /*S Supervisor Mode Execution Protection */ XEN_CPUFEATURE(BMI2, 5*32+ 8) /*A 2nd bit manipulation extensions */ XEN_CPUFEATURE(ERMS, 5*32+ 9) /*A Enhanced REP MOVSB/STOSB */ XEN_CPUFEATURE(INVPCID, 5*32+10) /*H Invalidate Process Context ID */ -XEN_CPUFEATURE(RTM, 5*32+11) /*A Restricted Transactional Memory */ +XEN_CPUFEATURE(RTM, 5*32+11) /*!A Restricted Transactional Memory */ XEN_CPUFEATURE(PQM, 5*32+12) /* Platform QoS Monitoring */ XEN_CPUFEATURE(NO_FPU_SEL, 5*32+13) /*! FPU CS/DS stored as zero */ XEN_CPUFEATURE(MPX, 5*32+14) /*S Memory Protection Extensions */ -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.13 From xen-changelog-bounces@lists.xenproject.org Fri Jun 04 13:11:57 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Fri, 04 Jun 2021 13:11:57 +0000 Received: from list by lists.xenproject.org with outflank-mailman.136836.253584 (Exim 4.92) (envelope-from ) id 1lp9cD-0007Ek-2f; Fri, 04 Jun 2021 13:11:57 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 136836.253584; Fri, 04 Jun 2021 13:11:57 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lp9cC-0007Ec-Vr; Fri, 04 Jun 2021 13:11:56 +0000 Received: by outflank-mailman (input) for mailman id 136836; Fri, 04 Jun 2021 13:11:55 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lp9cB-0007E6-NS for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 13:11:55 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lp9cB-0004f6-Mh for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 13:11:55 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lp9cB-0004hM-Lo for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 13:11:55 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=tRhdeX+BcobREIqltkysWLZjyYj9c0uKvUtNfo5t1HQ=; b=qUZdo8mD7//7rY6MqCshgr1Fan NUzh2KoaMpamlAdlczBSr3RfcnS+52X1zb00C0fqXTmq2/6wQ4GdpINAPosSPWSpZSOOIrhnRNiJ8 O8tKVeT/YTSeUVkGtDlyKZIyTt39bpgzP0gjd0zN2HsDk3aVMrmXdH2jKB6Qc130zv6k=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.13] x86: fix build race when generating temporary object files (take 2) Message-Id: Date: Fri, 04 Jun 2021 13:11:55 +0000 commit ef8b2357d83442b5d2e7607379a935d4f8b35416 Author: Jan Beulich AuthorDate: Fri Jun 4 15:05:38 2021 +0200 Commit: Jan Beulich CommitDate: Fri Jun 4 15:05:38 2021 +0200 x86: fix build race when generating temporary object files (take 2) The original commit wasn't quite sufficient: Emptying DEPS is helpful only when nothing will get added to it subsequently. xen/Rules.mk will, after including the local Makefile, amend DEPS by dependencies for objects living in sub-directories though. For the purpose of suppressing dependencies of the makefiles on the .*.d2 files (and thus to avoid their re-generation) it is, however, not necessary at all to play with DEPS. Instead we can override DEPS_INCLUDE (which generally is a late- expansion variable). Fixes: 761bb575ce97 ("x86: fix build race when generating temporary object files") Signed-off-by: Jan Beulich master commit: 8c90dbb99907f3b471d558775777a84daec7c3f6 master date: 2021-05-28 09:12:24 +0200 --- xen/arch/x86/Makefile | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/xen/arch/x86/Makefile b/xen/arch/x86/Makefile index c57ae581dd..3cca74d61b 100644 --- a/xen/arch/x86/Makefile +++ b/xen/arch/x86/Makefile @@ -269,5 +269,5 @@ clean:: # Suppress loading of DEPS files for internal, temporary target files. This # then also suppresses re-generation of the respective .*.d2 files. ifeq ($(filter-out .xen%.o,$(notdir $(MAKECMDGOALS))),) -DEPS:= +DEPS_INCLUDE:= endif -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.13 From xen-changelog-bounces@lists.xenproject.org Fri Jun 04 16:33:12 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Fri, 04 Jun 2021 16:33:12 +0000 Received: from list by lists.xenproject.org with outflank-mailman.136947.253748 (Exim 4.92) (envelope-from ) id 1lpCkr-0008CC-Vf; Fri, 04 Jun 2021 16:33:05 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 136947.253748; Fri, 04 Jun 2021 16:33:05 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpCkr-0008C4-Si; Fri, 04 Jun 2021 16:33:05 +0000 Received: by outflank-mailman (input) for mailman id 136947; Fri, 04 Jun 2021 16:33:05 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpCkq-0008By-Us for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 16:33:04 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpCkq-00009k-Tw for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 16:33:04 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lpCkq-0005BX-Sz for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 16:33:04 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=FYtbFrgiIbWKC4dfzlSG5qKGDwiiSor/lvxLkYHI5MM=; b=yFptuBDlH7BYQ9C+JZhoJt4Kmh 9Ez5Sy84qcWuNIjBw/ZauD2gEsGLv89koXig+7KphcII9BilMSNqwr01zX7whlYnOcJexZ9KS8kW2 oKzK4ZZj33Xeh1mk/+w+d4aucJXLgi+xtb7YdhCACr4V2/VeGa0MnPMaw/t65wyhQviE=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] SUPPORT.md: Un-shimmed 32-bit PV guests are no longer supported Message-Id: Date: Fri, 04 Jun 2021 16:33:04 +0000 commit 1a0f2fe2297d122a08fee2b26de5de995fdeca13 Author: George Dunlap AuthorDate: Thu May 6 13:38:02 2021 +0100 Commit: George Dunlap CommitDate: Fri Jun 4 17:24:05 2021 +0100 SUPPORT.md: Un-shimmed 32-bit PV guests are no longer supported The support status of 32-bit guests doesn't seem particularly useful. With it changed to fully unsupported outside of PV-shim, adjust the PV32 Kconfig default accordingly. Reported-by: Jann Horn Signed-off-by: George Dunlap Signed-off-by: Jan Beulich --- SUPPORT.md | 9 +-------- xen/arch/x86/Kconfig | 7 +++++-- 2 files changed, 6 insertions(+), 10 deletions(-) diff --git a/SUPPORT.md b/SUPPORT.md index 7a53635c9e..317392d8f3 100644 --- a/SUPPORT.md +++ b/SUPPORT.md @@ -86,14 +86,7 @@ No hardware requirements Status, x86_64: Supported Status, x86_32, shim: Supported - Status, x86_32, without shim: Supported, with caveats - -Due to architectural limitations, -32-bit PV guests must be assumed to be able to read arbitrary host memory -using speculative execution attacks. -Advisories will continue to be issued -for new vulnerabilities related to un-shimmed 32-bit PV guests -enabling denial-of-service attacks or privilege escalation attacks. + Status, x86_32, without shim: Supported, not security supported ### x86/HVM diff --git a/xen/arch/x86/Kconfig b/xen/arch/x86/Kconfig index e55e029b79..9b164db641 100644 --- a/xen/arch/x86/Kconfig +++ b/xen/arch/x86/Kconfig @@ -55,7 +55,7 @@ config PV config PV32 bool "Support for 32bit PV guests" depends on PV - default y + default PV_SHIM select COMPAT ---help--- The 32bit PV ABI uses Ring1, an area of the x86 architecture which @@ -67,7 +67,10 @@ config PV32 reduction, or performance reasons. Backwards compatibility can be provided via the PV Shim mechanism. - If unsure, say Y. + Note that outside of PV Shim, 32-bit PV guests are not security + supported anymore. + + If unsure, use the default setting. config PV_LINEAR_PT bool "Support for PV linear pagetables" -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Fri Jun 04 18:11:11 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Fri, 04 Jun 2021 18:11:11 +0000 Received: from list by lists.xenproject.org with outflank-mailman.136980.253802 (Exim 4.92) (envelope-from ) id 1lpEHi-0002KX-7Y; Fri, 04 Jun 2021 18:11:06 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 136980.253802; Fri, 04 Jun 2021 18:11:06 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpEHi-0002KP-4g; Fri, 04 Jun 2021 18:11:06 +0000 Received: by outflank-mailman (input) for mailman id 136980; Fri, 04 Jun 2021 18:11:04 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpEHg-0002KJ-KQ for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 18:11:04 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpEHg-0001uR-HZ for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 18:11:04 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lpEHg-0007zI-Ga for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 18:11:04 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=SPnCkcib2GjqIWC6yzVbnHizb5k7HWrqT9GocpW7b4Y=; b=Qbgq2Bpg96d258cqhJeCAXwPJG 0xcsSW11VQ7kVsnnFs6rUgGhhB2rd5UZ6oYQWWeRMOhy6ht19tFNDvpbeOLSWxyZlMZ11kLUzjR0T +kXSIz1YJBcY+bxQ680WVaKZI9lp/hOVgk9Fpa5LJ+aHvYu4WZncVDabVlv2KelGsc+Q=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] tools/libs/guest: fix max_pfn setting in map_p2m() Message-Id: Date: Fri, 04 Jun 2021 18:11:04 +0000 commit 7bd8989ab77b6ade3b7a5f4b640a55248d1791a3 Author: Juergen Gross AuthorDate: Fri Jun 4 08:02:09 2021 +0200 Commit: Julien Grall CommitDate: Fri Jun 4 18:56:46 2021 +0100 tools/libs/guest: fix max_pfn setting in map_p2m() When setting the highest pfn used in the guest, don't subtract 1 from the value read from the shared_info data. The value read already is the correct pfn. Fixes: 91e204d37f449 ("libxc: try to find last used pfn when migrating") Signed-off-by: Juergen Gross Acked-by: Wei Liu --- tools/libs/guest/xg_sr_save_x86_pv.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/tools/libs/guest/xg_sr_save_x86_pv.c b/tools/libs/guest/xg_sr_save_x86_pv.c index 4964f1f7b8..dae7f2817f 100644 --- a/tools/libs/guest/xg_sr_save_x86_pv.c +++ b/tools/libs/guest/xg_sr_save_x86_pv.c @@ -468,7 +468,7 @@ static int map_p2m(struct xc_sr_context *ctx) ctx->x86.pv.p2m_generation = ~0ULL; ctx->x86.pv.max_pfn = GET_FIELD(ctx->x86.pv.shinfo, arch.max_pfn, - ctx->x86.pv.width) - 1; + ctx->x86.pv.width); p2m_cr3 = GET_FIELD(ctx->x86.pv.shinfo, arch.p2m_cr3, ctx->x86.pv.width); return p2m_cr3 ? map_p2m_list(ctx, p2m_cr3) : map_p2m_tree(ctx); -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Fri Jun 04 18:11:16 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Fri, 04 Jun 2021 18:11:16 +0000 Received: from list by lists.xenproject.org with outflank-mailman.136981.253806 (Exim 4.92) (envelope-from ) id 1lpEHs-0002MS-9Q; Fri, 04 Jun 2021 18:11:16 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 136981.253806; Fri, 04 Jun 2021 18:11:16 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpEHs-0002MJ-6C; Fri, 04 Jun 2021 18:11:16 +0000 Received: by outflank-mailman (input) for mailman id 136981; Fri, 04 Jun 2021 18:11:14 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpEHq-0002MB-MD for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 18:11:14 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpEHq-0001ue-LJ for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 18:11:14 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lpEHq-000808-KA for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 18:11:14 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=+NC0ii3D1157/7qQPwgjiCW72qiVuHdej+XLqqy6jB8=; b=4QqPws2JFwFptneaRmPvt1ga5L YHr0GZAq+h6upOFjst/qEpqj/dG0yvUjQGc2s0g3ZTA8E1xThaWMd72mgM9u/BjziT9xnE/U9lM1F c8R1aimMzCDbeBAbx2PLHFn/rt5JcRFGbAt8tYXfUl8dk1xQGdKubguj8GWrY5swvvso=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] tools/libs/ctrl: fix xc_core_arch_map_p2m() to support linear p2m table Message-Id: Date: Fri, 04 Jun 2021 18:11:14 +0000 commit bd7a29c3d0b937ab542abea06ff1b575abe7247a Author: Juergen Gross AuthorDate: Fri Jun 4 08:02:10 2021 +0200 Commit: Julien Grall CommitDate: Fri Jun 4 18:56:46 2021 +0100 tools/libs/ctrl: fix xc_core_arch_map_p2m() to support linear p2m table The core of a pv linux guest produced via "xl dump-core" is nor usable as since kernel 4.14 only the linear p2m table is kept if Xen indicates it is supporting that. Unfortunately xc_core_arch_map_p2m() is still supporting the 3-level p2m tree only. Fix that by copying the functionality of map_p2m() from libxenguest to libxenctrl. Additionally the mapped p2m isn't of a fixed length now, so the interface to the mapping functions needs to be adapted. In order not to add even more parameters, expand struct domain_info_context and use a pointer to that as a parameter. Fixes: dc6d60937121 ("libxc: set flag for support of linear p2m list in domain builder") Signed-off-by: Juergen Gross Acked-by: Wei Liu --- tools/include/xenguest.h | 1 + tools/libs/ctrl/xc_core.c | 5 +- tools/libs/ctrl/xc_core.h | 8 +- tools/libs/ctrl/xc_core_arm.c | 23 ++-- tools/libs/ctrl/xc_core_x86.c | 256 ++++++++++++++++++++++++++++++++++-------- tools/libs/ctrl/xc_private.h | 1 + tools/libs/guest/xg_domain.c | 17 +-- 7 files changed, 233 insertions(+), 78 deletions(-) diff --git a/tools/include/xenguest.h b/tools/include/xenguest.h index a4492038cf..f9fb0449ad 100644 --- a/tools/include/xenguest.h +++ b/tools/include/xenguest.h @@ -700,6 +700,7 @@ struct xc_domain_meminfo { xen_pfn_t *pfn_type; xen_pfn_t *p2m_table; unsigned long p2m_size; + unsigned int p2m_frames; }; int xc_map_domain_meminfo(xc_interface *xch, uint32_t domid, diff --git a/tools/libs/ctrl/xc_core.c b/tools/libs/ctrl/xc_core.c index b47ab2f6d8..9576bec5a3 100644 --- a/tools/libs/ctrl/xc_core.c +++ b/tools/libs/ctrl/xc_core.c @@ -574,8 +574,7 @@ xc_domain_dumpcore_via_callback(xc_interface *xch, goto out; } - sts = xc_core_arch_map_p2m(xch, dinfo->guest_width, &info, live_shinfo, - &p2m, &dinfo->p2m_size); + sts = xc_core_arch_map_p2m(xch, dinfo, &info, live_shinfo, &p2m); if ( sts != 0 ) goto out; @@ -945,7 +944,7 @@ out: if ( memory_map != NULL ) free(memory_map); if ( p2m != NULL ) - munmap(p2m, PAGE_SIZE * P2M_FL_ENTRIES); + munmap(p2m, PAGE_SIZE * dinfo->p2m_frames); if ( p2m_array != NULL ) free(p2m_array); if ( pfn_array != NULL ) diff --git a/tools/libs/ctrl/xc_core.h b/tools/libs/ctrl/xc_core.h index 36fb755da2..8ea1f93a10 100644 --- a/tools/libs/ctrl/xc_core.h +++ b/tools/libs/ctrl/xc_core.h @@ -138,14 +138,14 @@ int xc_core_arch_memory_map_get(xc_interface *xch, xc_dominfo_t *info, shared_info_any_t *live_shinfo, xc_core_memory_map_t **mapp, unsigned int *nr_entries); -int xc_core_arch_map_p2m(xc_interface *xch, unsigned int guest_width, +int xc_core_arch_map_p2m(xc_interface *xch, struct domain_info_context *dinfo, xc_dominfo_t *info, shared_info_any_t *live_shinfo, - xen_pfn_t **live_p2m, unsigned long *pfnp); + xen_pfn_t **live_p2m); -int xc_core_arch_map_p2m_writable(xc_interface *xch, unsigned int guest_width, +int xc_core_arch_map_p2m_writable(xc_interface *xch, struct domain_info_context *dinfo, xc_dominfo_t *info, shared_info_any_t *live_shinfo, - xen_pfn_t **live_p2m, unsigned long *pfnp); + xen_pfn_t **live_p2m); int xc_core_arch_get_scratch_gpfn(xc_interface *xch, uint32_t domid, xen_pfn_t *gpfn); diff --git a/tools/libs/ctrl/xc_core_arm.c b/tools/libs/ctrl/xc_core_arm.c index 7b587b4cc5..93765a565f 100644 --- a/tools/libs/ctrl/xc_core_arm.c +++ b/tools/libs/ctrl/xc_core_arm.c @@ -66,33 +66,24 @@ xc_core_arch_memory_map_get(xc_interface *xch, struct xc_core_arch_context *unus static int xc_core_arch_map_p2m_rw(xc_interface *xch, struct domain_info_context *dinfo, xc_dominfo_t *info, - shared_info_any_t *live_shinfo, xen_pfn_t **live_p2m, - unsigned long *pfnp, int rw) + shared_info_any_t *live_shinfo, xen_pfn_t **live_p2m, int rw) { errno = ENOSYS; return -1; } int -xc_core_arch_map_p2m(xc_interface *xch, unsigned int guest_width, xc_dominfo_t *info, - shared_info_any_t *live_shinfo, xen_pfn_t **live_p2m, - unsigned long *pfnp) +xc_core_arch_map_p2m(xc_interface *xch, struct domain_info_context *dinfo, xc_dominfo_t *info, + shared_info_any_t *live_shinfo, xen_pfn_t **live_p2m) { - struct domain_info_context _dinfo = { .guest_width = guest_width }; - struct domain_info_context *dinfo = &_dinfo; - return xc_core_arch_map_p2m_rw(xch, dinfo, info, - live_shinfo, live_p2m, pfnp, 0); + return xc_core_arch_map_p2m_rw(xch, dinfo, info, live_shinfo, live_p2m, 0); } int -xc_core_arch_map_p2m_writable(xc_interface *xch, unsigned int guest_width, xc_dominfo_t *info, - shared_info_any_t *live_shinfo, xen_pfn_t **live_p2m, - unsigned long *pfnp) +xc_core_arch_map_p2m_writable(xc_interface *xch, struct domain_info_context *dinfo, xc_dominfo_t *info, + shared_info_any_t *live_shinfo, xen_pfn_t **live_p2m) { - struct domain_info_context _dinfo = { .guest_width = guest_width }; - struct domain_info_context *dinfo = &_dinfo; - return xc_core_arch_map_p2m_rw(xch, dinfo, info, - live_shinfo, live_p2m, pfnp, 1); + return xc_core_arch_map_p2m_rw(xch, dinfo, info, live_shinfo, live_p2m, 1); } int diff --git a/tools/libs/ctrl/xc_core_x86.c b/tools/libs/ctrl/xc_core_x86.c index cb76e6207b..c8f71d4b75 100644 --- a/tools/libs/ctrl/xc_core_x86.c +++ b/tools/libs/ctrl/xc_core_x86.c @@ -17,6 +17,7 @@ * */ +#include #include "xc_private.h" #include "xc_core.h" #include @@ -65,34 +66,169 @@ xc_core_arch_memory_map_get(xc_interface *xch, struct xc_core_arch_context *unus return 0; } -static int -xc_core_arch_map_p2m_rw(xc_interface *xch, struct domain_info_context *dinfo, xc_dominfo_t *info, - shared_info_any_t *live_shinfo, xen_pfn_t **live_p2m, - unsigned long *pfnp, int rw) +static inline bool is_canonical_address(uint64_t vaddr) { - /* Double and single indirect references to the live P2M table */ - xen_pfn_t *live_p2m_frame_list_list = NULL; - xen_pfn_t *live_p2m_frame_list = NULL; - /* Copies of the above. */ - xen_pfn_t *p2m_frame_list_list = NULL; - xen_pfn_t *p2m_frame_list = NULL; + return ((int64_t)vaddr >> 47) == ((int64_t)vaddr >> 63); +} - uint32_t dom = info->domid; - int ret = -1; - int err; - int i; +/* Virtual address ranges reserved for hypervisor. */ +#define HYPERVISOR_VIRT_START_X86_64 0xFFFF800000000000ULL +#define HYPERVISOR_VIRT_END_X86_64 0xFFFF87FFFFFFFFFFULL - if ( xc_domain_nr_gpfns(xch, info->domid, &dinfo->p2m_size) < 0 ) +#define HYPERVISOR_VIRT_START_X86_32 0x00000000F5800000ULL +#define HYPERVISOR_VIRT_END_X86_32 0x00000000FFFFFFFFULL + +static xen_pfn_t * +xc_core_arch_map_p2m_list_rw(xc_interface *xch, struct domain_info_context *dinfo, + uint32_t dom, shared_info_any_t *live_shinfo, + uint64_t p2m_cr3) +{ + uint64_t p2m_vaddr, p2m_end, mask, off; + xen_pfn_t p2m_mfn, mfn, saved_mfn, max_pfn; + uint64_t *ptes = NULL; + xen_pfn_t *mfns = NULL; + unsigned int fpp, n_pages, level, n_levels, shift, + idx_start, idx_end, idx, saved_idx; + + p2m_vaddr = GET_FIELD(live_shinfo, arch.p2m_vaddr, dinfo->guest_width); + fpp = PAGE_SIZE / dinfo->guest_width; + dinfo->p2m_frames = (dinfo->p2m_size - 1) / fpp + 1; + p2m_end = p2m_vaddr + dinfo->p2m_frames * PAGE_SIZE - 1; + + if ( dinfo->guest_width == 8 ) { - ERROR("Could not get maximum GPFN!"); - goto out; + mask = 0x0000ffffffffffffULL; + n_levels = 4; + p2m_mfn = p2m_cr3 >> 12; + if ( !is_canonical_address(p2m_vaddr) || + !is_canonical_address(p2m_end) || + p2m_end < p2m_vaddr || + (p2m_vaddr <= HYPERVISOR_VIRT_END_X86_64 && + p2m_end > HYPERVISOR_VIRT_START_X86_64) ) + { + ERROR("Bad virtual p2m address range %#" PRIx64 "-%#" PRIx64, + p2m_vaddr, p2m_end); + errno = ERANGE; + goto out; + } + } + else + { + mask = 0x00000000ffffffffULL; + n_levels = 3; + if ( p2m_cr3 & ~mask ) + p2m_mfn = ~0UL; + else + p2m_mfn = (uint32_t)((p2m_cr3 >> 12) | (p2m_cr3 << 20)); + if ( p2m_vaddr > mask || p2m_end > mask || p2m_end < p2m_vaddr || + (p2m_vaddr <= HYPERVISOR_VIRT_END_X86_32 && + p2m_end > HYPERVISOR_VIRT_START_X86_32) ) + { + ERROR("Bad virtual p2m address range %#" PRIx64 "-%#" PRIx64, + p2m_vaddr, p2m_end); + errno = ERANGE; + goto out; + } } - if ( dinfo->p2m_size < info->nr_pages ) + mfns = malloc(sizeof(*mfns)); + if ( !mfns ) { - ERROR("p2m_size < nr_pages -1 (%lx < %lx", dinfo->p2m_size, info->nr_pages - 1); + ERROR("Cannot allocate memory for array of %u mfns", 1); goto out; } + mfns[0] = p2m_mfn; + off = 0; + saved_mfn = 0; + idx_start = idx_end = saved_idx = 0; + + for ( level = n_levels; level > 0; level-- ) + { + n_pages = idx_end - idx_start + 1; + ptes = xc_map_foreign_pages(xch, dom, PROT_READ, mfns, n_pages); + if ( !ptes ) + { + PERROR("Failed to map %u page table pages for p2m list", n_pages); + goto out; + } + free(mfns); + + shift = level * 9 + 3; + idx_start = ((p2m_vaddr - off) & mask) >> shift; + idx_end = ((p2m_end - off) & mask) >> shift; + idx = idx_end - idx_start + 1; + mfns = malloc(sizeof(*mfns) * idx); + if ( !mfns ) + { + ERROR("Cannot allocate memory for array of %u mfns", idx); + goto out; + } + + for ( idx = idx_start; idx <= idx_end; idx++ ) + { + mfn = (ptes[idx] & 0x000ffffffffff000ULL) >> PAGE_SHIFT; + if ( mfn == 0 ) + { + ERROR("Bad mfn %#lx during page table walk for vaddr %#" PRIx64 " at level %d of p2m list", + mfn, off + ((uint64_t)idx << shift), level); + errno = ERANGE; + goto out; + } + mfns[idx - idx_start] = mfn; + + /* Maximum pfn check at level 2. Same reasoning as for p2m tree. */ + if ( level == 2 ) + { + if ( mfn != saved_mfn ) + { + saved_mfn = mfn; + saved_idx = idx - idx_start; + } + } + } + + if ( level == 2 ) + { + if ( saved_idx == idx_end ) + saved_idx++; + max_pfn = ((xen_pfn_t)saved_idx << 9) * fpp; + if ( max_pfn < dinfo->p2m_size ) + { + dinfo->p2m_size = max_pfn; + dinfo->p2m_frames = (dinfo->p2m_size + fpp - 1) / fpp; + p2m_end = p2m_vaddr + dinfo->p2m_frames * PAGE_SIZE - 1; + idx_end = idx_start + saved_idx; + } + } + + munmap(ptes, n_pages * PAGE_SIZE); + ptes = NULL; + off = p2m_vaddr & ((mask >> shift) << shift); + } + + return mfns; + + out: + free(mfns); + if ( ptes ) + munmap(ptes, n_pages * PAGE_SIZE); + + return NULL; +} + +static xen_pfn_t * +xc_core_arch_map_p2m_tree_rw(xc_interface *xch, struct domain_info_context *dinfo, + uint32_t dom, shared_info_any_t *live_shinfo) +{ + /* Double and single indirect references to the live P2M table */ + xen_pfn_t *live_p2m_frame_list_list; + xen_pfn_t *live_p2m_frame_list = NULL; + /* Copies of the above. */ + xen_pfn_t *p2m_frame_list_list = NULL; + xen_pfn_t *p2m_frame_list; + + int err; + int i; live_p2m_frame_list_list = xc_map_foreign_range(xch, dom, PAGE_SIZE, PROT_READ, @@ -151,10 +287,60 @@ xc_core_arch_map_p2m_rw(xc_interface *xch, struct domain_info_context *dinfo, xc for ( i = P2M_FL_ENTRIES - 1; i >= 0; i-- ) p2m_frame_list[i] = ((uint32_t *)p2m_frame_list)[i]; + dinfo->p2m_frames = P2M_FL_ENTRIES; + + return p2m_frame_list; + + out: + err = errno; + + if ( live_p2m_frame_list_list ) + munmap(live_p2m_frame_list_list, PAGE_SIZE); + + if ( live_p2m_frame_list ) + munmap(live_p2m_frame_list, P2M_FLL_ENTRIES * PAGE_SIZE); + + free(p2m_frame_list_list); + + errno = err; + + return NULL; +} + +static int +xc_core_arch_map_p2m_rw(xc_interface *xch, struct domain_info_context *dinfo, xc_dominfo_t *info, + shared_info_any_t *live_shinfo, xen_pfn_t **live_p2m, int rw) +{ + xen_pfn_t *p2m_frame_list = NULL; + uint64_t p2m_cr3; + uint32_t dom = info->domid; + int ret = -1; + int err; + + if ( xc_domain_nr_gpfns(xch, info->domid, &dinfo->p2m_size) < 0 ) + { + ERROR("Could not get maximum GPFN!"); + goto out; + } + + if ( dinfo->p2m_size < info->nr_pages ) + { + ERROR("p2m_size < nr_pages -1 (%lx < %lx", dinfo->p2m_size, info->nr_pages - 1); + goto out; + } + + p2m_cr3 = GET_FIELD(live_shinfo, arch.p2m_cr3, dinfo->guest_width); + + p2m_frame_list = p2m_cr3 ? xc_core_arch_map_p2m_list_rw(xch, dinfo, dom, live_shinfo, p2m_cr3) + : xc_core_arch_map_p2m_tree_rw(xch, dinfo, dom, live_shinfo); + + if ( !p2m_frame_list ) + goto out; + *live_p2m = xc_map_foreign_pages(xch, dom, rw ? (PROT_READ | PROT_WRITE) : PROT_READ, p2m_frame_list, - P2M_FL_ENTRIES); + dinfo->p2m_frames); if ( !*live_p2m ) { @@ -162,21 +348,11 @@ xc_core_arch_map_p2m_rw(xc_interface *xch, struct domain_info_context *dinfo, xc goto out; } - *pfnp = dinfo->p2m_size; - ret = 0; out: err = errno; - if ( live_p2m_frame_list_list ) - munmap(live_p2m_frame_list_list, PAGE_SIZE); - - if ( live_p2m_frame_list ) - munmap(live_p2m_frame_list, P2M_FLL_ENTRIES * PAGE_SIZE); - - free(p2m_frame_list_list); - free(p2m_frame_list); errno = err; @@ -184,25 +360,17 @@ out: } int -xc_core_arch_map_p2m(xc_interface *xch, unsigned int guest_width, xc_dominfo_t *info, - shared_info_any_t *live_shinfo, xen_pfn_t **live_p2m, - unsigned long *pfnp) +xc_core_arch_map_p2m(xc_interface *xch, struct domain_info_context *dinfo, xc_dominfo_t *info, + shared_info_any_t *live_shinfo, xen_pfn_t **live_p2m) { - struct domain_info_context _dinfo = { .guest_width = guest_width }; - struct domain_info_context *dinfo = &_dinfo; - return xc_core_arch_map_p2m_rw(xch, dinfo, info, - live_shinfo, live_p2m, pfnp, 0); + return xc_core_arch_map_p2m_rw(xch, dinfo, info, live_shinfo, live_p2m, 0); } int -xc_core_arch_map_p2m_writable(xc_interface *xch, unsigned int guest_width, xc_dominfo_t *info, - shared_info_any_t *live_shinfo, xen_pfn_t **live_p2m, - unsigned long *pfnp) +xc_core_arch_map_p2m_writable(xc_interface *xch, struct domain_info_context *dinfo, xc_dominfo_t *info, + shared_info_any_t *live_shinfo, xen_pfn_t **live_p2m) { - struct domain_info_context _dinfo = { .guest_width = guest_width }; - struct domain_info_context *dinfo = &_dinfo; - return xc_core_arch_map_p2m_rw(xch, dinfo, info, - live_shinfo, live_p2m, pfnp, 1); + return xc_core_arch_map_p2m_rw(xch, dinfo, info, live_shinfo, live_p2m, 1); } int diff --git a/tools/libs/ctrl/xc_private.h b/tools/libs/ctrl/xc_private.h index f0b5f83ac8..8ebc0b59da 100644 --- a/tools/libs/ctrl/xc_private.h +++ b/tools/libs/ctrl/xc_private.h @@ -79,6 +79,7 @@ struct iovec { struct domain_info_context { unsigned int guest_width; + unsigned int p2m_frames; unsigned long p2m_size; }; diff --git a/tools/libs/guest/xg_domain.c b/tools/libs/guest/xg_domain.c index 5019c84e0e..dd7db2cbd8 100644 --- a/tools/libs/guest/xg_domain.c +++ b/tools/libs/guest/xg_domain.c @@ -24,13 +24,9 @@ int xc_unmap_domain_meminfo(xc_interface *xch, struct xc_domain_meminfo *minfo) { - struct domain_info_context _di = { .guest_width = minfo->guest_width, - .p2m_size = minfo->p2m_size}; - struct domain_info_context *dinfo = &_di; - free(minfo->pfn_type); if ( minfo->p2m_table ) - munmap(minfo->p2m_table, P2M_FL_ENTRIES * PAGE_SIZE); + munmap(minfo->p2m_table, minfo->p2m_frames * PAGE_SIZE); minfo->p2m_table = NULL; return 0; @@ -40,7 +36,6 @@ int xc_map_domain_meminfo(xc_interface *xch, uint32_t domid, struct xc_domain_meminfo *minfo) { struct domain_info_context _di; - struct domain_info_context *dinfo = &_di; xc_dominfo_t info; shared_info_any_t *live_shinfo; @@ -96,16 +91,16 @@ int xc_map_domain_meminfo(xc_interface *xch, uint32_t domid, return -1; } - if ( xc_core_arch_map_p2m_writable(xch, minfo->guest_width, &info, - live_shinfo, &minfo->p2m_table, - &minfo->p2m_size) ) + if ( xc_core_arch_map_p2m_writable(xch, &_di, &info, + live_shinfo, &minfo->p2m_table) ) { PERROR("Could not map the P2M table"); munmap(live_shinfo, PAGE_SIZE); return -1; } munmap(live_shinfo, PAGE_SIZE); - _di.p2m_size = minfo->p2m_size; + minfo->p2m_size = _di.p2m_size; + minfo->p2m_frames = _di.p2m_frames; /* Make space and prepare for getting the PFN types */ minfo->pfn_type = calloc(sizeof(*minfo->pfn_type), minfo->p2m_size); @@ -141,7 +136,7 @@ failed: } if ( minfo->p2m_table ) { - munmap(minfo->p2m_table, P2M_FL_ENTRIES * PAGE_SIZE); + munmap(minfo->p2m_table, minfo->p2m_frames * PAGE_SIZE); minfo->p2m_table = NULL; } -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Fri Jun 04 18:11:26 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Fri, 04 Jun 2021 18:11:26 +0000 Received: from list by lists.xenproject.org with outflank-mailman.136982.253810 (Exim 4.92) (envelope-from ) id 1lpEI2-0002Pl-CU; Fri, 04 Jun 2021 18:11:26 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 136982.253810; Fri, 04 Jun 2021 18:11:26 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpEI2-0002Pb-9M; Fri, 04 Jun 2021 18:11:26 +0000 Received: by outflank-mailman (input) for mailman id 136982; Fri, 04 Jun 2021 18:11:24 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpEI0-0002PN-Px for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 18:11:24 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpEI0-0001uo-P6 for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 18:11:24 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lpEI0-00080h-Nf for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 18:11:24 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=KeE8SB8hcwuYnh4mXUSfZBson9KrWY7klOiy2pHa46o=; b=NS3dT+y5+eSu/EED2+fK/MEW/h EIp4r8jHoWWvShqvH3ob/TzmpH2tMSAztQ3M70aBWeyOYX80lD2FzkeEKxuTlz5uFURJmhCAQwjE7 3E5tAYrLrp5nyORVA7N9A43MVge5l9LHEDj5fJ51IEZF+cnfRIvFBDvE4Niaahb8i/XY=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] tools/libs/ctrl: use common p2m mapping code in xc_domain_resume_any() Message-Id: Date: Fri, 04 Jun 2021 18:11:24 +0000 commit f183854facad996fe891c086c024bca7cbcdc1e4 Author: Juergen Gross AuthorDate: Fri Jun 4 08:02:11 2021 +0200 Commit: Julien Grall CommitDate: Fri Jun 4 18:56:46 2021 +0100 tools/libs/ctrl: use common p2m mapping code in xc_domain_resume_any() Instead of open coding the mapping of the p2m list use the already existing xc_core_arch_map_p2m() call, especially as the current code does not support guests with the linear p2m map. It should be noted that this code is needed for colo/remus only. Switching to xc_core_arch_map_p2m() drops the need to bail out for bitness of tool stack and guest differing. Signed-off-by: Juergen Gross Acked-by: Christian Lindig Acked-by: Wei Liu --- tools/libs/ctrl/xc_resume.c | 66 ++++++++++++------------------------- tools/ocaml/libs/xc/xenctrl_stubs.c | 1 + 2 files changed, 22 insertions(+), 45 deletions(-) diff --git a/tools/libs/ctrl/xc_resume.c b/tools/libs/ctrl/xc_resume.c index 94c6c9fb31..e3c8e83aa9 100644 --- a/tools/libs/ctrl/xc_resume.c +++ b/tools/libs/ctrl/xc_resume.c @@ -20,6 +20,7 @@ #include #include #include +#include "xc_core.h" static int modify_returncode(xc_interface *xch, uint32_t domid) { @@ -137,12 +138,10 @@ static int xc_domain_resume_any(xc_interface *xch, uint32_t domid) struct domain_info_context _dinfo = { .guest_width = 0, .p2m_size = 0 }; struct domain_info_context *dinfo = &_dinfo; - unsigned long mfn; + xen_pfn_t mfn, store_mfn, console_mfn; vcpu_guest_context_any_t ctxt; - start_info_t *start_info; - shared_info_t *shinfo = NULL; - xen_pfn_t *p2m_frame_list_list = NULL; - xen_pfn_t *p2m_frame_list = NULL; + start_info_any_t *start_info; + shared_info_any_t *shinfo = NULL; xen_pfn_t *p2m = NULL; #endif @@ -164,11 +163,6 @@ static int xc_domain_resume_any(xc_interface *xch, uint32_t domid) PERROR("Could not get domain width"); return rc; } - if ( dinfo->guest_width != sizeof(long) ) - { - ERROR("Cannot resume uncooperative cross-address-size guests"); - return rc; - } /* Map the shared info frame */ shinfo = xc_map_foreign_range(xch, domid, PAGE_SIZE, @@ -179,34 +173,8 @@ static int xc_domain_resume_any(xc_interface *xch, uint32_t domid) goto out; } - dinfo->p2m_size = shinfo->arch.max_pfn; - - p2m_frame_list_list = - xc_map_foreign_range(xch, domid, PAGE_SIZE, PROT_READ, - shinfo->arch.pfn_to_mfn_frame_list_list); - if ( p2m_frame_list_list == NULL ) - { - ERROR("Couldn't map p2m_frame_list_list"); - goto out; - } - - p2m_frame_list = xc_map_foreign_pages(xch, domid, PROT_READ, - p2m_frame_list_list, - P2M_FLL_ENTRIES); - if ( p2m_frame_list == NULL ) - { - ERROR("Couldn't map p2m_frame_list"); - goto out; - } - - /* Map all the frames of the pfn->mfn table. For migrate to succeed, - the guest must not change which frames are used for this purpose. - (its not clear why it would want to change them, and we'll be OK - from a safety POV anyhow. */ - p2m = xc_map_foreign_pages(xch, domid, PROT_READ, - p2m_frame_list, - P2M_FL_ENTRIES); - if ( p2m == NULL ) + /* Map the p2m list */ + if ( xc_core_arch_map_p2m(xch, dinfo, &info, shinfo, &p2m) ) { ERROR("Couldn't map p2m table"); goto out; @@ -228,8 +196,20 @@ static int xc_domain_resume_any(xc_interface *xch, uint32_t domid) goto out; } - start_info->store_mfn = p2m[start_info->store_mfn]; - start_info->console.domU.mfn = p2m[start_info->console.domU.mfn]; + store_mfn = GET_FIELD(start_info, store_mfn, dinfo->guest_width); + console_mfn = GET_FIELD(start_info, console.domU.mfn, dinfo->guest_width); + if ( dinfo->guest_width == 4 ) + { + store_mfn = ((uint32_t *)p2m)[store_mfn]; + console_mfn = ((uint32_t *)p2m)[console_mfn]; + } + else + { + store_mfn = ((uint64_t *)p2m)[store_mfn]; + console_mfn = ((uint64_t *)p2m)[console_mfn]; + } + SET_FIELD(start_info, store_mfn, store_mfn, dinfo->guest_width); + SET_FIELD(start_info, console.domU.mfn, console_mfn, dinfo->guest_width); munmap(start_info, PAGE_SIZE); #endif /* defined(__i386__) || defined(__x86_64__) */ @@ -250,11 +230,7 @@ static int xc_domain_resume_any(xc_interface *xch, uint32_t domid) out: #if defined(__i386__) || defined(__x86_64__) if (p2m) - munmap(p2m, P2M_FL_ENTRIES*PAGE_SIZE); - if (p2m_frame_list) - munmap(p2m_frame_list, P2M_FLL_ENTRIES*PAGE_SIZE); - if (p2m_frame_list_list) - munmap(p2m_frame_list_list, PAGE_SIZE); + munmap(p2m, dinfo->p2m_frames * PAGE_SIZE); if (shinfo) munmap(shinfo, PAGE_SIZE); #endif diff --git a/tools/ocaml/libs/xc/xenctrl_stubs.c b/tools/ocaml/libs/xc/xenctrl_stubs.c index d05d7bb30e..6e4bc567f5 100644 --- a/tools/ocaml/libs/xc/xenctrl_stubs.c +++ b/tools/ocaml/libs/xc/xenctrl_stubs.c @@ -32,6 +32,7 @@ #define XC_WANT_COMPAT_MAP_FOREIGN_API #include +#include #include #include "mmap_stubs.h" -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Fri Jun 04 18:11:36 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Fri, 04 Jun 2021 18:11:36 +0000 Received: from list by lists.xenproject.org with outflank-mailman.136983.253813 (Exim 4.92) (envelope-from ) id 1lpEIC-0002T2-Ds; Fri, 04 Jun 2021 18:11:36 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 136983.253813; Fri, 04 Jun 2021 18:11:36 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpEIC-0002Sv-Au; Fri, 04 Jun 2021 18:11:36 +0000 Received: by outflank-mailman (input) for mailman id 136983; Fri, 04 Jun 2021 18:11:34 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpEIA-0002Sh-TF for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 18:11:34 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpEIA-0001uz-SQ for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 18:11:34 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lpEIA-00081M-RW for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 18:11:34 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=2CKWtYsRSMXPJEqizN3lddZ6PYL3SgybWj1kurzdPws=; b=V6dkhIpy1uCP99iQYVM4e+1EZp BD91LeId9ZKazCVZwrXX6x1XR+F9iLgnADgCakaLTl1ctaTkgdvzoww/dGbWWs+ABUNJeXQU9Ki/N 9NIRFdZEbHYZ55darNf4eGdwYlwzN2+rSMJIeCkCXoPs/XrmYdPh+ll6F4/xMPZuC5EU=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] tools/libs: move xc_resume.c to libxenguest Message-Id: Date: Fri, 04 Jun 2021 18:11:34 +0000 commit bf1fc18901dfea05a69f661493b934c0db7d3503 Author: Juergen Gross AuthorDate: Fri Jun 4 08:02:12 2021 +0200 Commit: Julien Grall CommitDate: Fri Jun 4 18:56:46 2021 +0100 tools/libs: move xc_resume.c to libxenguest The guest suspend functionality is already part of libxenguest. Move the resume functionality from libxenctrl to libxenguest, too. Signed-off-by: Juergen Gross Acked-by: Wei Liu --- tools/include/xenctrl.h | 63 ---------- tools/include/xenguest.h | 62 ++++++++++ tools/libs/ctrl/Makefile | 1 - tools/libs/ctrl/xc_resume.c | 264 ------------------------------------------ tools/libs/guest/Makefile | 1 + tools/libs/guest/xg_resume.c | 265 +++++++++++++++++++++++++++++++++++++++++++ 6 files changed, 328 insertions(+), 328 deletions(-) diff --git a/tools/include/xenctrl.h b/tools/include/xenctrl.h index 58d3377d6a..2a7c836a02 100644 --- a/tools/include/xenctrl.h +++ b/tools/include/xenctrl.h @@ -576,69 +576,6 @@ int xc_domain_destroy(xc_interface *xch, uint32_t domid); -/** - * This function resumes a suspended domain. The domain should have - * been previously suspended. - * - * Note that there are 'xc_domain_suspend' as suspending a domain - * is quite the endeavour. - * - * For the purpose of this explanation there are three guests: - * PV (using hypercalls for privilgied operations), HVM - * (fully hardware virtualized guests using emulated devices for everything), - * and PVHVM (PV aware with hardware virtualisation). - * - * HVM guest are the simplest - they suspend via S3 / S4 and resume from - * S3 / S4. Upon resume they have to re-negotiate with the emulated devices. - * - * PV and PVHVM communicate via hypercalls for suspend (and resume). - * For suspend the toolstack initiates the process by writing an value - * in XenBus "control/shutdown" with the string "suspend". - * - * The PV guest stashes anything it deems neccessary in 'struct - * start_info' in case of failure (PVHVM may ignore this) and calls - * the SCHEDOP_shutdown::SHUTDOWN_suspend hypercall (for PV as - * argument it passes the MFN to 'struct start_info'). - * - * And then the guest is suspended. - * - * The checkpointing or notifying a guest that the suspend failed or - * cancelled (in case of checkpoint) is by having the - * SCHEDOP_shutdown::SHUTDOWN_suspend hypercall return a non-zero - * value. - * - * The PV and PVHVM resume path are similar. For PV it would be - * similar to bootup - figure out where the 'struct start_info' is (or - * if the suspend was cancelled aka checkpointed - reuse the saved - * values). - * - * From here on they differ depending whether the guest is PV or PVHVM - * in specifics but follow overall the same path: - * - PV: Bringing up the vCPUS, - * - PVHVM: Setup vector callback, - * - Bring up vCPU runstates, - * - Remap the grant tables if checkpointing or setup from scratch, - * - * - * If the resume was not checkpointing (or if suspend was succesful) we would - * setup the PV timers and the different PV events. Lastly the PV drivers - * re-negotiate with the backend. - * - * This function would return before the guest started resuming. That is - * the guest would be in non-running state and its vCPU context would be - * in the the SCHEDOP_shutdown::SHUTDOWN_suspend hypercall return path - * (for PV and PVHVM). For HVM it would be in would be in QEMU emulated - * BIOS handling S3 suspend. - * - * @parm xch a handle to an open hypervisor interface - * @parm domid the domain id to resume - * @parm fast use cooperative resume (guest must support this) - * return 0 on success, -1 on failure - */ -int xc_domain_resume(xc_interface *xch, - uint32_t domid, - int fast); - /** * This function will shutdown a domain. This is intended for use in * fully-virtualized domains where this operation is analogous to the diff --git a/tools/include/xenguest.h b/tools/include/xenguest.h index f9fb0449ad..61d0a82f48 100644 --- a/tools/include/xenguest.h +++ b/tools/include/xenguest.h @@ -689,6 +689,68 @@ int xc_query_page_offline_status(xc_interface *xch, unsigned long start, int xc_exchange_page(xc_interface *xch, uint32_t domid, xen_pfn_t mfn); +/** + * This function resumes a suspended domain. The domain should have + * been previously suspended. + * + * Note that there are 'xc_domain_suspend' as suspending a domain + * is quite the endeavour. + * + * For the purpose of this explanation there are three guests: + * PV (using hypercalls for privilgied operations), HVM + * (fully hardware virtualized guests using emulated devices for everything), + * and PVHVM (PV aware with hardware virtualisation). + * + * HVM guest are the simplest - they suspend via S3 / S4 and resume from + * S3 / S4. Upon resume they have to re-negotiate with the emulated devices. + * + * PV and PVHVM communicate via hypercalls for suspend (and resume). + * For suspend the toolstack initiates the process by writing an value + * in XenBus "control/shutdown" with the string "suspend". + * + * The PV guest stashes anything it deems neccessary in 'struct + * start_info' in case of failure (PVHVM may ignore this) and calls + * the SCHEDOP_shutdown::SHUTDOWN_suspend hypercall (for PV as + * argument it passes the MFN to 'struct start_info'). + * + * And then the guest is suspended. + * + * The checkpointing or notifying a guest that the suspend failed or + * cancelled (in case of checkpoint) is by having the + * SCHEDOP_shutdown::SHUTDOWN_suspend hypercall return a non-zero + * value. + * + * The PV and PVHVM resume path are similar. For PV it would be + * similar to bootup - figure out where the 'struct start_info' is (or + * if the suspend was cancelled aka checkpointed - reuse the saved + * values). + * + * From here on they differ depending whether the guest is PV or PVHVM + * in specifics but follow overall the same path: + * - PV: Bringing up the vCPUS, + * - PVHVM: Setup vector callback, + * - Bring up vCPU runstates, + * - Remap the grant tables if checkpointing or setup from scratch, + * + * + * If the resume was not checkpointing (or if suspend was succesful) we would + * setup the PV timers and the different PV events. Lastly the PV drivers + * re-negotiate with the backend. + * + * This function would return before the guest started resuming. That is + * the guest would be in non-running state and its vCPU context would be + * in the the SCHEDOP_shutdown::SHUTDOWN_suspend hypercall return path + * (for PV and PVHVM). For HVM it would be in would be in QEMU emulated + * BIOS handling S3 suspend. + * + * @parm xch a handle to an open hypervisor interface + * @parm domid the domain id to resume + * @parm fast use cooperative resume (guest must support this) + * return 0 on success, -1 on failure + */ +int xc_domain_resume(xc_interface *xch, + uint32_t domid, + int fast); /** * Memory related information, such as PFN types, the P2M table, diff --git a/tools/libs/ctrl/Makefile b/tools/libs/ctrl/Makefile index ce9ecae710..fbeb3a3537 100644 --- a/tools/libs/ctrl/Makefile +++ b/tools/libs/ctrl/Makefile @@ -20,7 +20,6 @@ SRCS-y += xc_rt.c SRCS-y += xc_tbuf.c SRCS-y += xc_pm.c SRCS-y += xc_cpu_hotplug.c -SRCS-y += xc_resume.c SRCS-y += xc_vm_event.c SRCS-y += xc_vmtrace.c SRCS-y += xc_monitor.c diff --git a/tools/libs/ctrl/xc_resume.c b/tools/libs/ctrl/xc_resume.c deleted file mode 100644 index e3c8e83aa9..0000000000 --- a/tools/libs/ctrl/xc_resume.c +++ /dev/null @@ -1,264 +0,0 @@ -/* - * This library is free software; you can redistribute it and/or - * modify it under the terms of the GNU Lesser General Public - * License as published by the Free Software Foundation; - * version 2.1 of the License. - * - * This library is distributed in the hope that it will be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU - * Lesser General Public License for more details. - * - * You should have received a copy of the GNU Lesser General Public - * License along with this library; If not, see . - */ - -#include "xc_private.h" - -#if defined(__i386__) || defined(__x86_64__) - -#include -#include -#include -#include "xc_core.h" - -static int modify_returncode(xc_interface *xch, uint32_t domid) -{ - vcpu_guest_context_any_t ctxt; - xc_dominfo_t info; - xen_capabilities_info_t caps; - struct domain_info_context _dinfo = {}; - struct domain_info_context *dinfo = &_dinfo; - int rc; - - if ( xc_domain_getinfo(xch, domid, 1, &info) != 1 || - info.domid != domid ) - { - PERROR("Could not get domain info"); - return -1; - } - - if ( !info.shutdown || (info.shutdown_reason != SHUTDOWN_suspend) ) - { - ERROR("Dom %d not suspended: (shutdown %d, reason %d)", domid, - info.shutdown, info.shutdown_reason); - errno = EINVAL; - return -1; - } - - if ( info.hvm ) - { - /* HVM guests without PV drivers have no return code to modify. */ - uint64_t irq = 0; - xc_hvm_param_get(xch, domid, HVM_PARAM_CALLBACK_IRQ, &irq); - if ( !irq ) - return 0; - - /* HVM guests have host address width. */ - if ( xc_version(xch, XENVER_capabilities, &caps) != 0 ) - { - PERROR("Could not get Xen capabilities"); - return -1; - } - dinfo->guest_width = strstr(caps, "x86_64") ? 8 : 4; - } - else - { - /* Probe PV guest address width. */ - if ( xc_domain_get_guest_width(xch, domid, &dinfo->guest_width) ) - return -1; - } - - if ( (rc = xc_vcpu_getcontext(xch, domid, 0, &ctxt)) != 0 ) - return rc; - - SET_FIELD(&ctxt, user_regs.eax, 1, dinfo->guest_width); - - if ( (rc = xc_vcpu_setcontext(xch, domid, 0, &ctxt)) != 0 ) - return rc; - - return 0; -} - -#else - -static int modify_returncode(xc_interface *xch, uint32_t domid) -{ - return 0; - -} - -#endif - -static int xc_domain_resume_cooperative(xc_interface *xch, uint32_t domid) -{ - DECLARE_DOMCTL; - int rc; - - /* - * Set hypercall return code to indicate that suspend is cancelled - * (rather than resuming in a new domain context). - */ - if ( (rc = modify_returncode(xch, domid)) != 0 ) - return rc; - - domctl.cmd = XEN_DOMCTL_resumedomain; - domctl.domain = domid; - return do_domctl(xch, &domctl); -} - -#if defined(__i386__) || defined(__x86_64__) -static int xc_domain_resume_hvm(xc_interface *xch, uint32_t domid) -{ - DECLARE_DOMCTL; - - /* - * The domctl XEN_DOMCTL_resumedomain unpause each vcpu. After - * the domctl, the guest will run. - * - * If it is PVHVM, the guest called the hypercall - * SCHEDOP_shutdown:SHUTDOWN_suspend - * to suspend itself. We don't modify the return code, so the PV driver - * will disconnect and reconnect. - * - * If it is a HVM, the guest will continue running. - */ - domctl.cmd = XEN_DOMCTL_resumedomain; - domctl.domain = domid; - return do_domctl(xch, &domctl); -} -#endif - -static int xc_domain_resume_any(xc_interface *xch, uint32_t domid) -{ - DECLARE_DOMCTL; - xc_dominfo_t info; - int i, rc = -1; -#if defined(__i386__) || defined(__x86_64__) - struct domain_info_context _dinfo = { .guest_width = 0, - .p2m_size = 0 }; - struct domain_info_context *dinfo = &_dinfo; - xen_pfn_t mfn, store_mfn, console_mfn; - vcpu_guest_context_any_t ctxt; - start_info_any_t *start_info; - shared_info_any_t *shinfo = NULL; - xen_pfn_t *p2m = NULL; -#endif - - if ( xc_domain_getinfo(xch, domid, 1, &info) != 1 ) - { - PERROR("Could not get domain info"); - return rc; - } - - /* - * (x86 only) Rewrite store_mfn and console_mfn back to MFN (from PFN). - */ -#if defined(__i386__) || defined(__x86_64__) - if ( info.hvm ) - return xc_domain_resume_hvm(xch, domid); - - if ( xc_domain_get_guest_width(xch, domid, &dinfo->guest_width) != 0 ) - { - PERROR("Could not get domain width"); - return rc; - } - - /* Map the shared info frame */ - shinfo = xc_map_foreign_range(xch, domid, PAGE_SIZE, - PROT_READ, info.shared_info_frame); - if ( shinfo == NULL ) - { - ERROR("Couldn't map shared info"); - goto out; - } - - /* Map the p2m list */ - if ( xc_core_arch_map_p2m(xch, dinfo, &info, shinfo, &p2m) ) - { - ERROR("Couldn't map p2m table"); - goto out; - } - - if ( xc_vcpu_getcontext(xch, domid, 0, &ctxt) ) - { - ERROR("Could not get vcpu context"); - goto out; - } - - mfn = GET_FIELD(&ctxt, user_regs.edx, dinfo->guest_width); - - start_info = xc_map_foreign_range(xch, domid, PAGE_SIZE, - PROT_READ | PROT_WRITE, mfn); - if ( start_info == NULL ) - { - ERROR("Couldn't map start_info"); - goto out; - } - - store_mfn = GET_FIELD(start_info, store_mfn, dinfo->guest_width); - console_mfn = GET_FIELD(start_info, console.domU.mfn, dinfo->guest_width); - if ( dinfo->guest_width == 4 ) - { - store_mfn = ((uint32_t *)p2m)[store_mfn]; - console_mfn = ((uint32_t *)p2m)[console_mfn]; - } - else - { - store_mfn = ((uint64_t *)p2m)[store_mfn]; - console_mfn = ((uint64_t *)p2m)[console_mfn]; - } - SET_FIELD(start_info, store_mfn, store_mfn, dinfo->guest_width); - SET_FIELD(start_info, console.domU.mfn, console_mfn, dinfo->guest_width); - - munmap(start_info, PAGE_SIZE); -#endif /* defined(__i386__) || defined(__x86_64__) */ - - /* Reset all secondary CPU states. */ - for ( i = 1; i <= info.max_vcpu_id; i++ ) - if ( xc_vcpu_setcontext(xch, domid, i, NULL) != 0 ) - { - ERROR("Couldn't reset vcpu state"); - goto out; - } - - /* Ready to resume domain execution now. */ - domctl.cmd = XEN_DOMCTL_resumedomain; - domctl.domain = domid; - rc = do_domctl(xch, &domctl); - -out: -#if defined(__i386__) || defined(__x86_64__) - if (p2m) - munmap(p2m, dinfo->p2m_frames * PAGE_SIZE); - if (shinfo) - munmap(shinfo, PAGE_SIZE); -#endif - - return rc; -} - -/* - * Resume execution of a domain after suspend shutdown. - * This can happen in one of two ways: - * 1. (fast=1) Resume the guest without resetting the domain environment. - * The guests's call to SCHEDOP_shutdown(SHUTDOWN_suspend) will return 1. - * - * 2. (fast=0) Reset guest environment so it believes it is resumed in a new - * domain context. The guests's call to SCHEDOP_shutdown(SHUTDOWN_suspend) - * will return 0. - * - * (1) should only by used for guests which can handle the special return - * code. Also note that the insertion of the return code is quite interesting - * and that the guest MUST be paused - otherwise we would be corrupting - * the guest vCPU state. - * - * (2) should be used only for guests which cannot handle the special - * new return code - and it is always safe (but slower). - */ -int xc_domain_resume(xc_interface *xch, uint32_t domid, int fast) -{ - return (fast - ? xc_domain_resume_cooperative(xch, domid) - : xc_domain_resume_any(xch, domid)); -} diff --git a/tools/libs/guest/Makefile b/tools/libs/guest/Makefile index 6d2a1d5bbc..2a2323ff09 100644 --- a/tools/libs/guest/Makefile +++ b/tools/libs/guest/Makefile @@ -9,6 +9,7 @@ endif SRCS-y += xg_private.c SRCS-y += xg_domain.c SRCS-y += xg_suspend.c +SRCS-y += xg_resume.c ifeq ($(CONFIG_MIGRATE),y) SRCS-y += xg_sr_common.c SRCS-$(CONFIG_X86) += xg_sr_common_x86.c diff --git a/tools/libs/guest/xg_resume.c b/tools/libs/guest/xg_resume.c new file mode 100644 index 0000000000..3bdefb2eef --- /dev/null +++ b/tools/libs/guest/xg_resume.c @@ -0,0 +1,265 @@ +/* + * This library is free software; you can redistribute it and/or + * modify it under the terms of the GNU Lesser General Public + * License as published by the Free Software Foundation; + * version 2.1 of the License. + * + * This library is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public + * License along with this library; If not, see . + */ + +#include "xc_private.h" +#include "xenguest.h" + +#if defined(__i386__) || defined(__x86_64__) + +#include +#include +#include +#include "xc_core.h" + +static int modify_returncode(xc_interface *xch, uint32_t domid) +{ + vcpu_guest_context_any_t ctxt; + xc_dominfo_t info; + xen_capabilities_info_t caps; + struct domain_info_context _dinfo = {}; + struct domain_info_context *dinfo = &_dinfo; + int rc; + + if ( xc_domain_getinfo(xch, domid, 1, &info) != 1 || + info.domid != domid ) + { + PERROR("Could not get domain info"); + return -1; + } + + if ( !info.shutdown || (info.shutdown_reason != SHUTDOWN_suspend) ) + { + ERROR("Dom %d not suspended: (shutdown %d, reason %d)", domid, + info.shutdown, info.shutdown_reason); + errno = EINVAL; + return -1; + } + + if ( info.hvm ) + { + /* HVM guests without PV drivers have no return code to modify. */ + uint64_t irq = 0; + xc_hvm_param_get(xch, domid, HVM_PARAM_CALLBACK_IRQ, &irq); + if ( !irq ) + return 0; + + /* HVM guests have host address width. */ + if ( xc_version(xch, XENVER_capabilities, &caps) != 0 ) + { + PERROR("Could not get Xen capabilities"); + return -1; + } + dinfo->guest_width = strstr(caps, "x86_64") ? 8 : 4; + } + else + { + /* Probe PV guest address width. */ + if ( xc_domain_get_guest_width(xch, domid, &dinfo->guest_width) ) + return -1; + } + + if ( (rc = xc_vcpu_getcontext(xch, domid, 0, &ctxt)) != 0 ) + return rc; + + SET_FIELD(&ctxt, user_regs.eax, 1, dinfo->guest_width); + + if ( (rc = xc_vcpu_setcontext(xch, domid, 0, &ctxt)) != 0 ) + return rc; + + return 0; +} + +#else + +static int modify_returncode(xc_interface *xch, uint32_t domid) +{ + return 0; + +} + +#endif + +static int xc_domain_resume_cooperative(xc_interface *xch, uint32_t domid) +{ + DECLARE_DOMCTL; + int rc; + + /* + * Set hypercall return code to indicate that suspend is cancelled + * (rather than resuming in a new domain context). + */ + if ( (rc = modify_returncode(xch, domid)) != 0 ) + return rc; + + domctl.cmd = XEN_DOMCTL_resumedomain; + domctl.domain = domid; + return do_domctl(xch, &domctl); +} + +#if defined(__i386__) || defined(__x86_64__) +static int xc_domain_resume_hvm(xc_interface *xch, uint32_t domid) +{ + DECLARE_DOMCTL; + + /* + * The domctl XEN_DOMCTL_resumedomain unpause each vcpu. After + * the domctl, the guest will run. + * + * If it is PVHVM, the guest called the hypercall + * SCHEDOP_shutdown:SHUTDOWN_suspend + * to suspend itself. We don't modify the return code, so the PV driver + * will disconnect and reconnect. + * + * If it is a HVM, the guest will continue running. + */ + domctl.cmd = XEN_DOMCTL_resumedomain; + domctl.domain = domid; + return do_domctl(xch, &domctl); +} +#endif + +static int xc_domain_resume_any(xc_interface *xch, uint32_t domid) +{ + DECLARE_DOMCTL; + xc_dominfo_t info; + int i, rc = -1; +#if defined(__i386__) || defined(__x86_64__) + struct domain_info_context _dinfo = { .guest_width = 0, + .p2m_size = 0 }; + struct domain_info_context *dinfo = &_dinfo; + xen_pfn_t mfn, store_mfn, console_mfn; + vcpu_guest_context_any_t ctxt; + start_info_any_t *start_info; + shared_info_any_t *shinfo = NULL; + xen_pfn_t *p2m = NULL; +#endif + + if ( xc_domain_getinfo(xch, domid, 1, &info) != 1 ) + { + PERROR("Could not get domain info"); + return rc; + } + + /* + * (x86 only) Rewrite store_mfn and console_mfn back to MFN (from PFN). + */ +#if defined(__i386__) || defined(__x86_64__) + if ( info.hvm ) + return xc_domain_resume_hvm(xch, domid); + + if ( xc_domain_get_guest_width(xch, domid, &dinfo->guest_width) != 0 ) + { + PERROR("Could not get domain width"); + return rc; + } + + /* Map the shared info frame */ + shinfo = xc_map_foreign_range(xch, domid, PAGE_SIZE, + PROT_READ, info.shared_info_frame); + if ( shinfo == NULL ) + { + ERROR("Couldn't map shared info"); + goto out; + } + + /* Map the p2m list */ + if ( xc_core_arch_map_p2m(xch, dinfo, &info, shinfo, &p2m) ) + { + ERROR("Couldn't map p2m table"); + goto out; + } + + if ( xc_vcpu_getcontext(xch, domid, 0, &ctxt) ) + { + ERROR("Could not get vcpu context"); + goto out; + } + + mfn = GET_FIELD(&ctxt, user_regs.edx, dinfo->guest_width); + + start_info = xc_map_foreign_range(xch, domid, PAGE_SIZE, + PROT_READ | PROT_WRITE, mfn); + if ( start_info == NULL ) + { + ERROR("Couldn't map start_info"); + goto out; + } + + store_mfn = GET_FIELD(start_info, store_mfn, dinfo->guest_width); + console_mfn = GET_FIELD(start_info, console.domU.mfn, dinfo->guest_width); + if ( dinfo->guest_width == 4 ) + { + store_mfn = ((uint32_t *)p2m)[store_mfn]; + console_mfn = ((uint32_t *)p2m)[console_mfn]; + } + else + { + store_mfn = ((uint64_t *)p2m)[store_mfn]; + console_mfn = ((uint64_t *)p2m)[console_mfn]; + } + SET_FIELD(start_info, store_mfn, store_mfn, dinfo->guest_width); + SET_FIELD(start_info, console.domU.mfn, console_mfn, dinfo->guest_width); + + munmap(start_info, PAGE_SIZE); +#endif /* defined(__i386__) || defined(__x86_64__) */ + + /* Reset all secondary CPU states. */ + for ( i = 1; i <= info.max_vcpu_id; i++ ) + if ( xc_vcpu_setcontext(xch, domid, i, NULL) != 0 ) + { + ERROR("Couldn't reset vcpu state"); + goto out; + } + + /* Ready to resume domain execution now. */ + domctl.cmd = XEN_DOMCTL_resumedomain; + domctl.domain = domid; + rc = do_domctl(xch, &domctl); + +out: +#if defined(__i386__) || defined(__x86_64__) + if (p2m) + munmap(p2m, dinfo->p2m_frames * PAGE_SIZE); + if (shinfo) + munmap(shinfo, PAGE_SIZE); +#endif + + return rc; +} + +/* + * Resume execution of a domain after suspend shutdown. + * This can happen in one of two ways: + * 1. (fast=1) Resume the guest without resetting the domain environment. + * The guests's call to SCHEDOP_shutdown(SHUTDOWN_suspend) will return 1. + * + * 2. (fast=0) Reset guest environment so it believes it is resumed in a new + * domain context. The guests's call to SCHEDOP_shutdown(SHUTDOWN_suspend) + * will return 0. + * + * (1) should only by used for guests which can handle the special return + * code. Also note that the insertion of the return code is quite interesting + * and that the guest MUST be paused - otherwise we would be corrupting + * the guest vCPU state. + * + * (2) should be used only for guests which cannot handle the special + * new return code - and it is always safe (but slower). + */ +int xc_domain_resume(xc_interface *xch, uint32_t domid, int fast) +{ + return (fast + ? xc_domain_resume_cooperative(xch, domid) + : xc_domain_resume_any(xch, domid)); +} -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Fri Jun 04 18:11:46 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Fri, 04 Jun 2021 18:11:46 +0000 Received: from list by lists.xenproject.org with outflank-mailman.136984.253817 (Exim 4.92) (envelope-from ) id 1lpEIM-0002Wl-Hr; Fri, 04 Jun 2021 18:11:46 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 136984.253817; Fri, 04 Jun 2021 18:11:46 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpEIM-0002We-EP; Fri, 04 Jun 2021 18:11:46 +0000 Received: by outflank-mailman (input) for mailman id 136984; Fri, 04 Jun 2021 18:11:45 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpEIL-0002WM-0i for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 18:11:45 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpEIK-0001vY-W0 for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 18:11:44 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lpEIK-00082D-V3 for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 18:11:44 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=mL//WOPR3CslskCSLz22EVJCFnSVWdh22+KrZklJwlk=; b=WJTq/y7Z5qCxxa6nROIZr4ckyC 4Ds4dJ5/2kYV7NK4uM4UWsoga7D9DuDNZVlDW7mObKBzM7pD05mqgIVewYqcL/VJbCIEGS2IMUNlR EhJek1zNTadwjKPKH0ZswLI3R9qVV8N7/725vfbcnuTtTrcGNHQglbwdsRbyrzsefQsw=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] tools/libs: move xc_core* from libxenctrl to libxenguest Message-Id: Date: Fri, 04 Jun 2021 18:11:44 +0000 commit 455790573d3bbad6d5a1bb7e9d28b6dd71075693 Author: Juergen Gross AuthorDate: Fri Jun 4 08:02:13 2021 +0200 Commit: Julien Grall CommitDate: Fri Jun 4 18:56:46 2021 +0100 tools/libs: move xc_core* from libxenctrl to libxenguest The functionality in xc_core* should be part of libxenguest instead of libxenctrl. Users are already either in libxenguest, or in xl. There is one single exception: xc_core_arch_auto_translated_physmap() is being used by xc_domain_memory_mapping(), which is used by qemu. So leave the xc_core_arch_auto_translated_physmap() functionality in libxenctrl. This will make it easier to merge common functionality of xc_core* and xg_sr_save*. Signed-off-by: Juergen Gross Acked-by: Wei Liu --- tools/libs/ctrl/Makefile | 3 - tools/libs/ctrl/xc_core.c | 1027 ------------------------------------ tools/libs/ctrl/xc_core.h | 176 ------ tools/libs/ctrl/xc_core_arm.c | 113 ---- tools/libs/ctrl/xc_core_arm.h | 59 --- tools/libs/ctrl/xc_core_x86.c | 391 -------------- tools/libs/ctrl/xc_core_x86.h | 60 --- tools/libs/ctrl/xc_domain.c | 2 - tools/libs/ctrl/xc_private.h | 12 + tools/libs/guest/Makefile | 3 + tools/libs/guest/xg_core.c | 1027 ++++++++++++++++++++++++++++++++++++ tools/libs/guest/xg_core.h | 175 ++++++ tools/libs/guest/xg_core_arm.c | 107 ++++ tools/libs/guest/xg_core_arm.h | 59 +++ tools/libs/guest/xg_core_x86.c | 385 ++++++++++++++ tools/libs/guest/xg_core_x86.h | 60 +++ tools/libs/guest/xg_dom_boot.c | 2 +- tools/libs/guest/xg_domain.c | 2 +- tools/libs/guest/xg_offline_page.c | 2 +- tools/libs/guest/xg_resume.c | 2 +- 20 files changed, 1832 insertions(+), 1835 deletions(-) diff --git a/tools/libs/ctrl/Makefile b/tools/libs/ctrl/Makefile index fbeb3a3537..519246b0d6 100644 --- a/tools/libs/ctrl/Makefile +++ b/tools/libs/ctrl/Makefile @@ -2,9 +2,6 @@ XEN_ROOT = $(CURDIR)/../../.. include $(XEN_ROOT)/tools/Rules.mk SRCS-y += xc_altp2m.c -SRCS-y += xc_core.c -SRCS-$(CONFIG_X86) += xc_core_x86.c -SRCS-$(CONFIG_ARM) += xc_core_arm.c SRCS-y += xc_cpupool.c SRCS-y += xc_domain.c SRCS-y += xc_evtchn.c diff --git a/tools/libs/ctrl/xc_core.c b/tools/libs/ctrl/xc_core.c deleted file mode 100644 index 9576bec5a3..0000000000 --- a/tools/libs/ctrl/xc_core.c +++ /dev/null @@ -1,1027 +0,0 @@ -/* - * Elf format, (pfn, gmfn) table, IA64 support. - * Copyright (c) 2007 Isaku Yamahata - * VA Linux Systems Japan K.K. - * - * This library is free software; you can redistribute it and/or - * modify it under the terms of the GNU Lesser General Public - * License as published by the Free Software Foundation; - * version 2.1 of the License. - * - * This library is distributed in the hope that it will be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU - * Lesser General Public License for more details. - * - * You should have received a copy of the GNU Lesser General Public - * License along with this library; If not, see . - */ - -/* - * xen dump-core file format follows ELF format specification. - * Analisys tools shouldn't depends on the order of sections. - * They should follow elf header and check section names. - * - * +--------------------------------------------------------+ - * |ELF header | - * +--------------------------------------------------------+ - * |section headers | - * | null section header | - * | .shstrtab | - * | .note.Xen | - * | .xen_prstatus | - * | .xen_shared_info if present | - * | .xen_pages | - * | .xen_p2m or .xen_pfn | - * +--------------------------------------------------------+ - * |.note.Xen:note section | - * | "Xen" is used as note name, | - * | types are defined in xen/include/public/elfnote.h | - * | and descriptors are defined in xc_core.h. | - * | dumpcore none | - * | dumpcore header | - * | dumpcore xen version | - * | dumpcore format version | - * +--------------------------------------------------------+ - * |.xen_prstatus | - * | vcpu_guest_context_t[nr_vcpus] | - * +--------------------------------------------------------+ - * |.xen_shared_info if possible | - * +--------------------------------------------------------+ - * |.xen_pages | - * | page * nr_pages | - * +--------------------------------------------------------+ - * |.xen_p2m or .xen_pfn | - * | .xen_p2m: struct xen_dumpcore_p2m[nr_pages] | - * | .xen_pfn: uint64_t[nr_pages] | - * +--------------------------------------------------------+ - * |.shstrtab: section header string table | - * +--------------------------------------------------------+ - * - */ - -#include "xc_private.h" -#include "xc_core.h" -#include -#include - -#include - -/* number of pages to write at a time */ -#define DUMP_INCREMENT (4 * 1024) - -/* string table */ -struct xc_core_strtab { - char *strings; - uint16_t length; - uint16_t max; -}; - -static struct xc_core_strtab* -xc_core_strtab_init(xc_interface *xch) -{ - struct xc_core_strtab *strtab; - char *strings; - strtab = malloc(sizeof(*strtab)); - if ( strtab == NULL ) - return NULL; - - strings = malloc(PAGE_SIZE); - if ( strings == NULL ) - { - PERROR("Could not allocate string table init"); - free(strtab); - return NULL; - } - strtab->strings = strings; - strtab->max = PAGE_SIZE; - - /* index 0 represents none */ - strtab->strings[0] = '\0'; - strtab->length = 1; - - return strtab; -} - -static void -xc_core_strtab_free(struct xc_core_strtab *strtab) -{ - free(strtab->strings); - free(strtab); -} - -static uint16_t -xc_core_strtab_get(xc_interface *xch, struct xc_core_strtab *strtab, const char *name) -{ - uint16_t ret = 0; - uint16_t len = strlen(name) + 1; - - if ( strtab->length > UINT16_MAX - len ) - { - PERROR("too long string table"); - errno = E2BIG; - return ret; - } - - if ( strtab->length + len > strtab->max ) - { - char *tmp; - if ( strtab->max > UINT16_MAX / 2 ) - { - PERROR("too long string table"); - errno = ENOMEM; - return ret; - } - - tmp = realloc(strtab->strings, strtab->max * 2); - if ( tmp == NULL ) - { - PERROR("Could not allocate string table"); - return ret; - } - - strtab->strings = tmp; - strtab->max *= 2; - } - - ret = strtab->length; - strcpy(strtab->strings + strtab->length, name); - strtab->length += len; - return ret; -} - - -/* section headers */ -struct xc_core_section_headers { - uint16_t num; - uint16_t num_max; - - Elf64_Shdr *shdrs; -}; -#define SHDR_INIT ((uint16_t)16) -#define SHDR_INC ((uint16_t)4) - -static struct xc_core_section_headers* -xc_core_shdr_init(xc_interface *xch) -{ - struct xc_core_section_headers *sheaders; - sheaders = malloc(sizeof(*sheaders)); - if ( sheaders == NULL ) - return NULL; - - sheaders->num = 0; - sheaders->num_max = SHDR_INIT; - sheaders->shdrs = malloc(sizeof(sheaders->shdrs[0]) * sheaders->num_max); - if ( sheaders->shdrs == NULL ) - { - free(sheaders); - return NULL; - } - return sheaders; -} - -static void -xc_core_shdr_free(struct xc_core_section_headers *sheaders) -{ - free(sheaders->shdrs); - free(sheaders); -} - -Elf64_Shdr* -xc_core_shdr_get(xc_interface *xch, - struct xc_core_section_headers *sheaders) -{ - Elf64_Shdr *shdr; - - if ( sheaders->num == sheaders->num_max ) - { - Elf64_Shdr *shdrs; - if ( sheaders->num_max > UINT16_MAX - SHDR_INC ) - { - errno = E2BIG; - return NULL; - } - sheaders->num_max += SHDR_INC; - shdrs = realloc(sheaders->shdrs, - sizeof(sheaders->shdrs[0]) * sheaders->num_max); - if ( shdrs == NULL ) - return NULL; - sheaders->shdrs = shdrs; - } - - shdr = &sheaders->shdrs[sheaders->num]; - sheaders->num++; - memset(shdr, 0, sizeof(*shdr)); - return shdr; -} - -int -xc_core_shdr_set(xc_interface *xch, - Elf64_Shdr *shdr, - struct xc_core_strtab *strtab, - const char *name, uint32_t type, - uint64_t offset, uint64_t size, - uint64_t addralign, uint64_t entsize) -{ - uint64_t name_idx = xc_core_strtab_get(xch, strtab, name); - if ( name_idx == 0 ) - return -1; - - shdr->sh_name = name_idx; - shdr->sh_type = type; - shdr->sh_offset = offset; - shdr->sh_size = size; - shdr->sh_addralign = addralign; - shdr->sh_entsize = entsize; - return 0; -} - -static void -xc_core_ehdr_init(Elf64_Ehdr *ehdr) -{ - memset(ehdr, 0, sizeof(*ehdr)); - ehdr->e_ident[EI_MAG0] = ELFMAG0; - ehdr->e_ident[EI_MAG1] = ELFMAG1; - ehdr->e_ident[EI_MAG2] = ELFMAG2; - ehdr->e_ident[EI_MAG3] = ELFMAG3; - ehdr->e_ident[EI_CLASS] = ELFCLASS64; - ehdr->e_ident[EI_DATA] = ELF_ARCH_DATA; - ehdr->e_ident[EI_VERSION] = EV_CURRENT; - ehdr->e_ident[EI_OSABI] = ELFOSABI_SYSV; - ehdr->e_ident[EI_ABIVERSION] = EV_CURRENT; - - ehdr->e_type = ET_CORE; - /* e_machine will be filled in later */ - ehdr->e_version = EV_CURRENT; - ehdr->e_entry = 0; - ehdr->e_phoff = 0; - ehdr->e_shoff = sizeof(*ehdr); - ehdr->e_flags = ELF_CORE_EFLAGS; - ehdr->e_ehsize = sizeof(*ehdr); - ehdr->e_phentsize = sizeof(Elf64_Phdr); - ehdr->e_phnum = 0; - ehdr->e_shentsize = sizeof(Elf64_Shdr); - /* ehdr->e_shnum and ehdr->e_shstrndx aren't known here yet. - * fill it later */ -} - -static int -elfnote_fill_xen_version(xc_interface *xch, - struct xen_dumpcore_elfnote_xen_version_desc - *xen_version) -{ - int rc; - memset(xen_version, 0, sizeof(*xen_version)); - - rc = xc_version(xch, XENVER_version, NULL); - if ( rc < 0 ) - return rc; - xen_version->major_version = rc >> 16; - xen_version->minor_version = rc & ((1 << 16) - 1); - - rc = xc_version(xch, XENVER_extraversion, - &xen_version->extra_version); - if ( rc < 0 ) - return rc; - - rc = xc_version(xch, XENVER_compile_info, - &xen_version->compile_info); - if ( rc < 0 ) - return rc; - - rc = xc_version(xch, - XENVER_capabilities, &xen_version->capabilities); - if ( rc < 0 ) - return rc; - - rc = xc_version(xch, XENVER_changeset, &xen_version->changeset); - if ( rc < 0 ) - return rc; - - rc = xc_version(xch, XENVER_platform_parameters, - &xen_version->platform_parameters); - if ( rc < 0 ) - return rc; - - rc = xc_version(xch, XENVER_pagesize, NULL); - if ( rc < 0 ) - return rc; - xen_version->pagesize = rc; - - return 0; -} - -static void -elfnote_fill_format_version(struct xen_dumpcore_elfnote_format_version_desc - *format_version) -{ - format_version->version = XEN_DUMPCORE_FORMAT_VERSION_CURRENT; -} - -static void -elfnote_init(struct elfnote *elfnote) -{ - /* elf note section */ - memset(elfnote, 0, sizeof(*elfnote)); - elfnote->namesz = strlen(XEN_DUMPCORE_ELFNOTE_NAME) + 1; - strncpy(elfnote->name, XEN_DUMPCORE_ELFNOTE_NAME, sizeof(elfnote->name)); -} - -static int -elfnote_dump_none(xc_interface *xch, void *args, dumpcore_rtn_t dump_rtn) -{ - int sts; - struct elfnote elfnote; - struct xen_dumpcore_elfnote_none_desc none; - - elfnote_init(&elfnote); - /* Avoid compile warning about constant-zero-sized memset(). */ - /*memset(&none, 0, sizeof(none));*/ - - elfnote.descsz = sizeof(none); - elfnote.type = XEN_ELFNOTE_DUMPCORE_NONE; - sts = dump_rtn(xch, args, (char*)&elfnote, sizeof(elfnote)); - if ( sts != 0 ) - return sts; - return dump_rtn(xch, args, (char*)&none, sizeof(none)); -} - -static int -elfnote_dump_core_header( - xc_interface *xch, - void *args, dumpcore_rtn_t dump_rtn, const xc_dominfo_t *info, - int nr_vcpus, unsigned long nr_pages) -{ - int sts; - struct elfnote elfnote; - struct xen_dumpcore_elfnote_header_desc header; - - elfnote_init(&elfnote); - memset(&header, 0, sizeof(header)); - - elfnote.descsz = sizeof(header); - elfnote.type = XEN_ELFNOTE_DUMPCORE_HEADER; - header.xch_magic = info->hvm ? XC_CORE_MAGIC_HVM : XC_CORE_MAGIC; - header.xch_nr_vcpus = nr_vcpus; - header.xch_nr_pages = nr_pages; - header.xch_page_size = PAGE_SIZE; - sts = dump_rtn(xch, args, (char*)&elfnote, sizeof(elfnote)); - if ( sts != 0 ) - return sts; - return dump_rtn(xch, args, (char*)&header, sizeof(header)); -} - -static int -elfnote_dump_xen_version(xc_interface *xch, void *args, - dumpcore_rtn_t dump_rtn, unsigned int guest_width) -{ - int sts; - struct elfnote elfnote; - struct xen_dumpcore_elfnote_xen_version_desc xen_version; - - elfnote_init(&elfnote); - memset(&xen_version, 0, sizeof(xen_version)); - - elfnote.descsz = sizeof(xen_version); - elfnote.type = XEN_ELFNOTE_DUMPCORE_XEN_VERSION; - elfnote_fill_xen_version(xch, &xen_version); - if (guest_width < sizeof(unsigned long)) - { - // 32 bit elf file format differs in pagesize's alignment - char *p = (char *)&xen_version.pagesize; - memmove(p - 4, p, sizeof(xen_version.pagesize)); - } - sts = dump_rtn(xch, args, (char*)&elfnote, sizeof(elfnote)); - if ( sts != 0 ) - return sts; - return dump_rtn(xch, args, (char*)&xen_version, sizeof(xen_version)); -} - -static int -elfnote_dump_format_version(xc_interface *xch, - void *args, dumpcore_rtn_t dump_rtn) -{ - int sts; - struct elfnote elfnote; - struct xen_dumpcore_elfnote_format_version_desc format_version; - - elfnote_init(&elfnote); - memset(&format_version, 0, sizeof(format_version)); - - elfnote.descsz = sizeof(format_version); - elfnote.type = XEN_ELFNOTE_DUMPCORE_FORMAT_VERSION; - elfnote_fill_format_version(&format_version); - sts = dump_rtn(xch, args, (char*)&elfnote, sizeof(elfnote)); - if ( sts != 0 ) - return sts; - return dump_rtn(xch, args, (char*)&format_version, sizeof(format_version)); -} - -int -xc_domain_dumpcore_via_callback(xc_interface *xch, - uint32_t domid, - void *args, - dumpcore_rtn_t dump_rtn) -{ - xc_dominfo_t info; - shared_info_any_t *live_shinfo = NULL; - struct domain_info_context _dinfo = {}; - struct domain_info_context *dinfo = &_dinfo; - - int nr_vcpus = 0; - char *dump_mem, *dump_mem_start = NULL; - vcpu_guest_context_any_t *ctxt = NULL; - struct xc_core_arch_context arch_ctxt; - char dummy[PAGE_SIZE]; - int dummy_len; - int sts = -1; - - unsigned long i; - unsigned long j; - unsigned long nr_pages; - unsigned long max_mfn; - - xc_core_memory_map_t *memory_map = NULL; - unsigned int nr_memory_map; - unsigned int map_idx; - - int auto_translated_physmap; - xen_pfn_t *p2m = NULL; - struct xen_dumpcore_p2m *p2m_array = NULL; - - uint64_t *pfn_array = NULL; - - Elf64_Ehdr ehdr; - uint64_t filesz; - uint64_t offset; - uint64_t fixup; - - struct xc_core_strtab *strtab = NULL; - uint16_t strtab_idx; - struct xc_core_section_headers *sheaders = NULL; - Elf64_Shdr *shdr; - - xc_core_arch_context_init(&arch_ctxt); - if ( (dump_mem_start = malloc(DUMP_INCREMENT*PAGE_SIZE)) == NULL ) - { - PERROR("Could not allocate dump_mem"); - goto out; - } - - if ( xc_domain_getinfo(xch, domid, 1, &info) != 1 ) - { - PERROR("Could not get info for domain"); - goto out; - } - /* Map the shared info frame */ - live_shinfo = xc_map_foreign_range(xch, domid, PAGE_SIZE, - PROT_READ, info.shared_info_frame); - if ( !live_shinfo && !info.hvm ) - { - PERROR("Couldn't map live_shinfo"); - goto out; - } - auto_translated_physmap = xc_core_arch_auto_translated_physmap(&info); - - if ( !auto_translated_physmap ) - - { - if ( xc_domain_get_guest_width(xch, domid, &dinfo->guest_width) != 0 ) - { - PERROR("Could not get address size for domain"); - goto out; - } - } - else - { - /* - * Autotranslated guest never sets guest width in the first - * place. Force guest_width to be sizeof(unsigned long) so - * code below functions properly. - * - * Here is why this is correct. - * - * 1. Before f969bc9fc, xc_domain_get_guest_width for HVM (x86 - * and ARM) always returned hypervisor's idea of - * sizeof(unsigned long). - * - * 2. There has never been a situation in which hypervisor's - * word width is smaller than toolstack domain's (i.e. no - * 32bit hypervisor + 64bit toolstack). - * - * Predicates in code test guest_width against toolstack - * domain's sizeof(unsigned long), so setting guest_width to - * toolstack domain's idea of sizeof(unsigned long) matches - * the original behaviour for HVM guests. - */ - dinfo->guest_width = sizeof(unsigned long); - } - - if ( domid != info.domid ) - { - PERROR("Domain %d does not exist", domid); - goto out; - } - - ctxt = calloc(sizeof(*ctxt), info.max_vcpu_id + 1); - if ( !ctxt ) - { - PERROR("Could not allocate vcpu context array"); - goto out; - } - - for ( i = 0; i <= info.max_vcpu_id; i++ ) - { - if ( xc_vcpu_getcontext(xch, domid, i, &ctxt[nr_vcpus]) == 0 ) - { - if ( xc_core_arch_context_get(&arch_ctxt, &ctxt[nr_vcpus], - xch, domid) ) - continue; - nr_vcpus++; - } - } - if ( nr_vcpus == 0 ) - { - PERROR("No VCPU context could be grabbed"); - goto out; - } - - /* obtain memory map */ - sts = xc_core_arch_memory_map_get(xch, &arch_ctxt, &info, - live_shinfo, &memory_map, - &nr_memory_map); - if ( sts != 0 ) - goto out; - - /* - * Note: this is the *current* number of pages and may change under - * a live dump-core. We'll just take this value, and if more pages - * exist, we'll skip them. If there's less, then we'll just not use - * all the array... - * - * We don't want to use the total potential size of the memory map - * since that is usually much higher than info.nr_pages. - */ - nr_pages = info.nr_pages; - - if ( !auto_translated_physmap ) - { - /* obtain p2m table */ - p2m_array = malloc(nr_pages * sizeof(p2m_array[0])); - if ( p2m_array == NULL ) - { - PERROR("Could not allocate p2m array"); - goto out; - } - - sts = xc_core_arch_map_p2m(xch, dinfo, &info, live_shinfo, &p2m); - if ( sts != 0 ) - goto out; - - sts = xc_maximum_ram_page(xch, &max_mfn); - if ( sts != 0 ) - goto out; - } - else - { - pfn_array = malloc(nr_pages * sizeof(pfn_array[0])); - if ( pfn_array == NULL ) - { - PERROR("Could not allocate pfn array"); - goto out; - } - } - - /* ehdr.e_shnum and ehdr.e_shstrndx aren't known here yet. fill it later*/ - xc_core_ehdr_init(&ehdr); - - /* create section header */ - strtab = xc_core_strtab_init(xch); - if ( strtab == NULL ) - { - PERROR("Could not allocate string table"); - goto out; - } - sheaders = xc_core_shdr_init(xch); - if ( sheaders == NULL ) - { - PERROR("Could not allocate section headers"); - goto out; - } - /* null section */ - shdr = xc_core_shdr_get(xch,sheaders); - if ( shdr == NULL ) - { - PERROR("Could not get section header for null section"); - goto out; - } - - /* .shstrtab */ - shdr = xc_core_shdr_get(xch,sheaders); - if ( shdr == NULL ) - { - PERROR("Could not get section header for shstrtab"); - goto out; - } - strtab_idx = shdr - sheaders->shdrs; - /* strtab_shdr.sh_offset, strtab_shdr.sh_size aren't unknown. - * fill it later - */ - sts = xc_core_shdr_set(xch, shdr, strtab, ELF_SHSTRTAB, SHT_STRTAB, 0, 0, 0, 0); - if ( sts != 0 ) - goto out; - - /* elf note section */ - /* here the number of section header is unknown. fix up offset later. */ - offset = sizeof(ehdr); - filesz = - sizeof(struct xen_dumpcore_elfnote_none) + /* none */ - sizeof(struct xen_dumpcore_elfnote_header) + /* core header */ - sizeof(struct xen_dumpcore_elfnote_xen_version) + /* xen version */ - sizeof(struct xen_dumpcore_elfnote_format_version);/* format version */ - shdr = xc_core_shdr_get(xch,sheaders); - if ( shdr == NULL ) - { - PERROR("Could not get section header for note section"); - goto out; - } - sts = xc_core_shdr_set(xch, shdr, strtab, XEN_DUMPCORE_SEC_NOTE, SHT_NOTE, - offset, filesz, 0, 0); - if ( sts != 0 ) - goto out; - offset += filesz; - - /* prstatus */ - shdr = xc_core_shdr_get(xch,sheaders); - if ( shdr == NULL ) - { - PERROR("Could not get section header for .xen_prstatus"); - goto out; - } - filesz = sizeof(*ctxt) * nr_vcpus; - sts = xc_core_shdr_set(xch, shdr, strtab, XEN_DUMPCORE_SEC_PRSTATUS, - SHT_PROGBITS, offset, filesz, - __alignof__(*ctxt), sizeof(*ctxt)); - if ( sts != 0 ) - goto out; - offset += filesz; - - /* arch context */ - sts = xc_core_arch_context_get_shdr(xch, &arch_ctxt, sheaders, strtab, - &filesz, offset); - if ( sts != 0 ) - goto out; - offset += filesz; - - /* shared_info */ - if ( live_shinfo != NULL ) - { - shdr = xc_core_shdr_get(xch,sheaders); - if ( shdr == NULL ) - { - PERROR("Could not get section header for .xen_shared_info"); - goto out; - } - filesz = PAGE_SIZE; - sts = xc_core_shdr_set(xch, shdr, strtab, XEN_DUMPCORE_SEC_SHARED_INFO, - SHT_PROGBITS, offset, filesz, - __alignof__(*live_shinfo), PAGE_SIZE); - if ( sts != 0 ) - goto out; - offset += filesz; - } - - /* - * pages and p2m/pfn are the last section to allocate section headers - * so that we know the number of section headers here. - * 2 = pages section and p2m/pfn table section - */ - fixup = (sheaders->num + 2) * sizeof(*shdr); - /* zeroth section should have zero offset */ - for ( i = 1; i < sheaders->num; i++ ) - sheaders->shdrs[i].sh_offset += fixup; - offset += fixup; - dummy_len = ROUNDUP(offset, PAGE_SHIFT) - offset; /* padding length */ - offset += dummy_len; - - /* pages */ - shdr = xc_core_shdr_get(xch,sheaders); - if ( shdr == NULL ) - { - PERROR("could not get section headers for .xen_pages"); - goto out; - } - filesz = (uint64_t)nr_pages * PAGE_SIZE; - sts = xc_core_shdr_set(xch, shdr, strtab, XEN_DUMPCORE_SEC_PAGES, SHT_PROGBITS, - offset, filesz, PAGE_SIZE, PAGE_SIZE); - if ( sts != 0 ) - goto out; - offset += filesz; - - /* p2m/pfn table */ - shdr = xc_core_shdr_get(xch,sheaders); - if ( shdr == NULL ) - { - PERROR("Could not get section header for .xen_{p2m, pfn} table"); - goto out; - } - if ( !auto_translated_physmap ) - { - filesz = (uint64_t)nr_pages * sizeof(p2m_array[0]); - sts = xc_core_shdr_set(xch, shdr, strtab, XEN_DUMPCORE_SEC_P2M, - SHT_PROGBITS, - offset, filesz, __alignof__(p2m_array[0]), - sizeof(p2m_array[0])); - } - else - { - filesz = (uint64_t)nr_pages * sizeof(pfn_array[0]); - sts = xc_core_shdr_set(xch, shdr, strtab, XEN_DUMPCORE_SEC_PFN, - SHT_PROGBITS, - offset, filesz, __alignof__(pfn_array[0]), - sizeof(pfn_array[0])); - } - if ( sts != 0 ) - goto out; - offset += filesz; - - /* fixing up section header string table section header */ - filesz = strtab->length; - sheaders->shdrs[strtab_idx].sh_offset = offset; - sheaders->shdrs[strtab_idx].sh_size = filesz; - - /* write out elf header */ - ehdr.e_shnum = sheaders->num; - ehdr.e_shstrndx = strtab_idx; - ehdr.e_machine = ELF_ARCH_MACHINE; - sts = dump_rtn(xch, args, (char*)&ehdr, sizeof(ehdr)); - if ( sts != 0 ) - goto out; - - /* section headers */ - sts = dump_rtn(xch, args, (char*)sheaders->shdrs, - sheaders->num * sizeof(sheaders->shdrs[0])); - if ( sts != 0 ) - goto out; - - /* elf note section: xen core header */ - sts = elfnote_dump_none(xch, args, dump_rtn); - if ( sts != 0 ) - goto out; - - /* elf note section: xen core header */ - sts = elfnote_dump_core_header(xch, args, dump_rtn, &info, nr_vcpus, nr_pages); - if ( sts != 0 ) - goto out; - - /* elf note section: xen version */ - sts = elfnote_dump_xen_version(xch, args, dump_rtn, dinfo->guest_width); - if ( sts != 0 ) - goto out; - - /* elf note section: format version */ - sts = elfnote_dump_format_version(xch, args, dump_rtn); - if ( sts != 0 ) - goto out; - - /* prstatus: .xen_prstatus */ - sts = dump_rtn(xch, args, (char *)ctxt, sizeof(*ctxt) * nr_vcpus); - if ( sts != 0 ) - goto out; - - if ( live_shinfo != NULL ) - { - /* shared_info: .xen_shared_info */ - sts = dump_rtn(xch, args, (char*)live_shinfo, PAGE_SIZE); - if ( sts != 0 ) - goto out; - } - - /* arch specific context */ - sts = xc_core_arch_context_dump(xch, &arch_ctxt, args, dump_rtn); - if ( sts != 0 ) - goto out; - - /* Pad the output data to page alignment. */ - memset(dummy, 0, PAGE_SIZE); - sts = dump_rtn(xch, args, dummy, dummy_len); - if ( sts != 0 ) - goto out; - - /* dump pages: .xen_pages */ - j = 0; - dump_mem = dump_mem_start; - for ( map_idx = 0; map_idx < nr_memory_map; map_idx++ ) - { - uint64_t pfn_start; - uint64_t pfn_end; - - pfn_start = memory_map[map_idx].addr >> PAGE_SHIFT; - pfn_end = pfn_start + (memory_map[map_idx].size >> PAGE_SHIFT); - for ( i = pfn_start; i < pfn_end; i++ ) - { - uint64_t gmfn; - void *vaddr; - - if ( !auto_translated_physmap ) - { - if ( i >= dinfo->p2m_size ) - break; - - if ( dinfo->guest_width >= sizeof(unsigned long) ) - { - if ( dinfo->guest_width == sizeof(unsigned long) ) - gmfn = p2m[i]; - else - gmfn = ((uint64_t *)p2m)[i]; - if ( gmfn == INVALID_PFN ) - continue; - } - else - { - gmfn = ((uint32_t *)p2m)[i]; - if ( gmfn == (uint32_t)INVALID_PFN ) - continue; - } - if ( gmfn > max_mfn ) - continue; - - if ( j >= nr_pages ) - { - j++; - continue; - } - - p2m_array[j].pfn = i; - p2m_array[j].gmfn = gmfn; - } - else - { - if ( !xc_core_arch_gpfn_may_present(&arch_ctxt, i) ) - continue; - - if ( j >= nr_pages ) - { - j++; - continue; - } - - gmfn = i; - pfn_array[j] = i; - } - - vaddr = xc_map_foreign_range( - xch, domid, PAGE_SIZE, PROT_READ, gmfn); - if ( vaddr == NULL ) - continue; - memcpy(dump_mem, vaddr, PAGE_SIZE); - munmap(vaddr, PAGE_SIZE); - dump_mem += PAGE_SIZE; - if ( (j + 1) % DUMP_INCREMENT == 0 ) - { - sts = dump_rtn( - xch, args, dump_mem_start, dump_mem - dump_mem_start); - if ( sts != 0 ) - goto out; - dump_mem = dump_mem_start; - } - - j++; - } - } - - if ( j > nr_pages ) - { - /* - * When live dump-mode (-L option) is specified, - * guest domain may increase memory. - */ - IPRINTF("exceeded nr_pages (%ld) losing %ld pages", nr_pages, j - nr_pages); - } - - sts = dump_rtn(xch, args, dump_mem_start, dump_mem - dump_mem_start); - if ( sts != 0 ) - goto out; - if ( j < nr_pages ) - { - /* When live dump-mode (-L option) is specified, - * guest domain may reduce memory. pad with zero pages. - */ - DPRINTF("j (%ld) != nr_pages (%ld)", j, nr_pages); - memset(dump_mem_start, 0, PAGE_SIZE); - for (; j < nr_pages; j++) { - sts = dump_rtn(xch, args, dump_mem_start, PAGE_SIZE); - if ( sts != 0 ) - goto out; - if ( !auto_translated_physmap ) - { - p2m_array[j].pfn = XC_CORE_INVALID_PFN; - p2m_array[j].gmfn = XC_CORE_INVALID_GMFN; - } - else - pfn_array[j] = XC_CORE_INVALID_PFN; - } - } - - /* p2m/pfn table: .xen_p2m/.xen_pfn */ - if ( !auto_translated_physmap ) - sts = dump_rtn( - xch, args, (char *)p2m_array, sizeof(p2m_array[0]) * nr_pages); - else - sts = dump_rtn( - xch, args, (char *)pfn_array, sizeof(pfn_array[0]) * nr_pages); - if ( sts != 0 ) - goto out; - - /* elf section header string table: .shstrtab */ - sts = dump_rtn(xch, args, strtab->strings, strtab->length); - if ( sts != 0 ) - goto out; - - sts = 0; - -out: - if ( memory_map != NULL ) - free(memory_map); - if ( p2m != NULL ) - munmap(p2m, PAGE_SIZE * dinfo->p2m_frames); - if ( p2m_array != NULL ) - free(p2m_array); - if ( pfn_array != NULL ) - free(pfn_array); - if ( sheaders != NULL ) - xc_core_shdr_free(sheaders); - if ( strtab != NULL ) - xc_core_strtab_free(strtab); - if ( ctxt != NULL ) - free(ctxt); - if ( dump_mem_start != NULL ) - free(dump_mem_start); - if ( live_shinfo != NULL ) - munmap(live_shinfo, PAGE_SIZE); - xc_core_arch_context_free(&arch_ctxt); - - return sts; -} - -/* Callback args for writing to a local dump file. */ -struct dump_args { - int fd; -}; - -/* Callback routine for writing to a local dump file. */ -static int local_file_dump(xc_interface *xch, - void *args, char *buffer, unsigned int length) -{ - struct dump_args *da = args; - - if ( write_exact(da->fd, buffer, length) == -1 ) - { - PERROR("Failed to write buffer"); - return -errno; - } - - if ( length >= (DUMP_INCREMENT * PAGE_SIZE) ) - { - // Now dumping pages -- make sure we discard clean pages from - // the cache after each write - discard_file_cache(xch, da->fd, 0 /* no flush */); - } - - return 0; -} - -int -xc_domain_dumpcore(xc_interface *xch, - uint32_t domid, - const char *corename) -{ - struct dump_args da; - int sts; - - if ( (da.fd = open(corename, O_CREAT|O_RDWR|O_TRUNC, S_IWUSR|S_IRUSR)) < 0 ) - { - PERROR("Could not open corefile %s", corename); - return -errno; - } - - sts = xc_domain_dumpcore_via_callback( - xch, domid, &da, &local_file_dump); - - /* flush and discard any remaining portion of the file from cache */ - discard_file_cache(xch, da.fd, 1/* flush first*/); - - close(da.fd); - - return sts; -} - -/* - * Local variables: - * mode: C - * c-file-style: "BSD" - * c-basic-offset: 4 - * tab-width: 4 - * indent-tabs-mode: nil - * End: - */ diff --git a/tools/libs/ctrl/xc_core.h b/tools/libs/ctrl/xc_core.h deleted file mode 100644 index 8ea1f93a10..0000000000 --- a/tools/libs/ctrl/xc_core.h +++ /dev/null @@ -1,176 +0,0 @@ -/* - * Copyright (c) 2006 Isaku Yamahata - * VA Linux Systems Japan K.K. - * - * This library is free software; you can redistribute it and/or - * modify it under the terms of the GNU Lesser General Public - * License as published by the Free Software Foundation; either - * version 2.1 of the License, or (at your option) any later version. - * - * This library is distributed in the hope that it will be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU - * Lesser General Public License for more details. - * - * You should have received a copy of the GNU Lesser General Public - * License along with this library; If not, see . - * - */ - -#ifndef XC_CORE_H -#define XC_CORE_H - -#include "xen/version.h" -#include "xc_private.h" -#include "xen/libelf/elfstructs.h" - -/* section names */ -#define XEN_DUMPCORE_SEC_NOTE ".note.Xen" -#define XEN_DUMPCORE_SEC_PRSTATUS ".xen_prstatus" -#define XEN_DUMPCORE_SEC_SHARED_INFO ".xen_shared_info" -#define XEN_DUMPCORE_SEC_P2M ".xen_p2m" -#define XEN_DUMPCORE_SEC_PFN ".xen_pfn" -#define XEN_DUMPCORE_SEC_PAGES ".xen_pages" - -/* elf note name */ -#define XEN_DUMPCORE_ELFNOTE_NAME "Xen" -/* note numbers are defined in xen/elfnote.h */ - -struct elfnote { - uint32_t namesz; /* Elf_Note note; */ - uint32_t descsz; - uint32_t type; - char name[4]; /* sizeof("Xen") = 4 - * Fotunately this is 64bit aligned so that - * we can use same structore for both 32/64bit - */ -}; - -struct xen_dumpcore_elfnote_none_desc { - /* nothing */ -}; - -struct xen_dumpcore_elfnote_header_desc { - uint64_t xch_magic; - uint64_t xch_nr_vcpus; - uint64_t xch_nr_pages; - uint64_t xch_page_size; -}; - -struct xen_dumpcore_elfnote_xen_version_desc { - uint64_t major_version; - uint64_t minor_version; - xen_extraversion_t extra_version; - xen_compile_info_t compile_info; - xen_capabilities_info_t capabilities; - xen_changeset_info_t changeset; - xen_platform_parameters_t platform_parameters; - uint64_t pagesize; -}; - -#define XEN_DUMPCORE_FORMAT_VERSION(major, minor) \ - ((major) << 32) | ((minor) & 0xffffffff) -#define XEN_DUMPCORE_FORMAT_MAJOR(version) ((major) >> 32) -#define XEN_DUMPCORE_FORMAT_MINOR(version) ((minor) & 0xffffffff) - -#define XEN_DUMPCORE_FORMAT_MAJOR_CURRENT ((uint64_t)0) -#define XEN_DUMPCORE_FORMAT_MINOR_CURRENT ((uint64_t)1) -#define XEN_DUMPCORE_FORMAT_VERSION_CURRENT \ - XEN_DUMPCORE_FORMAT_VERSION(XEN_DUMPCORE_FORMAT_MAJOR_CURRENT, \ - XEN_DUMPCORE_FORMAT_MINOR_CURRENT) - -struct xen_dumpcore_elfnote_format_version_desc { - uint64_t version; -}; - - -struct xen_dumpcore_elfnote_none { - struct elfnote elfnote; - struct xen_dumpcore_elfnote_none_desc none; -}; - -struct xen_dumpcore_elfnote_header { - struct elfnote elfnote; - struct xen_dumpcore_elfnote_header_desc header; -}; - -struct xen_dumpcore_elfnote_xen_version { - struct elfnote elfnote; - struct xen_dumpcore_elfnote_xen_version_desc xen_version; -}; - -struct xen_dumpcore_elfnote_format_version { - struct elfnote elfnote; - struct xen_dumpcore_elfnote_format_version_desc format_version; -}; - -#define XC_CORE_INVALID_PFN (~(uint64_t)0) -#define XC_CORE_INVALID_GMFN (~(uint64_t)0) -struct xen_dumpcore_p2m { - uint64_t pfn; - uint64_t gmfn; -}; - - -struct xc_core_strtab; -struct xc_core_section_headers; - -Elf64_Shdr* -xc_core_shdr_get(xc_interface *xch, - struct xc_core_section_headers *sheaders); -int -xc_core_shdr_set(xc_interface *xch, - Elf64_Shdr *shdr, - struct xc_core_strtab *strtab, - const char *name, uint32_t type, - uint64_t offset, uint64_t size, - uint64_t addralign, uint64_t entsize); - -struct xc_core_memory_map { - uint64_t addr; - uint64_t size; -}; -typedef struct xc_core_memory_map xc_core_memory_map_t; -int xc_core_arch_auto_translated_physmap(const xc_dominfo_t *info); -struct xc_core_arch_context; -int xc_core_arch_memory_map_get(xc_interface *xch, - struct xc_core_arch_context *arch_ctxt, - xc_dominfo_t *info, shared_info_any_t *live_shinfo, - xc_core_memory_map_t **mapp, - unsigned int *nr_entries); -int xc_core_arch_map_p2m(xc_interface *xch, struct domain_info_context *dinfo, - xc_dominfo_t *info, shared_info_any_t *live_shinfo, - xen_pfn_t **live_p2m); - -int xc_core_arch_map_p2m_writable(xc_interface *xch, struct domain_info_context *dinfo, - xc_dominfo_t *info, - shared_info_any_t *live_shinfo, - xen_pfn_t **live_p2m); - -int xc_core_arch_get_scratch_gpfn(xc_interface *xch, uint32_t domid, - xen_pfn_t *gpfn); - - -#if defined (__i386__) || defined (__x86_64__) -# include "xc_core_x86.h" -#elif defined (__arm__) || defined(__aarch64__) -# include "xc_core_arm.h" -#else -# error "unsupported architecture" -#endif - -#ifndef ELF_CORE_EFLAGS -# define ELF_CORE_EFLAGS 0 -#endif - -#endif /* XC_CORE_H */ - -/* - * Local variables: - * mode: C - * c-file-style: "BSD" - * c-basic-offset: 4 - * tab-width: 4 - * indent-tabs-mode: nil - * End: - */ diff --git a/tools/libs/ctrl/xc_core_arm.c b/tools/libs/ctrl/xc_core_arm.c deleted file mode 100644 index 93765a565f..0000000000 --- a/tools/libs/ctrl/xc_core_arm.c +++ /dev/null @@ -1,113 +0,0 @@ -/* - * This library is free software; you can redistribute it and/or - * modify it under the terms of the GNU Lesser General Public - * License as published by the Free Software Foundation; either - * version 2.1 of the License, or (at your option) any later version. - * - * This library is distributed in the hope that it will be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU - * Lesser General Public License for more details. - * - * You should have received a copy of the GNU Lesser General Public - * License along with this library; If not, see . - * - * Copyright (c) 2011 Citrix Systems - * - */ - -#include "xc_private.h" -#include "xc_core.h" - -#include - -int -xc_core_arch_gpfn_may_present(struct xc_core_arch_context *arch_ctxt, - unsigned long pfn) -{ - /* TODO: memory from DT */ - if (pfn >= 0x80000 && pfn < 0x88000) - return 1; - return 0; -} - -int -xc_core_arch_auto_translated_physmap(const xc_dominfo_t *info) -{ - return 1; -} - -int -xc_core_arch_memory_map_get(xc_interface *xch, struct xc_core_arch_context *unused, - xc_dominfo_t *info, shared_info_any_t *live_shinfo, - xc_core_memory_map_t **mapp, - unsigned int *nr_entries) -{ - xen_pfn_t p2m_size = 0; - xc_core_memory_map_t *map; - - if ( xc_domain_nr_gpfns(xch, info->domid, &p2m_size) < 0 ) - return -1; - - map = malloc(sizeof(*map)); - if ( map == NULL ) - { - PERROR("Could not allocate memory"); - return -1; - } - - map->addr = 0; - map->size = ((uint64_t)p2m_size) << PAGE_SHIFT; - - *mapp = map; - *nr_entries = 1; - return 0; -} - -static int -xc_core_arch_map_p2m_rw(xc_interface *xch, struct domain_info_context *dinfo, xc_dominfo_t *info, - shared_info_any_t *live_shinfo, xen_pfn_t **live_p2m, int rw) -{ - errno = ENOSYS; - return -1; -} - -int -xc_core_arch_map_p2m(xc_interface *xch, struct domain_info_context *dinfo, xc_dominfo_t *info, - shared_info_any_t *live_shinfo, xen_pfn_t **live_p2m) -{ - return xc_core_arch_map_p2m_rw(xch, dinfo, info, live_shinfo, live_p2m, 0); -} - -int -xc_core_arch_map_p2m_writable(xc_interface *xch, struct domain_info_context *dinfo, xc_dominfo_t *info, - shared_info_any_t *live_shinfo, xen_pfn_t **live_p2m) -{ - return xc_core_arch_map_p2m_rw(xch, dinfo, info, live_shinfo, live_p2m, 1); -} - -int -xc_core_arch_get_scratch_gpfn(xc_interface *xch, uint32_t domid, - xen_pfn_t *gpfn) -{ - /* - * The Grant Table region space is not used until the guest is - * booting. Use the first page for the scratch pfn. - */ - BUILD_BUG_ON(GUEST_GNTTAB_SIZE < XC_PAGE_SIZE); - - *gpfn = GUEST_GNTTAB_BASE >> XC_PAGE_SHIFT; - - return 0; -} - - -/* - * Local variables: - * mode: C - * c-file-style: "BSD" - * c-basic-offset: 4 - * tab-width: 4 - * indent-tabs-mode: nil - * End: - */ diff --git a/tools/libs/ctrl/xc_core_arm.h b/tools/libs/ctrl/xc_core_arm.h deleted file mode 100644 index 162f7a7569..0000000000 --- a/tools/libs/ctrl/xc_core_arm.h +++ /dev/null @@ -1,59 +0,0 @@ -/* - * This library is free software; you can redistribute it and/or - * modify it under the terms of the GNU Lesser General Public - * License as published by the Free Software Foundation; either - * version 2.1 of the License, or (at your option) any later version. - * - * This library is distributed in the hope that it will be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU - * Lesser General Public License for more details. - * - * You should have received a copy of the GNU Lesser General Public - * License along with this library; If not, see . - * - * Copyright (c) 2012 Citrix Systems - * - */ - -#ifndef XC_CORE_ARM_H -#define XC_CORE_ARM_H - -#define ELF_ARCH_DATA ELFDATA2LSB -#define ELF_ARCH_MACHINE EM_ARM - -struct xc_core_arch_context { - /* nothing */ -}; - -#define xc_core_arch_context_init(arch_ctxt) do {} while (0) -#define xc_core_arch_context_free(arch_ctxt) do {} while (0) -#define xc_core_arch_context_get(arch_ctxt, ctxt, xch, domid) \ - (0) -#define xc_core_arch_context_dump(xch, arch_ctxt, args, dump_rtn) (0) - -int -xc_core_arch_gpfn_may_present(struct xc_core_arch_context *arch_ctxt, - unsigned long pfn); -static inline int -xc_core_arch_context_get_shdr(xc_interface *xch, - struct xc_core_arch_context *arch_ctxt, - struct xc_core_section_headers *sheaders, - struct xc_core_strtab *strtab, - uint64_t *filesz, uint64_t offset) -{ - *filesz = 0; - return 0; -} - -#endif /* XC_CORE_ARM_H */ - -/* - * Local variables: - * mode: C - * c-file-style: "BSD" - * c-basic-offset: 4 - * tab-width: 4 - * indent-tabs-mode: nil - * End: - */ diff --git a/tools/libs/ctrl/xc_core_x86.c b/tools/libs/ctrl/xc_core_x86.c deleted file mode 100644 index c8f71d4b75..0000000000 --- a/tools/libs/ctrl/xc_core_x86.c +++ /dev/null @@ -1,391 +0,0 @@ -/* - * This library is free software; you can redistribute it and/or - * modify it under the terms of the GNU Lesser General Public - * License as published by the Free Software Foundation; either - * version 2.1 of the License, or (at your option) any later version. - * - * This library is distributed in the hope that it will be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU - * Lesser General Public License for more details. - * - * You should have received a copy of the GNU Lesser General Public - * License along with this library; If not, see . - * - * Copyright (c) 2007 Isaku Yamahata - * VA Linux Systems Japan K.K. - * - */ - -#include -#include "xc_private.h" -#include "xc_core.h" -#include - -int -xc_core_arch_gpfn_may_present(struct xc_core_arch_context *arch_ctxt, - unsigned long pfn) -{ - if ((pfn >= 0xa0 && pfn < 0xc0) /* VGA hole */ - || (pfn >= (HVM_BELOW_4G_MMIO_START >> PAGE_SHIFT) - && pfn < (1ULL<<32) >> PAGE_SHIFT)) /* MMIO */ - return 0; - return 1; -} - -int -xc_core_arch_auto_translated_physmap(const xc_dominfo_t *info) -{ - return info->hvm; -} - -int -xc_core_arch_memory_map_get(xc_interface *xch, struct xc_core_arch_context *unused, - xc_dominfo_t *info, shared_info_any_t *live_shinfo, - xc_core_memory_map_t **mapp, - unsigned int *nr_entries) -{ - xen_pfn_t p2m_size = 0; - xc_core_memory_map_t *map; - - if ( xc_domain_nr_gpfns(xch, info->domid, &p2m_size) < 0 ) - return -1; - - map = malloc(sizeof(*map)); - if ( map == NULL ) - { - PERROR("Could not allocate memory"); - return -1; - } - - map->addr = 0; - map->size = ((uint64_t)p2m_size) << PAGE_SHIFT; - - *mapp = map; - *nr_entries = 1; - return 0; -} - -static inline bool is_canonical_address(uint64_t vaddr) -{ - return ((int64_t)vaddr >> 47) == ((int64_t)vaddr >> 63); -} - -/* Virtual address ranges reserved for hypervisor. */ -#define HYPERVISOR_VIRT_START_X86_64 0xFFFF800000000000ULL -#define HYPERVISOR_VIRT_END_X86_64 0xFFFF87FFFFFFFFFFULL - -#define HYPERVISOR_VIRT_START_X86_32 0x00000000F5800000ULL -#define HYPERVISOR_VIRT_END_X86_32 0x00000000FFFFFFFFULL - -static xen_pfn_t * -xc_core_arch_map_p2m_list_rw(xc_interface *xch, struct domain_info_context *dinfo, - uint32_t dom, shared_info_any_t *live_shinfo, - uint64_t p2m_cr3) -{ - uint64_t p2m_vaddr, p2m_end, mask, off; - xen_pfn_t p2m_mfn, mfn, saved_mfn, max_pfn; - uint64_t *ptes = NULL; - xen_pfn_t *mfns = NULL; - unsigned int fpp, n_pages, level, n_levels, shift, - idx_start, idx_end, idx, saved_idx; - - p2m_vaddr = GET_FIELD(live_shinfo, arch.p2m_vaddr, dinfo->guest_width); - fpp = PAGE_SIZE / dinfo->guest_width; - dinfo->p2m_frames = (dinfo->p2m_size - 1) / fpp + 1; - p2m_end = p2m_vaddr + dinfo->p2m_frames * PAGE_SIZE - 1; - - if ( dinfo->guest_width == 8 ) - { - mask = 0x0000ffffffffffffULL; - n_levels = 4; - p2m_mfn = p2m_cr3 >> 12; - if ( !is_canonical_address(p2m_vaddr) || - !is_canonical_address(p2m_end) || - p2m_end < p2m_vaddr || - (p2m_vaddr <= HYPERVISOR_VIRT_END_X86_64 && - p2m_end > HYPERVISOR_VIRT_START_X86_64) ) - { - ERROR("Bad virtual p2m address range %#" PRIx64 "-%#" PRIx64, - p2m_vaddr, p2m_end); - errno = ERANGE; - goto out; - } - } - else - { - mask = 0x00000000ffffffffULL; - n_levels = 3; - if ( p2m_cr3 & ~mask ) - p2m_mfn = ~0UL; - else - p2m_mfn = (uint32_t)((p2m_cr3 >> 12) | (p2m_cr3 << 20)); - if ( p2m_vaddr > mask || p2m_end > mask || p2m_end < p2m_vaddr || - (p2m_vaddr <= HYPERVISOR_VIRT_END_X86_32 && - p2m_end > HYPERVISOR_VIRT_START_X86_32) ) - { - ERROR("Bad virtual p2m address range %#" PRIx64 "-%#" PRIx64, - p2m_vaddr, p2m_end); - errno = ERANGE; - goto out; - } - } - - mfns = malloc(sizeof(*mfns)); - if ( !mfns ) - { - ERROR("Cannot allocate memory for array of %u mfns", 1); - goto out; - } - mfns[0] = p2m_mfn; - off = 0; - saved_mfn = 0; - idx_start = idx_end = saved_idx = 0; - - for ( level = n_levels; level > 0; level-- ) - { - n_pages = idx_end - idx_start + 1; - ptes = xc_map_foreign_pages(xch, dom, PROT_READ, mfns, n_pages); - if ( !ptes ) - { - PERROR("Failed to map %u page table pages for p2m list", n_pages); - goto out; - } - free(mfns); - - shift = level * 9 + 3; - idx_start = ((p2m_vaddr - off) & mask) >> shift; - idx_end = ((p2m_end - off) & mask) >> shift; - idx = idx_end - idx_start + 1; - mfns = malloc(sizeof(*mfns) * idx); - if ( !mfns ) - { - ERROR("Cannot allocate memory for array of %u mfns", idx); - goto out; - } - - for ( idx = idx_start; idx <= idx_end; idx++ ) - { - mfn = (ptes[idx] & 0x000ffffffffff000ULL) >> PAGE_SHIFT; - if ( mfn == 0 ) - { - ERROR("Bad mfn %#lx during page table walk for vaddr %#" PRIx64 " at level %d of p2m list", - mfn, off + ((uint64_t)idx << shift), level); - errno = ERANGE; - goto out; - } - mfns[idx - idx_start] = mfn; - - /* Maximum pfn check at level 2. Same reasoning as for p2m tree. */ - if ( level == 2 ) - { - if ( mfn != saved_mfn ) - { - saved_mfn = mfn; - saved_idx = idx - idx_start; - } - } - } - - if ( level == 2 ) - { - if ( saved_idx == idx_end ) - saved_idx++; - max_pfn = ((xen_pfn_t)saved_idx << 9) * fpp; - if ( max_pfn < dinfo->p2m_size ) - { - dinfo->p2m_size = max_pfn; - dinfo->p2m_frames = (dinfo->p2m_size + fpp - 1) / fpp; - p2m_end = p2m_vaddr + dinfo->p2m_frames * PAGE_SIZE - 1; - idx_end = idx_start + saved_idx; - } - } - - munmap(ptes, n_pages * PAGE_SIZE); - ptes = NULL; - off = p2m_vaddr & ((mask >> shift) << shift); - } - - return mfns; - - out: - free(mfns); - if ( ptes ) - munmap(ptes, n_pages * PAGE_SIZE); - - return NULL; -} - -static xen_pfn_t * -xc_core_arch_map_p2m_tree_rw(xc_interface *xch, struct domain_info_context *dinfo, - uint32_t dom, shared_info_any_t *live_shinfo) -{ - /* Double and single indirect references to the live P2M table */ - xen_pfn_t *live_p2m_frame_list_list; - xen_pfn_t *live_p2m_frame_list = NULL; - /* Copies of the above. */ - xen_pfn_t *p2m_frame_list_list = NULL; - xen_pfn_t *p2m_frame_list; - - int err; - int i; - - live_p2m_frame_list_list = - xc_map_foreign_range(xch, dom, PAGE_SIZE, PROT_READ, - GET_FIELD(live_shinfo, arch.pfn_to_mfn_frame_list_list, dinfo->guest_width)); - - if ( !live_p2m_frame_list_list ) - { - PERROR("Couldn't map p2m_frame_list_list (errno %d)", errno); - goto out; - } - - /* Get a local copy of the live_P2M_frame_list_list */ - if ( !(p2m_frame_list_list = malloc(PAGE_SIZE)) ) - { - ERROR("Couldn't allocate p2m_frame_list_list array"); - goto out; - } - memcpy(p2m_frame_list_list, live_p2m_frame_list_list, PAGE_SIZE); - - /* Canonicalize guest's unsigned long vs ours */ - if ( dinfo->guest_width > sizeof(unsigned long) ) - for ( i = 0; i < PAGE_SIZE/sizeof(unsigned long); i++ ) - if ( i < PAGE_SIZE/dinfo->guest_width ) - p2m_frame_list_list[i] = ((uint64_t *)p2m_frame_list_list)[i]; - else - p2m_frame_list_list[i] = 0; - else if ( dinfo->guest_width < sizeof(unsigned long) ) - for ( i = PAGE_SIZE/sizeof(unsigned long) - 1; i >= 0; i-- ) - p2m_frame_list_list[i] = ((uint32_t *)p2m_frame_list_list)[i]; - - live_p2m_frame_list = - xc_map_foreign_pages(xch, dom, PROT_READ, - p2m_frame_list_list, - P2M_FLL_ENTRIES); - - if ( !live_p2m_frame_list ) - { - PERROR("Couldn't map p2m_frame_list"); - goto out; - } - - /* Get a local copy of the live_P2M_frame_list */ - if ( !(p2m_frame_list = malloc(P2M_TOOLS_FL_SIZE)) ) - { - ERROR("Couldn't allocate p2m_frame_list array"); - goto out; - } - memset(p2m_frame_list, 0, P2M_TOOLS_FL_SIZE); - memcpy(p2m_frame_list, live_p2m_frame_list, P2M_GUEST_FL_SIZE); - - /* Canonicalize guest's unsigned long vs ours */ - if ( dinfo->guest_width > sizeof(unsigned long) ) - for ( i = 0; i < P2M_FL_ENTRIES; i++ ) - p2m_frame_list[i] = ((uint64_t *)p2m_frame_list)[i]; - else if ( dinfo->guest_width < sizeof(unsigned long) ) - for ( i = P2M_FL_ENTRIES - 1; i >= 0; i-- ) - p2m_frame_list[i] = ((uint32_t *)p2m_frame_list)[i]; - - dinfo->p2m_frames = P2M_FL_ENTRIES; - - return p2m_frame_list; - - out: - err = errno; - - if ( live_p2m_frame_list_list ) - munmap(live_p2m_frame_list_list, PAGE_SIZE); - - if ( live_p2m_frame_list ) - munmap(live_p2m_frame_list, P2M_FLL_ENTRIES * PAGE_SIZE); - - free(p2m_frame_list_list); - - errno = err; - - return NULL; -} - -static int -xc_core_arch_map_p2m_rw(xc_interface *xch, struct domain_info_context *dinfo, xc_dominfo_t *info, - shared_info_any_t *live_shinfo, xen_pfn_t **live_p2m, int rw) -{ - xen_pfn_t *p2m_frame_list = NULL; - uint64_t p2m_cr3; - uint32_t dom = info->domid; - int ret = -1; - int err; - - if ( xc_domain_nr_gpfns(xch, info->domid, &dinfo->p2m_size) < 0 ) - { - ERROR("Could not get maximum GPFN!"); - goto out; - } - - if ( dinfo->p2m_size < info->nr_pages ) - { - ERROR("p2m_size < nr_pages -1 (%lx < %lx", dinfo->p2m_size, info->nr_pages - 1); - goto out; - } - - p2m_cr3 = GET_FIELD(live_shinfo, arch.p2m_cr3, dinfo->guest_width); - - p2m_frame_list = p2m_cr3 ? xc_core_arch_map_p2m_list_rw(xch, dinfo, dom, live_shinfo, p2m_cr3) - : xc_core_arch_map_p2m_tree_rw(xch, dinfo, dom, live_shinfo); - - if ( !p2m_frame_list ) - goto out; - - *live_p2m = xc_map_foreign_pages(xch, dom, - rw ? (PROT_READ | PROT_WRITE) : PROT_READ, - p2m_frame_list, - dinfo->p2m_frames); - - if ( !*live_p2m ) - { - PERROR("Couldn't map p2m table"); - goto out; - } - - ret = 0; - -out: - err = errno; - - free(p2m_frame_list); - - errno = err; - return ret; -} - -int -xc_core_arch_map_p2m(xc_interface *xch, struct domain_info_context *dinfo, xc_dominfo_t *info, - shared_info_any_t *live_shinfo, xen_pfn_t **live_p2m) -{ - return xc_core_arch_map_p2m_rw(xch, dinfo, info, live_shinfo, live_p2m, 0); -} - -int -xc_core_arch_map_p2m_writable(xc_interface *xch, struct domain_info_context *dinfo, xc_dominfo_t *info, - shared_info_any_t *live_shinfo, xen_pfn_t **live_p2m) -{ - return xc_core_arch_map_p2m_rw(xch, dinfo, info, live_shinfo, live_p2m, 1); -} - -int -xc_core_arch_get_scratch_gpfn(xc_interface *xch, uint32_t domid, - xen_pfn_t *gpfn) -{ - return xc_domain_nr_gpfns(xch, domid, gpfn); -} - -/* - * Local variables: - * mode: C - * c-file-style: "BSD" - * c-basic-offset: 4 - * tab-width: 4 - * indent-tabs-mode: nil - * End: - */ diff --git a/tools/libs/ctrl/xc_core_x86.h b/tools/libs/ctrl/xc_core_x86.h deleted file mode 100644 index 867146b1d9..0000000000 --- a/tools/libs/ctrl/xc_core_x86.h +++ /dev/null @@ -1,60 +0,0 @@ -/* - * This library is free software; you can redistribute it and/or - * modify it under the terms of the GNU Lesser General Public - * License as published by the Free Software Foundation; either - * version 2.1 of the License, or (at your option) any later version. - * - * This library is distributed in the hope that it will be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU - * Lesser General Public License for more details. - * - * You should have received a copy of the GNU Lesser General Public - * License along with this library; If not, see . - * - * Copyright (c) 2007 Isaku Yamahata - * VA Linux Systems Japan K.K. - * - */ - -#ifndef XC_CORE_X86_H -#define XC_CORE_X86_H - -#define ELF_ARCH_DATA ELFDATA2LSB -#define ELF_ARCH_MACHINE (dinfo->guest_width == 8 ? EM_X86_64 : EM_386) - -struct xc_core_arch_context { - /* nothing */ -}; - -#define xc_core_arch_context_init(arch_ctxt) do {} while (0) -#define xc_core_arch_context_free(arch_ctxt) do {} while (0) -#define xc_core_arch_context_get(arch_ctxt, ctxt, xch, domid) \ - (0) -#define xc_core_arch_context_dump(xch, arch_ctxt, args, dump_rtn) (0) - -int -xc_core_arch_gpfn_may_present(struct xc_core_arch_context *arch_ctxt, - unsigned long pfn); -static inline int -xc_core_arch_context_get_shdr(xc_interface *xch, - struct xc_core_arch_context *arch_ctxt, - struct xc_core_section_headers *sheaders, - struct xc_core_strtab *strtab, - uint64_t *filesz, uint64_t offset) -{ - *filesz = 0; - return 0; -} - -#endif /* XC_CORE_X86_H */ - -/* - * Local variables: - * mode: C - * c-file-style: "BSD" - * c-basic-offset: 4 - * tab-width: 4 - * indent-tabs-mode: nil - * End: - */ diff --git a/tools/libs/ctrl/xc_domain.c b/tools/libs/ctrl/xc_domain.c index e7cea4a17d..7d118848f1 100644 --- a/tools/libs/ctrl/xc_domain.c +++ b/tools/libs/ctrl/xc_domain.c @@ -19,8 +19,6 @@ * Copyright (c) 2003, K A Fraser. */ -#include "xc_private.h" -#include "xc_core.h" #include "xc_private.h" #include #include diff --git a/tools/libs/ctrl/xc_private.h b/tools/libs/ctrl/xc_private.h index 8ebc0b59da..dff0f0289b 100644 --- a/tools/libs/ctrl/xc_private.h +++ b/tools/libs/ctrl/xc_private.h @@ -467,6 +467,18 @@ void *xc_vm_event_enable(xc_interface *xch, uint32_t domain_id, int param, int do_dm_op(xc_interface *xch, uint32_t domid, unsigned int nr_bufs, ...); +#if defined (__i386__) || defined (__x86_64__) +static inline int xc_core_arch_auto_translated_physmap(const xc_dominfo_t *info) +{ + return info->hvm; +} +#elif defined (__arm__) || defined(__aarch64__) +static inline int xc_core_arch_auto_translated_physmap(const xc_dominfo_t *info) +{ + return 1; +} +#endif + #endif /* __XC_PRIVATE_H__ */ /* diff --git a/tools/libs/guest/Makefile b/tools/libs/guest/Makefile index 2a2323ff09..2ce92d247e 100644 --- a/tools/libs/guest/Makefile +++ b/tools/libs/guest/Makefile @@ -24,6 +24,9 @@ SRCS-y += xg_offline_page.c else SRCS-y += xg_nomigrate.c endif +SRCS-y += xg_core.c +SRCS-$(CONFIG_X86) += xg_core_x86.c +SRCS-$(CONFIG_ARM) += xg_core_arm.c CFLAGS += -I$(XEN_libxenctrl) diff --git a/tools/libs/guest/xg_core.c b/tools/libs/guest/xg_core.c new file mode 100644 index 0000000000..c52f1161c1 --- /dev/null +++ b/tools/libs/guest/xg_core.c @@ -0,0 +1,1027 @@ +/* + * Elf format, (pfn, gmfn) table, IA64 support. + * Copyright (c) 2007 Isaku Yamahata + * VA Linux Systems Japan K.K. + * + * This library is free software; you can redistribute it and/or + * modify it under the terms of the GNU Lesser General Public + * License as published by the Free Software Foundation; + * version 2.1 of the License. + * + * This library is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public + * License along with this library; If not, see . + */ + +/* + * xen dump-core file format follows ELF format specification. + * Analisys tools shouldn't depends on the order of sections. + * They should follow elf header and check section names. + * + * +--------------------------------------------------------+ + * |ELF header | + * +--------------------------------------------------------+ + * |section headers | + * | null section header | + * | .shstrtab | + * | .note.Xen | + * | .xen_prstatus | + * | .xen_shared_info if present | + * | .xen_pages | + * | .xen_p2m or .xen_pfn | + * +--------------------------------------------------------+ + * |.note.Xen:note section | + * | "Xen" is used as note name, | + * | types are defined in xen/include/public/elfnote.h | + * | and descriptors are defined in xc_core.h. | + * | dumpcore none | + * | dumpcore header | + * | dumpcore xen version | + * | dumpcore format version | + * +--------------------------------------------------------+ + * |.xen_prstatus | + * | vcpu_guest_context_t[nr_vcpus] | + * +--------------------------------------------------------+ + * |.xen_shared_info if possible | + * +--------------------------------------------------------+ + * |.xen_pages | + * | page * nr_pages | + * +--------------------------------------------------------+ + * |.xen_p2m or .xen_pfn | + * | .xen_p2m: struct xen_dumpcore_p2m[nr_pages] | + * | .xen_pfn: uint64_t[nr_pages] | + * +--------------------------------------------------------+ + * |.shstrtab: section header string table | + * +--------------------------------------------------------+ + * + */ + +#include "xc_private.h" +#include "xg_core.h" +#include +#include + +#include + +/* number of pages to write at a time */ +#define DUMP_INCREMENT (4 * 1024) + +/* string table */ +struct xc_core_strtab { + char *strings; + uint16_t length; + uint16_t max; +}; + +static struct xc_core_strtab* +xc_core_strtab_init(xc_interface *xch) +{ + struct xc_core_strtab *strtab; + char *strings; + strtab = malloc(sizeof(*strtab)); + if ( strtab == NULL ) + return NULL; + + strings = malloc(PAGE_SIZE); + if ( strings == NULL ) + { + PERROR("Could not allocate string table init"); + free(strtab); + return NULL; + } + strtab->strings = strings; + strtab->max = PAGE_SIZE; + + /* index 0 represents none */ + strtab->strings[0] = '\0'; + strtab->length = 1; + + return strtab; +} + +static void +xc_core_strtab_free(struct xc_core_strtab *strtab) +{ + free(strtab->strings); + free(strtab); +} + +static uint16_t +xc_core_strtab_get(xc_interface *xch, struct xc_core_strtab *strtab, const char *name) +{ + uint16_t ret = 0; + uint16_t len = strlen(name) + 1; + + if ( strtab->length > UINT16_MAX - len ) + { + PERROR("too long string table"); + errno = E2BIG; + return ret; + } + + if ( strtab->length + len > strtab->max ) + { + char *tmp; + if ( strtab->max > UINT16_MAX / 2 ) + { + PERROR("too long string table"); + errno = ENOMEM; + return ret; + } + + tmp = realloc(strtab->strings, strtab->max * 2); + if ( tmp == NULL ) + { + PERROR("Could not allocate string table"); + return ret; + } + + strtab->strings = tmp; + strtab->max *= 2; + } + + ret = strtab->length; + strcpy(strtab->strings + strtab->length, name); + strtab->length += len; + return ret; +} + + +/* section headers */ +struct xc_core_section_headers { + uint16_t num; + uint16_t num_max; + + Elf64_Shdr *shdrs; +}; +#define SHDR_INIT ((uint16_t)16) +#define SHDR_INC ((uint16_t)4) + +static struct xc_core_section_headers* +xc_core_shdr_init(xc_interface *xch) +{ + struct xc_core_section_headers *sheaders; + sheaders = malloc(sizeof(*sheaders)); + if ( sheaders == NULL ) + return NULL; + + sheaders->num = 0; + sheaders->num_max = SHDR_INIT; + sheaders->shdrs = malloc(sizeof(sheaders->shdrs[0]) * sheaders->num_max); + if ( sheaders->shdrs == NULL ) + { + free(sheaders); + return NULL; + } + return sheaders; +} + +static void +xc_core_shdr_free(struct xc_core_section_headers *sheaders) +{ + free(sheaders->shdrs); + free(sheaders); +} + +Elf64_Shdr* +xc_core_shdr_get(xc_interface *xch, + struct xc_core_section_headers *sheaders) +{ + Elf64_Shdr *shdr; + + if ( sheaders->num == sheaders->num_max ) + { + Elf64_Shdr *shdrs; + if ( sheaders->num_max > UINT16_MAX - SHDR_INC ) + { + errno = E2BIG; + return NULL; + } + sheaders->num_max += SHDR_INC; + shdrs = realloc(sheaders->shdrs, + sizeof(sheaders->shdrs[0]) * sheaders->num_max); + if ( shdrs == NULL ) + return NULL; + sheaders->shdrs = shdrs; + } + + shdr = &sheaders->shdrs[sheaders->num]; + sheaders->num++; + memset(shdr, 0, sizeof(*shdr)); + return shdr; +} + +int +xc_core_shdr_set(xc_interface *xch, + Elf64_Shdr *shdr, + struct xc_core_strtab *strtab, + const char *name, uint32_t type, + uint64_t offset, uint64_t size, + uint64_t addralign, uint64_t entsize) +{ + uint64_t name_idx = xc_core_strtab_get(xch, strtab, name); + if ( name_idx == 0 ) + return -1; + + shdr->sh_name = name_idx; + shdr->sh_type = type; + shdr->sh_offset = offset; + shdr->sh_size = size; + shdr->sh_addralign = addralign; + shdr->sh_entsize = entsize; + return 0; +} + +static void +xc_core_ehdr_init(Elf64_Ehdr *ehdr) +{ + memset(ehdr, 0, sizeof(*ehdr)); + ehdr->e_ident[EI_MAG0] = ELFMAG0; + ehdr->e_ident[EI_MAG1] = ELFMAG1; + ehdr->e_ident[EI_MAG2] = ELFMAG2; + ehdr->e_ident[EI_MAG3] = ELFMAG3; + ehdr->e_ident[EI_CLASS] = ELFCLASS64; + ehdr->e_ident[EI_DATA] = ELF_ARCH_DATA; + ehdr->e_ident[EI_VERSION] = EV_CURRENT; + ehdr->e_ident[EI_OSABI] = ELFOSABI_SYSV; + ehdr->e_ident[EI_ABIVERSION] = EV_CURRENT; + + ehdr->e_type = ET_CORE; + /* e_machine will be filled in later */ + ehdr->e_version = EV_CURRENT; + ehdr->e_entry = 0; + ehdr->e_phoff = 0; + ehdr->e_shoff = sizeof(*ehdr); + ehdr->e_flags = ELF_CORE_EFLAGS; + ehdr->e_ehsize = sizeof(*ehdr); + ehdr->e_phentsize = sizeof(Elf64_Phdr); + ehdr->e_phnum = 0; + ehdr->e_shentsize = sizeof(Elf64_Shdr); + /* ehdr->e_shnum and ehdr->e_shstrndx aren't known here yet. + * fill it later */ +} + +static int +elfnote_fill_xen_version(xc_interface *xch, + struct xen_dumpcore_elfnote_xen_version_desc + *xen_version) +{ + int rc; + memset(xen_version, 0, sizeof(*xen_version)); + + rc = xc_version(xch, XENVER_version, NULL); + if ( rc < 0 ) + return rc; + xen_version->major_version = rc >> 16; + xen_version->minor_version = rc & ((1 << 16) - 1); + + rc = xc_version(xch, XENVER_extraversion, + &xen_version->extra_version); + if ( rc < 0 ) + return rc; + + rc = xc_version(xch, XENVER_compile_info, + &xen_version->compile_info); + if ( rc < 0 ) + return rc; + + rc = xc_version(xch, + XENVER_capabilities, &xen_version->capabilities); + if ( rc < 0 ) + return rc; + + rc = xc_version(xch, XENVER_changeset, &xen_version->changeset); + if ( rc < 0 ) + return rc; + + rc = xc_version(xch, XENVER_platform_parameters, + &xen_version->platform_parameters); + if ( rc < 0 ) + return rc; + + rc = xc_version(xch, XENVER_pagesize, NULL); + if ( rc < 0 ) + return rc; + xen_version->pagesize = rc; + + return 0; +} + +static void +elfnote_fill_format_version(struct xen_dumpcore_elfnote_format_version_desc + *format_version) +{ + format_version->version = XEN_DUMPCORE_FORMAT_VERSION_CURRENT; +} + +static void +elfnote_init(struct elfnote *elfnote) +{ + /* elf note section */ + memset(elfnote, 0, sizeof(*elfnote)); + elfnote->namesz = strlen(XEN_DUMPCORE_ELFNOTE_NAME) + 1; + strncpy(elfnote->name, XEN_DUMPCORE_ELFNOTE_NAME, sizeof(elfnote->name)); +} + +static int +elfnote_dump_none(xc_interface *xch, void *args, dumpcore_rtn_t dump_rtn) +{ + int sts; + struct elfnote elfnote; + struct xen_dumpcore_elfnote_none_desc none; + + elfnote_init(&elfnote); + /* Avoid compile warning about constant-zero-sized memset(). */ + /*memset(&none, 0, sizeof(none));*/ + + elfnote.descsz = sizeof(none); + elfnote.type = XEN_ELFNOTE_DUMPCORE_NONE; + sts = dump_rtn(xch, args, (char*)&elfnote, sizeof(elfnote)); + if ( sts != 0 ) + return sts; + return dump_rtn(xch, args, (char*)&none, sizeof(none)); +} + +static int +elfnote_dump_core_header( + xc_interface *xch, + void *args, dumpcore_rtn_t dump_rtn, const xc_dominfo_t *info, + int nr_vcpus, unsigned long nr_pages) +{ + int sts; + struct elfnote elfnote; + struct xen_dumpcore_elfnote_header_desc header; + + elfnote_init(&elfnote); + memset(&header, 0, sizeof(header)); + + elfnote.descsz = sizeof(header); + elfnote.type = XEN_ELFNOTE_DUMPCORE_HEADER; + header.xch_magic = info->hvm ? XC_CORE_MAGIC_HVM : XC_CORE_MAGIC; + header.xch_nr_vcpus = nr_vcpus; + header.xch_nr_pages = nr_pages; + header.xch_page_size = PAGE_SIZE; + sts = dump_rtn(xch, args, (char*)&elfnote, sizeof(elfnote)); + if ( sts != 0 ) + return sts; + return dump_rtn(xch, args, (char*)&header, sizeof(header)); +} + +static int +elfnote_dump_xen_version(xc_interface *xch, void *args, + dumpcore_rtn_t dump_rtn, unsigned int guest_width) +{ + int sts; + struct elfnote elfnote; + struct xen_dumpcore_elfnote_xen_version_desc xen_version; + + elfnote_init(&elfnote); + memset(&xen_version, 0, sizeof(xen_version)); + + elfnote.descsz = sizeof(xen_version); + elfnote.type = XEN_ELFNOTE_DUMPCORE_XEN_VERSION; + elfnote_fill_xen_version(xch, &xen_version); + if (guest_width < sizeof(unsigned long)) + { + // 32 bit elf file format differs in pagesize's alignment + char *p = (char *)&xen_version.pagesize; + memmove(p - 4, p, sizeof(xen_version.pagesize)); + } + sts = dump_rtn(xch, args, (char*)&elfnote, sizeof(elfnote)); + if ( sts != 0 ) + return sts; + return dump_rtn(xch, args, (char*)&xen_version, sizeof(xen_version)); +} + +static int +elfnote_dump_format_version(xc_interface *xch, + void *args, dumpcore_rtn_t dump_rtn) +{ + int sts; + struct elfnote elfnote; + struct xen_dumpcore_elfnote_format_version_desc format_version; + + elfnote_init(&elfnote); + memset(&format_version, 0, sizeof(format_version)); + + elfnote.descsz = sizeof(format_version); + elfnote.type = XEN_ELFNOTE_DUMPCORE_FORMAT_VERSION; + elfnote_fill_format_version(&format_version); + sts = dump_rtn(xch, args, (char*)&elfnote, sizeof(elfnote)); + if ( sts != 0 ) + return sts; + return dump_rtn(xch, args, (char*)&format_version, sizeof(format_version)); +} + +int +xc_domain_dumpcore_via_callback(xc_interface *xch, + uint32_t domid, + void *args, + dumpcore_rtn_t dump_rtn) +{ + xc_dominfo_t info; + shared_info_any_t *live_shinfo = NULL; + struct domain_info_context _dinfo = {}; + struct domain_info_context *dinfo = &_dinfo; + + int nr_vcpus = 0; + char *dump_mem, *dump_mem_start = NULL; + vcpu_guest_context_any_t *ctxt = NULL; + struct xc_core_arch_context arch_ctxt; + char dummy[PAGE_SIZE]; + int dummy_len; + int sts = -1; + + unsigned long i; + unsigned long j; + unsigned long nr_pages; + unsigned long max_mfn; + + xc_core_memory_map_t *memory_map = NULL; + unsigned int nr_memory_map; + unsigned int map_idx; + + int auto_translated_physmap; + xen_pfn_t *p2m = NULL; + struct xen_dumpcore_p2m *p2m_array = NULL; + + uint64_t *pfn_array = NULL; + + Elf64_Ehdr ehdr; + uint64_t filesz; + uint64_t offset; + uint64_t fixup; + + struct xc_core_strtab *strtab = NULL; + uint16_t strtab_idx; + struct xc_core_section_headers *sheaders = NULL; + Elf64_Shdr *shdr; + + xc_core_arch_context_init(&arch_ctxt); + if ( (dump_mem_start = malloc(DUMP_INCREMENT*PAGE_SIZE)) == NULL ) + { + PERROR("Could not allocate dump_mem"); + goto out; + } + + if ( xc_domain_getinfo(xch, domid, 1, &info) != 1 ) + { + PERROR("Could not get info for domain"); + goto out; + } + /* Map the shared info frame */ + live_shinfo = xc_map_foreign_range(xch, domid, PAGE_SIZE, + PROT_READ, info.shared_info_frame); + if ( !live_shinfo && !info.hvm ) + { + PERROR("Couldn't map live_shinfo"); + goto out; + } + auto_translated_physmap = xc_core_arch_auto_translated_physmap(&info); + + if ( !auto_translated_physmap ) + + { + if ( xc_domain_get_guest_width(xch, domid, &dinfo->guest_width) != 0 ) + { + PERROR("Could not get address size for domain"); + goto out; + } + } + else + { + /* + * Autotranslated guest never sets guest width in the first + * place. Force guest_width to be sizeof(unsigned long) so + * code below functions properly. + * + * Here is why this is correct. + * + * 1. Before f969bc9fc, xc_domain_get_guest_width for HVM (x86 + * and ARM) always returned hypervisor's idea of + * sizeof(unsigned long). + * + * 2. There has never been a situation in which hypervisor's + * word width is smaller than toolstack domain's (i.e. no + * 32bit hypervisor + 64bit toolstack). + * + * Predicates in code test guest_width against toolstack + * domain's sizeof(unsigned long), so setting guest_width to + * toolstack domain's idea of sizeof(unsigned long) matches + * the original behaviour for HVM guests. + */ + dinfo->guest_width = sizeof(unsigned long); + } + + if ( domid != info.domid ) + { + PERROR("Domain %d does not exist", domid); + goto out; + } + + ctxt = calloc(sizeof(*ctxt), info.max_vcpu_id + 1); + if ( !ctxt ) + { + PERROR("Could not allocate vcpu context array"); + goto out; + } + + for ( i = 0; i <= info.max_vcpu_id; i++ ) + { + if ( xc_vcpu_getcontext(xch, domid, i, &ctxt[nr_vcpus]) == 0 ) + { + if ( xc_core_arch_context_get(&arch_ctxt, &ctxt[nr_vcpus], + xch, domid) ) + continue; + nr_vcpus++; + } + } + if ( nr_vcpus == 0 ) + { + PERROR("No VCPU context could be grabbed"); + goto out; + } + + /* obtain memory map */ + sts = xc_core_arch_memory_map_get(xch, &arch_ctxt, &info, + live_shinfo, &memory_map, + &nr_memory_map); + if ( sts != 0 ) + goto out; + + /* + * Note: this is the *current* number of pages and may change under + * a live dump-core. We'll just take this value, and if more pages + * exist, we'll skip them. If there's less, then we'll just not use + * all the array... + * + * We don't want to use the total potential size of the memory map + * since that is usually much higher than info.nr_pages. + */ + nr_pages = info.nr_pages; + + if ( !auto_translated_physmap ) + { + /* obtain p2m table */ + p2m_array = malloc(nr_pages * sizeof(p2m_array[0])); + if ( p2m_array == NULL ) + { + PERROR("Could not allocate p2m array"); + goto out; + } + + sts = xc_core_arch_map_p2m(xch, dinfo, &info, live_shinfo, &p2m); + if ( sts != 0 ) + goto out; + + sts = xc_maximum_ram_page(xch, &max_mfn); + if ( sts != 0 ) + goto out; + } + else + { + pfn_array = malloc(nr_pages * sizeof(pfn_array[0])); + if ( pfn_array == NULL ) + { + PERROR("Could not allocate pfn array"); + goto out; + } + } + + /* ehdr.e_shnum and ehdr.e_shstrndx aren't known here yet. fill it later*/ + xc_core_ehdr_init(&ehdr); + + /* create section header */ + strtab = xc_core_strtab_init(xch); + if ( strtab == NULL ) + { + PERROR("Could not allocate string table"); + goto out; + } + sheaders = xc_core_shdr_init(xch); + if ( sheaders == NULL ) + { + PERROR("Could not allocate section headers"); + goto out; + } + /* null section */ + shdr = xc_core_shdr_get(xch,sheaders); + if ( shdr == NULL ) + { + PERROR("Could not get section header for null section"); + goto out; + } + + /* .shstrtab */ + shdr = xc_core_shdr_get(xch,sheaders); + if ( shdr == NULL ) + { + PERROR("Could not get section header for shstrtab"); + goto out; + } + strtab_idx = shdr - sheaders->shdrs; + /* strtab_shdr.sh_offset, strtab_shdr.sh_size aren't unknown. + * fill it later + */ + sts = xc_core_shdr_set(xch, shdr, strtab, ELF_SHSTRTAB, SHT_STRTAB, 0, 0, 0, 0); + if ( sts != 0 ) + goto out; + + /* elf note section */ + /* here the number of section header is unknown. fix up offset later. */ + offset = sizeof(ehdr); + filesz = + sizeof(struct xen_dumpcore_elfnote_none) + /* none */ + sizeof(struct xen_dumpcore_elfnote_header) + /* core header */ + sizeof(struct xen_dumpcore_elfnote_xen_version) + /* xen version */ + sizeof(struct xen_dumpcore_elfnote_format_version);/* format version */ + shdr = xc_core_shdr_get(xch,sheaders); + if ( shdr == NULL ) + { + PERROR("Could not get section header for note section"); + goto out; + } + sts = xc_core_shdr_set(xch, shdr, strtab, XEN_DUMPCORE_SEC_NOTE, SHT_NOTE, + offset, filesz, 0, 0); + if ( sts != 0 ) + goto out; + offset += filesz; + + /* prstatus */ + shdr = xc_core_shdr_get(xch,sheaders); + if ( shdr == NULL ) + { + PERROR("Could not get section header for .xen_prstatus"); + goto out; + } + filesz = sizeof(*ctxt) * nr_vcpus; + sts = xc_core_shdr_set(xch, shdr, strtab, XEN_DUMPCORE_SEC_PRSTATUS, + SHT_PROGBITS, offset, filesz, + __alignof__(*ctxt), sizeof(*ctxt)); + if ( sts != 0 ) + goto out; + offset += filesz; + + /* arch context */ + sts = xc_core_arch_context_get_shdr(xch, &arch_ctxt, sheaders, strtab, + &filesz, offset); + if ( sts != 0 ) + goto out; + offset += filesz; + + /* shared_info */ + if ( live_shinfo != NULL ) + { + shdr = xc_core_shdr_get(xch,sheaders); + if ( shdr == NULL ) + { + PERROR("Could not get section header for .xen_shared_info"); + goto out; + } + filesz = PAGE_SIZE; + sts = xc_core_shdr_set(xch, shdr, strtab, XEN_DUMPCORE_SEC_SHARED_INFO, + SHT_PROGBITS, offset, filesz, + __alignof__(*live_shinfo), PAGE_SIZE); + if ( sts != 0 ) + goto out; + offset += filesz; + } + + /* + * pages and p2m/pfn are the last section to allocate section headers + * so that we know the number of section headers here. + * 2 = pages section and p2m/pfn table section + */ + fixup = (sheaders->num + 2) * sizeof(*shdr); + /* zeroth section should have zero offset */ + for ( i = 1; i < sheaders->num; i++ ) + sheaders->shdrs[i].sh_offset += fixup; + offset += fixup; + dummy_len = ROUNDUP(offset, PAGE_SHIFT) - offset; /* padding length */ + offset += dummy_len; + + /* pages */ + shdr = xc_core_shdr_get(xch,sheaders); + if ( shdr == NULL ) + { + PERROR("could not get section headers for .xen_pages"); + goto out; + } + filesz = (uint64_t)nr_pages * PAGE_SIZE; + sts = xc_core_shdr_set(xch, shdr, strtab, XEN_DUMPCORE_SEC_PAGES, SHT_PROGBITS, + offset, filesz, PAGE_SIZE, PAGE_SIZE); + if ( sts != 0 ) + goto out; + offset += filesz; + + /* p2m/pfn table */ + shdr = xc_core_shdr_get(xch,sheaders); + if ( shdr == NULL ) + { + PERROR("Could not get section header for .xen_{p2m, pfn} table"); + goto out; + } + if ( !auto_translated_physmap ) + { + filesz = (uint64_t)nr_pages * sizeof(p2m_array[0]); + sts = xc_core_shdr_set(xch, shdr, strtab, XEN_DUMPCORE_SEC_P2M, + SHT_PROGBITS, + offset, filesz, __alignof__(p2m_array[0]), + sizeof(p2m_array[0])); + } + else + { + filesz = (uint64_t)nr_pages * sizeof(pfn_array[0]); + sts = xc_core_shdr_set(xch, shdr, strtab, XEN_DUMPCORE_SEC_PFN, + SHT_PROGBITS, + offset, filesz, __alignof__(pfn_array[0]), + sizeof(pfn_array[0])); + } + if ( sts != 0 ) + goto out; + offset += filesz; + + /* fixing up section header string table section header */ + filesz = strtab->length; + sheaders->shdrs[strtab_idx].sh_offset = offset; + sheaders->shdrs[strtab_idx].sh_size = filesz; + + /* write out elf header */ + ehdr.e_shnum = sheaders->num; + ehdr.e_shstrndx = strtab_idx; + ehdr.e_machine = ELF_ARCH_MACHINE; + sts = dump_rtn(xch, args, (char*)&ehdr, sizeof(ehdr)); + if ( sts != 0 ) + goto out; + + /* section headers */ + sts = dump_rtn(xch, args, (char*)sheaders->shdrs, + sheaders->num * sizeof(sheaders->shdrs[0])); + if ( sts != 0 ) + goto out; + + /* elf note section: xen core header */ + sts = elfnote_dump_none(xch, args, dump_rtn); + if ( sts != 0 ) + goto out; + + /* elf note section: xen core header */ + sts = elfnote_dump_core_header(xch, args, dump_rtn, &info, nr_vcpus, nr_pages); + if ( sts != 0 ) + goto out; + + /* elf note section: xen version */ + sts = elfnote_dump_xen_version(xch, args, dump_rtn, dinfo->guest_width); + if ( sts != 0 ) + goto out; + + /* elf note section: format version */ + sts = elfnote_dump_format_version(xch, args, dump_rtn); + if ( sts != 0 ) + goto out; + + /* prstatus: .xen_prstatus */ + sts = dump_rtn(xch, args, (char *)ctxt, sizeof(*ctxt) * nr_vcpus); + if ( sts != 0 ) + goto out; + + if ( live_shinfo != NULL ) + { + /* shared_info: .xen_shared_info */ + sts = dump_rtn(xch, args, (char*)live_shinfo, PAGE_SIZE); + if ( sts != 0 ) + goto out; + } + + /* arch specific context */ + sts = xc_core_arch_context_dump(xch, &arch_ctxt, args, dump_rtn); + if ( sts != 0 ) + goto out; + + /* Pad the output data to page alignment. */ + memset(dummy, 0, PAGE_SIZE); + sts = dump_rtn(xch, args, dummy, dummy_len); + if ( sts != 0 ) + goto out; + + /* dump pages: .xen_pages */ + j = 0; + dump_mem = dump_mem_start; + for ( map_idx = 0; map_idx < nr_memory_map; map_idx++ ) + { + uint64_t pfn_start; + uint64_t pfn_end; + + pfn_start = memory_map[map_idx].addr >> PAGE_SHIFT; + pfn_end = pfn_start + (memory_map[map_idx].size >> PAGE_SHIFT); + for ( i = pfn_start; i < pfn_end; i++ ) + { + uint64_t gmfn; + void *vaddr; + + if ( !auto_translated_physmap ) + { + if ( i >= dinfo->p2m_size ) + break; + + if ( dinfo->guest_width >= sizeof(unsigned long) ) + { + if ( dinfo->guest_width == sizeof(unsigned long) ) + gmfn = p2m[i]; + else + gmfn = ((uint64_t *)p2m)[i]; + if ( gmfn == INVALID_PFN ) + continue; + } + else + { + gmfn = ((uint32_t *)p2m)[i]; + if ( gmfn == (uint32_t)INVALID_PFN ) + continue; + } + if ( gmfn > max_mfn ) + continue; + + if ( j >= nr_pages ) + { + j++; + continue; + } + + p2m_array[j].pfn = i; + p2m_array[j].gmfn = gmfn; + } + else + { + if ( !xc_core_arch_gpfn_may_present(&arch_ctxt, i) ) + continue; + + if ( j >= nr_pages ) + { + j++; + continue; + } + + gmfn = i; + pfn_array[j] = i; + } + + vaddr = xc_map_foreign_range( + xch, domid, PAGE_SIZE, PROT_READ, gmfn); + if ( vaddr == NULL ) + continue; + memcpy(dump_mem, vaddr, PAGE_SIZE); + munmap(vaddr, PAGE_SIZE); + dump_mem += PAGE_SIZE; + if ( (j + 1) % DUMP_INCREMENT == 0 ) + { + sts = dump_rtn( + xch, args, dump_mem_start, dump_mem - dump_mem_start); + if ( sts != 0 ) + goto out; + dump_mem = dump_mem_start; + } + + j++; + } + } + + if ( j > nr_pages ) + { + /* + * When live dump-mode (-L option) is specified, + * guest domain may increase memory. + */ + IPRINTF("exceeded nr_pages (%ld) losing %ld pages", nr_pages, j - nr_pages); + } + + sts = dump_rtn(xch, args, dump_mem_start, dump_mem - dump_mem_start); + if ( sts != 0 ) + goto out; + if ( j < nr_pages ) + { + /* When live dump-mode (-L option) is specified, + * guest domain may reduce memory. pad with zero pages. + */ + DPRINTF("j (%ld) != nr_pages (%ld)", j, nr_pages); + memset(dump_mem_start, 0, PAGE_SIZE); + for (; j < nr_pages; j++) { + sts = dump_rtn(xch, args, dump_mem_start, PAGE_SIZE); + if ( sts != 0 ) + goto out; + if ( !auto_translated_physmap ) + { + p2m_array[j].pfn = XC_CORE_INVALID_PFN; + p2m_array[j].gmfn = XC_CORE_INVALID_GMFN; + } + else + pfn_array[j] = XC_CORE_INVALID_PFN; + } + } + + /* p2m/pfn table: .xen_p2m/.xen_pfn */ + if ( !auto_translated_physmap ) + sts = dump_rtn( + xch, args, (char *)p2m_array, sizeof(p2m_array[0]) * nr_pages); + else + sts = dump_rtn( + xch, args, (char *)pfn_array, sizeof(pfn_array[0]) * nr_pages); + if ( sts != 0 ) + goto out; + + /* elf section header string table: .shstrtab */ + sts = dump_rtn(xch, args, strtab->strings, strtab->length); + if ( sts != 0 ) + goto out; + + sts = 0; + +out: + if ( memory_map != NULL ) + free(memory_map); + if ( p2m != NULL ) + munmap(p2m, PAGE_SIZE * dinfo->p2m_frames); + if ( p2m_array != NULL ) + free(p2m_array); + if ( pfn_array != NULL ) + free(pfn_array); + if ( sheaders != NULL ) + xc_core_shdr_free(sheaders); + if ( strtab != NULL ) + xc_core_strtab_free(strtab); + if ( ctxt != NULL ) + free(ctxt); + if ( dump_mem_start != NULL ) + free(dump_mem_start); + if ( live_shinfo != NULL ) + munmap(live_shinfo, PAGE_SIZE); + xc_core_arch_context_free(&arch_ctxt); + + return sts; +} + +/* Callback args for writing to a local dump file. */ +struct dump_args { + int fd; +}; + +/* Callback routine for writing to a local dump file. */ +static int local_file_dump(xc_interface *xch, + void *args, char *buffer, unsigned int length) +{ + struct dump_args *da = args; + + if ( write_exact(da->fd, buffer, length) == -1 ) + { + PERROR("Failed to write buffer"); + return -errno; + } + + if ( length >= (DUMP_INCREMENT * PAGE_SIZE) ) + { + // Now dumping pages -- make sure we discard clean pages from + // the cache after each write + discard_file_cache(xch, da->fd, 0 /* no flush */); + } + + return 0; +} + +int +xc_domain_dumpcore(xc_interface *xch, + uint32_t domid, + const char *corename) +{ + struct dump_args da; + int sts; + + if ( (da.fd = open(corename, O_CREAT|O_RDWR|O_TRUNC, S_IWUSR|S_IRUSR)) < 0 ) + { + PERROR("Could not open corefile %s", corename); + return -errno; + } + + sts = xc_domain_dumpcore_via_callback( + xch, domid, &da, &local_file_dump); + + /* flush and discard any remaining portion of the file from cache */ + discard_file_cache(xch, da.fd, 1/* flush first*/); + + close(da.fd); + + return sts; +} + +/* + * Local variables: From xen-changelog-bounces@lists.xenproject.org Fri Jun 04 18:11:56 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Fri, 04 Jun 2021 18:11:56 +0000 Received: from list by lists.xenproject.org with outflank-mailman.136985.253822 (Exim 4.92) (envelope-from ) id 1lpEIW-0002Zi-LC; Fri, 04 Jun 2021 18:11:56 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 136985.253822; Fri, 04 Jun 2021 18:11:56 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpEIW-0002Za-IH; Fri, 04 Jun 2021 18:11:56 +0000 Received: by outflank-mailman (input) for mailman id 136985; Fri, 04 Jun 2021 18:11:55 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpEIV-0002ZS-3h for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 18:11:55 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpEIV-0001vg-2x for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 18:11:55 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lpEIV-00082m-20 for xen-changelog@lists.xenproject.org; Fri, 04 Jun 2021 18:11:55 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=LdABfB5nYe0WomB2bzzFF9VeIZO/Gc4wUwiyzeht2Z0=; b=SB1dj+nSwd9MEY4Yvb/vyCT0nN AfjfAYPkD+hAbaRAVkhIuTFFmKMcKAy4CNfeg6+B70NFsuu2mOnLh4haNj9HR7On7tP+VNNjTCzAY fXsnkS5QuoCP6w6WXPmsBEQoDce+WjeJWXlMuHfx03TgLIZ87Jm+D2Rqc6SWsulXRyAg=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] tools/libs/guest: make some definitions private to libxenguest Message-Id: Date: Fri, 04 Jun 2021 18:11:55 +0000 commit 75f13e9b221e2c8603f15ee1d53318526cf56113 Author: Juergen Gross AuthorDate: Fri Jun 4 08:02:14 2021 +0200 Commit: Julien Grall CommitDate: Fri Jun 4 18:56:46 2021 +0100 tools/libs/guest: make some definitions private to libxenguest There are some definitions which are used in libxenguest only now. Move them from libxenctrl over to libxenguest. Remove an unused macro. Signed-off-by: Juergen Gross Acked-by: Wei Liu --- tools/libs/ctrl/xc_private.h | 32 -------------------------------- tools/libs/guest/xg_core.h | 2 +- tools/libs/guest/xg_core_x86.c | 16 +++++++++++++++- tools/libs/guest/xg_private.h | 16 +++++++++++++++- tools/libs/guest/xg_resume.c | 2 +- 5 files changed, 32 insertions(+), 36 deletions(-) diff --git a/tools/libs/ctrl/xc_private.h b/tools/libs/ctrl/xc_private.h index dff0f0289b..3e299b943f 100644 --- a/tools/libs/ctrl/xc_private.h +++ b/tools/libs/ctrl/xc_private.h @@ -65,38 +65,6 @@ struct iovec { #define ROUNDUP(_x,_w) (((unsigned long)(_x)+(1UL<<(_w))-1) & ~((1UL<<(_w))-1)) -#define GET_FIELD(_p, _f, _w) (((_w) == 8) ? ((_p)->x64._f) : ((_p)->x32._f)) - -#define SET_FIELD(_p, _f, _v, _w) do { \ - if ((_w) == 8) \ - (_p)->x64._f = (_v); \ - else \ - (_p)->x32._f = (_v); \ -} while (0) - -/* XXX SMH: following skanky macros rely on variable p2m_size being set */ -/* XXX TJD: also, "guest_width" should be the guest's sizeof(unsigned long) */ - -struct domain_info_context { - unsigned int guest_width; - unsigned int p2m_frames; - unsigned long p2m_size; -}; - -/* Number of xen_pfn_t in a page */ -#define FPP (PAGE_SIZE/(dinfo->guest_width)) - -/* Number of entries in the pfn_to_mfn_frame_list_list */ -#define P2M_FLL_ENTRIES (((dinfo->p2m_size)+(FPP*FPP)-1)/(FPP*FPP)) - -/* Number of entries in the pfn_to_mfn_frame_list */ -#define P2M_FL_ENTRIES (((dinfo->p2m_size)+FPP-1)/FPP) - -/* Size in bytes of the pfn_to_mfn_frame_list */ -#define P2M_GUEST_FL_SIZE ((P2M_FL_ENTRIES) * (dinfo->guest_width)) -#define P2M_TOOLS_FL_SIZE ((P2M_FL_ENTRIES) * \ - max_t(size_t, sizeof(xen_pfn_t), dinfo->guest_width)) - #define DECLARE_DOMCTL struct xen_domctl domctl #define DECLARE_SYSCTL struct xen_sysctl sysctl #define DECLARE_PHYSDEV_OP struct physdev_op physdev_op diff --git a/tools/libs/guest/xg_core.h b/tools/libs/guest/xg_core.h index f07584aaa6..aaca9e0a8b 100644 --- a/tools/libs/guest/xg_core.h +++ b/tools/libs/guest/xg_core.h @@ -21,7 +21,7 @@ #define XC_CORE_H #include "xen/version.h" -#include "xc_private.h" +#include "xg_private.h" #include "xen/libelf/elfstructs.h" /* section names */ diff --git a/tools/libs/guest/xg_core_x86.c b/tools/libs/guest/xg_core_x86.c index 09f5d696ce..61106b98b8 100644 --- a/tools/libs/guest/xg_core_x86.c +++ b/tools/libs/guest/xg_core_x86.c @@ -18,10 +18,24 @@ */ #include -#include "xc_private.h" +#include "xg_private.h" #include "xg_core.h" #include +/* Number of xen_pfn_t in a page */ +#define FPP (PAGE_SIZE/(dinfo->guest_width)) + +/* Number of entries in the pfn_to_mfn_frame_list_list */ +#define P2M_FLL_ENTRIES (((dinfo->p2m_size)+(FPP*FPP)-1)/(FPP*FPP)) + +/* Number of entries in the pfn_to_mfn_frame_list */ +#define P2M_FL_ENTRIES (((dinfo->p2m_size)+FPP-1)/FPP) + +/* Size in bytes of the pfn_to_mfn_frame_list */ +#define P2M_GUEST_FL_SIZE ((P2M_FL_ENTRIES) * (dinfo->guest_width)) +#define P2M_TOOLS_FL_SIZE ((P2M_FL_ENTRIES) * \ + max_t(size_t, sizeof(xen_pfn_t), dinfo->guest_width)) + int xc_core_arch_gpfn_may_present(struct xc_core_arch_context *arch_ctxt, unsigned long pfn) diff --git a/tools/libs/guest/xg_private.h b/tools/libs/guest/xg_private.h index 25e46d7ce1..03d765da21 100644 --- a/tools/libs/guest/xg_private.h +++ b/tools/libs/guest/xg_private.h @@ -42,6 +42,21 @@ #endif #endif +#define GET_FIELD(_p, _f, _w) (((_w) == 8) ? ((_p)->x64._f) : ((_p)->x32._f)) + +#define SET_FIELD(_p, _f, _v, _w) do { \ + if ((_w) == 8) \ + (_p)->x64._f = (_v); \ + else \ + (_p)->x32._f = (_v); \ +} while (0) + +struct domain_info_context { + unsigned int guest_width; + unsigned int p2m_frames; + unsigned long p2m_size; +}; + struct xc_dom_loader { const char *name; /* Sadly the error returns from these functions are not consistent: */ @@ -139,7 +154,6 @@ static inline xen_pfn_t xc_pfn_to_mfn(xen_pfn_t pfn, xen_pfn_t *p2m, /* Masks for PTE<->PFN conversions */ #define MADDR_BITS_X86 ((dinfo->guest_width == 8) ? 52 : 44) #define MFN_MASK_X86 ((1ULL << (MADDR_BITS_X86 - PAGE_SHIFT_X86)) - 1) -#define MADDR_MASK_X86 (MFN_MASK_X86 << PAGE_SHIFT_X86) int pin_table(xc_interface *xch, unsigned int type, unsigned long mfn, uint32_t dom); diff --git a/tools/libs/guest/xg_resume.c b/tools/libs/guest/xg_resume.c index d201c1488d..77e2451a3c 100644 --- a/tools/libs/guest/xg_resume.c +++ b/tools/libs/guest/xg_resume.c @@ -13,7 +13,7 @@ * License along with this library; If not, see . */ -#include "xc_private.h" +#include "xg_private.h" #include "xenguest.h" #if defined(__i386__) || defined(__x86_64__) -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Sat Jun 05 02:22:09 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Sat, 05 Jun 2021 02:22:09 +0000 Received: from list by lists.xenproject.org with outflank-mailman.137097.253994 (Exim 4.92) (envelope-from ) id 1lpLwq-0007GG-6G; Sat, 05 Jun 2021 02:22:04 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 137097.253994; Sat, 05 Jun 2021 02:22:04 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpLwq-0007G7-2w; Sat, 05 Jun 2021 02:22:04 +0000 Received: by outflank-mailman (input) for mailman id 137097; Sat, 05 Jun 2021 02:22:02 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpLwo-0007G1-HY for xen-changelog@lists.xenproject.org; Sat, 05 Jun 2021 02:22:02 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpLwo-00085Z-Gi for xen-changelog@lists.xenproject.org; Sat, 05 Jun 2021 02:22:02 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lpLwo-0006ff-FQ for xen-changelog@lists.xenproject.org; Sat, 05 Jun 2021 02:22:02 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=cTX/qPGUQqGPBPji6/a9p9K92Q/PkXVjQ1uFVCJEswg=; b=5SfeYztALiVpd0df7UxsCPSJ5+ 9+4OsPr/qlzGGymwtVcMlMphtyGmMNtIgcmYqR1RN2ieUioxh3MWWoj+qf+scl7LmSx1g6hDC68DJ jAmK4FyIrcWskyidlo+Tgs0YJBerYUPcXucjszgInsBFCOFCRxLZxTQJiCnAKa7MuX58=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.15] VT-d: Don't assume register-based invalidation is always supported Message-Id: Date: Sat, 05 Jun 2021 02:22:02 +0000 commit 78a7c3b1f2e80be44cfcb27c102edfc8e7741555 Author: Chao Gao AuthorDate: Fri Jun 4 14:43:42 2021 +0200 Commit: Jan Beulich CommitDate: Fri Jun 4 14:43:42 2021 +0200 VT-d: Don't assume register-based invalidation is always supported According to Intel VT-d SPEC rev3.3 Section 6.5, Register-based Invalidation isn't supported by Intel VT-d version 6 and beyond. This hardware change impacts following two scenarios: admin can disable queued invalidation via 'qinval' cmdline and use register-based interface; VT-d switches to register-based invalidation when queued invalidation needs to be disabled, for example, during disabling x2apic or during system suspension or after enabling queued invalidation fails. To deal with this hardware change, if register-based invalidation isn't supported, queued invalidation cannot be disabled through Xen cmdline; and if queued invalidation has to be disabled temporarily in some scenarios, VT-d won't switch to register-based interface but use some dummy functions to catch errors in case there is any invalidation request issued when queued invalidation is disabled. Signed-off-by: Chao Gao Reviewed-by: Jan Beulich Reviewed-by: Kevin Tian master commit: 6773b1a7584a75a486e9774541ad5bd84c9aa5ee master date: 2021-04-26 10:16:50 +0200 --- docs/misc/xen-command-line.pandoc | 4 +++- xen/drivers/passthrough/vtd/iommu.c | 27 +++++++++++++++++++++++++-- xen/drivers/passthrough/vtd/iommu.h | 7 +++++++ xen/drivers/passthrough/vtd/qinval.c | 33 +++++++++++++++++++++++++++++++-- 4 files changed, 66 insertions(+), 5 deletions(-) diff --git a/docs/misc/xen-command-line.pandoc b/docs/misc/xen-command-line.pandoc index a4bd3f12c5..c32a397a12 100644 --- a/docs/misc/xen-command-line.pandoc +++ b/docs/misc/xen-command-line.pandoc @@ -1458,7 +1458,9 @@ The following options are specific to Intel VT-d hardware: * The `qinval` boolean controls the Queued Invalidation sub-feature, and is active by default on compatible hardware. Queued Invalidation is a feature in second-generation IOMMUs and is a functional prerequisite for - Interrupt Remapping. + Interrupt Remapping. Note that Xen disregards this setting for Intel VT-d + version 6 and greater as Registered-Based Invalidation isn't supported + by them. * The `igfx` boolean is active by default, and controls whether the IOMMU in front of an Intel Graphics Device is enabled or not. diff --git a/xen/drivers/passthrough/vtd/iommu.c b/xen/drivers/passthrough/vtd/iommu.c index 6428c8fe3e..d14fb43182 100644 --- a/xen/drivers/passthrough/vtd/iommu.c +++ b/xen/drivers/passthrough/vtd/iommu.c @@ -1193,6 +1193,14 @@ int __init iommu_alloc(struct acpi_drhd_unit *drhd) iommu->cap = dmar_readq(iommu->reg, DMAR_CAP_REG); iommu->ecap = dmar_readq(iommu->reg, DMAR_ECAP_REG); + iommu->version = dmar_readl(iommu->reg, DMAR_VER_REG); + + if ( !iommu_qinval && !has_register_based_invalidation(iommu) ) + { + printk(XENLOG_WARNING VTDPREFIX "IOMMU %d: cannot disable Queued Invalidation\n", + iommu->index); + iommu_qinval = true; + } if ( iommu_verbose ) { @@ -2141,6 +2149,10 @@ static int __must_check init_vtd_hw(bool resume) */ if ( enable_qinval(iommu) != 0 ) { + /* Ensure register-based invalidation is available */ + if ( !has_register_based_invalidation(iommu) ) + return -EIO; + iommu->flush.context = vtd_flush_context_reg; iommu->flush.iotlb = vtd_flush_iotlb_reg; } @@ -2231,6 +2243,7 @@ static int __init vtd_setup(void) struct acpi_drhd_unit *drhd; struct vtd_iommu *iommu; int ret; + bool reg_inval_supported = true; if ( list_empty(&acpi_drhd_units) ) { @@ -2252,8 +2265,8 @@ static int __init vtd_setup(void) } /* We enable the following features only if they are supported by all VT-d - * engines: Snoop Control, DMA passthrough, Queued Invalidation, Interrupt - * Remapping, and Posted Interrupt + * engines: Snoop Control, DMA passthrough, Register-based Invalidation, + * Queued Invalidation, Interrupt Remapping, and Posted Interrupt. */ for_each_drhd_unit ( drhd ) { @@ -2275,6 +2288,9 @@ static int __init vtd_setup(void) if ( iommu_qinval && !ecap_queued_inval(iommu->ecap) ) iommu_qinval = 0; + if ( !has_register_based_invalidation(iommu) ) + reg_inval_supported = false; + if ( iommu_intremap && !ecap_intr_remap(iommu->ecap) ) iommu_intremap = iommu_intremap_off; @@ -2301,6 +2317,13 @@ static int __init vtd_setup(void) softirq_tasklet_init(&vtd_fault_tasklet, do_iommu_page_fault, NULL); + if ( !iommu_qinval && !reg_inval_supported ) + { + dprintk(XENLOG_ERR VTDPREFIX, "No available invalidation interface\n"); + ret = -ENODEV; + goto error; + } + if ( !iommu_qinval && iommu_intremap ) { iommu_intremap = iommu_intremap_off; diff --git a/xen/drivers/passthrough/vtd/iommu.h b/xen/drivers/passthrough/vtd/iommu.h index 216791b3d6..57737b1a4a 100644 --- a/xen/drivers/passthrough/vtd/iommu.h +++ b/xen/drivers/passthrough/vtd/iommu.h @@ -540,6 +540,7 @@ struct vtd_iommu { struct list_head ats_devices; unsigned long *domid_bitmap; /* domain id bitmap */ u16 *domid_map; /* domain id mapping array */ + uint32_t version; }; #define INTEL_IOMMU_DEBUG(fmt, args...) \ @@ -549,4 +550,10 @@ struct vtd_iommu { dprintk(XENLOG_WARNING VTDPREFIX, fmt, ## args); \ } while(0) +/* Register-based invalidation isn't supported by VT-d version 6 and beyond. */ +static inline bool has_register_based_invalidation(const struct vtd_iommu *vtd) +{ + return VER_MAJOR(vtd->version) < 6; +} + #endif diff --git a/xen/drivers/passthrough/vtd/qinval.c b/xen/drivers/passthrough/vtd/qinval.c index 764ef9f0fc..dfcbb66c8a 100644 --- a/xen/drivers/passthrough/vtd/qinval.c +++ b/xen/drivers/passthrough/vtd/qinval.c @@ -442,6 +442,23 @@ int enable_qinval(struct vtd_iommu *iommu) return 0; } +static int vtd_flush_context_noop(struct vtd_iommu *iommu, uint16_t did, + uint16_t source_id, uint8_t function_mask, + uint64_t type, bool flush_non_present_entry) +{ + WARN(); + return -EIO; +} + +static int vtd_flush_iotlb_noop(struct vtd_iommu *iommu, uint16_t did, + uint64_t addr, unsigned int size_order, + uint64_t type, bool flush_non_present_entry, + bool flush_dev_iotlb) +{ + WARN(); + return -EIO; +} + void disable_qinval(struct vtd_iommu *iommu) { u32 sts; @@ -463,6 +480,18 @@ void disable_qinval(struct vtd_iommu *iommu) out: spin_unlock_irqrestore(&iommu->register_lock, flags); - iommu->flush.context = vtd_flush_context_reg; - iommu->flush.iotlb = vtd_flush_iotlb_reg; + /* + * Assign callbacks to noop to catch errors if register-based invalidation + * isn't supported. + */ + if ( has_register_based_invalidation(iommu) ) + { + iommu->flush.context = vtd_flush_context_reg; + iommu->flush.iotlb = vtd_flush_iotlb_reg; + } + else + { + iommu->flush.context = vtd_flush_context_noop; + iommu->flush.iotlb = vtd_flush_iotlb_noop; + } } -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.15 From xen-changelog-bounces@lists.xenproject.org Sat Jun 05 02:22:14 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Sat, 05 Jun 2021 02:22:14 +0000 Received: from list by lists.xenproject.org with outflank-mailman.137098.253998 (Exim 4.92) (envelope-from ) id 1lpLx0-0007IW-7R; Sat, 05 Jun 2021 02:22:14 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 137098.253998; Sat, 05 Jun 2021 02:22:14 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpLx0-0007IO-4P; Sat, 05 Jun 2021 02:22:14 +0000 Received: by outflank-mailman (input) for mailman id 137098; Sat, 05 Jun 2021 02:22:12 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpLwy-0007I7-LZ for xen-changelog@lists.xenproject.org; Sat, 05 Jun 2021 02:22:12 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpLwy-00085d-Kh for xen-changelog@lists.xenproject.org; Sat, 05 Jun 2021 02:22:12 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lpLwy-0006gg-JP for xen-changelog@lists.xenproject.org; Sat, 05 Jun 2021 02:22:12 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=E7icLi5Kmb1D3+B8emgrd9ZE+VMSqUrK5XjMKvOInEg=; b=CuknbGkgeduDQryM/M/QvEROFh AKEzsDywjMhtbqL4UmotO36bgkSuGW8ua1EbJCl54ioQz384z7vPNGq82glNizyycUmE1TG/a9X8i /lTh/qSFMHHLk6L1VwWRjEjSpfo5/pkpfL+/3U0TmUpgdHy+vdzUMGOsVIQq9Tg5Z3Rk=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.15] x86/vtx: add LBR_SELECT to the list of LBR MSRs Message-Id: Date: Sat, 05 Jun 2021 02:22:12 +0000 commit 3e6c1b6cccd8c7bc63871836c14ac104d2a6e86a Author: Igor Druzhinin AuthorDate: Fri Jun 4 14:45:00 2021 +0200 Commit: Jan Beulich CommitDate: Fri Jun 4 14:45:00 2021 +0200 x86/vtx: add LBR_SELECT to the list of LBR MSRs This MSR exists since Nehalem / Silvermont and is actively used by Linux, for instance, to improve sampling efficiency. Signed-off-by: Igor Druzhinin Reviewed-by: Jan Beulich Reviewed-by: Kevin Tian master commit: 08693c03e00ea3448adc4406c891e707f0068eb6 master date: 2021-04-26 10:22:04 +0200 --- xen/arch/x86/hvm/vmx/vmx.c | 20 ++++++++++++++++---- xen/include/asm-x86/msr-index.h | 10 ++++++++-- 2 files changed, 24 insertions(+), 6 deletions(-) diff --git a/xen/arch/x86/hvm/vmx/vmx.c b/xen/arch/x86/hvm/vmx/vmx.c index b52824677e..8e0aa976c2 100644 --- a/xen/arch/x86/hvm/vmx/vmx.c +++ b/xen/arch/x86/hvm/vmx/vmx.c @@ -2898,14 +2898,16 @@ static const struct lbr_info { }, nh_lbr[] = { { MSR_IA32_LASTINTFROMIP, 1 }, { MSR_IA32_LASTINTTOIP, 1 }, - { MSR_C2_LASTBRANCH_TOS, 1 }, + { MSR_NHL_LBR_SELECT, 1 }, + { MSR_NHL_LASTBRANCH_TOS, 1 }, { MSR_P4_LASTBRANCH_0_FROM_LIP, NUM_MSR_P4_LASTBRANCH_FROM_TO }, { MSR_P4_LASTBRANCH_0_TO_LIP, NUM_MSR_P4_LASTBRANCH_FROM_TO }, { 0, 0 } }, sk_lbr[] = { { MSR_IA32_LASTINTFROMIP, 1 }, { MSR_IA32_LASTINTTOIP, 1 }, - { MSR_SKL_LASTBRANCH_TOS, 1 }, + { MSR_NHL_LBR_SELECT, 1 }, + { MSR_NHL_LASTBRANCH_TOS, 1 }, { MSR_SKL_LASTBRANCH_0_FROM_IP, NUM_MSR_SKL_LASTBRANCH }, { MSR_SKL_LASTBRANCH_0_TO_IP, NUM_MSR_SKL_LASTBRANCH }, { MSR_SKL_LASTBRANCH_0_INFO, NUM_MSR_SKL_LASTBRANCH }, @@ -2917,10 +2919,19 @@ static const struct lbr_info { { MSR_C2_LASTBRANCH_0_FROM_IP, NUM_MSR_ATOM_LASTBRANCH_FROM_TO }, { MSR_C2_LASTBRANCH_0_TO_IP, NUM_MSR_ATOM_LASTBRANCH_FROM_TO }, { 0, 0 } +}, sm_lbr[] = { + { MSR_IA32_LASTINTFROMIP, 1 }, + { MSR_IA32_LASTINTTOIP, 1 }, + { MSR_SM_LBR_SELECT, 1 }, + { MSR_SM_LASTBRANCH_TOS, 1 }, + { MSR_C2_LASTBRANCH_0_FROM_IP, NUM_MSR_ATOM_LASTBRANCH_FROM_TO }, + { MSR_C2_LASTBRANCH_0_TO_IP, NUM_MSR_ATOM_LASTBRANCH_FROM_TO }, + { 0, 0 } }, gm_lbr[] = { { MSR_IA32_LASTINTFROMIP, 1 }, { MSR_IA32_LASTINTTOIP, 1 }, - { MSR_GM_LASTBRANCH_TOS, 1 }, + { MSR_SM_LBR_SELECT, 1 }, + { MSR_SM_LASTBRANCH_TOS, 1 }, { MSR_GM_LASTBRANCH_0_FROM_IP, NUM_MSR_GM_LASTBRANCH_FROM_TO }, { MSR_GM_LASTBRANCH_0_TO_IP, NUM_MSR_GM_LASTBRANCH_FROM_TO }, { 0, 0 } @@ -2974,6 +2985,7 @@ static const struct lbr_info *last_branch_msr_get(void) return sk_lbr; /* Atom */ case 0x1c: case 0x26: case 0x27: case 0x35: case 0x36: + return at_lbr; /* Silvermont */ case 0x37: case 0x4a: case 0x4d: case 0x5a: case 0x5d: /* Xeon Phi Knights Landing */ @@ -2982,7 +2994,7 @@ static const struct lbr_info *last_branch_msr_get(void) case 0x85: /* Airmont */ case 0x4c: - return at_lbr; + return sm_lbr; /* Goldmont */ case 0x5c: case 0x5f: return gm_lbr; diff --git a/xen/include/asm-x86/msr-index.h b/xen/include/asm-x86/msr-index.h index f2e34dd22b..bd3a3a1e7f 100644 --- a/xen/include/asm-x86/msr-index.h +++ b/xen/include/asm-x86/msr-index.h @@ -603,15 +603,21 @@ #define NUM_MSR_C2_LASTBRANCH_FROM_TO 4 #define NUM_MSR_ATOM_LASTBRANCH_FROM_TO 8 +/* Nehalem (and newer) last-branch recording */ +#define MSR_NHL_LBR_SELECT 0x000001c8 +#define MSR_NHL_LASTBRANCH_TOS 0x000001c9 + /* Skylake (and newer) last-branch recording */ -#define MSR_SKL_LASTBRANCH_TOS 0x000001c9 #define MSR_SKL_LASTBRANCH_0_FROM_IP 0x00000680 #define MSR_SKL_LASTBRANCH_0_TO_IP 0x000006c0 #define MSR_SKL_LASTBRANCH_0_INFO 0x00000dc0 #define NUM_MSR_SKL_LASTBRANCH 32 +/* Silvermont (and newer) last-branch recording */ +#define MSR_SM_LBR_SELECT 0x000001c8 +#define MSR_SM_LASTBRANCH_TOS 0x000001c9 + /* Goldmont last-branch recording */ -#define MSR_GM_LASTBRANCH_TOS 0x000001c9 #define MSR_GM_LASTBRANCH_0_FROM_IP 0x00000680 #define MSR_GM_LASTBRANCH_0_TO_IP 0x000006c0 #define NUM_MSR_GM_LASTBRANCH_FROM_TO 32 -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.15 From xen-changelog-bounces@lists.xenproject.org Sat Jun 05 02:22:24 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Sat, 05 Jun 2021 02:22:24 +0000 Received: from list by lists.xenproject.org with outflank-mailman.137099.254002 (Exim 4.92) (envelope-from ) id 1lpLxA-0007L7-8t; Sat, 05 Jun 2021 02:22:24 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 137099.254002; Sat, 05 Jun 2021 02:22:24 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpLxA-0007Kx-5q; Sat, 05 Jun 2021 02:22:24 +0000 Received: by outflank-mailman (input) for mailman id 137099; Sat, 05 Jun 2021 02:22:22 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpLx8-0007Kl-On for xen-changelog@lists.xenproject.org; Sat, 05 Jun 2021 02:22:22 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpLx8-00085v-O1 for xen-changelog@lists.xenproject.org; Sat, 05 Jun 2021 02:22:22 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lpLx8-0006hJ-N7 for xen-changelog@lists.xenproject.org; Sat, 05 Jun 2021 02:22:22 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=h6ake5wxum4niVqheh467GltOkkPb+SHSUx2F1d7izE=; b=RM31wUg0Z+zcZjmZB6PG6L9d2M 1PuqpBcponxaWeq3spdsJKKVI6iQzhADYessJ+FN3dIQSZfQg5QWKEhv2sND+L6sISiY8+9tihykC 4/hrZQqbHs0LCmS/flvfEnvfJlOEdqHYO6uIAlZaqKCiEVO+BC/XBDyzYDbfY0C0xw/k=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.15] x86/intel: insert Ice Lake-SP and Ice Lake-D model numbers Message-Id: Date: Sat, 05 Jun 2021 02:22:22 +0000 commit cacad0cdfcf2763e3377316e779532a4a115ae8c Author: Igor Druzhinin AuthorDate: Fri Jun 4 14:45:39 2021 +0200 Commit: Jan Beulich CommitDate: Fri Jun 4 14:45:39 2021 +0200 x86/intel: insert Ice Lake-SP and Ice Lake-D model numbers LBR, C-state MSRs should correspond to Ice Lake desktop according to SDM rev. 74 for both models. Ice Lake-SP is known to expose IF_PSCHANGE_MC_NO in IA32_ARCH_CAPABILITIES MSR (as advisory tells and Whitley SDP confirms) which means the erratum is fixed in hardware for that model and therefore it shouldn't be present in has_if_pschange_mc list. Provisionally assume the same to be the case for Ice Lake-D. Signed-off-by: Igor Druzhinin Reviewed-by: Jan Beulich Reviewed-by: Kevin Tian master commit: 95419adfd4b275cffe24b96edcc2f15bc4db8907 master date: 2021-04-26 10:22:48 +0200 --- xen/arch/x86/acpi/cpu_idle.c | 2 ++ xen/arch/x86/hvm/vmx/vmx.c | 2 +- 2 files changed, 3 insertions(+), 1 deletion(-) diff --git a/xen/arch/x86/acpi/cpu_idle.c b/xen/arch/x86/acpi/cpu_idle.c index c092086b33..d788c8bffc 100644 --- a/xen/arch/x86/acpi/cpu_idle.c +++ b/xen/arch/x86/acpi/cpu_idle.c @@ -181,6 +181,8 @@ static void do_get_hw_residencies(void *arg) case 0x55: case 0x5E: /* Ice Lake */ + case 0x6A: + case 0x6C: case 0x7D: case 0x7E: /* Tiger Lake */ diff --git a/xen/arch/x86/hvm/vmx/vmx.c b/xen/arch/x86/hvm/vmx/vmx.c index 8e0aa976c2..dde4f3b70d 100644 --- a/xen/arch/x86/hvm/vmx/vmx.c +++ b/xen/arch/x86/hvm/vmx/vmx.c @@ -2973,7 +2973,7 @@ static const struct lbr_info *last_branch_msr_get(void) /* Goldmont Plus */ case 0x7a: /* Ice Lake */ - case 0x7d: case 0x7e: + case 0x6a: case 0x6c: case 0x7d: case 0x7e: /* Tiger Lake */ case 0x8c: case 0x8d: /* Tremont */ -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.15 From xen-changelog-bounces@lists.xenproject.org Sat Jun 05 02:22:34 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Sat, 05 Jun 2021 02:22:34 +0000 Received: from list by lists.xenproject.org with outflank-mailman.137100.254005 (Exim 4.92) (envelope-from ) id 1lpLxK-0007Oa-Al; Sat, 05 Jun 2021 02:22:34 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 137100.254005; Sat, 05 Jun 2021 02:22:34 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpLxK-0007OU-7S; Sat, 05 Jun 2021 02:22:34 +0000 Received: by outflank-mailman (input) for mailman id 137100; Sat, 05 Jun 2021 02:22:32 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpLxI-0007OA-TH for xen-changelog@lists.xenproject.org; Sat, 05 Jun 2021 02:22:32 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpLxI-00086A-R2 for xen-changelog@lists.xenproject.org; Sat, 05 Jun 2021 02:22:32 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lpLxI-0006hn-Q8 for xen-changelog@lists.xenproject.org; Sat, 05 Jun 2021 02:22:32 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=CGwMdmFDGb57W1b5vxaXvqi2fkyUJHoEz1K2+cqbtvg=; b=PCLjSRXWjPO8PaEv0/gaZiNGpB qWxoYggHLHXR857VvlMyMvXzM3i8nrEXsjMoC/dq93dZ4AD34BrmOPrIbPmf9TMRY+mTC8aHRfXsW 1Olihm6TE9SvWAAraQwsw30Gsb4V+BH7rKqZosKcKDCiGLxBLTCbZRK5dAuAPcPW2vgs=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.15] x86/vhpet: fix RTC special casing Message-Id: Date: Sat, 05 Jun 2021 02:22:32 +0000 commit edeaa04ab895a7283a57532ee349df19fe34c56b Author: Roger Pau Monné AuthorDate: Fri Jun 4 14:46:09 2021 +0200 Commit: Jan Beulich CommitDate: Fri Jun 4 14:46:09 2021 +0200 x86/vhpet: fix RTC special casing Restore setting the virtual timer callback private data to NULL if the timer is not level triggered. This fixes the special casing done in pt_update_irq so that the RTC interrupt when originating from the HPET is suspended if the interrupt source is masked. Note the RTC special casing done in pt_update_irq should only apply to the RTC interrupt originating from the emulated RTC device (which does set the callback private data), as in that case the callback itself will destroy the virtual timer if the interrupt is ignored. While there also use RTC_IRQ instead of 8 when the HPET is configured in LegacyReplacement Mode. Fixes: be07023be115 ("x86/vhpet: add support for level triggered interrupts") Reported-by: Jan Beulich Signed-off-by: Roger Pau Monné Reviewed-by: Jan Beulich master commit: 472a13988a051e5ae84b95815c6caf4378062abe master date: 2021-05-07 10:43:29 +0200 --- xen/arch/x86/hvm/hpet.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/xen/arch/x86/hvm/hpet.c b/xen/arch/x86/hvm/hpet.c index ca94e8b453..ee756abb82 100644 --- a/xen/arch/x86/hvm/hpet.c +++ b/xen/arch/x86/hvm/hpet.c @@ -22,6 +22,7 @@ #include #include #include +#include #include #include #include @@ -290,7 +291,7 @@ static void hpet_set_timer(HPETState *h, unsigned int tn, /* if LegacyReplacementRoute bit is set, HPET specification requires timer0 be routed to IRQ0 in NON-APIC or IRQ2 in the I/O APIC, timer1 be routed to IRQ8 in NON-APIC or IRQ8 in the I/O APIC. */ - irq = (tn == 0) ? 0 : 8; + irq = (tn == 0) ? 0 : RTC_IRQ; h->pt[tn].source = PTSRC_isa; } else @@ -318,7 +319,8 @@ static void hpet_set_timer(HPETState *h, unsigned int tn, hpet_tick_to_ns(h, diff), oneshot ? 0 : hpet_tick_to_ns(h, h->hpet.period[tn]), irq, timer_level(h, tn) ? hpet_timer_fired : NULL, - (void *)(unsigned long)tn, timer_level(h, tn)); + timer_level(h, tn) ? (void *)(unsigned long)tn : NULL, + timer_level(h, tn)); } static inline uint64_t hpet_fixup_reg( -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.15 From xen-changelog-bounces@lists.xenproject.org Sat Jun 05 02:22:43 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Sat, 05 Jun 2021 02:22:43 +0000 Received: from list by lists.xenproject.org with outflank-mailman.137101.254009 (Exim 4.92) (envelope-from ) id 1lpLxT-0007SL-Ds; Sat, 05 Jun 2021 02:22:43 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 137101.254009; Sat, 05 Jun 2021 02:22:43 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpLxT-0007SF-Ar; Sat, 05 Jun 2021 02:22:43 +0000 Received: by outflank-mailman (input) for mailman id 137101; Sat, 05 Jun 2021 02:22:43 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpLxS-0007S1-Uf for xen-changelog@lists.xenproject.org; Sat, 05 Jun 2021 02:22:42 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpLxS-00086K-Tz for xen-changelog@lists.xenproject.org; Sat, 05 Jun 2021 02:22:42 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lpLxS-0006iQ-T5 for xen-changelog@lists.xenproject.org; Sat, 05 Jun 2021 02:22:42 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=Yp8f+Oc9JqznHQWFLW0c7d7ge71je7rk9Pf4qQaGBXY=; b=jqHIu1/XzWIl6klL7AD4jAOw+M ZbkxBONuCgymVpssBZ7ZI3ssGyK+UVt3QY1Un18/usjLKN1dPhCy/HD8xTGMDPTOVjTEXeFTyefG0 bMSU+jFYnpLgmsZuxZwygroO3WlLV2K23KXAzIwWCa2S+ljAgCk/KJTX4Huvp7n2ticw=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.15] x86emul: fix test harness build for gas 2.36 Message-Id: Date: Sat, 05 Jun 2021 02:22:42 +0000 commit ee2b1d616e8dda8edf1946c403c781983d1888b7 Author: Jan Beulich AuthorDate: Fri Jun 4 14:46:42 2021 +0200 Commit: Jan Beulich CommitDate: Fri Jun 4 14:46:42 2021 +0200 x86emul: fix test harness build for gas 2.36 All of the sudden, besides .text and .rodata and alike, an always present .note.gnu.property section has appeared. This section, when converting to binary format output, gets placed according to its linked address, causing the resulting blobs to be about 128Mb in size. The resulting headers with a C representation of the binary blobs then are, of course all a multiple of that size (and take accordingly long to create). I didn't bother waiting to see what size the final test_x86_emulator binary then would have had. See also https://sourceware.org/bugzilla/show_bug.cgi?id=27753. Rather than figuring out whether gas supports -mx86-used-note=, simply remove the section while creating *.bin. Signed-off-by: Jan Beulich master commit: aa803ba38a867551917d11059eaa044955556e05 master date: 2021-05-17 15:41:28 +0200 --- tools/tests/x86_emulator/testcase.mk | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/tools/tests/x86_emulator/testcase.mk b/tools/tests/x86_emulator/testcase.mk index dafeb6caf7..d4aa82e3f3 100644 --- a/tools/tests/x86_emulator/testcase.mk +++ b/tools/tests/x86_emulator/testcase.mk @@ -12,11 +12,11 @@ all: $(TESTCASE).bin %.bin: %.c $(CC) $(filter-out -M% .%,$(CFLAGS)) -c $< $(LD) $(LDFLAGS_DIRECT) -N -Ttext 0x100000 -o $*.tmp $*.o - $(OBJCOPY) -O binary $*.tmp $@ + $(OBJCOPY) -O binary -R .note.gnu.property $*.tmp $@ rm -f $*.tmp %-opmask.bin: opmask.S $(CC) $(filter-out -M% .%,$(CFLAGS)) -c $< -o $(basename $@).o $(LD) $(LDFLAGS_DIRECT) -N -Ttext 0x100000 -o $(basename $@).tmp $(basename $@).o - $(OBJCOPY) -O binary $(basename $@).tmp $@ + $(OBJCOPY) -O binary -R .note.gnu.property $(basename $@).tmp $@ rm -f $(basename $@).tmp -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.15 From xen-changelog-bounces@lists.xenproject.org Sat Jun 05 02:22:53 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Sat, 05 Jun 2021 02:22:53 +0000 Received: from list by lists.xenproject.org with outflank-mailman.137102.254014 (Exim 4.92) (envelope-from ) id 1lpLxd-0007VV-FP; Sat, 05 Jun 2021 02:22:53 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 137102.254014; Sat, 05 Jun 2021 02:22:53 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpLxd-0007VN-CM; Sat, 05 Jun 2021 02:22:53 +0000 Received: by outflank-mailman (input) for mailman id 137102; Sat, 05 Jun 2021 02:22:53 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpLxd-0007VD-1T for xen-changelog@lists.xenproject.org; Sat, 05 Jun 2021 02:22:53 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpLxd-00086q-0j for xen-changelog@lists.xenproject.org; Sat, 05 Jun 2021 02:22:53 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lpLxc-0006j5-W0 for xen-changelog@lists.xenproject.org; Sat, 05 Jun 2021 02:22:52 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=AfK4T2nWY3jvqt0XfRD+v6BxTO+7+q6ZcC1QGhyCXmA=; b=lbmeHruYwm9kr+2M/TA8GE+u50 eWezmcSeMENqQcrIWIlzjj92XxFl+DLxDHNe/gPWzsEZsVM+r3S3y+gZ3P7K9LYD9muqmTMpSBntL EqJh4C8Bhb6ZDiyWpCbS75wCmnwkYxZ1GwWbE9mKYLaWlHrHtIwkLN8OmjwI4i4lSHKU=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.15] firmware/shim: update linkfarm exclusions Message-Id: Date: Sat, 05 Jun 2021 02:22:52 +0000 commit 6a7e21a135148ef5bf6a62177dfbd5c104e3db13 Author: Jan Beulich AuthorDate: Fri Jun 4 14:47:14 2021 +0200 Commit: Jan Beulich CommitDate: Fri Jun 4 14:47:14 2021 +0200 firmware/shim: update linkfarm exclusions Some intermediate files weren't considered at all at the time. Also after its introduction, various changes to the build environment have rendered the exclusion sets stale. For example, we now have some .*.cmd files in the build tree. Combine all respective patterns into a single .* one, seeing that we don't have any actual source files matching this pattern in the tree. Add other patterns as well as individual files. Also introduce LINK_EXCLUDE_PATHS to deal with entire directories full of generated headers as well as a few specific files the names of which are too generic to list under LINK_EXCLUDES. Signed-off-by: Jan Beulich Acked-by: Roger Pau Monné master commit: 022da003747a7f4336f8c581fdf898a7d5cd47e4 master date: 2021-05-26 09:34:07 +0200 --- tools/firmware/xen-dir/Makefile | 26 +++++++++++++++++++------- 1 file changed, 19 insertions(+), 7 deletions(-) diff --git a/tools/firmware/xen-dir/Makefile b/tools/firmware/xen-dir/Makefile index 5413f86790..afc651f861 100644 --- a/tools/firmware/xen-dir/Makefile +++ b/tools/firmware/xen-dir/Makefile @@ -15,9 +15,19 @@ DEP_DIRS=$(foreach i, $(LINK_DIRS), $(XEN_ROOT)/$(i)) DEP_FILES=$(foreach i, $(LINK_FILES), $(XEN_ROOT)/$(i)) # Exclude some intermediate files and final build products -LINK_EXCLUDES := '*.[isoa]' '.*.d' '.*.d2' '.config' -LINK_EXCLUDES += '*.map' 'xen' 'xen.gz' 'xen.efi' 'xen-syms' -LINK_EXCLUDES += '.*.tmp' +LINK_EXCLUDES := '*.[isoa]' '*.bin' '*.chk' '*.lnk' '*.gz' '.*' +LINK_EXCLUDES += lexer.lex.? parser.tab.? conf +LINK_EXCLUDES += asm-offsets.h asm-macros.h compile.h '*-autogen.h' +LINK_EXCLUDES += mkelf32 mkreloc symbols config_data.S xen.lds efi.lds +LINK_EXCLUDES += '*.map' xen xen.gz xen.efi xen-syms check.efi + +# To exclude full subtrees or individual files of not sufficiently specific +# names, regular expressions are used: +LINK_EXCLUDE_PATHS := xen/include/compat/.* +LINK_EXCLUDE_PATHS += xen/include/config/.* +LINK_EXCLUDE_PATHS += xen/include/generated/.* +LINK_EXCLUDE_PATHS += xen/arch/x86/boot/reloc[.]S +LINK_EXCLUDE_PATHS += xen/arch/x86/boot/cmdline[.]S # This is all a giant mess and doesn't really work. # @@ -32,9 +42,10 @@ LINK_EXCLUDES += '.*.tmp' # support easy development of the shim, but has a side effect of clobbering # the already-built shim. # -# $(LINK_EXCLUDES) should be set such that a parallel build of shim and xen/ -# doesn't cause a subsequent `make install` to decide to regenerate the -# linkfarm. This means that all final build artefacts must be excluded. +# $(LINK_EXCLUDES) and $(LINK_EXCLUDE_DIRS) should be set such that a parallel +# build of shim and xen/ doesn't cause a subsequent `make install` to decide to +# regenerate the linkfarm. This means that all intermediate and final build +# artefacts must be excluded. linkfarm.stamp: $(DEP_DIRS) $(DEP_FILES) FORCE mkdir -p $(D) rm -f linkfarm.stamp.tmp @@ -46,7 +57,8 @@ linkfarm.stamp: $(DEP_DIRS) $(DEP_FILES) FORCE sed 's,^$(XEN_ROOT)/$(d)/,,g' | xargs mkdir -p .);) \ $(foreach d, $(LINK_DIRS), \ (cd $(XEN_ROOT); \ - find $(d) ! -type l -type f $(addprefix ! -name ,$(LINK_EXCLUDES))) \ + find $(d) ! -type l -type f $(addprefix ! -name ,$(LINK_EXCLUDES)) \ + | grep -v $(patsubst %,-e '^%$$',$(LINK_EXCLUDE_PATHS))) \ >> linkfarm.stamp.tmp ; ) \ $(foreach f, $(LINK_FILES), \ echo $(f) >> linkfarm.stamp.tmp ;) -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.15 From xen-changelog-bounces@lists.xenproject.org Sat Jun 05 02:23:03 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Sat, 05 Jun 2021 02:23:03 +0000 Received: from list by lists.xenproject.org with outflank-mailman.137103.254018 (Exim 4.92) (envelope-from ) id 1lpLxn-0007YF-Gm; Sat, 05 Jun 2021 02:23:03 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 137103.254018; Sat, 05 Jun 2021 02:23:03 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpLxn-0007Y7-Dx; Sat, 05 Jun 2021 02:23:03 +0000 Received: by outflank-mailman (input) for mailman id 137103; Sat, 05 Jun 2021 02:23:03 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpLxn-0007Xy-5K for xen-changelog@lists.xenproject.org; Sat, 05 Jun 2021 02:23:03 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpLxn-00087M-4V for xen-changelog@lists.xenproject.org; Sat, 05 Jun 2021 02:23:03 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lpLxn-0006jt-3Y for xen-changelog@lists.xenproject.org; Sat, 05 Jun 2021 02:23:03 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=5nYaIsl1J6o+9dsTmL5SEeKJ8u1rtJPAOj0A5wocdyY=; b=JCuPFUljLcFqNjbE84dFiFUyTu ZGuJhoVlX5qIObf27njyEg9Sb2hJe0wn+ICDWBZyYUbD/BxNxBvb/BLE6DwVmlfOqmdADqheLEmU9 uc0ksWJ2EfZzq+adqDgL3IFqKqXqIr1J5WOwkCyJl8I6lCqNwhs8rcqOC3jUHUFE0hAQ=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.15] x86: make hypervisor build with gcc11 Message-Id: Date: Sat, 05 Jun 2021 02:23:03 +0000 commit 7c3c98497f8b5e413ca1858f71f369170ead8d5b Author: Jan Beulich AuthorDate: Fri Jun 4 14:47:36 2021 +0200 Commit: Jan Beulich CommitDate: Fri Jun 4 14:47:36 2021 +0200 x86: make hypervisor build with gcc11 Gcc 11 looks to make incorrect assumptions about valid ranges that pointers may be used for addressing when they are derived from e.g. a plain constant. See https://gcc.gnu.org/bugzilla/show_bug.cgi?id=100680. Utilize RELOC_HIDE() to work around the issue, which for x86 manifests in at least - mpparse.c:efi_check_config(), - tboot.c:tboot_probe(), - tboot.c:tboot_gen_frametable_integrity(), - x86_emulate.c:x86_emulate() (at -O2 only). The last case is particularly odd not just because it only triggers at higher optimization levels, but also because it only affects one of at least three similar constructs. Various "note" diagnostics claim the valid index range to be [0, 2⁶³-1]. Signed-off-by: Jan Beulich Tested-by: Jason Andryuk Acked-by: Roger Pau Monné master commit: 722f59d38c710a940ab05e542a83020eb5546dea master date: 2021-05-27 14:40:29 +0200 --- tools/tests/x86_emulator/x86-emulate.c | 7 +++++++ xen/arch/x86/x86_emulate/x86_emulate.c | 2 +- xen/include/asm-x86/fixmap.h | 2 +- xen/include/xen/compiler.h | 6 ++++++ xen/include/xen/pdx.h | 2 +- 5 files changed, 16 insertions(+), 3 deletions(-) diff --git a/tools/tests/x86_emulator/x86-emulate.c b/tools/tests/x86_emulator/x86-emulate.c index fb34933e82..adb4f7aca4 100644 --- a/tools/tests/x86_emulator/x86-emulate.c +++ b/tools/tests/x86_emulator/x86-emulate.c @@ -8,6 +8,13 @@ #define ERR_PTR(val) NULL +/* See gcc bug 100680, but here don't bother making this version dependent. */ +#define gcc11_wrap(x) ({ \ + unsigned long x_; \ + __asm__ ( "" : "=g" (x_) : "0" (x) ); \ + (typeof(x))x_; \ +}) + #define cpu_has_amd_erratum(nr) 0 #define cpu_has_mpx false #define read_bndcfgu() 0 diff --git a/xen/arch/x86/x86_emulate/x86_emulate.c b/xen/arch/x86/x86_emulate/x86_emulate.c index 4a33fe9613..a8000cbcbd 100644 --- a/xen/arch/x86/x86_emulate/x86_emulate.c +++ b/xen/arch/x86/x86_emulate/x86_emulate.c @@ -726,7 +726,7 @@ union vex { #define copy_VEX(ptr, vex) ({ \ if ( !mode_64bit() ) \ (vex).reg |= 8; \ - (ptr)[0 - PFX_BYTES] = ext < ext_8f08 ? 0xc4 : 0x8f; \ + gcc11_wrap(ptr)[0 - PFX_BYTES] = ext < ext_8f08 ? 0xc4 : 0x8f; \ (ptr)[1 - PFX_BYTES] = (vex).raw[0]; \ (ptr)[2 - PFX_BYTES] = (vex).raw[1]; \ container_of((ptr) + 1 - PFX_BYTES, typeof(vex), raw[0]); \ diff --git a/xen/include/asm-x86/fixmap.h b/xen/include/asm-x86/fixmap.h index 0db314baeb..20746afd0a 100644 --- a/xen/include/asm-x86/fixmap.h +++ b/xen/include/asm-x86/fixmap.h @@ -78,7 +78,7 @@ extern void __set_fixmap( #define clear_fixmap(idx) __set_fixmap(idx, 0, 0) -#define __fix_to_virt(x) (FIXADDR_TOP - ((x) << PAGE_SHIFT)) +#define __fix_to_virt(x) gcc11_wrap(FIXADDR_TOP - ((x) << PAGE_SHIFT)) #define __virt_to_fix(x) ((FIXADDR_TOP - ((x)&PAGE_MASK)) >> PAGE_SHIFT) #define fix_to_virt(x) ((void *)__fix_to_virt(x)) diff --git a/xen/include/xen/compiler.h b/xen/include/xen/compiler.h index 17cf00e1ec..696c7eb89e 100644 --- a/xen/include/xen/compiler.h +++ b/xen/include/xen/compiler.h @@ -140,6 +140,12 @@ __asm__ ("" : "=r"(__ptr) : "0"(ptr)); \ (typeof(ptr)) (__ptr + (off)); }) +#if CONFIG_GCC_VERSION >= 110000 /* See gcc bug 100680. */ +# define gcc11_wrap(x) RELOC_HIDE(x, 0) +#else +# define gcc11_wrap(x) (x) +#endif + #ifdef __GCC_ASM_FLAG_OUTPUTS__ # define ASM_FLAG_OUT(yes, no) yes #else diff --git a/xen/include/xen/pdx.h b/xen/include/xen/pdx.h index 770fadc06c..9fcfb0ce52 100644 --- a/xen/include/xen/pdx.h +++ b/xen/include/xen/pdx.h @@ -19,7 +19,7 @@ extern u64 pdx_region_mask(u64 base, u64 len); extern void set_pdx_range(unsigned long smfn, unsigned long emfn); #define page_to_pdx(pg) ((pg) - frame_table) -#define pdx_to_page(pdx) (frame_table + (pdx)) +#define pdx_to_page(pdx) gcc11_wrap(frame_table + (pdx)) bool __mfn_valid(unsigned long mfn); -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.15 From xen-changelog-bounces@lists.xenproject.org Sat Jun 05 02:23:14 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Sat, 05 Jun 2021 02:23:14 +0000 Received: from list by lists.xenproject.org with outflank-mailman.137104.254022 (Exim 4.92) (envelope-from ) id 1lpLxy-0007bd-Ig; Sat, 05 Jun 2021 02:23:14 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 137104.254022; Sat, 05 Jun 2021 02:23:14 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpLxy-0007bV-FS; Sat, 05 Jun 2021 02:23:14 +0000 Received: by outflank-mailman (input) for mailman id 137104; Sat, 05 Jun 2021 02:23:13 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpLxx-0007bG-8M for xen-changelog@lists.xenproject.org; Sat, 05 Jun 2021 02:23:13 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpLxx-00087W-7g for xen-changelog@lists.xenproject.org; Sat, 05 Jun 2021 02:23:13 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lpLxx-0006kr-6s for xen-changelog@lists.xenproject.org; Sat, 05 Jun 2021 02:23:13 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=xy2JzABH5ig8nusUmNFZlRHyV4Q3uvB1A11GVEC58W4=; b=TxwGC8NRx6hhcY2dk5SnP3FmD7 dOqdvWwyPXFE5gDO2UklWVUobRViTWX+qrYBWTU0kQ1n3LRLxAPAIg4BLysOk6GS0RVSY96iG/RvW HqnuGwd9/he7IsmDZBoXLyP4+9tQ4JvMjwcqwWdQV8ESEbezpPtrW83yI8t7zVw6LD/A=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.15] x86/cpuid: Rework HLE and RTM handling Message-Id: Date: Sat, 05 Jun 2021 02:23:13 +0000 commit 89c6e84f6dd31a73ed9037194c9a61081e000c38 Author: Andrew Cooper AuthorDate: Fri Jun 4 14:49:05 2021 +0200 Commit: Jan Beulich CommitDate: Fri Jun 4 14:49:05 2021 +0200 x86/cpuid: Rework HLE and RTM handling The TAA mitigation offered the option to hide the HLE and RTM CPUID bits, which has caused some migration compatibility problems. These two bits are special. Annotate them with ! to emphasise this point. Hardware Lock Elision (HLE) may or may not be visible in CPUID, but is disabled in microcode on all CPUs, and has been removed from the architecture. Do not advertise it to VMs by default. Restricted Transactional Memory (RTM) may or may not be visible in CPUID, and may or may not be configured in force-abort mode. Have tsx_init() note whether RTM has been configured into force-abort mode, so guest_common_feature_adjustments() can conditionally hide it from VMs by default. The host policy values for HLE/RTM may or may not be set, depending on any previous running kernel's choice of visibility, and Xen's choice. TSX is available on any CPU which enumerates a TSX-hiding mechanism, so instead of doing a two-step to clobber any hiding, scan CPUID, then set the visibility, just force visibility of the bits in the first place. With the HLE/RTM bits now unilaterally visible in the host policy, xc_cpuid_apply_policy() can construct a more appropriate policy out of thin air for pre-4.13 VMs with no CPUID data in their migration stream, and specifically one where HLE/RTM doesn't potentially disappear behind the back of a running VM. Fixes: 8c4330818f6 ("x86/spec-ctrl: Mitigate the TSX Asynchronous Abort sidechannel") Signed-off-by: Andrew Cooper Reviewed-by: Jan Beulich Reviewed-by: Roger Pau Monné master commit: 8fe24090d940d760145ccd5e234290be7418b175 master date: 2021-05-27 19:34:00 +0100 --- tools/libs/guest/xg_cpuid_x86.c | 2 ++ xen/arch/x86/cpuid.c | 24 ++++++++++------------ xen/arch/x86/spec_ctrl.c | 3 --- xen/arch/x86/tsx.c | 31 +++++++++++++++++++++++++++-- xen/include/asm-x86/processor.h | 1 + xen/include/public/arch-x86/cpufeatureset.h | 4 ++-- 6 files changed, 44 insertions(+), 21 deletions(-) diff --git a/tools/libs/guest/xg_cpuid_x86.c b/tools/libs/guest/xg_cpuid_x86.c index aae6931a11..c6c806884a 100644 --- a/tools/libs/guest/xg_cpuid_x86.c +++ b/tools/libs/guest/xg_cpuid_x86.c @@ -504,6 +504,8 @@ int xc_cpuid_apply_policy(xc_interface *xch, uint32_t domid, bool restore, if ( restore ) { p->basic.rdrand = test_bit(X86_FEATURE_RDRAND, host_featureset); + p->feat.hle = test_bit(X86_FEATURE_HLE, host_featureset); + p->feat.rtm = test_bit(X86_FEATURE_RTM, host_featureset); if ( di.hvm ) { diff --git a/xen/arch/x86/cpuid.c b/xen/arch/x86/cpuid.c index dde3e95284..7ac6636efc 100644 --- a/xen/arch/x86/cpuid.c +++ b/xen/arch/x86/cpuid.c @@ -357,6 +357,16 @@ static void __init guest_common_default_feature_adjustments(uint32_t *fs) boot_cpu_data.x86 == 6 && boot_cpu_data.x86_model == 0x3a && cpu_has_rdrand && !is_forced_cpu_cap(X86_FEATURE_RDRAND) ) __clear_bit(X86_FEATURE_RDRAND, fs); + + /* + * On certain hardware, speculative or errata workarounds can result in + * TSX being placed in "force-abort" mode, where it doesn't actually + * function as expected, but is technically compatible with the ISA. + * + * Do not advertise RTM to guests by default if it won't actually work. + */ + if ( rtm_disabled ) + __clear_bit(X86_FEATURE_RTM, fs); } static void __init guest_common_feature_adjustments(uint32_t *fs) @@ -620,20 +630,6 @@ void recalculate_cpuid_policy(struct domain *d) __clear_bit(X86_FEATURE_SYSCALL, max_fs); } - /* - * On hardware with MSR_TSX_CTRL, the admin may have elected to disable - * TSX and hide the feature bits. Migrating-in VMs may have been booted - * pre-mitigation when the TSX features were visbile. - * - * This situation is compatible (albeit with a perf hit to any TSX code in - * the guest), so allow the feature bits to remain set. - */ - if ( cpu_has_tsx_ctrl ) - { - __set_bit(X86_FEATURE_HLE, max_fs); - __set_bit(X86_FEATURE_RTM, max_fs); - } - /* Clamp the toolstacks choices to reality. */ for ( i = 0; i < ARRAY_SIZE(fs); i++ ) fs[i] &= max_fs[i]; diff --git a/xen/arch/x86/spec_ctrl.c b/xen/arch/x86/spec_ctrl.c index c9f78ead62..bea4c3e8b4 100644 --- a/xen/arch/x86/spec_ctrl.c +++ b/xen/arch/x86/spec_ctrl.c @@ -1158,9 +1158,6 @@ void __init init_speculation_mitigations(void) ((hw_smt_enabled && opt_smt) || !boot_cpu_has(X86_FEATURE_SC_VERW_IDLE)) ) { - setup_clear_cpu_cap(X86_FEATURE_HLE); - setup_clear_cpu_cap(X86_FEATURE_RTM); - opt_tsx = 0; tsx_init(); } diff --git a/xen/arch/x86/tsx.c b/xen/arch/x86/tsx.c index 39e483640a..e09e819dce 100644 --- a/xen/arch/x86/tsx.c +++ b/xen/arch/x86/tsx.c @@ -15,6 +15,7 @@ */ int8_t __read_mostly opt_tsx = -1; int8_t __read_mostly cpu_has_tsx_ctrl = -1; +bool __read_mostly rtm_disabled; static int __init parse_tsx(const char *s) { @@ -45,6 +46,30 @@ void tsx_init(void) rdmsrl(MSR_ARCH_CAPABILITIES, caps); cpu_has_tsx_ctrl = !!(caps & ARCH_CAPS_TSX_CTRL); + + /* + * The TSX features (HLE/RTM) are handled specially. They both + * enumerate features but, on certain parts, have mechanisms to be + * hidden without disrupting running software. + * + * At the moment, we're running in an unknown context (WRT hiding - + * particularly if another fully fledged kernel ran before us) and + * depending on user settings, may elect to continue hiding them from + * native CPUID instructions. + * + * Xen doesn't use TSX itself, but use cpu_has_{hle,rtm} for various + * system reasons, mostly errata detection, so the meaning is more + * useful as "TSX infrastructure available", as opposed to "features + * advertised and working". + * + * Force the features to be visible in Xen's view if we see any of the + * infrastructure capable of hiding them. + */ + if ( cpu_has_tsx_ctrl ) + { + setup_force_cpu_cap(X86_FEATURE_HLE); + setup_force_cpu_cap(X86_FEATURE_RTM); + } } if ( cpu_has_tsx_ctrl ) @@ -53,9 +78,11 @@ void tsx_init(void) rdmsrl(MSR_TSX_CTRL, val); + /* Check bottom bit only. Higher bits are various sentinels. */ + rtm_disabled = !(opt_tsx & 1); + val &= ~(TSX_CTRL_RTM_DISABLE | TSX_CTRL_CPUID_CLEAR); - /* Check bottom bit only. Higher bits are various sentinals. */ - if ( !(opt_tsx & 1) ) + if ( rtm_disabled ) val |= TSX_CTRL_RTM_DISABLE | TSX_CTRL_CPUID_CLEAR; wrmsrl(MSR_TSX_CTRL, val); diff --git a/xen/include/asm-x86/processor.h b/xen/include/asm-x86/processor.h index d5f467d245..44752bfdf6 100644 --- a/xen/include/asm-x86/processor.h +++ b/xen/include/asm-x86/processor.h @@ -629,6 +629,7 @@ static inline uint8_t get_cpu_family(uint32_t raw, uint8_t *model, } extern int8_t opt_tsx, cpu_has_tsx_ctrl; +extern bool rtm_disabled; void tsx_init(void); enum ap_boot_method { diff --git a/xen/include/public/arch-x86/cpufeatureset.h b/xen/include/public/arch-x86/cpufeatureset.h index a501479820..6477f75505 100644 --- a/xen/include/public/arch-x86/cpufeatureset.h +++ b/xen/include/public/arch-x86/cpufeatureset.h @@ -197,14 +197,14 @@ XEN_CPUFEATURE(FSGSBASE, 5*32+ 0) /*A {RD,WR}{FS,GS}BASE instructions */ XEN_CPUFEATURE(TSC_ADJUST, 5*32+ 1) /*S TSC_ADJUST MSR available */ XEN_CPUFEATURE(SGX, 5*32+ 2) /* Software Guard extensions */ XEN_CPUFEATURE(BMI1, 5*32+ 3) /*A 1st bit manipulation extensions */ -XEN_CPUFEATURE(HLE, 5*32+ 4) /*A Hardware Lock Elision */ +XEN_CPUFEATURE(HLE, 5*32+ 4) /*!a Hardware Lock Elision */ XEN_CPUFEATURE(AVX2, 5*32+ 5) /*A AVX2 instructions */ XEN_CPUFEATURE(FDP_EXCP_ONLY, 5*32+ 6) /*! x87 FDP only updated on exception. */ XEN_CPUFEATURE(SMEP, 5*32+ 7) /*S Supervisor Mode Execution Protection */ XEN_CPUFEATURE(BMI2, 5*32+ 8) /*A 2nd bit manipulation extensions */ XEN_CPUFEATURE(ERMS, 5*32+ 9) /*A Enhanced REP MOVSB/STOSB */ XEN_CPUFEATURE(INVPCID, 5*32+10) /*H Invalidate Process Context ID */ -XEN_CPUFEATURE(RTM, 5*32+11) /*A Restricted Transactional Memory */ +XEN_CPUFEATURE(RTM, 5*32+11) /*!A Restricted Transactional Memory */ XEN_CPUFEATURE(PQM, 5*32+12) /* Platform QoS Monitoring */ XEN_CPUFEATURE(NO_FPU_SEL, 5*32+13) /*! FPU CS/DS stored as zero */ XEN_CPUFEATURE(MPX, 5*32+14) /*s Memory Protection Extensions */ -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.15 From xen-changelog-bounces@lists.xenproject.org Sat Jun 05 02:23:24 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Sat, 05 Jun 2021 02:23:24 +0000 Received: from list by lists.xenproject.org with outflank-mailman.137105.254026 (Exim 4.92) (envelope-from ) id 1lpLy8-0007ed-K0; Sat, 05 Jun 2021 02:23:24 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 137105.254026; Sat, 05 Jun 2021 02:23:24 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpLy8-0007eV-H7; Sat, 05 Jun 2021 02:23:24 +0000 Received: by outflank-mailman (input) for mailman id 137105; Sat, 05 Jun 2021 02:23:23 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpLy7-0007eK-BS for xen-changelog@lists.xenproject.org; Sat, 05 Jun 2021 02:23:23 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpLy7-00087e-Af for xen-changelog@lists.xenproject.org; Sat, 05 Jun 2021 02:23:23 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lpLy7-0006n0-9p for xen-changelog@lists.xenproject.org; Sat, 05 Jun 2021 02:23:23 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=1RN0wn5Hx9rbR3jTuDa+lVbN6cVddwUz30xNsucf21Y=; b=TACx+twfKJofAIOwCokdJWnkG4 INMkVcg5HP7IcT3tL/KoZvXYy5CsZ5UrqBYuvcahHAWTDi1cOsTW5VZgCtTk99vUsGYqyx8W2fV+I GisJLoUNjT+yAA/ehfAtrWlrcfyw9svEYdQrvipgCuiqRTasjlYBPc4yiV+Ofe1Q+a9Y=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.15] x86: fix build race when generating temporary object files (take 2) Message-Id: Date: Sat, 05 Jun 2021 02:23:23 +0000 commit eae0dfac891f521ceb6c4733e22a0cd718f336c0 Author: Jan Beulich AuthorDate: Fri Jun 4 14:49:40 2021 +0200 Commit: Jan Beulich CommitDate: Fri Jun 4 14:49:40 2021 +0200 x86: fix build race when generating temporary object files (take 2) The original commit wasn't quite sufficient: Emptying DEPS is helpful only when nothing will get added to it subsequently. xen/Rules.mk will, after including the local Makefile, amend DEPS by dependencies for objects living in sub-directories though. For the purpose of suppressing dependencies of the makefiles on the .*.d2 files (and thus to avoid their re-generation) it is, however, not necessary at all to play with DEPS. Instead we can override DEPS_INCLUDE (which generally is a late- expansion variable). Fixes: 761bb575ce97 ("x86: fix build race when generating temporary object files") Signed-off-by: Jan Beulich master commit: 8c90dbb99907f3b471d558775777a84daec7c3f6 master date: 2021-05-28 09:12:24 +0200 --- xen/arch/x86/Makefile | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/xen/arch/x86/Makefile b/xen/arch/x86/Makefile index b6567c4127..b560c12a58 100644 --- a/xen/arch/x86/Makefile +++ b/xen/arch/x86/Makefile @@ -286,5 +286,5 @@ clean:: # Suppress loading of DEPS files for internal, temporary target files. This # then also suppresses re-generation of the respective .*.d2 files. ifeq ($(filter-out .xen%.o,$(notdir $(MAKECMDGOALS))),) -DEPS:= +DEPS_INCLUDE:= endif -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.15 From xen-changelog-bounces@lists.xenproject.org Sat Jun 05 03:33:11 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Sat, 05 Jun 2021 03:33:11 +0000 Received: from list by lists.xenproject.org with outflank-mailman.137121.254061 (Exim 4.92) (envelope-from ) id 1lpN3Y-0005sI-70; Sat, 05 Jun 2021 03:33:04 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 137121.254061; Sat, 05 Jun 2021 03:33:04 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpN3Y-0005sB-40; Sat, 05 Jun 2021 03:33:04 +0000 Received: by outflank-mailman (input) for mailman id 137121; Sat, 05 Jun 2021 03:33:02 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpN3W-0005s5-5H for xen-changelog@lists.xenproject.org; Sat, 05 Jun 2021 03:33:02 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpN3W-0000oY-3L for xen-changelog@lists.xenproject.org; Sat, 05 Jun 2021 03:33:02 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lpN3W-0002iH-2F for xen-changelog@lists.xenproject.org; Sat, 05 Jun 2021 03:33:02 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=g5fMmlbYh1PQyqEEo1dlMQboK6LiIUqoS+NWSZiKH7M=; b=f/PNNoY8NmWV/diuD6V/u5Xiqm xd1Ix8hVL6u0F9egILwhj9CQj5nWpLYbxIFR51i3tN6U40dStDNozK9ZdRZbL/AJ9UavmIH3ZMJhz jJ1e7QRJ1jlHwKbV8CUoU9CGp7rI6JHTUNk2bQNcl21aN4LYQHA1pK+fmM4a191hmWDo=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.14] x86/Intel: insert Tiger Lake model numbers Message-Id: Date: Sat, 05 Jun 2021 03:33:02 +0000 commit 02f9760498ce1c4fb320045f3f3b8f515d124ad4 Author: Jan Beulich AuthorDate: Fri Jun 4 14:51:25 2021 +0200 Commit: Jan Beulich CommitDate: Fri Jun 4 14:51:25 2021 +0200 x86/Intel: insert Tiger Lake model numbers Both match prior generation processors as far as LBR and C-state MSRs go (SDM rev 073). The if_pschange_mc erratum, according to the spec update, is not applicable. Signed-off-by: Jan Beulich Acked-by: Andrew Cooper master commit: e93c3712d67098453760fd61c338cbf62dd08da1 master date: 2020-12-22 09:00:03 +0100 --- xen/arch/x86/acpi/cpu_idle.c | 3 +++ xen/arch/x86/hvm/vmx/vmx.c | 2 ++ 2 files changed, 5 insertions(+) diff --git a/xen/arch/x86/acpi/cpu_idle.c b/xen/arch/x86/acpi/cpu_idle.c index 27e0b52621..c092086b33 100644 --- a/xen/arch/x86/acpi/cpu_idle.c +++ b/xen/arch/x86/acpi/cpu_idle.c @@ -183,6 +183,9 @@ static void do_get_hw_residencies(void *arg) /* Ice Lake */ case 0x7D: case 0x7E: + /* Tiger Lake */ + case 0x8C: + case 0x8D: /* Kaby Lake */ case 0x8E: case 0x9E: diff --git a/xen/arch/x86/hvm/vmx/vmx.c b/xen/arch/x86/hvm/vmx/vmx.c index cc6d4ece22..ca47f83cd4 100644 --- a/xen/arch/x86/hvm/vmx/vmx.c +++ b/xen/arch/x86/hvm/vmx/vmx.c @@ -2787,6 +2787,8 @@ static const struct lbr_info *last_branch_msr_get(void) case 0x7a: /* Ice Lake */ case 0x7d: case 0x7e: + /* Tiger Lake */ + case 0x8c: case 0x8d: /* Tremont */ case 0x86: /* Kaby Lake */ -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.14 From xen-changelog-bounces@lists.xenproject.org Sat Jun 05 03:33:13 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Sat, 05 Jun 2021 03:33:13 +0000 Received: from list by lists.xenproject.org with outflank-mailman.137122.254067 (Exim 4.92) (envelope-from ) id 1lpN3h-0005uU-9N; Sat, 05 Jun 2021 03:33:13 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 137122.254067; Sat, 05 Jun 2021 03:33:13 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpN3h-0005uM-5q; Sat, 05 Jun 2021 03:33:13 +0000 Received: by outflank-mailman (input) for mailman id 137122; Sat, 05 Jun 2021 03:33:12 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpN3g-0005uA-7H for xen-changelog@lists.xenproject.org; Sat, 05 Jun 2021 03:33:12 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpN3g-0000oc-6N for xen-changelog@lists.xenproject.org; Sat, 05 Jun 2021 03:33:12 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lpN3g-0002j1-5N for xen-changelog@lists.xenproject.org; Sat, 05 Jun 2021 03:33:12 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=ckVdhJMbwSKB0pZVtLaoDBzNoFW4eJHbLCb0mxy4AnI=; b=GsmQQIJElbFgDaqhRp4xnP2pVi rKxkzIfXRhTTnOjbYryvXnjIuVBlhJspEH5q4Y/4k+rtT5rqDnRHq3nnQLW5Jqgpw/5p3tn/ju0Cv +9/1r4nNeR+Qk0ffzX0ziac0nGm6Eg8sLK1jLQenV8PCgaNhcH2tgtSbUPOs+SkqNzRk=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.14] update Xen version to 4.14.3-pre Message-Id: Date: Sat, 05 Jun 2021 03:33:12 +0000 commit a351751d88973da710a3a902a7ced2ede9df78b0 Author: Jan Beulich AuthorDate: Fri Jun 4 14:52:04 2021 +0200 Commit: Jan Beulich CommitDate: Fri Jun 4 14:52:04 2021 +0200 update Xen version to 4.14.3-pre --- xen/Makefile | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/xen/Makefile b/xen/Makefile index 4ffa2ed7d5..22362d7348 100644 --- a/xen/Makefile +++ b/xen/Makefile @@ -2,7 +2,7 @@ # All other places this is stored (eg. compile.h) should be autogenerated. export XEN_VERSION = 4 export XEN_SUBVERSION = 14 -export XEN_EXTRAVERSION ?= .2$(XEN_VENDORVERSION) +export XEN_EXTRAVERSION ?= .3-pre$(XEN_VENDORVERSION) export XEN_FULLVERSION = $(XEN_VERSION).$(XEN_SUBVERSION)$(XEN_EXTRAVERSION) -include xen-version -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.14 From xen-changelog-bounces@lists.xenproject.org Sat Jun 05 03:33:23 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Sat, 05 Jun 2021 03:33:23 +0000 Received: from list by lists.xenproject.org with outflank-mailman.137123.254070 (Exim 4.92) (envelope-from ) id 1lpN3r-0005xe-Ag; Sat, 05 Jun 2021 03:33:23 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 137123.254070; Sat, 05 Jun 2021 03:33:23 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpN3r-0005xU-7P; Sat, 05 Jun 2021 03:33:23 +0000 Received: by outflank-mailman (input) for mailman id 137123; Sat, 05 Jun 2021 03:33:22 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpN3q-0005x5-AT for xen-changelog@lists.xenproject.org; Sat, 05 Jun 2021 03:33:22 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpN3q-0000on-9g for xen-changelog@lists.xenproject.org; Sat, 05 Jun 2021 03:33:22 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lpN3q-0002lB-8o for xen-changelog@lists.xenproject.org; Sat, 05 Jun 2021 03:33:22 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=1m3m0tyRF+EK536N2AK3B3EVuDBzkDGltnhhTeyp1Bk=; b=2FX4iCs8CyUPf2p210NY/tRUkp t7u37Gj/gm/Rw8bCGOoRszIuyjb4MBoJdgg7Rs/1PDIFwM7SrtntiSz3b8uAeKzNZtHSyGlp4j9Fm Ut/X2va7fOEtqE3vB1PkR5x39+qOS1Jidx/BKJC0z9FBIb++fLV4huH5UdkU5vsGmS6U=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.14] VT-d: Don't assume register-based invalidation is always supported Message-Id: Date: Sat, 05 Jun 2021 03:33:22 +0000 commit c2f78b49050335f9a0871263e622d7a10fdccbc4 Author: Chao Gao AuthorDate: Fri Jun 4 14:53:25 2021 +0200 Commit: Jan Beulich CommitDate: Fri Jun 4 14:53:25 2021 +0200 VT-d: Don't assume register-based invalidation is always supported According to Intel VT-d SPEC rev3.3 Section 6.5, Register-based Invalidation isn't supported by Intel VT-d version 6 and beyond. This hardware change impacts following two scenarios: admin can disable queued invalidation via 'qinval' cmdline and use register-based interface; VT-d switches to register-based invalidation when queued invalidation needs to be disabled, for example, during disabling x2apic or during system suspension or after enabling queued invalidation fails. To deal with this hardware change, if register-based invalidation isn't supported, queued invalidation cannot be disabled through Xen cmdline; and if queued invalidation has to be disabled temporarily in some scenarios, VT-d won't switch to register-based interface but use some dummy functions to catch errors in case there is any invalidation request issued when queued invalidation is disabled. Signed-off-by: Chao Gao Reviewed-by: Jan Beulich Reviewed-by: Kevin Tian master commit: 6773b1a7584a75a486e9774541ad5bd84c9aa5ee master date: 2021-04-26 10:16:50 +0200 --- docs/misc/xen-command-line.pandoc | 4 +++- xen/drivers/passthrough/vtd/iommu.c | 27 +++++++++++++++++++++++++-- xen/drivers/passthrough/vtd/iommu.h | 7 +++++++ xen/drivers/passthrough/vtd/qinval.c | 33 +++++++++++++++++++++++++++++++-- 4 files changed, 66 insertions(+), 5 deletions(-) diff --git a/docs/misc/xen-command-line.pandoc b/docs/misc/xen-command-line.pandoc index f63fe9680a..b9824d443f 100644 --- a/docs/misc/xen-command-line.pandoc +++ b/docs/misc/xen-command-line.pandoc @@ -1405,7 +1405,9 @@ The following options are specific to Intel VT-d hardware: * The `qinval` boolean controls the Queued Invalidation sub-feature, and is active by default on compatible hardware. Queued Invalidation is a feature in second-generation IOMMUs and is a functional prerequisite for - Interrupt Remapping. + Interrupt Remapping. Note that Xen disregards this setting for Intel VT-d + version 6 and greater as Registered-Based Invalidation isn't supported + by them. * The `igfx` boolean is active by default, and controls whether the IOMMU in front of an Intel Graphics Device is enabled or not. diff --git a/xen/drivers/passthrough/vtd/iommu.c b/xen/drivers/passthrough/vtd/iommu.c index 74c7aca24b..32ff60bfe3 100644 --- a/xen/drivers/passthrough/vtd/iommu.c +++ b/xen/drivers/passthrough/vtd/iommu.c @@ -1188,6 +1188,14 @@ int __init iommu_alloc(struct acpi_drhd_unit *drhd) iommu->cap = dmar_readq(iommu->reg, DMAR_CAP_REG); iommu->ecap = dmar_readq(iommu->reg, DMAR_ECAP_REG); + iommu->version = dmar_readl(iommu->reg, DMAR_VER_REG); + + if ( !iommu_qinval && !has_register_based_invalidation(iommu) ) + { + printk(XENLOG_WARNING VTDPREFIX "IOMMU %d: cannot disable Queued Invalidation\n", + iommu->index); + iommu_qinval = true; + } if ( iommu_verbose ) { @@ -2161,6 +2169,10 @@ static int __must_check init_vtd_hw(bool resume) */ if ( enable_qinval(iommu) != 0 ) { + /* Ensure register-based invalidation is available */ + if ( !has_register_based_invalidation(iommu) ) + return -EIO; + iommu->flush.context = vtd_flush_context_reg; iommu->flush.iotlb = vtd_flush_iotlb_reg; } @@ -2251,6 +2263,7 @@ static int __init vtd_setup(void) struct acpi_drhd_unit *drhd; struct vtd_iommu *iommu; int ret; + bool reg_inval_supported = true; if ( list_empty(&acpi_drhd_units) ) { @@ -2272,8 +2285,8 @@ static int __init vtd_setup(void) } /* We enable the following features only if they are supported by all VT-d - * engines: Snoop Control, DMA passthrough, Queued Invalidation, Interrupt - * Remapping, and Posted Interrupt + * engines: Snoop Control, DMA passthrough, Register-based Invalidation, + * Queued Invalidation, Interrupt Remapping, and Posted Interrupt. */ for_each_drhd_unit ( drhd ) { @@ -2295,6 +2308,9 @@ static int __init vtd_setup(void) if ( iommu_qinval && !ecap_queued_inval(iommu->ecap) ) iommu_qinval = 0; + if ( !has_register_based_invalidation(iommu) ) + reg_inval_supported = false; + if ( iommu_intremap && !ecap_intr_remap(iommu->ecap) ) iommu_intremap = iommu_intremap_off; @@ -2321,6 +2337,13 @@ static int __init vtd_setup(void) softirq_tasklet_init(&vtd_fault_tasklet, do_iommu_page_fault, NULL); + if ( !iommu_qinval && !reg_inval_supported ) + { + dprintk(XENLOG_ERR VTDPREFIX, "No available invalidation interface\n"); + ret = -ENODEV; + goto error; + } + if ( !iommu_qinval && iommu_intremap ) { iommu_intremap = iommu_intremap_off; diff --git a/xen/drivers/passthrough/vtd/iommu.h b/xen/drivers/passthrough/vtd/iommu.h index 216791b3d6..57737b1a4a 100644 --- a/xen/drivers/passthrough/vtd/iommu.h +++ b/xen/drivers/passthrough/vtd/iommu.h @@ -540,6 +540,7 @@ struct vtd_iommu { struct list_head ats_devices; unsigned long *domid_bitmap; /* domain id bitmap */ u16 *domid_map; /* domain id mapping array */ + uint32_t version; }; #define INTEL_IOMMU_DEBUG(fmt, args...) \ @@ -549,4 +550,10 @@ struct vtd_iommu { dprintk(XENLOG_WARNING VTDPREFIX, fmt, ## args); \ } while(0) +/* Register-based invalidation isn't supported by VT-d version 6 and beyond. */ +static inline bool has_register_based_invalidation(const struct vtd_iommu *vtd) +{ + return VER_MAJOR(vtd->version) < 6; +} + #endif diff --git a/xen/drivers/passthrough/vtd/qinval.c b/xen/drivers/passthrough/vtd/qinval.c index 764ef9f0fc..dfcbb66c8a 100644 --- a/xen/drivers/passthrough/vtd/qinval.c +++ b/xen/drivers/passthrough/vtd/qinval.c @@ -442,6 +442,23 @@ int enable_qinval(struct vtd_iommu *iommu) return 0; } +static int vtd_flush_context_noop(struct vtd_iommu *iommu, uint16_t did, + uint16_t source_id, uint8_t function_mask, + uint64_t type, bool flush_non_present_entry) +{ + WARN(); + return -EIO; +} + +static int vtd_flush_iotlb_noop(struct vtd_iommu *iommu, uint16_t did, + uint64_t addr, unsigned int size_order, + uint64_t type, bool flush_non_present_entry, + bool flush_dev_iotlb) +{ + WARN(); + return -EIO; +} + void disable_qinval(struct vtd_iommu *iommu) { u32 sts; @@ -463,6 +480,18 @@ void disable_qinval(struct vtd_iommu *iommu) out: spin_unlock_irqrestore(&iommu->register_lock, flags); - iommu->flush.context = vtd_flush_context_reg; - iommu->flush.iotlb = vtd_flush_iotlb_reg; + /* + * Assign callbacks to noop to catch errors if register-based invalidation + * isn't supported. + */ + if ( has_register_based_invalidation(iommu) ) + { + iommu->flush.context = vtd_flush_context_reg; + iommu->flush.iotlb = vtd_flush_iotlb_reg; + } + else + { + iommu->flush.context = vtd_flush_context_noop; + iommu->flush.iotlb = vtd_flush_iotlb_noop; + } } -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.14 From xen-changelog-bounces@lists.xenproject.org Sat Jun 05 03:33:33 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Sat, 05 Jun 2021 03:33:33 +0000 Received: from list by lists.xenproject.org with outflank-mailman.137124.254074 (Exim 4.92) (envelope-from ) id 1lpN41-00061z-Bl; Sat, 05 Jun 2021 03:33:33 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 137124.254074; Sat, 05 Jun 2021 03:33:33 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpN41-00061r-8y; Sat, 05 Jun 2021 03:33:33 +0000 Received: by outflank-mailman (input) for mailman id 137124; Sat, 05 Jun 2021 03:33:32 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpN40-00061h-Dr for xen-changelog@lists.xenproject.org; Sat, 05 Jun 2021 03:33:32 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpN40-0000ox-Ct for xen-changelog@lists.xenproject.org; Sat, 05 Jun 2021 03:33:32 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lpN40-0002lq-Bx for xen-changelog@lists.xenproject.org; Sat, 05 Jun 2021 03:33:32 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=8lNAXpw7euH9XgY45CEMVQMmhWALccbN6AB4X8CT+6c=; b=NauYc3fPTPIr2INfnMFbcT+PQ7 C0ItZlvZmN6oah1rbFX9jnXbnje4N/BhSNP9ITJ2Ie3ZTi90Agby2bLXD+SbPpA4cwWk22rGm7NpT WWsMRYJWg/DHLHXtITV00f33H6vXo2kkGUftoCuv+tSr2am4MvWtYwSXUxwtwqFFzupc=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.14] x86/vtx: add LBR_SELECT to the list of LBR MSRs Message-Id: Date: Sat, 05 Jun 2021 03:33:32 +0000 commit f23cb474e8183f1e3e27c5d20342173d8869b234 Author: Igor Druzhinin AuthorDate: Fri Jun 4 14:54:10 2021 +0200 Commit: Jan Beulich CommitDate: Fri Jun 4 14:54:10 2021 +0200 x86/vtx: add LBR_SELECT to the list of LBR MSRs This MSR exists since Nehalem / Silvermont and is actively used by Linux, for instance, to improve sampling efficiency. Signed-off-by: Igor Druzhinin Reviewed-by: Jan Beulich Reviewed-by: Kevin Tian master commit: 08693c03e00ea3448adc4406c891e707f0068eb6 master date: 2021-04-26 10:22:04 +0200 --- xen/arch/x86/hvm/vmx/vmx.c | 20 ++++++++++++++++---- xen/include/asm-x86/msr-index.h | 10 ++++++++-- 2 files changed, 24 insertions(+), 6 deletions(-) diff --git a/xen/arch/x86/hvm/vmx/vmx.c b/xen/arch/x86/hvm/vmx/vmx.c index ca47f83cd4..09221c4811 100644 --- a/xen/arch/x86/hvm/vmx/vmx.c +++ b/xen/arch/x86/hvm/vmx/vmx.c @@ -2722,14 +2722,16 @@ static const struct lbr_info { }, nh_lbr[] = { { MSR_IA32_LASTINTFROMIP, 1 }, { MSR_IA32_LASTINTTOIP, 1 }, - { MSR_C2_LASTBRANCH_TOS, 1 }, + { MSR_NHL_LBR_SELECT, 1 }, + { MSR_NHL_LASTBRANCH_TOS, 1 }, { MSR_P4_LASTBRANCH_0_FROM_LIP, NUM_MSR_P4_LASTBRANCH_FROM_TO }, { MSR_P4_LASTBRANCH_0_TO_LIP, NUM_MSR_P4_LASTBRANCH_FROM_TO }, { 0, 0 } }, sk_lbr[] = { { MSR_IA32_LASTINTFROMIP, 1 }, { MSR_IA32_LASTINTTOIP, 1 }, - { MSR_SKL_LASTBRANCH_TOS, 1 }, + { MSR_NHL_LBR_SELECT, 1 }, + { MSR_NHL_LASTBRANCH_TOS, 1 }, { MSR_SKL_LASTBRANCH_0_FROM_IP, NUM_MSR_SKL_LASTBRANCH }, { MSR_SKL_LASTBRANCH_0_TO_IP, NUM_MSR_SKL_LASTBRANCH }, { MSR_SKL_LASTBRANCH_0_INFO, NUM_MSR_SKL_LASTBRANCH }, @@ -2741,10 +2743,19 @@ static const struct lbr_info { { MSR_C2_LASTBRANCH_0_FROM_IP, NUM_MSR_ATOM_LASTBRANCH_FROM_TO }, { MSR_C2_LASTBRANCH_0_TO_IP, NUM_MSR_ATOM_LASTBRANCH_FROM_TO }, { 0, 0 } +}, sm_lbr[] = { + { MSR_IA32_LASTINTFROMIP, 1 }, + { MSR_IA32_LASTINTTOIP, 1 }, + { MSR_SM_LBR_SELECT, 1 }, + { MSR_SM_LASTBRANCH_TOS, 1 }, + { MSR_C2_LASTBRANCH_0_FROM_IP, NUM_MSR_ATOM_LASTBRANCH_FROM_TO }, + { MSR_C2_LASTBRANCH_0_TO_IP, NUM_MSR_ATOM_LASTBRANCH_FROM_TO }, + { 0, 0 } }, gm_lbr[] = { { MSR_IA32_LASTINTFROMIP, 1 }, { MSR_IA32_LASTINTTOIP, 1 }, - { MSR_GM_LASTBRANCH_TOS, 1 }, + { MSR_SM_LBR_SELECT, 1 }, + { MSR_SM_LASTBRANCH_TOS, 1 }, { MSR_GM_LASTBRANCH_0_FROM_IP, NUM_MSR_GM_LASTBRANCH_FROM_TO }, { MSR_GM_LASTBRANCH_0_TO_IP, NUM_MSR_GM_LASTBRANCH_FROM_TO }, { 0, 0 } @@ -2798,6 +2809,7 @@ static const struct lbr_info *last_branch_msr_get(void) return sk_lbr; /* Atom */ case 0x1c: case 0x26: case 0x27: case 0x35: case 0x36: + return at_lbr; /* Silvermont */ case 0x37: case 0x4a: case 0x4d: case 0x5a: case 0x5d: /* Xeon Phi Knights Landing */ @@ -2806,7 +2818,7 @@ static const struct lbr_info *last_branch_msr_get(void) case 0x85: /* Airmont */ case 0x4c: - return at_lbr; + return sm_lbr; /* Goldmont */ case 0x5c: case 0x5f: return gm_lbr; diff --git a/xen/include/asm-x86/msr-index.h b/xen/include/asm-x86/msr-index.h index 5e64ecff91..d66c92b946 100644 --- a/xen/include/asm-x86/msr-index.h +++ b/xen/include/asm-x86/msr-index.h @@ -616,15 +616,21 @@ #define NUM_MSR_C2_LASTBRANCH_FROM_TO 4 #define NUM_MSR_ATOM_LASTBRANCH_FROM_TO 8 +/* Nehalem (and newer) last-branch recording */ +#define MSR_NHL_LBR_SELECT 0x000001c8 +#define MSR_NHL_LASTBRANCH_TOS 0x000001c9 + /* Skylake (and newer) last-branch recording */ -#define MSR_SKL_LASTBRANCH_TOS 0x000001c9 #define MSR_SKL_LASTBRANCH_0_FROM_IP 0x00000680 #define MSR_SKL_LASTBRANCH_0_TO_IP 0x000006c0 #define MSR_SKL_LASTBRANCH_0_INFO 0x00000dc0 #define NUM_MSR_SKL_LASTBRANCH 32 +/* Silvermont (and newer) last-branch recording */ +#define MSR_SM_LBR_SELECT 0x000001c8 +#define MSR_SM_LASTBRANCH_TOS 0x000001c9 + /* Goldmont last-branch recording */ -#define MSR_GM_LASTBRANCH_TOS 0x000001c9 #define MSR_GM_LASTBRANCH_0_FROM_IP 0x00000680 #define MSR_GM_LASTBRANCH_0_TO_IP 0x000006c0 #define NUM_MSR_GM_LASTBRANCH_FROM_TO 32 -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.14 From xen-changelog-bounces@lists.xenproject.org Sat Jun 05 03:33:43 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Sat, 05 Jun 2021 03:33:43 +0000 Received: from list by lists.xenproject.org with outflank-mailman.137125.254078 (Exim 4.92) (envelope-from ) id 1lpN4B-000653-De; Sat, 05 Jun 2021 03:33:43 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 137125.254078; Sat, 05 Jun 2021 03:33:43 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpN4B-00064v-Aj; Sat, 05 Jun 2021 03:33:43 +0000 Received: by outflank-mailman (input) for mailman id 137125; Sat, 05 Jun 2021 03:33:42 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpN4A-00064e-Gr for xen-changelog@lists.xenproject.org; Sat, 05 Jun 2021 03:33:42 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpN4A-0000pL-Fz for xen-changelog@lists.xenproject.org; Sat, 05 Jun 2021 03:33:42 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lpN4A-0002md-FB for xen-changelog@lists.xenproject.org; Sat, 05 Jun 2021 03:33:42 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=sOmCZgCe0oHI5e99Bv+mQMbwnG003ENa4bwqaJbHZDg=; b=vSAMwGPJz+bWNfqYlbtYrnJWLC Itao7MewkIrktH9HFJTs5wtYAjz1ljVFXeE7hVDPs5w9rP4VALv8Hfbc2/WRQ7dNTQoA7fy3Cd20B c1YUzhbImuCPaOoi1Ahye8FAU8gd2vVNzuxnX/YEisGX6ucAJ5wkuNlHwnA0K8eHRySw=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.14] x86/intel: insert Ice Lake-SP and Ice Lake-D model numbers Message-Id: Date: Sat, 05 Jun 2021 03:33:42 +0000 commit b15c24a70c8be8b32f61b1962a6dc9df3d65ce78 Author: Igor Druzhinin AuthorDate: Fri Jun 4 14:54:43 2021 +0200 Commit: Jan Beulich CommitDate: Fri Jun 4 14:54:43 2021 +0200 x86/intel: insert Ice Lake-SP and Ice Lake-D model numbers LBR, C-state MSRs should correspond to Ice Lake desktop according to SDM rev. 74 for both models. Ice Lake-SP is known to expose IF_PSCHANGE_MC_NO in IA32_ARCH_CAPABILITIES MSR (as advisory tells and Whitley SDP confirms) which means the erratum is fixed in hardware for that model and therefore it shouldn't be present in has_if_pschange_mc list. Provisionally assume the same to be the case for Ice Lake-D. Signed-off-by: Igor Druzhinin Reviewed-by: Jan Beulich Reviewed-by: Kevin Tian master commit: 95419adfd4b275cffe24b96edcc2f15bc4db8907 master date: 2021-04-26 10:22:48 +0200 --- xen/arch/x86/acpi/cpu_idle.c | 2 ++ xen/arch/x86/hvm/vmx/vmx.c | 2 +- 2 files changed, 3 insertions(+), 1 deletion(-) diff --git a/xen/arch/x86/acpi/cpu_idle.c b/xen/arch/x86/acpi/cpu_idle.c index c092086b33..d788c8bffc 100644 --- a/xen/arch/x86/acpi/cpu_idle.c +++ b/xen/arch/x86/acpi/cpu_idle.c @@ -181,6 +181,8 @@ static void do_get_hw_residencies(void *arg) case 0x55: case 0x5E: /* Ice Lake */ + case 0x6A: + case 0x6C: case 0x7D: case 0x7E: /* Tiger Lake */ diff --git a/xen/arch/x86/hvm/vmx/vmx.c b/xen/arch/x86/hvm/vmx/vmx.c index 09221c4811..6972d4ab23 100644 --- a/xen/arch/x86/hvm/vmx/vmx.c +++ b/xen/arch/x86/hvm/vmx/vmx.c @@ -2797,7 +2797,7 @@ static const struct lbr_info *last_branch_msr_get(void) /* Goldmont Plus */ case 0x7a: /* Ice Lake */ - case 0x7d: case 0x7e: + case 0x6a: case 0x6c: case 0x7d: case 0x7e: /* Tiger Lake */ case 0x8c: case 0x8d: /* Tremont */ -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.14 From xen-changelog-bounces@lists.xenproject.org Sat Jun 05 03:33:53 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Sat, 05 Jun 2021 03:33:53 +0000 Received: from list by lists.xenproject.org with outflank-mailman.137126.254082 (Exim 4.92) (envelope-from ) id 1lpN4L-00067q-Fc; Sat, 05 Jun 2021 03:33:53 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 137126.254082; Sat, 05 Jun 2021 03:33:53 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpN4L-00067i-CI; Sat, 05 Jun 2021 03:33:53 +0000 Received: by outflank-mailman (input) for mailman id 137126; Sat, 05 Jun 2021 03:33:52 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpN4K-00067V-Jo for xen-changelog@lists.xenproject.org; Sat, 05 Jun 2021 03:33:52 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpN4K-0000po-J4 for xen-changelog@lists.xenproject.org; Sat, 05 Jun 2021 03:33:52 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lpN4K-0002nD-I6 for xen-changelog@lists.xenproject.org; Sat, 05 Jun 2021 03:33:52 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=YXK0hl8pO0MkSTpBFuMvdRsQ/Ds8evOy8W/Mvphqsz8=; b=smt+ZOHrHoQzrp4drPG8xZuwi2 5dUTEYjfHdpvrvXm97/SkD0OWutF3VY+fot1sdnvqxgAvVJeaYuDEl+TdIHoDQlofaLOFIPXXL1al rMDyCvzqDSiBMNlPPEmzhTr7ildI6pQ81ttTag93lTJdBTX+SsHTr49CQUxlkR6j2oNY=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.14] x86/vhpet: fix RTC special casing Message-Id: Date: Sat, 05 Jun 2021 03:33:52 +0000 commit 9d963a7169a3b6e2e0fb9c8abc31c48064b2906d Author: Roger Pau Monné AuthorDate: Fri Jun 4 14:55:12 2021 +0200 Commit: Jan Beulich CommitDate: Fri Jun 4 14:55:12 2021 +0200 x86/vhpet: fix RTC special casing Restore setting the virtual timer callback private data to NULL if the timer is not level triggered. This fixes the special casing done in pt_update_irq so that the RTC interrupt when originating from the HPET is suspended if the interrupt source is masked. Note the RTC special casing done in pt_update_irq should only apply to the RTC interrupt originating from the emulated RTC device (which does set the callback private data), as in that case the callback itself will destroy the virtual timer if the interrupt is ignored. While there also use RTC_IRQ instead of 8 when the HPET is configured in LegacyReplacement Mode. Fixes: be07023be115 ("x86/vhpet: add support for level triggered interrupts") Reported-by: Jan Beulich Signed-off-by: Roger Pau Monné Reviewed-by: Jan Beulich master commit: 472a13988a051e5ae84b95815c6caf4378062abe master date: 2021-05-07 10:43:29 +0200 --- xen/arch/x86/hvm/hpet.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/xen/arch/x86/hvm/hpet.c b/xen/arch/x86/hvm/hpet.c index ca94e8b453..ee756abb82 100644 --- a/xen/arch/x86/hvm/hpet.c +++ b/xen/arch/x86/hvm/hpet.c @@ -22,6 +22,7 @@ #include #include #include +#include #include #include #include @@ -290,7 +291,7 @@ static void hpet_set_timer(HPETState *h, unsigned int tn, /* if LegacyReplacementRoute bit is set, HPET specification requires timer0 be routed to IRQ0 in NON-APIC or IRQ2 in the I/O APIC, timer1 be routed to IRQ8 in NON-APIC or IRQ8 in the I/O APIC. */ - irq = (tn == 0) ? 0 : 8; + irq = (tn == 0) ? 0 : RTC_IRQ; h->pt[tn].source = PTSRC_isa; } else @@ -318,7 +319,8 @@ static void hpet_set_timer(HPETState *h, unsigned int tn, hpet_tick_to_ns(h, diff), oneshot ? 0 : hpet_tick_to_ns(h, h->hpet.period[tn]), irq, timer_level(h, tn) ? hpet_timer_fired : NULL, - (void *)(unsigned long)tn, timer_level(h, tn)); + timer_level(h, tn) ? (void *)(unsigned long)tn : NULL, + timer_level(h, tn)); } static inline uint64_t hpet_fixup_reg( -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.14 From xen-changelog-bounces@lists.xenproject.org Sat Jun 05 03:34:03 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Sat, 05 Jun 2021 03:34:03 +0000 Received: from list by lists.xenproject.org with outflank-mailman.137127.254086 (Exim 4.92) (envelope-from ) id 1lpN4V-0006BX-JR; Sat, 05 Jun 2021 03:34:03 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 137127.254086; Sat, 05 Jun 2021 03:34:03 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpN4V-0006BP-GB; Sat, 05 Jun 2021 03:34:03 +0000 Received: by outflank-mailman (input) for mailman id 137127; Sat, 05 Jun 2021 03:34:02 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpN4U-0006BE-Md for xen-changelog@lists.xenproject.org; Sat, 05 Jun 2021 03:34:02 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpN4U-0000qB-Lv for xen-changelog@lists.xenproject.org; Sat, 05 Jun 2021 03:34:02 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lpN4U-0002oE-L8 for xen-changelog@lists.xenproject.org; Sat, 05 Jun 2021 03:34:02 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=c7oDgSnbuWeIraUmOIrM1xtOV2cy32InwPY81B5XX6U=; b=UFeVBoX3JZOVjt7IcgvRQwHjut bh9112TPSke2wqduhtLbmioC8HJATvZ7VVKDgdzu/AcWUpwCxZ5Paz4aQMyg/pEPKUoD4kWm62BVA sv/ZxB05+Dc0jnv58q1rd0bzqmUzAB8xlrnnqN0VPqYd9lYzjPlKVJvR5usxJjUymydA=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.14] x86emul: fix test harness build for gas 2.36 Message-Id: Date: Sat, 05 Jun 2021 03:34:02 +0000 commit ebfdf0c7eb0ba26a9e83f3f39bdf51452fa11b03 Author: Jan Beulich AuthorDate: Fri Jun 4 14:55:55 2021 +0200 Commit: Jan Beulich CommitDate: Fri Jun 4 14:55:55 2021 +0200 x86emul: fix test harness build for gas 2.36 All of the sudden, besides .text and .rodata and alike, an always present .note.gnu.property section has appeared. This section, when converting to binary format output, gets placed according to its linked address, causing the resulting blobs to be about 128Mb in size. The resulting headers with a C representation of the binary blobs then are, of course all a multiple of that size (and take accordingly long to create). I didn't bother waiting to see what size the final test_x86_emulator binary then would have had. See also https://sourceware.org/bugzilla/show_bug.cgi?id=27753. Rather than figuring out whether gas supports -mx86-used-note=, simply remove the section while creating *.bin. Signed-off-by: Jan Beulich master commit: aa803ba38a867551917d11059eaa044955556e05 master date: 2021-05-17 15:41:28 +0200 --- tools/tests/x86_emulator/testcase.mk | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/tools/tests/x86_emulator/testcase.mk b/tools/tests/x86_emulator/testcase.mk index dafeb6caf7..d4aa82e3f3 100644 --- a/tools/tests/x86_emulator/testcase.mk +++ b/tools/tests/x86_emulator/testcase.mk @@ -12,11 +12,11 @@ all: $(TESTCASE).bin %.bin: %.c $(CC) $(filter-out -M% .%,$(CFLAGS)) -c $< $(LD) $(LDFLAGS_DIRECT) -N -Ttext 0x100000 -o $*.tmp $*.o - $(OBJCOPY) -O binary $*.tmp $@ + $(OBJCOPY) -O binary -R .note.gnu.property $*.tmp $@ rm -f $*.tmp %-opmask.bin: opmask.S $(CC) $(filter-out -M% .%,$(CFLAGS)) -c $< -o $(basename $@).o $(LD) $(LDFLAGS_DIRECT) -N -Ttext 0x100000 -o $(basename $@).tmp $(basename $@).o - $(OBJCOPY) -O binary $(basename $@).tmp $@ + $(OBJCOPY) -O binary -R .note.gnu.property $(basename $@).tmp $@ rm -f $(basename $@).tmp -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.14 From xen-changelog-bounces@lists.xenproject.org Sat Jun 05 03:34:14 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Sat, 05 Jun 2021 03:34:14 +0000 Received: from list by lists.xenproject.org with outflank-mailman.137130.254104 (Exim 4.92) (envelope-from ) id 1lpN4g-0006Ye-7O; Sat, 05 Jun 2021 03:34:14 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 137130.254104; Sat, 05 Jun 2021 03:34:14 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpN4g-0006YW-3s; Sat, 05 Jun 2021 03:34:14 +0000 Received: by outflank-mailman (input) for mailman id 137130; Sat, 05 Jun 2021 03:34:12 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpN4e-0006Xv-QM for xen-changelog@lists.xenproject.org; Sat, 05 Jun 2021 03:34:12 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpN4e-0000qW-Pf for xen-changelog@lists.xenproject.org; Sat, 05 Jun 2021 03:34:12 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lpN4e-0002p7-Ok for xen-changelog@lists.xenproject.org; Sat, 05 Jun 2021 03:34:12 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=Psi6uotisGlBaqfMac55lTI0EkCKkRk8zgh71jjduUQ=; b=kii/5Ant9Ntp1/6UlHzwm8zNkY jrVBZeCb5dFgy+oD9E+wyHCxpTBn8SXVbPvqpErJvcBLR/75Yrq1CWkIVvdR+x7f/Gvz7o80HvEc7 BXJ8p6CEUTIAR8we2gIAHIerYSUsb0t//ccFoq9B212tPYtqYzuw4ejGoNsGkJl6QuuA=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.14] x86: make hypervisor build with gcc11 Message-Id: Date: Sat, 05 Jun 2021 03:34:12 +0000 commit ac507e078edcc18215d86c436d6c26dd76ffa15c Author: Jan Beulich AuthorDate: Fri Jun 4 14:56:19 2021 +0200 Commit: Jan Beulich CommitDate: Fri Jun 4 14:56:19 2021 +0200 x86: make hypervisor build with gcc11 Gcc 11 looks to make incorrect assumptions about valid ranges that pointers may be used for addressing when they are derived from e.g. a plain constant. See https://gcc.gnu.org/bugzilla/show_bug.cgi?id=100680. Utilize RELOC_HIDE() to work around the issue, which for x86 manifests in at least - mpparse.c:efi_check_config(), - tboot.c:tboot_probe(), - tboot.c:tboot_gen_frametable_integrity(), - x86_emulate.c:x86_emulate() (at -O2 only). The last case is particularly odd not just because it only triggers at higher optimization levels, but also because it only affects one of at least three similar constructs. Various "note" diagnostics claim the valid index range to be [0, 2⁶³-1]. Signed-off-by: Jan Beulich Tested-by: Jason Andryuk Acked-by: Roger Pau Monné master commit: 722f59d38c710a940ab05e542a83020eb5546dea master date: 2021-05-27 14:40:29 +0200 --- tools/tests/x86_emulator/x86-emulate.c | 7 +++++++ xen/arch/x86/x86_emulate/x86_emulate.c | 2 +- xen/include/asm-x86/fixmap.h | 2 +- xen/include/xen/compiler.h | 6 ++++++ xen/include/xen/pdx.h | 2 +- 5 files changed, 16 insertions(+), 3 deletions(-) diff --git a/tools/tests/x86_emulator/x86-emulate.c b/tools/tests/x86_emulator/x86-emulate.c index 572989078a..c32d9a55b4 100644 --- a/tools/tests/x86_emulator/x86-emulate.c +++ b/tools/tests/x86_emulator/x86-emulate.c @@ -8,6 +8,13 @@ #define ERR_PTR(val) NULL +/* See gcc bug 100680, but here don't bother making this version dependent. */ +#define gcc11_wrap(x) ({ \ + unsigned long x_; \ + __asm__ ( "" : "=g" (x_) : "0" (x) ); \ + (typeof(x))x_; \ +}) + #define cpu_has_amd_erratum(nr) 0 #define cpu_has_mpx false #define read_bndcfgu() 0 diff --git a/xen/arch/x86/x86_emulate/x86_emulate.c b/xen/arch/x86/x86_emulate/x86_emulate.c index 8ef66d824d..16d164904c 100644 --- a/xen/arch/x86/x86_emulate/x86_emulate.c +++ b/xen/arch/x86/x86_emulate/x86_emulate.c @@ -725,7 +725,7 @@ union vex { #define copy_VEX(ptr, vex) ({ \ if ( !mode_64bit() ) \ (vex).reg |= 8; \ - (ptr)[0 - PFX_BYTES] = ext < ext_8f08 ? 0xc4 : 0x8f; \ + gcc11_wrap(ptr)[0 - PFX_BYTES] = ext < ext_8f08 ? 0xc4 : 0x8f; \ (ptr)[1 - PFX_BYTES] = (vex).raw[0]; \ (ptr)[2 - PFX_BYTES] = (vex).raw[1]; \ container_of((ptr) + 1 - PFX_BYTES, typeof(vex), raw[0]); \ diff --git a/xen/include/asm-x86/fixmap.h b/xen/include/asm-x86/fixmap.h index 8330097a74..22afb92ff0 100644 --- a/xen/include/asm-x86/fixmap.h +++ b/xen/include/asm-x86/fixmap.h @@ -79,7 +79,7 @@ extern void __set_fixmap( #define clear_fixmap(idx) __set_fixmap(idx, 0, 0) -#define __fix_to_virt(x) (FIXADDR_TOP - ((x) << PAGE_SHIFT)) +#define __fix_to_virt(x) gcc11_wrap(FIXADDR_TOP - ((x) << PAGE_SHIFT)) #define __virt_to_fix(x) ((FIXADDR_TOP - ((x)&PAGE_MASK)) >> PAGE_SHIFT) #define fix_to_virt(x) ((void *)__fix_to_virt(x)) diff --git a/xen/include/xen/compiler.h b/xen/include/xen/compiler.h index c0e0ee9f27..eb819fab46 100644 --- a/xen/include/xen/compiler.h +++ b/xen/include/xen/compiler.h @@ -105,6 +105,12 @@ __asm__ ("" : "=r"(__ptr) : "0"(ptr)); \ (typeof(ptr)) (__ptr + (off)); }) +#if CONFIG_GCC_VERSION >= 110000 /* See gcc bug 100680. */ +# define gcc11_wrap(x) RELOC_HIDE(x, 0) +#else +# define gcc11_wrap(x) (x) +#endif + #ifdef __GCC_ASM_FLAG_OUTPUTS__ # define ASM_FLAG_OUT(yes, no) yes #else diff --git a/xen/include/xen/pdx.h b/xen/include/xen/pdx.h index 770fadc06c..9fcfb0ce52 100644 --- a/xen/include/xen/pdx.h +++ b/xen/include/xen/pdx.h @@ -19,7 +19,7 @@ extern u64 pdx_region_mask(u64 base, u64 len); extern void set_pdx_range(unsigned long smfn, unsigned long emfn); #define page_to_pdx(pg) ((pg) - frame_table) -#define pdx_to_page(pdx) (frame_table + (pdx)) +#define pdx_to_page(pdx) gcc11_wrap(frame_table + (pdx)) bool __mfn_valid(unsigned long mfn); -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.14 From xen-changelog-bounces@lists.xenproject.org Sat Jun 05 03:34:24 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Sat, 05 Jun 2021 03:34:24 +0000 Received: from list by lists.xenproject.org with outflank-mailman.137131.254107 (Exim 4.92) (envelope-from ) id 1lpN4q-0006fC-95; Sat, 05 Jun 2021 03:34:24 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 137131.254107; Sat, 05 Jun 2021 03:34:24 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpN4q-0006f4-5f; Sat, 05 Jun 2021 03:34:24 +0000 Received: by outflank-mailman (input) for mailman id 137131; Sat, 05 Jun 2021 03:34:22 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpN4o-0006eJ-Tv for xen-changelog@lists.xenproject.org; Sat, 05 Jun 2021 03:34:22 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpN4o-0000qh-TB for xen-changelog@lists.xenproject.org; Sat, 05 Jun 2021 03:34:22 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lpN4o-0002qJ-SE for xen-changelog@lists.xenproject.org; Sat, 05 Jun 2021 03:34:22 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=fwg7loo74ekj3uFrxqxJVJifhYjo7lUq62DFKeBrCDo=; b=wpKM3i3i8BbUNenFVbTegHj6bc lJIVZq18phJvcDOJHywgNorIkmE5emQBErJnOmJZHuIcnDyogCcVqIstxPNlr7eA41DKE5mDmvhd8 5w7QyS6g3NodFSeu/m/yC8/Cl0OYPHCXEYDTYWh4gpMryHVWKpG6iQjSNYDfl64r+DPI=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.14] x86/cpuid: Rework HLE and RTM handling Message-Id: Date: Sat, 05 Jun 2021 03:34:22 +0000 commit 3f8549386bc72b0ddc7020eec97f0b497785193e Author: Andrew Cooper AuthorDate: Fri Jun 4 14:56:56 2021 +0200 Commit: Jan Beulich CommitDate: Fri Jun 4 14:56:56 2021 +0200 x86/cpuid: Rework HLE and RTM handling The TAA mitigation offered the option to hide the HLE and RTM CPUID bits, which has caused some migration compatibility problems. These two bits are special. Annotate them with ! to emphasise this point. Hardware Lock Elision (HLE) may or may not be visible in CPUID, but is disabled in microcode on all CPUs, and has been removed from the architecture. Do not advertise it to VMs by default. Restricted Transactional Memory (RTM) may or may not be visible in CPUID, and may or may not be configured in force-abort mode. Have tsx_init() note whether RTM has been configured into force-abort mode, so guest_common_feature_adjustments() can conditionally hide it from VMs by default. The host policy values for HLE/RTM may or may not be set, depending on any previous running kernel's choice of visibility, and Xen's choice. TSX is available on any CPU which enumerates a TSX-hiding mechanism, so instead of doing a two-step to clobber any hiding, scan CPUID, then set the visibility, just force visibility of the bits in the first place. With the HLE/RTM bits now unilaterally visible in the host policy, xc_cpuid_apply_policy() can construct a more appropriate policy out of thin air for pre-4.13 VMs with no CPUID data in their migration stream, and specifically one where HLE/RTM doesn't potentially disappear behind the back of a running VM. Fixes: 8c4330818f6 ("x86/spec-ctrl: Mitigate the TSX Asynchronous Abort sidechannel") Signed-off-by: Andrew Cooper Reviewed-by: Jan Beulich Reviewed-by: Roger Pau Monné master commit: 8fe24090d940d760145ccd5e234290be7418b175 master date: 2021-05-27 19:34:00 +0100 --- tools/libxc/xc_cpuid_x86.c | 2 ++ xen/arch/x86/cpuid.c | 24 ++++++++++------------ xen/arch/x86/spec_ctrl.c | 3 --- xen/arch/x86/tsx.c | 31 +++++++++++++++++++++++++++-- xen/include/asm-x86/processor.h | 1 + xen/include/public/arch-x86/cpufeatureset.h | 4 ++-- 6 files changed, 44 insertions(+), 21 deletions(-) diff --git a/tools/libxc/xc_cpuid_x86.c b/tools/libxc/xc_cpuid_x86.c index 0f24d6dd08..512eb1f78f 100644 --- a/tools/libxc/xc_cpuid_x86.c +++ b/tools/libxc/xc_cpuid_x86.c @@ -504,6 +504,8 @@ int xc_cpuid_apply_policy(xc_interface *xch, uint32_t domid, bool restore, if ( restore ) { p->basic.rdrand = test_bit(X86_FEATURE_RDRAND, host_featureset); + p->feat.hle = test_bit(X86_FEATURE_HLE, host_featureset); + p->feat.rtm = test_bit(X86_FEATURE_RTM, host_featureset); if ( di.hvm ) { diff --git a/xen/arch/x86/cpuid.c b/xen/arch/x86/cpuid.c index 4b424fac95..425dce7bc7 100644 --- a/xen/arch/x86/cpuid.c +++ b/xen/arch/x86/cpuid.c @@ -357,6 +357,16 @@ static void __init guest_common_default_feature_adjustments(uint32_t *fs) boot_cpu_data.x86 == 6 && boot_cpu_data.x86_model == 0x3a && cpu_has_rdrand && !is_forced_cpu_cap(X86_FEATURE_RDRAND) ) __clear_bit(X86_FEATURE_RDRAND, fs); + + /* + * On certain hardware, speculative or errata workarounds can result in + * TSX being placed in "force-abort" mode, where it doesn't actually + * function as expected, but is technically compatible with the ISA. + * + * Do not advertise RTM to guests by default if it won't actually work. + */ + if ( rtm_disabled ) + __clear_bit(X86_FEATURE_RTM, fs); } static void __init guest_common_feature_adjustments(uint32_t *fs) @@ -628,20 +638,6 @@ void recalculate_cpuid_policy(struct domain *d) if ( cpu_has_itsc && (d->disable_migrate || d->arch.vtsc) ) __set_bit(X86_FEATURE_ITSC, max_fs); - /* - * On hardware with MSR_TSX_CTRL, the admin may have elected to disable - * TSX and hide the feature bits. Migrating-in VMs may have been booted - * pre-mitigation when the TSX features were visbile. - * - * This situation is compatible (albeit with a perf hit to any TSX code in - * the guest), so allow the feature bits to remain set. - */ - if ( cpu_has_tsx_ctrl ) - { - __set_bit(X86_FEATURE_HLE, max_fs); - __set_bit(X86_FEATURE_RTM, max_fs); - } - /* Clamp the toolstacks choices to reality. */ for ( i = 0; i < ARRAY_SIZE(fs); i++ ) fs[i] &= max_fs[i]; diff --git a/xen/arch/x86/spec_ctrl.c b/xen/arch/x86/spec_ctrl.c index c9f78ead62..bea4c3e8b4 100644 --- a/xen/arch/x86/spec_ctrl.c +++ b/xen/arch/x86/spec_ctrl.c @@ -1158,9 +1158,6 @@ void __init init_speculation_mitigations(void) ((hw_smt_enabled && opt_smt) || !boot_cpu_has(X86_FEATURE_SC_VERW_IDLE)) ) { - setup_clear_cpu_cap(X86_FEATURE_HLE); - setup_clear_cpu_cap(X86_FEATURE_RTM); - opt_tsx = 0; tsx_init(); } diff --git a/xen/arch/x86/tsx.c b/xen/arch/x86/tsx.c index 39e483640a..e09e819dce 100644 --- a/xen/arch/x86/tsx.c +++ b/xen/arch/x86/tsx.c @@ -15,6 +15,7 @@ */ int8_t __read_mostly opt_tsx = -1; int8_t __read_mostly cpu_has_tsx_ctrl = -1; +bool __read_mostly rtm_disabled; static int __init parse_tsx(const char *s) { @@ -45,6 +46,30 @@ void tsx_init(void) rdmsrl(MSR_ARCH_CAPABILITIES, caps); cpu_has_tsx_ctrl = !!(caps & ARCH_CAPS_TSX_CTRL); + + /* + * The TSX features (HLE/RTM) are handled specially. They both + * enumerate features but, on certain parts, have mechanisms to be + * hidden without disrupting running software. + * + * At the moment, we're running in an unknown context (WRT hiding - + * particularly if another fully fledged kernel ran before us) and + * depending on user settings, may elect to continue hiding them from + * native CPUID instructions. + * + * Xen doesn't use TSX itself, but use cpu_has_{hle,rtm} for various + * system reasons, mostly errata detection, so the meaning is more + * useful as "TSX infrastructure available", as opposed to "features + * advertised and working". + * + * Force the features to be visible in Xen's view if we see any of the + * infrastructure capable of hiding them. + */ + if ( cpu_has_tsx_ctrl ) + { + setup_force_cpu_cap(X86_FEATURE_HLE); + setup_force_cpu_cap(X86_FEATURE_RTM); + } } if ( cpu_has_tsx_ctrl ) @@ -53,9 +78,11 @@ void tsx_init(void) rdmsrl(MSR_TSX_CTRL, val); + /* Check bottom bit only. Higher bits are various sentinels. */ + rtm_disabled = !(opt_tsx & 1); + val &= ~(TSX_CTRL_RTM_DISABLE | TSX_CTRL_CPUID_CLEAR); - /* Check bottom bit only. Higher bits are various sentinals. */ - if ( !(opt_tsx & 1) ) + if ( rtm_disabled ) val |= TSX_CTRL_RTM_DISABLE | TSX_CTRL_CPUID_CLEAR; wrmsrl(MSR_TSX_CTRL, val); diff --git a/xen/include/asm-x86/processor.h b/xen/include/asm-x86/processor.h index 9acb80fdcd..f8e1e4d523 100644 --- a/xen/include/asm-x86/processor.h +++ b/xen/include/asm-x86/processor.h @@ -629,6 +629,7 @@ static inline uint8_t get_cpu_family(uint32_t raw, uint8_t *model, } extern int8_t opt_tsx, cpu_has_tsx_ctrl; +extern bool rtm_disabled; void tsx_init(void); #endif /* !__ASSEMBLY__ */ diff --git a/xen/include/public/arch-x86/cpufeatureset.h b/xen/include/public/arch-x86/cpufeatureset.h index d06a2fd4c8..02b3a2a265 100644 --- a/xen/include/public/arch-x86/cpufeatureset.h +++ b/xen/include/public/arch-x86/cpufeatureset.h @@ -197,14 +197,14 @@ XEN_CPUFEATURE(FSGSBASE, 5*32+ 0) /*A {RD,WR}{FS,GS}BASE instructions */ XEN_CPUFEATURE(TSC_ADJUST, 5*32+ 1) /*S TSC_ADJUST MSR available */ XEN_CPUFEATURE(SGX, 5*32+ 2) /* Software Guard extensions */ XEN_CPUFEATURE(BMI1, 5*32+ 3) /*A 1st bit manipulation extensions */ -XEN_CPUFEATURE(HLE, 5*32+ 4) /*A Hardware Lock Elision */ +XEN_CPUFEATURE(HLE, 5*32+ 4) /*!a Hardware Lock Elision */ XEN_CPUFEATURE(AVX2, 5*32+ 5) /*A AVX2 instructions */ XEN_CPUFEATURE(FDP_EXCP_ONLY, 5*32+ 6) /*! x87 FDP only updated on exception. */ XEN_CPUFEATURE(SMEP, 5*32+ 7) /*S Supervisor Mode Execution Protection */ XEN_CPUFEATURE(BMI2, 5*32+ 8) /*A 2nd bit manipulation extensions */ XEN_CPUFEATURE(ERMS, 5*32+ 9) /*A Enhanced REP MOVSB/STOSB */ XEN_CPUFEATURE(INVPCID, 5*32+10) /*H Invalidate Process Context ID */ -XEN_CPUFEATURE(RTM, 5*32+11) /*A Restricted Transactional Memory */ +XEN_CPUFEATURE(RTM, 5*32+11) /*!A Restricted Transactional Memory */ XEN_CPUFEATURE(PQM, 5*32+12) /* Platform QoS Monitoring */ XEN_CPUFEATURE(NO_FPU_SEL, 5*32+13) /*! FPU CS/DS stored as zero */ XEN_CPUFEATURE(MPX, 5*32+14) /*s Memory Protection Extensions */ -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.14 From xen-changelog-bounces@lists.xenproject.org Sat Jun 05 03:34:34 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Sat, 05 Jun 2021 03:34:34 +0000 Received: from list by lists.xenproject.org with outflank-mailman.137136.254114 (Exim 4.92) (envelope-from ) id 1lpN50-0006rH-Bm; Sat, 05 Jun 2021 03:34:34 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 137136.254114; Sat, 05 Jun 2021 03:34:34 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpN50-0006r9-7Y; Sat, 05 Jun 2021 03:34:34 +0000 Received: by outflank-mailman (input) for mailman id 137136; Sat, 05 Jun 2021 03:34:33 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpN4z-0006q9-0b for xen-changelog@lists.xenproject.org; Sat, 05 Jun 2021 03:34:33 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpN4y-0000qp-W8 for xen-changelog@lists.xenproject.org; Sat, 05 Jun 2021 03:34:32 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lpN4y-0002rL-VM for xen-changelog@lists.xenproject.org; Sat, 05 Jun 2021 03:34:32 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=IiW7QAsGsNnMyhdqIL2x4Z3y8s51LOM+k3hMPHzZmik=; b=bAg+4a+O8DJ883S9MCXzZcuSue bVZ7Ts1pybG7RPY8ZcC6SlUB0E2X4PhQlA7R7Xrffuy/YldwnZB3WqR2jliLkaKX2NId0UDQqE4Mq lQJUmEBjg8OCzAVdIbGDD/YWKoPzyXm3bgHZOJP1E1QYONPUFkPNWjlCRUg/jB4t17C0=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.14] x86: fix build race when generating temporary object files (take 2) Message-Id: Date: Sat, 05 Jun 2021 03:34:32 +0000 commit b046e05736deecbd8254540c5e45444115fb1c98 Author: Jan Beulich AuthorDate: Fri Jun 4 14:57:42 2021 +0200 Commit: Jan Beulich CommitDate: Fri Jun 4 14:57:42 2021 +0200 x86: fix build race when generating temporary object files (take 2) The original commit wasn't quite sufficient: Emptying DEPS is helpful only when nothing will get added to it subsequently. xen/Rules.mk will, after including the local Makefile, amend DEPS by dependencies for objects living in sub-directories though. For the purpose of suppressing dependencies of the makefiles on the .*.d2 files (and thus to avoid their re-generation) it is, however, not necessary at all to play with DEPS. Instead we can override DEPS_INCLUDE (which generally is a late- expansion variable). Fixes: 761bb575ce97 ("x86: fix build race when generating temporary object files") Signed-off-by: Jan Beulich master commit: 8c90dbb99907f3b471d558775777a84daec7c3f6 master date: 2021-05-28 09:12:24 +0200 --- xen/arch/x86/Makefile | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/xen/arch/x86/Makefile b/xen/arch/x86/Makefile index d45b213ab6..bbd11a4726 100644 --- a/xen/arch/x86/Makefile +++ b/xen/arch/x86/Makefile @@ -271,5 +271,5 @@ clean:: # Suppress loading of DEPS files for internal, temporary target files. This # then also suppresses re-generation of the respective .*.d2 files. ifeq ($(filter-out .xen%.o,$(notdir $(MAKECMDGOALS))),) -DEPS:= +DEPS_INCLUDE:= endif -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.14 From xen-changelog-bounces@lists.xenproject.org Sat Jun 05 09:22:12 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Sat, 05 Jun 2021 09:22:12 +0000 Received: from list by lists.xenproject.org with outflank-mailman.137159.254153 (Exim 4.92) (envelope-from ) id 1lpSVI-0004tB-3k; Sat, 05 Jun 2021 09:22:04 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 137159.254153; Sat, 05 Jun 2021 09:22:04 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpSVI-0004t4-0W; Sat, 05 Jun 2021 09:22:04 +0000 Received: by outflank-mailman (input) for mailman id 137159; Sat, 05 Jun 2021 09:22:02 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpSVG-0004sy-Ko for xen-changelog@lists.xenproject.org; Sat, 05 Jun 2021 09:22:02 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpSVG-0007Yw-Hs for xen-changelog@lists.xenproject.org; Sat, 05 Jun 2021 09:22:02 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lpSVG-0002sd-Gr for xen-changelog@lists.xenproject.org; Sat, 05 Jun 2021 09:22:02 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=6p73jMlVLO+63ltM/ucCTDpXYcVq7SXA25PfxOz1dig=; b=xo28pULR/8+AUcB8iE4y54MEX1 ojk+X+HwM+Y/qgABpYBSYM+ZrzzFDrJtAraidAnPIZQ6B69zfNQfeciut+8JUHswDxFMLFSEp5Oy2 MaY30gJrpea6Og8UDdOg7NQ8XznPMvHB7ZSn1cdEE0fafawiFt5mRo65XsBrAiAbBTJI=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.13] VT-d: Don't assume register-based invalidation is always supported Message-Id: Date: Sat, 05 Jun 2021 09:22:02 +0000 commit e289ed670712782f119d18a9f7b56c93680b4b43 Author: Chao Gao AuthorDate: Fri Jun 4 15:00:37 2021 +0200 Commit: Jan Beulich CommitDate: Fri Jun 4 15:00:37 2021 +0200 VT-d: Don't assume register-based invalidation is always supported According to Intel VT-d SPEC rev3.3 Section 6.5, Register-based Invalidation isn't supported by Intel VT-d version 6 and beyond. This hardware change impacts following two scenarios: admin can disable queued invalidation via 'qinval' cmdline and use register-based interface; VT-d switches to register-based invalidation when queued invalidation needs to be disabled, for example, during disabling x2apic or during system suspension or after enabling queued invalidation fails. To deal with this hardware change, if register-based invalidation isn't supported, queued invalidation cannot be disabled through Xen cmdline; and if queued invalidation has to be disabled temporarily in some scenarios, VT-d won't switch to register-based interface but use some dummy functions to catch errors in case there is any invalidation request issued when queued invalidation is disabled. Signed-off-by: Chao Gao Reviewed-by: Jan Beulich Reviewed-by: Kevin Tian master commit: 6773b1a7584a75a486e9774541ad5bd84c9aa5ee master date: 2021-04-26 10:16:50 +0200 --- docs/misc/xen-command-line.pandoc | 4 +++- xen/drivers/passthrough/vtd/iommu.c | 27 +++++++++++++++++++++++++-- xen/drivers/passthrough/vtd/iommu.h | 7 +++++++ xen/drivers/passthrough/vtd/qinval.c | 33 +++++++++++++++++++++++++++++++-- 4 files changed, 66 insertions(+), 5 deletions(-) diff --git a/docs/misc/xen-command-line.pandoc b/docs/misc/xen-command-line.pandoc index 0b6ec8b6a1..cf9dea62db 100644 --- a/docs/misc/xen-command-line.pandoc +++ b/docs/misc/xen-command-line.pandoc @@ -1370,7 +1370,9 @@ The following options are specific to Intel VT-d hardware: * The `qinval` boolean controls the Queued Invalidation sub-feature, and is active by default on compatible hardware. Queued Invalidation is a feature in second-generation IOMMUs and is a functional prerequisite for - Interrupt Remapping. + Interrupt Remapping. Note that Xen disregards this setting for Intel VT-d + version 6 and greater as Registered-Based Invalidation isn't supported + by them. * The `igfx` boolean is active by default, and controls whether the IOMMU in front of an Intel Graphics Device is enabled or not. diff --git a/xen/drivers/passthrough/vtd/iommu.c b/xen/drivers/passthrough/vtd/iommu.c index 01a2a461ba..34ee5d22f1 100644 --- a/xen/drivers/passthrough/vtd/iommu.c +++ b/xen/drivers/passthrough/vtd/iommu.c @@ -1185,6 +1185,14 @@ int __init iommu_alloc(struct acpi_drhd_unit *drhd) iommu->cap = dmar_readq(iommu->reg, DMAR_CAP_REG); iommu->ecap = dmar_readq(iommu->reg, DMAR_ECAP_REG); + iommu->version = dmar_readl(iommu->reg, DMAR_VER_REG); + + if ( !iommu_qinval && !has_register_based_invalidation(iommu) ) + { + printk(XENLOG_WARNING VTDPREFIX "IOMMU %d: cannot disable Queued Invalidation\n", + iommu->index); + iommu_qinval = true; + } if ( iommu_verbose ) { @@ -2158,6 +2166,10 @@ static int __must_check init_vtd_hw(bool resume) */ if ( enable_qinval(iommu) != 0 ) { + /* Ensure register-based invalidation is available */ + if ( !has_register_based_invalidation(iommu) ) + return -EIO; + iommu->flush.context = vtd_flush_context_reg; iommu->flush.iotlb = vtd_flush_iotlb_reg; } @@ -2248,6 +2260,7 @@ static int __init vtd_setup(void) struct acpi_drhd_unit *drhd; struct vtd_iommu *iommu; int ret; + bool reg_inval_supported = true; if ( list_empty(&acpi_drhd_units) ) { @@ -2269,8 +2282,8 @@ static int __init vtd_setup(void) } /* We enable the following features only if they are supported by all VT-d - * engines: Snoop Control, DMA passthrough, Queued Invalidation, Interrupt - * Remapping, and Posted Interrupt + * engines: Snoop Control, DMA passthrough, Register-based Invalidation, + * Queued Invalidation, Interrupt Remapping, and Posted Interrupt. */ for_each_drhd_unit ( drhd ) { @@ -2290,6 +2303,9 @@ static int __init vtd_setup(void) if ( iommu_qinval && !ecap_queued_inval(iommu->ecap) ) iommu_qinval = 0; + if ( !has_register_based_invalidation(iommu) ) + reg_inval_supported = false; + if ( iommu_intremap && !ecap_intr_remap(iommu->ecap) ) iommu_intremap = 0; @@ -2314,6 +2330,13 @@ static int __init vtd_setup(void) softirq_tasklet_init(&vtd_fault_tasklet, do_iommu_page_fault, 0); + if ( !iommu_qinval && !reg_inval_supported ) + { + dprintk(XENLOG_ERR VTDPREFIX, "No available invalidation interface\n"); + ret = -ENODEV; + goto error; + } + if ( !iommu_qinval && iommu_intremap ) { iommu_intremap = 0; diff --git a/xen/drivers/passthrough/vtd/iommu.h b/xen/drivers/passthrough/vtd/iommu.h index 216791b3d6..57737b1a4a 100644 --- a/xen/drivers/passthrough/vtd/iommu.h +++ b/xen/drivers/passthrough/vtd/iommu.h @@ -540,6 +540,7 @@ struct vtd_iommu { struct list_head ats_devices; unsigned long *domid_bitmap; /* domain id bitmap */ u16 *domid_map; /* domain id mapping array */ + uint32_t version; }; #define INTEL_IOMMU_DEBUG(fmt, args...) \ @@ -549,4 +550,10 @@ struct vtd_iommu { dprintk(XENLOG_WARNING VTDPREFIX, fmt, ## args); \ } while(0) +/* Register-based invalidation isn't supported by VT-d version 6 and beyond. */ +static inline bool has_register_based_invalidation(const struct vtd_iommu *vtd) +{ + return VER_MAJOR(vtd->version) < 6; +} + #endif diff --git a/xen/drivers/passthrough/vtd/qinval.c b/xen/drivers/passthrough/vtd/qinval.c index 764ef9f0fc..dfcbb66c8a 100644 --- a/xen/drivers/passthrough/vtd/qinval.c +++ b/xen/drivers/passthrough/vtd/qinval.c @@ -442,6 +442,23 @@ int enable_qinval(struct vtd_iommu *iommu) return 0; } +static int vtd_flush_context_noop(struct vtd_iommu *iommu, uint16_t did, + uint16_t source_id, uint8_t function_mask, + uint64_t type, bool flush_non_present_entry) +{ + WARN(); + return -EIO; +} + +static int vtd_flush_iotlb_noop(struct vtd_iommu *iommu, uint16_t did, + uint64_t addr, unsigned int size_order, + uint64_t type, bool flush_non_present_entry, + bool flush_dev_iotlb) +{ + WARN(); + return -EIO; +} + void disable_qinval(struct vtd_iommu *iommu) { u32 sts; @@ -463,6 +480,18 @@ void disable_qinval(struct vtd_iommu *iommu) out: spin_unlock_irqrestore(&iommu->register_lock, flags); - iommu->flush.context = vtd_flush_context_reg; - iommu->flush.iotlb = vtd_flush_iotlb_reg; + /* + * Assign callbacks to noop to catch errors if register-based invalidation + * isn't supported. + */ + if ( has_register_based_invalidation(iommu) ) + { + iommu->flush.context = vtd_flush_context_reg; + iommu->flush.iotlb = vtd_flush_iotlb_reg; + } + else + { + iommu->flush.context = vtd_flush_context_noop; + iommu->flush.iotlb = vtd_flush_iotlb_noop; + } } -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.13 From xen-changelog-bounces@lists.xenproject.org Sat Jun 05 09:22:14 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Sat, 05 Jun 2021 09:22:14 +0000 Received: from list by lists.xenproject.org with outflank-mailman.137160.254158 (Exim 4.92) (envelope-from ) id 1lpSVS-0004vb-7L; Sat, 05 Jun 2021 09:22:14 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 137160.254158; Sat, 05 Jun 2021 09:22:14 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpSVS-0004vU-4D; Sat, 05 Jun 2021 09:22:14 +0000 Received: by outflank-mailman (input) for mailman id 137160; Sat, 05 Jun 2021 09:22:12 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpSVQ-0004vC-Lp for xen-changelog@lists.xenproject.org; Sat, 05 Jun 2021 09:22:12 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpSVQ-0007Z0-L0 for xen-changelog@lists.xenproject.org; Sat, 05 Jun 2021 09:22:12 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lpSVQ-0002te-K6 for xen-changelog@lists.xenproject.org; Sat, 05 Jun 2021 09:22:12 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=LGcpVw/JkZbd0QDYBLJZ5a5OhNR/Yb5jWgsvV3Q9Jc0=; b=uRdzF2lUbiozNp/4P4X60XFNdI WnFPgEQu6UAmTpCQ/bcl224wu0qSXT4MAY9vTrrRJhcs3niFXEUQgvFqHBsbefemXLlmkWgJMYUCl TuuvI4ct4WYbG8yUFPTcjbejpm/Dghj751WDM2Ev2AD2A9CVVwGJaA1ext+vC2ozVmgc=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.13] x86/vhpet: fix RTC special casing Message-Id: Date: Sat, 05 Jun 2021 09:22:12 +0000 commit 287f2292c38575dcf059358ce3d9a70294017554 Author: Roger Pau Monné AuthorDate: Fri Jun 4 15:01:33 2021 +0200 Commit: Jan Beulich CommitDate: Fri Jun 4 15:01:33 2021 +0200 x86/vhpet: fix RTC special casing Restore setting the virtual timer callback private data to NULL if the timer is not level triggered. This fixes the special casing done in pt_update_irq so that the RTC interrupt when originating from the HPET is suspended if the interrupt source is masked. Note the RTC special casing done in pt_update_irq should only apply to the RTC interrupt originating from the emulated RTC device (which does set the callback private data), as in that case the callback itself will destroy the virtual timer if the interrupt is ignored. While there also use RTC_IRQ instead of 8 when the HPET is configured in LegacyReplacement Mode. Fixes: be07023be115 ("x86/vhpet: add support for level triggered interrupts") Reported-by: Jan Beulich Signed-off-by: Roger Pau Monné Reviewed-by: Jan Beulich master commit: 472a13988a051e5ae84b95815c6caf4378062abe master date: 2021-05-07 10:43:29 +0200 --- xen/arch/x86/hvm/hpet.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/xen/arch/x86/hvm/hpet.c b/xen/arch/x86/hvm/hpet.c index 9fbdc5302b..1dcd38cebc 100644 --- a/xen/arch/x86/hvm/hpet.c +++ b/xen/arch/x86/hvm/hpet.c @@ -22,6 +22,7 @@ #include #include #include +#include #include #include #include @@ -289,7 +290,7 @@ static void hpet_set_timer(HPETState *h, unsigned int tn, /* if LegacyReplacementRoute bit is set, HPET specification requires timer0 be routed to IRQ0 in NON-APIC or IRQ2 in the I/O APIC, timer1 be routed to IRQ8 in NON-APIC or IRQ8 in the I/O APIC. */ - irq = (tn == 0) ? 0 : 8; + irq = (tn == 0) ? 0 : RTC_IRQ; h->pt[tn].source = PTSRC_isa; } else @@ -317,7 +318,8 @@ static void hpet_set_timer(HPETState *h, unsigned int tn, hpet_tick_to_ns(h, diff), oneshot ? 0 : hpet_tick_to_ns(h, h->hpet.period[tn]), irq, timer_level(h, tn) ? hpet_timer_fired : NULL, - (void *)(unsigned long)tn, timer_level(h, tn)); + timer_level(h, tn) ? (void *)(unsigned long)tn : NULL, + timer_level(h, tn)); } static inline uint64_t hpet_fixup_reg( -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.13 From xen-changelog-bounces@lists.xenproject.org Sat Jun 05 09:22:24 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Sat, 05 Jun 2021 09:22:24 +0000 Received: from list by lists.xenproject.org with outflank-mailman.137161.254162 (Exim 4.92) (envelope-from ) id 1lpSVc-0004ye-9a; Sat, 05 Jun 2021 09:22:24 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 137161.254162; Sat, 05 Jun 2021 09:22:24 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpSVc-0004yW-62; Sat, 05 Jun 2021 09:22:24 +0000 Received: by outflank-mailman (input) for mailman id 137161; Sat, 05 Jun 2021 09:22:22 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpSVa-0004xu-Ox for xen-changelog@lists.xenproject.org; Sat, 05 Jun 2021 09:22:22 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpSVa-0007ZI-OB for xen-changelog@lists.xenproject.org; Sat, 05 Jun 2021 09:22:22 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lpSVa-0002uP-N5 for xen-changelog@lists.xenproject.org; Sat, 05 Jun 2021 09:22:22 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=SEa/TQBoaIaVicRHEu+mtDUSIkZH6F11VunE3RQ2oBs=; b=TcmFR2Ed/iViQKEd5R3Td9LvLr OdlCmN8TaIHnMYeNhBmdJd1xUUEJ7vhtX2DEGRbxoD7rkwocrCJ0eGjnblKieeTUPtBfKdXaJ7Lw4 kY5nROBQrJB4j3kJVAPOhbkWSMihswbjHk+7vMrNGZKvnKagaA49m9exYfjsZ7RtPdi8=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.13] x86emul: fix test harness build for gas 2.36 Message-Id: Date: Sat, 05 Jun 2021 09:22:22 +0000 commit 4d544141ebab31dabb542c7dbacad658c5d3757c Author: Jan Beulich AuthorDate: Fri Jun 4 15:02:13 2021 +0200 Commit: Jan Beulich CommitDate: Fri Jun 4 15:02:13 2021 +0200 x86emul: fix test harness build for gas 2.36 All of the sudden, besides .text and .rodata and alike, an always present .note.gnu.property section has appeared. This section, when converting to binary format output, gets placed according to its linked address, causing the resulting blobs to be about 128Mb in size. The resulting headers with a C representation of the binary blobs then are, of course all a multiple of that size (and take accordingly long to create). I didn't bother waiting to see what size the final test_x86_emulator binary then would have had. See also https://sourceware.org/bugzilla/show_bug.cgi?id=27753. Rather than figuring out whether gas supports -mx86-used-note=, simply remove the section while creating *.bin. Signed-off-by: Jan Beulich master commit: aa803ba38a867551917d11059eaa044955556e05 master date: 2021-05-17 15:41:28 +0200 --- tools/tests/x86_emulator/testcase.mk | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/tools/tests/x86_emulator/testcase.mk b/tools/tests/x86_emulator/testcase.mk index dafeb6caf7..d4aa82e3f3 100644 --- a/tools/tests/x86_emulator/testcase.mk +++ b/tools/tests/x86_emulator/testcase.mk @@ -12,11 +12,11 @@ all: $(TESTCASE).bin %.bin: %.c $(CC) $(filter-out -M% .%,$(CFLAGS)) -c $< $(LD) $(LDFLAGS_DIRECT) -N -Ttext 0x100000 -o $*.tmp $*.o - $(OBJCOPY) -O binary $*.tmp $@ + $(OBJCOPY) -O binary -R .note.gnu.property $*.tmp $@ rm -f $*.tmp %-opmask.bin: opmask.S $(CC) $(filter-out -M% .%,$(CFLAGS)) -c $< -o $(basename $@).o $(LD) $(LDFLAGS_DIRECT) -N -Ttext 0x100000 -o $(basename $@).tmp $(basename $@).o - $(OBJCOPY) -O binary $(basename $@).tmp $@ + $(OBJCOPY) -O binary -R .note.gnu.property $(basename $@).tmp $@ rm -f $(basename $@).tmp -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.13 From xen-changelog-bounces@lists.xenproject.org Sat Jun 05 09:22:34 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Sat, 05 Jun 2021 09:22:34 +0000 Received: from list by lists.xenproject.org with outflank-mailman.137162.254166 (Exim 4.92) (envelope-from ) id 1lpSVm-00051y-B2; Sat, 05 Jun 2021 09:22:34 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 137162.254166; Sat, 05 Jun 2021 09:22:34 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpSVm-00051q-82; Sat, 05 Jun 2021 09:22:34 +0000 Received: by outflank-mailman (input) for mailman id 137162; Sat, 05 Jun 2021 09:22:32 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpSVk-00051e-Sm for xen-changelog@lists.xenproject.org; Sat, 05 Jun 2021 09:22:32 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpSVk-0007ZV-Rw for xen-changelog@lists.xenproject.org; Sat, 05 Jun 2021 09:22:32 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lpSVk-0002v4-Qt for xen-changelog@lists.xenproject.org; Sat, 05 Jun 2021 09:22:32 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=OtrFUrsKvK3jByknve7byS6v530B0o6HVZBFoefsGmU=; b=DRf0sj5Fk4CwaNEoyfs5fJjihv c2GvwGY6YRYWFeVTFkz0+y63de08iol/nKit6wTemCr8y8F4TzEnxwnuaaeYiSUcYpEawammSl1G6 Xckb5JZHwOkndCH721Zy4PiPU8E4V0NaaNxt2Zv2/XryrqY0YBZAKUEUNJv8S7UlHfmA=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.13] x86: make hypervisor build with gcc11 Message-Id: Date: Sat, 05 Jun 2021 09:22:32 +0000 commit fa5afbbc20ef3577c5338f9d0b24dad45cef59cd Author: Jan Beulich AuthorDate: Fri Jun 4 15:02:33 2021 +0200 Commit: Jan Beulich CommitDate: Fri Jun 4 15:02:33 2021 +0200 x86: make hypervisor build with gcc11 Gcc 11 looks to make incorrect assumptions about valid ranges that pointers may be used for addressing when they are derived from e.g. a plain constant. See https://gcc.gnu.org/bugzilla/show_bug.cgi?id=100680. Utilize RELOC_HIDE() to work around the issue, which for x86 manifests in at least - mpparse.c:efi_check_config(), - tboot.c:tboot_probe(), - tboot.c:tboot_gen_frametable_integrity(), - x86_emulate.c:x86_emulate() (at -O2 only). The last case is particularly odd not just because it only triggers at higher optimization levels, but also because it only affects one of at least three similar constructs. Various "note" diagnostics claim the valid index range to be [0, 2⁶³-1]. Signed-off-by: Jan Beulich Tested-by: Jason Andryuk Acked-by: Roger Pau Monné master commit: 722f59d38c710a940ab05e542a83020eb5546dea master date: 2021-05-27 14:40:29 +0200 --- tools/tests/x86_emulator/x86-emulate.c | 7 +++++++ xen/arch/x86/x86_emulate/x86_emulate.c | 2 +- xen/include/asm-x86/fixmap.h | 2 +- xen/include/xen/compiler.h | 6 ++++++ xen/include/xen/pdx.h | 2 +- 5 files changed, 16 insertions(+), 3 deletions(-) diff --git a/tools/tests/x86_emulator/x86-emulate.c b/tools/tests/x86_emulator/x86-emulate.c index 23dcbe4247..ecb4b6245e 100644 --- a/tools/tests/x86_emulator/x86-emulate.c +++ b/tools/tests/x86_emulator/x86-emulate.c @@ -2,6 +2,13 @@ #include +/* See gcc bug 100680, but here don't bother making this version dependent. */ +#define gcc11_wrap(x) ({ \ + unsigned long x_; \ + __asm__ ( "" : "=g" (x_) : "0" (x) ); \ + (typeof(x))x_; \ +}) + #define cpu_has_amd_erratum(nr) 0 #define cpu_has_mpx false #define read_bndcfgu() 0 diff --git a/xen/arch/x86/x86_emulate/x86_emulate.c b/xen/arch/x86/x86_emulate/x86_emulate.c index bba6dd0187..2dee940adb 100644 --- a/xen/arch/x86/x86_emulate/x86_emulate.c +++ b/xen/arch/x86/x86_emulate/x86_emulate.c @@ -723,7 +723,7 @@ union vex { #define copy_VEX(ptr, vex) ({ \ if ( !mode_64bit() ) \ (vex).reg |= 8; \ - (ptr)[0 - PFX_BYTES] = ext < ext_8f08 ? 0xc4 : 0x8f; \ + gcc11_wrap(ptr)[0 - PFX_BYTES] = ext < ext_8f08 ? 0xc4 : 0x8f; \ (ptr)[1 - PFX_BYTES] = (vex).raw[0]; \ (ptr)[2 - PFX_BYTES] = (vex).raw[1]; \ container_of((ptr) + 1 - PFX_BYTES, typeof(vex), raw[0]); \ diff --git a/xen/include/asm-x86/fixmap.h b/xen/include/asm-x86/fixmap.h index 9fb2f47946..b6b981b621 100644 --- a/xen/include/asm-x86/fixmap.h +++ b/xen/include/asm-x86/fixmap.h @@ -78,7 +78,7 @@ extern void __set_fixmap( #define clear_fixmap(idx) __set_fixmap(idx, 0, 0) -#define __fix_to_virt(x) (FIXADDR_TOP - ((x) << PAGE_SHIFT)) +#define __fix_to_virt(x) gcc11_wrap(FIXADDR_TOP - ((x) << PAGE_SHIFT)) #define __virt_to_fix(x) ((FIXADDR_TOP - ((x)&PAGE_MASK)) >> PAGE_SHIFT) #define fix_to_virt(x) ((void *)__fix_to_virt(x)) diff --git a/xen/include/xen/compiler.h b/xen/include/xen/compiler.h index ff6c0f5cdd..85cbd1ab00 100644 --- a/xen/include/xen/compiler.h +++ b/xen/include/xen/compiler.h @@ -99,6 +99,12 @@ __asm__ ("" : "=r"(__ptr) : "0"(ptr)); \ (typeof(ptr)) (__ptr + (off)); }) +#if CONFIG_GCC_VERSION >= 110000 /* See gcc bug 100680. */ +# define gcc11_wrap(x) RELOC_HIDE(x, 0) +#else +# define gcc11_wrap(x) (x) +#endif + #ifdef __GCC_ASM_FLAG_OUTPUTS__ # define ASM_FLAG_OUT(yes, no) yes #else diff --git a/xen/include/xen/pdx.h b/xen/include/xen/pdx.h index 770fadc06c..9fcfb0ce52 100644 --- a/xen/include/xen/pdx.h +++ b/xen/include/xen/pdx.h @@ -19,7 +19,7 @@ extern u64 pdx_region_mask(u64 base, u64 len); extern void set_pdx_range(unsigned long smfn, unsigned long emfn); #define page_to_pdx(pg) ((pg) - frame_table) -#define pdx_to_page(pdx) (frame_table + (pdx)) +#define pdx_to_page(pdx) gcc11_wrap(frame_table + (pdx)) bool __mfn_valid(unsigned long mfn); -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.13 From xen-changelog-bounces@lists.xenproject.org Sat Jun 05 09:22:44 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Sat, 05 Jun 2021 09:22:44 +0000 Received: from list by lists.xenproject.org with outflank-mailman.137163.254170 (Exim 4.92) (envelope-from ) id 1lpSVw-00055X-CY; Sat, 05 Jun 2021 09:22:44 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 137163.254170; Sat, 05 Jun 2021 09:22:44 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpSVw-00055P-9Z; Sat, 05 Jun 2021 09:22:44 +0000 Received: by outflank-mailman (input) for mailman id 137163; Sat, 05 Jun 2021 09:22:43 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpSVu-00055A-Ve for xen-changelog@lists.xenproject.org; Sat, 05 Jun 2021 09:22:42 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpSVu-0007Zf-Ut for xen-changelog@lists.xenproject.org; Sat, 05 Jun 2021 09:22:42 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lpSVu-0002vh-U1 for xen-changelog@lists.xenproject.org; Sat, 05 Jun 2021 09:22:42 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=UubSxj7Jvx6DusGwjYClmZUZWaewdsrZ2zsegkfdtr4=; b=l8xd/WzDVD4P8lXpev0fCokacF 5Ek8SYRI1m8RjFRvVj2WSYH4H8f0J4/O2O05lU0zEk0OU8/WEMCUNcGLxM4pMGQSD19iu8B4tH4kc 79O5uy0/CgbjWHsAwXrH453d/qyRf8yN2EAemF5hJEMPNgzqUpM16gtpJcmymGfGtYjo=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.13] x86/cpuid: Rework HLE and RTM handling Message-Id: Date: Sat, 05 Jun 2021 09:22:42 +0000 commit f17d848c4caa4fa562f3121c430667b6df838700 Author: Andrew Cooper AuthorDate: Fri Jun 4 15:04:56 2021 +0200 Commit: Jan Beulich CommitDate: Fri Jun 4 15:04:56 2021 +0200 x86/cpuid: Rework HLE and RTM handling The TAA mitigation offered the option to hide the HLE and RTM CPUID bits, which has caused some migration compatibility problems. These two bits are special. Annotate them with ! to emphasise this point. Hardware Lock Elision (HLE) may or may not be visible in CPUID, but is disabled in microcode on all CPUs, and has been removed from the architecture. Do not advertise it to VMs by default. Restricted Transactional Memory (RTM) may or may not be visible in CPUID, and may or may not be configured in force-abort mode. Have tsx_init() note whether RTM has been configured into force-abort mode, so guest_common_feature_adjustments() can conditionally hide it from VMs by default. The host policy values for HLE/RTM may or may not be set, depending on any previous running kernel's choice of visibility, and Xen's choice. TSX is available on any CPU which enumerates a TSX-hiding mechanism, so instead of doing a two-step to clobber any hiding, scan CPUID, then set the visibility, just force visibility of the bits in the first place. With the HLE/RTM bits now unilaterally visible in the host policy, xc_cpuid_apply_policy() can construct a more appropriate policy out of thin air for pre-4.13 VMs with no CPUID data in their migration stream, and specifically one where HLE/RTM doesn't potentially disappear behind the back of a running VM. Fixes: 8c4330818f6 ("x86/spec-ctrl: Mitigate the TSX Asynchronous Abort sidechannel") Signed-off-by: Andrew Cooper Reviewed-by: Jan Beulich Reviewed-by: Roger Pau Monné master commit: 8fe24090d940d760145ccd5e234290be7418b175 master date: 2021-05-27 19:34:00 +0100 --- xen/arch/x86/tsx.c | 24 ++++++++++++++++++++++++ xen/include/public/arch-x86/cpufeatureset.h | 4 ++-- 2 files changed, 26 insertions(+), 2 deletions(-) diff --git a/xen/arch/x86/tsx.c b/xen/arch/x86/tsx.c index 2d202a0d4e..a5d5944fe8 100644 --- a/xen/arch/x86/tsx.c +++ b/xen/arch/x86/tsx.c @@ -44,6 +44,30 @@ void tsx_init(void) rdmsrl(MSR_ARCH_CAPABILITIES, caps); cpu_has_tsx_ctrl = !!(caps & ARCH_CAPS_TSX_CTRL); + + /* + * The TSX features (HLE/RTM) are handled specially. They both + * enumerate features but, on certain parts, have mechanisms to be + * hidden without disrupting running software. + * + * At the moment, we're running in an unknown context (WRT hiding - + * particularly if another fully fledged kernel ran before us) and + * depending on user settings, may elect to continue hiding them from + * native CPUID instructions. + * + * Xen doesn't use TSX itself, but use cpu_has_{hle,rtm} for various + * system reasons, mostly errata detection, so the meaning is more + * useful as "TSX infrastructure available", as opposed to "features + * advertised and working". + * + * Force the features to be visible in Xen's view if we see any of the + * infrastructure capable of hiding them. + */ + if ( cpu_has_tsx_ctrl ) + { + setup_force_cpu_cap(X86_FEATURE_HLE); + setup_force_cpu_cap(X86_FEATURE_RTM); + } } if ( cpu_has_tsx_ctrl ) diff --git a/xen/include/public/arch-x86/cpufeatureset.h b/xen/include/public/arch-x86/cpufeatureset.h index 3162ee0378..a1d1619643 100644 --- a/xen/include/public/arch-x86/cpufeatureset.h +++ b/xen/include/public/arch-x86/cpufeatureset.h @@ -195,14 +195,14 @@ XEN_CPUFEATURE(FSGSBASE, 5*32+ 0) /*A {RD,WR}{FS,GS}BASE instructions */ XEN_CPUFEATURE(TSC_ADJUST, 5*32+ 1) /*S TSC_ADJUST MSR available */ XEN_CPUFEATURE(SGX, 5*32+ 2) /* Software Guard extensions */ XEN_CPUFEATURE(BMI1, 5*32+ 3) /*A 1st bit manipulation extensions */ -XEN_CPUFEATURE(HLE, 5*32+ 4) /*A Hardware Lock Elision */ +XEN_CPUFEATURE(HLE, 5*32+ 4) /*!A Hardware Lock Elision */ XEN_CPUFEATURE(AVX2, 5*32+ 5) /*A AVX2 instructions */ XEN_CPUFEATURE(FDP_EXCP_ONLY, 5*32+ 6) /*! x87 FDP only updated on exception. */ XEN_CPUFEATURE(SMEP, 5*32+ 7) /*S Supervisor Mode Execution Protection */ XEN_CPUFEATURE(BMI2, 5*32+ 8) /*A 2nd bit manipulation extensions */ XEN_CPUFEATURE(ERMS, 5*32+ 9) /*A Enhanced REP MOVSB/STOSB */ XEN_CPUFEATURE(INVPCID, 5*32+10) /*H Invalidate Process Context ID */ -XEN_CPUFEATURE(RTM, 5*32+11) /*A Restricted Transactional Memory */ +XEN_CPUFEATURE(RTM, 5*32+11) /*!A Restricted Transactional Memory */ XEN_CPUFEATURE(PQM, 5*32+12) /* Platform QoS Monitoring */ XEN_CPUFEATURE(NO_FPU_SEL, 5*32+13) /*! FPU CS/DS stored as zero */ XEN_CPUFEATURE(MPX, 5*32+14) /*S Memory Protection Extensions */ -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.13 From xen-changelog-bounces@lists.xenproject.org Sat Jun 05 09:22:54 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Sat, 05 Jun 2021 09:22:54 +0000 Received: from list by lists.xenproject.org with outflank-mailman.137166.254182 (Exim 4.92) (envelope-from ) id 1lpSW6-0005Fd-1U; Sat, 05 Jun 2021 09:22:54 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 137166.254182; Sat, 05 Jun 2021 09:22:53 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpSW5-0005Eg-Sr; Sat, 05 Jun 2021 09:22:53 +0000 Received: by outflank-mailman (input) for mailman id 137166; Sat, 05 Jun 2021 09:22:53 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpSW5-00058A-2O for xen-changelog@lists.xenproject.org; Sat, 05 Jun 2021 09:22:53 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lpSW5-0007aI-1d for xen-changelog@lists.xenproject.org; Sat, 05 Jun 2021 09:22:53 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lpSW5-0002wK-0b for xen-changelog@lists.xenproject.org; Sat, 05 Jun 2021 09:22:53 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=aD9cHZtQAGrmZGsPGBn5ENluAh5RHx5/iIYX6jXM+n4=; b=mMK59dB8hZS7SAZGHneSUeDFRe 7AUu3viWGXpu8k6Bftf5l5Hg7ZP5QTJv6mt0xnsQjD87/esafHGXMg22ylzZ95w4izjE3aWD+49yp gMLtZqTk9d01MC9bVv1YsHleEVfxta4mhS3Atlyz3qoRKbopbcL8iGIDfNsUlYZ4WODI=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.13] x86: fix build race when generating temporary object files (take 2) Message-Id: Date: Sat, 05 Jun 2021 09:22:53 +0000 commit ef8b2357d83442b5d2e7607379a935d4f8b35416 Author: Jan Beulich AuthorDate: Fri Jun 4 15:05:38 2021 +0200 Commit: Jan Beulich CommitDate: Fri Jun 4 15:05:38 2021 +0200 x86: fix build race when generating temporary object files (take 2) The original commit wasn't quite sufficient: Emptying DEPS is helpful only when nothing will get added to it subsequently. xen/Rules.mk will, after including the local Makefile, amend DEPS by dependencies for objects living in sub-directories though. For the purpose of suppressing dependencies of the makefiles on the .*.d2 files (and thus to avoid their re-generation) it is, however, not necessary at all to play with DEPS. Instead we can override DEPS_INCLUDE (which generally is a late- expansion variable). Fixes: 761bb575ce97 ("x86: fix build race when generating temporary object files") Signed-off-by: Jan Beulich master commit: 8c90dbb99907f3b471d558775777a84daec7c3f6 master date: 2021-05-28 09:12:24 +0200 --- xen/arch/x86/Makefile | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/xen/arch/x86/Makefile b/xen/arch/x86/Makefile index c57ae581dd..3cca74d61b 100644 --- a/xen/arch/x86/Makefile +++ b/xen/arch/x86/Makefile @@ -269,5 +269,5 @@ clean:: # Suppress loading of DEPS files for internal, temporary target files. This # then also suppresses re-generation of the respective .*.d2 files. ifeq ($(filter-out .xen%.o,$(notdir $(MAKECMDGOALS))),) -DEPS:= +DEPS_INCLUDE:= endif -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.13 From xen-changelog-bounces@lists.xenproject.org Mon Jun 07 12:22:14 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Mon, 07 Jun 2021 12:22:14 +0000 Received: from list by lists.xenproject.org with outflank-mailman.137907.255403 (Exim 4.92) (envelope-from ) id 1lqEGg-0006Ei-2b; Mon, 07 Jun 2021 12:22:10 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 137907.255403; Mon, 07 Jun 2021 12:22:10 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqEGf-0006Ea-Vr; Mon, 07 Jun 2021 12:22:09 +0000 Received: by outflank-mailman (input) for mailman id 137907; Mon, 07 Jun 2021 12:22:09 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqEGf-0006EU-54 for xen-changelog@lists.xenproject.org; Mon, 07 Jun 2021 12:22:09 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqEGf-0007sv-1r for xen-changelog@lists.xenproject.org; Mon, 07 Jun 2021 12:22:09 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqEGe-0003V9-V1 for xen-changelog@lists.xenproject.org; Mon, 07 Jun 2021 12:22:08 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=VQu61LZPqnd+zu0OntZxJMEEG8UqVHXman36v4Hor1E=; b=g8cPZhXjDrGO88EtAavFHjRoo7 4ZPCkBEMQnxO/9KSTflFl0z62GhA0eaGrEHHyVnFOUiWDMaziC12868t0AqdfNC305ufVVretd49q 67xxDDZZvuuJs5nLsLs6xTqaJnlu7EyL9pQGlZAIePSkX7nOExFN1Y7Z2TWIXeZIKmng=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] credit2: make sure we pick a runnable unit from the runq if there is one Message-Id: Date: Mon, 07 Jun 2021 12:22:08 +0000 commit 07b0eb5d0ef0be154606aa46b5b4c5c59b158505 Author: Dario Faggioli AuthorDate: Fri May 28 17:12:48 2021 +0200 Commit: George Dunlap CommitDate: Mon Jun 7 13:16:36 2021 +0100 credit2: make sure we pick a runnable unit from the runq if there is one A !runnable unit (temporarily) present in the runq may cause us to stop scanning the runq itself too early. Of course, we don't run any non-runnable vCPUs, but we end the scan and we fallback to picking the idle unit. In other word, this prevent us to find there and pick the actual unit that we're meant to start running (which might be further ahead in the runq). Depending on the vCPU pinning configuration, this may lead to such unit to be stuck in the runq for long time, causing malfunctioning inside the guest. Fix this by checking runnable/non-runnable status up-front, in the runq scanning function. Reported-by: Michał Leszczyński Reported-by: Dion Kant Signed-off-by: Dario Faggioli Reviewed-by: George Dunlap --- xen/common/sched/credit2.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) diff --git a/xen/common/sched/credit2.c b/xen/common/sched/credit2.c index eb5e5a78c5..f5c1e5b944 100644 --- a/xen/common/sched/credit2.c +++ b/xen/common/sched/credit2.c @@ -3463,6 +3463,10 @@ runq_candidate(struct csched2_runqueue_data *rqd, (unsigned char *)&d); } + /* Skip non runnable units that we (temporarily) have in the runq */ + if ( unlikely(!unit_runnable_state(svc->unit)) ) + continue; + /* Only consider vcpus that are allowed to run on this processor. */ if ( !cpumask_test_cpu(cpu, svc->unit->cpu_hard_affinity) ) continue; @@ -3496,8 +3500,7 @@ runq_candidate(struct csched2_runqueue_data *rqd, * some budget, then choose it. */ if ( (yield || svc->credit > snext->credit) && - (!has_cap(svc) || unit_grab_budget(svc)) && - unit_runnable_state(svc->unit) ) + (!has_cap(svc) || unit_grab_budget(svc)) ) snext = svc; /* In any case, if we got this far, break. */ -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Mon Jun 07 12:22:21 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Mon, 07 Jun 2021 12:22:21 +0000 Received: from list by lists.xenproject.org with outflank-mailman.137908.255407 (Exim 4.92) (envelope-from ) id 1lqEGr-0006HE-46; Mon, 07 Jun 2021 12:22:21 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 137908.255407; Mon, 07 Jun 2021 12:22:21 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqEGr-0006H6-1F; Mon, 07 Jun 2021 12:22:21 +0000 Received: by outflank-mailman (input) for mailman id 137908; Mon, 07 Jun 2021 12:22:19 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqEGp-0006Gq-7P for xen-changelog@lists.xenproject.org; Mon, 07 Jun 2021 12:22:19 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqEGp-0007t4-6S for xen-changelog@lists.xenproject.org; Mon, 07 Jun 2021 12:22:19 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqEGp-0003WC-5B for xen-changelog@lists.xenproject.org; Mon, 07 Jun 2021 12:22:19 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=hUj+eYnFco1sDn75LlBw9WHycD7R1PlsDW2FJIiXeGY=; b=TsVQuIv/D/BXF8Poll9SPEhnPd 3AGiYzcvSkyoLUgJs1VFZcdKLlAPEEoijfTWag20a7opG1Y9kxZxe7Xkd1g8UXZs30nPGm/xVyi8e n1fLscTESWDl7WZqCc0o4pD1+KJBzZWi79umsfa/WljbjjvdC1hHDzpPXa9juF5w0Rws=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] xen: credit2: fix per-entity load tracking when continuing running Message-Id: Date: Mon, 07 Jun 2021 12:22:19 +0000 commit 89052b9fa24bf976924e40918fc9fa3b1b940e17 Author: Dario Faggioli AuthorDate: Fri Mar 19 12:14:17 2021 +0000 Commit: George Dunlap CommitDate: Mon Jun 7 13:17:06 2021 +0100 xen: credit2: fix per-entity load tracking when continuing running If we schedule, and the current vCPU continues to run, its statistical load is not properly updated, resulting in something like this, even if all the 8 vCPUs are 100% busy: (XEN) Runqueue 0: (XEN) [...] (XEN) aveload = 2097152 (~800%) (XEN) [...] (XEN) Domain: 0 w 256 c 0 v 8 (XEN) 1: [0.0] flags=2 cpu=4 credit=9996885 [w=256] load=35 (~0%) (XEN) 2: [0.1] flags=2 cpu=2 credit=9993725 [w=256] load=796 (~0%) (XEN) 3: [0.2] flags=2 cpu=1 credit=9995885 [w=256] load=883 (~0%) (XEN) 4: [0.3] flags=2 cpu=5 credit=9998833 [w=256] load=487 (~0%) (XEN) 5: [0.4] flags=2 cpu=6 credit=9998942 [w=256] load=1595 (~0%) (XEN) 6: [0.5] flags=2 cpu=0 credit=9994669 [w=256] load=22 (~0%) (XEN) 7: [0.6] flags=2 cpu=7 credit=9997706 [w=256] load=0 (~0%) (XEN) 8: [0.7] flags=2 cpu=3 credit=9992440 [w=256] load=0 (~0%) As we can see, the average load of the runqueue as a whole is, instead, computed properly. This issue would, in theory, potentially affect Credit2 load balancing logic. In practice, however, the problem only manifests (at least with these characteristics) when there is only 1 runqueue active in the cpupool, which also means there is no need to do any load-balancing. Hence its real impact is pretty much limited to wrong per-vCPU load percentages, when looking at the output of the 'r' debug-key. With this patch, the load is updated and displayed correctly: (XEN) Runqueue 0: (XEN) [...] (XEN) aveload = 2097152 (~800%) (XEN) [...] (XEN) Domain info: (XEN) Domain: 0 w 256 c 0 v 8 (XEN) 1: [0.0] flags=2 cpu=4 credit=9995584 [w=256] load=262144 (~100%) (XEN) 2: [0.1] flags=2 cpu=6 credit=9992992 [w=256] load=262144 (~100%) (XEN) 3: [0.2] flags=2 cpu=3 credit=9998918 [w=256] load=262118 (~99%) (XEN) 4: [0.3] flags=2 cpu=5 credit=9996867 [w=256] load=262144 (~100%) (XEN) 5: [0.4] flags=2 cpu=1 credit=9998912 [w=256] load=262144 (~100%) (XEN) 6: [0.5] flags=2 cpu=2 credit=9997842 [w=256] load=262144 (~100%) (XEN) 7: [0.6] flags=2 cpu=7 credit=9994623 [w=256] load=262144 (~100%) (XEN) 8: [0.7] flags=2 cpu=0 credit=9991815 [w=256] load=262144 (~100%) Signed-off-by: Dario Faggioli Reviewed-by: George Dunlap --- xen/common/sched/credit2.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/xen/common/sched/credit2.c b/xen/common/sched/credit2.c index f5c1e5b944..ebb09ea43a 100644 --- a/xen/common/sched/credit2.c +++ b/xen/common/sched/credit2.c @@ -3649,6 +3649,8 @@ static void csched2_schedule( runq_remove(snext); __set_bit(__CSFLAG_scheduled, &snext->flags); } + else + update_load(ops, rqd, snext, 0, now); /* Clear the idle mask if necessary */ if ( cpumask_test_cpu(sched_cpu, &rqd->idle) ) -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Mon Jun 07 13:44:11 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Mon, 07 Jun 2021 13:44:11 +0000 Received: from list by lists.xenproject.org with outflank-mailman.137983.255530 (Exim 4.92) (envelope-from ) id 1lqFXx-0001ty-Qb; Mon, 07 Jun 2021 13:44:05 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 137983.255530; Mon, 07 Jun 2021 13:44:05 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqFXx-0001tp-NC; Mon, 07 Jun 2021 13:44:05 +0000 Received: by outflank-mailman (input) for mailman id 137983; Mon, 07 Jun 2021 13:44:04 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqFXw-0001sy-RC for xen-changelog@lists.xenproject.org; Mon, 07 Jun 2021 13:44:04 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqFXw-0000uI-NL for xen-changelog@lists.xenproject.org; Mon, 07 Jun 2021 13:44:04 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqFXw-0002b0-MD for xen-changelog@lists.xenproject.org; Mon, 07 Jun 2021 13:44:04 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=kI/G9wy12zGNxN1Vbgh9qt0bVAsKxPe/cKbh8luVb6M=; b=SNXxXeVCUIMwsVJEXA/ys5MsL2 m9OK0H6Kr/r5xrXbPKiYl7x0zd+g9Ybf0GZtYMOeqL4HzIv1CtwMfamuGz3GWMYe0QssuGi5+hugl kTZGyvLhmC2EfYrnD+FdJCHisLD7m/2YkvIJknfqpHiYMaHdXJCxt1bl26ga6fCV4dSo=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] docs: release-technician-checklist: update to leaf tree version pinning Message-Id: Date: Mon, 07 Jun 2021 13:44:04 +0000 commit c4beefcada0a406681dcfb6e89f6cbe4aa368c2d Author: Jan Beulich AuthorDate: Mon Jun 7 15:40:55 2021 +0200 Commit: Jan Beulich CommitDate: Mon Jun 7 15:40:55 2021 +0200 docs: release-technician-checklist: update to leaf tree version pinning Our releases look to flip-flop between keeping or discarding the date and title of the referenced qemu-trad commit. I think with the hash replaced by a tag, the commit's date and title would better also be purged. Signed-off-by: Jan Beulich Acked-by: Ian Jackson --- docs/process/release-technician-checklist.txt | 1 + 1 file changed, 1 insertion(+) diff --git a/docs/process/release-technician-checklist.txt b/docs/process/release-technician-checklist.txt index 880630c1ed..7515da202c 100644 --- a/docs/process/release-technician-checklist.txt +++ b/docs/process/release-technician-checklist.txt @@ -53,6 +53,7 @@ t=RELEASE-$r # QEMU_UPSTREAM_REVISION, # QEMU_TRADITIONAL_REVISION # MINIOS_UPSTREAM_REVISION +# (drop any references to the specific commits, e.g. date or title) * change SUPPORT.md heading version number; -unstable or -rc tag * (empty in stable branches after .0 release). * SUPPORT.md: insert correct version number in release-notes link -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Mon Jun 07 14:44:10 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Mon, 07 Jun 2021 14:44:10 +0000 Received: from list by lists.xenproject.org with outflank-mailman.138009.255574 (Exim 4.92) (envelope-from ) id 1lqGU2-0000XA-5o; Mon, 07 Jun 2021 14:44:06 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 138009.255574; Mon, 07 Jun 2021 14:44:06 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqGU2-0000X1-1o; Mon, 07 Jun 2021 14:44:06 +0000 Received: by outflank-mailman (input) for mailman id 138009; Mon, 07 Jun 2021 14:44:05 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqGU1-0000WA-FA for xen-changelog@lists.xenproject.org; Mon, 07 Jun 2021 14:44:05 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqGU1-000208-AR for xen-changelog@lists.xenproject.org; Mon, 07 Jun 2021 14:44:05 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqGU1-0007f0-9W for xen-changelog@lists.xenproject.org; Mon, 07 Jun 2021 14:44:05 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=Ye5GRf7BSo1/HJsy7WWi+kO7uPJnQTWQLlPlqsKbYXM=; b=BPzpLdsRA3M1aaayvGT+UGmmPF 9t0zkuy7iSPI/VIF1F9SScHW5z4cZUfzsz0aaSnxr+RFkDW/Mj1AaKjMkykh6vWyZJf9KCSDOmZ0t qnbp2O6uQHsc6JYQEnFZpiJK4ROSGHyJyKalkuD82hQGU0REC7C4RujTlHx+H6J6nmxg=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] x86/cpuid: Fix HLE and RTM handling (again) Message-Id: Date: Mon, 07 Jun 2021 14:44:05 +0000 commit 60fa12dbf1d4d2c4ffe1ef34b495b24aa7e41aa0 Author: Andrew Cooper AuthorDate: Mon Jun 7 13:25:09 2021 +0100 Commit: Andrew Cooper CommitDate: Mon Jun 7 15:43:35 2021 +0100 x86/cpuid: Fix HLE and RTM handling (again) For reasons which are my fault, but I don't recall why, the FDP_EXCP_ONLY/NO_FPU_SEL adjustment uses the whole special_features[] array element, not the two relevant bits. HLE and RTM were recently added to the list of special features, causing them to be always set in guest view, irrespective of the toolstacks choice on the matter. Rewrite the logic to refer to the features specifically, rather than relying on the contents of the special_features[] array. Fixes: 8fe24090d9 ("x86/cpuid: Rework HLE and RTM handling") Reported-by: Edwin Török Signed-off-by: Andrew Cooper Reviewed-by: Jan Beulich --- xen/arch/x86/cpuid.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/xen/arch/x86/cpuid.c b/xen/arch/x86/cpuid.c index f3c8950aa3..958caf35da 100644 --- a/xen/arch/x86/cpuid.c +++ b/xen/arch/x86/cpuid.c @@ -672,9 +672,11 @@ void recalculate_cpuid_policy(struct domain *d) sanitise_featureset(fs); /* Fold host's FDP_EXCP_ONLY and NO_FPU_SEL into guest's view. */ - fs[FEATURESET_7b0] &= ~special_features[FEATURESET_7b0]; + fs[FEATURESET_7b0] &= ~(cpufeat_mask(X86_FEATURE_FDP_EXCP_ONLY) | + cpufeat_mask(X86_FEATURE_NO_FPU_SEL)); fs[FEATURESET_7b0] |= (host_cpuid_policy.feat._7b0 & - special_features[FEATURESET_7b0]); + (cpufeat_mask(X86_FEATURE_FDP_EXCP_ONLY) | + cpufeat_mask(X86_FEATURE_NO_FPU_SEL))); cpuid_featureset_to_policy(fs, p); -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Mon Jun 07 14:44:18 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Mon, 07 Jun 2021 14:44:18 +0000 Received: from list by lists.xenproject.org with outflank-mailman.138010.255578 (Exim 4.92) (envelope-from ) id 1lqGUC-0000aU-6P; Mon, 07 Jun 2021 14:44:16 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 138010.255578; Mon, 07 Jun 2021 14:44:16 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqGUC-0000aM-3Q; Mon, 07 Jun 2021 14:44:16 +0000 Received: by outflank-mailman (input) for mailman id 138010; Mon, 07 Jun 2021 14:44:15 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqGUB-0000aE-EK for xen-changelog@lists.xenproject.org; Mon, 07 Jun 2021 14:44:15 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqGUB-00020G-DQ for xen-changelog@lists.xenproject.org; Mon, 07 Jun 2021 14:44:15 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqGUB-0007fc-CZ for xen-changelog@lists.xenproject.org; Mon, 07 Jun 2021 14:44:15 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=M8SmffVpLD1v3aIOdyivdutRsWCZvG9UtNkohVMUA5E=; b=m4YI9c92TJ9Iz2H5+7zvd53MTW +QRHLnO+ZQjN3RbGx/DGHHQ7x6DMu1LD+yoUcZYyczIbfloMxoWpkUcDmoX6dlfrXn3YCfDBYTn2h ZA14qiequ+VrSSYHfihtcQgrEMO9kpxZ7gvB+0R+CwUqDzI1giqtEInFIBwJYahqm9fc=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] x86/cpuid: Drop special_features[] Message-Id: Date: Mon, 07 Jun 2021 14:44:15 +0000 commit 69e1472d21cf7e5cf0795ef38b99d00de78a910e Author: Andrew Cooper AuthorDate: Mon Jun 7 13:38:53 2021 +0100 Commit: Andrew Cooper CommitDate: Mon Jun 7 15:43:35 2021 +0100 x86/cpuid: Drop special_features[] While the ! annotation is useful to indicate that something special is happening, an array of bits is not. Drop it, to prevent mistakes. Signed-off-by: Andrew Cooper Reviewed-by: Jan Beulich --- xen/arch/x86/cpuid.c | 2 -- xen/include/asm-x86/cpuid.h | 1 - xen/tools/gen-cpuid.py | 3 --- 3 files changed, 6 deletions(-) diff --git a/xen/arch/x86/cpuid.c b/xen/arch/x86/cpuid.c index 958caf35da..2079a30ae4 100644 --- a/xen/arch/x86/cpuid.c +++ b/xen/arch/x86/cpuid.c @@ -14,7 +14,6 @@ #include const uint32_t known_features[] = INIT_KNOWN_FEATURES; -const uint32_t special_features[] = INIT_SPECIAL_FEATURES; static const uint32_t __initconst pv_max_featuremask[] = INIT_PV_MAX_FEATURES; static const uint32_t hvm_shadow_max_featuremask[] = INIT_HVM_SHADOW_MAX_FEATURES; @@ -1132,7 +1131,6 @@ void guest_cpuid(const struct vcpu *v, uint32_t leaf, static void __init __maybe_unused build_assertions(void) { BUILD_BUG_ON(ARRAY_SIZE(known_features) != FSCAPINTS); - BUILD_BUG_ON(ARRAY_SIZE(special_features) != FSCAPINTS); BUILD_BUG_ON(ARRAY_SIZE(pv_max_featuremask) != FSCAPINTS); BUILD_BUG_ON(ARRAY_SIZE(hvm_shadow_max_featuremask) != FSCAPINTS); BUILD_BUG_ON(ARRAY_SIZE(hvm_hap_max_featuremask) != FSCAPINTS); diff --git a/xen/include/asm-x86/cpuid.h b/xen/include/asm-x86/cpuid.h index 7baf6c9628..46904061d0 100644 --- a/xen/include/asm-x86/cpuid.h +++ b/xen/include/asm-x86/cpuid.h @@ -14,7 +14,6 @@ #include extern const uint32_t known_features[FSCAPINTS]; -extern const uint32_t special_features[FSCAPINTS]; void init_guest_cpuid(void); diff --git a/xen/tools/gen-cpuid.py b/xen/tools/gen-cpuid.py index b953648b65..c6b5056a8d 100755 --- a/xen/tools/gen-cpuid.py +++ b/xen/tools/gen-cpuid.py @@ -362,8 +362,6 @@ def write_results(state): #define INIT_KNOWN_FEATURES { \\\n%s\n} -#define INIT_SPECIAL_FEATURES { \\\n%s\n} - #define INIT_PV_DEF_FEATURES { \\\n%s\n} #define INIT_PV_MAX_FEATURES { \\\n%s\n} @@ -384,7 +382,6 @@ def write_results(state): """ % (state.nr_entries, next(featureset_to_uint32s(state.common_1d, 1)), format_uint32s(state, state.names.keys(), 4), - format_uint32s(state, state.raw['!'], 4), format_uint32s(state, state.pv_def, 4), format_uint32s(state, state.pv_max, 4), format_uint32s(state, state.hvm_shadow_def, 4), -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Mon Jun 07 14:44:26 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Mon, 07 Jun 2021 14:44:26 +0000 Received: from list by lists.xenproject.org with outflank-mailman.138011.255582 (Exim 4.92) (envelope-from ) id 1lqGUM-0000dR-8E; Mon, 07 Jun 2021 14:44:26 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 138011.255582; Mon, 07 Jun 2021 14:44:26 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqGUM-0000dJ-5G; Mon, 07 Jun 2021 14:44:26 +0000 Received: by outflank-mailman (input) for mailman id 138011; Mon, 07 Jun 2021 14:44:25 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqGUL-0000d9-HQ for xen-changelog@lists.xenproject.org; Mon, 07 Jun 2021 14:44:25 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqGUL-00020k-Gb for xen-changelog@lists.xenproject.org; Mon, 07 Jun 2021 14:44:25 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqGUL-0007hL-FT for xen-changelog@lists.xenproject.org; Mon, 07 Jun 2021 14:44:25 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=b02BY40qdpfDzICv3uB0KfaOvL6nCTDmP7/ICq5XAZE=; b=kTIESyUhxKW58dvVCCL/ogBnHG 8f1PRro7Q9//kTy9u0ZmEGgYX+VFAMfDyBqhBmPTSTNrdTj03gl9pKhm6CwqLdqZ9RUSVLmKhDn7b If/YxuC5a6wks0oXux10hFp8uZfFc+Nfl/cmlxmAO5YU/xrf+MGTHvuJCKKObgKL53BU=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] tools/libs/guest: fix save and restore of pv domains after 32-bit de-support Message-Id: Date: Mon, 07 Jun 2021 14:44:25 +0000 commit d21121685fac829c988e432407fb0e4ef9b19331 Author: Juergen Gross AuthorDate: Mon Jun 7 15:00:05 2021 +0200 Commit: Andrew Cooper CommitDate: Mon Jun 7 15:43:35 2021 +0100 tools/libs/guest: fix save and restore of pv domains after 32-bit de-support After 32-bit PV-guests have been security de-supported when not running under PV-shim, the hypervisor will no longer be configured to support those domains per default when not being built as PV-shim. Unfortunately libxenguest will fail saving or restoring a PV domain due to this restriction, as it is trying to get the compat MFN list even for 64 bit guests. Fix that by obtaining the compat MFN list only for 32-bit PV guests. Fixes: 1a0f2fe2297d122a08fe ("SUPPORT.md: Un-shimmed 32-bit PV guests are no longer supported") Signed-off-by: Juergen Gross Reviewed-by: Jan Beulich Reviewed-by: Andrew Cooper --- tools/libs/guest/xg_sr_common.h | 2 +- tools/libs/guest/xg_sr_common_x86_pv.c | 13 ++++++++----- 2 files changed, 9 insertions(+), 6 deletions(-) diff --git a/tools/libs/guest/xg_sr_common.h b/tools/libs/guest/xg_sr_common.h index cc3ad1c394..e2994e18ac 100644 --- a/tools/libs/guest/xg_sr_common.h +++ b/tools/libs/guest/xg_sr_common.h @@ -325,7 +325,7 @@ struct xc_sr_context xen_pfn_t max_mfn; /* Read-only machine to phys map */ xen_pfn_t *m2p; - /* first mfn of the compat m2p (Only needed for 32bit PV guests) */ + /* first mfn of the compat m2p (Only set for 32bit PV guests) */ xen_pfn_t compat_m2p_mfn0; /* Number of m2p frames mapped */ unsigned long nr_m2p_frames; diff --git a/tools/libs/guest/xg_sr_common_x86_pv.c b/tools/libs/guest/xg_sr_common_x86_pv.c index cd33406aab..c0acf00f90 100644 --- a/tools/libs/guest/xg_sr_common_x86_pv.c +++ b/tools/libs/guest/xg_sr_common_x86_pv.c @@ -149,12 +149,13 @@ int x86_pv_map_m2p(struct xc_sr_context *ctx) ctx->x86.pv.nr_m2p_frames = (M2P_CHUNK_SIZE >> PAGE_SHIFT) * m2p_chunks; + if ( ctx->x86.pv.levels == 3 ) + { #ifdef __i386__ - /* 32 bit toolstacks automatically get the compat m2p */ - ctx->x86.pv.compat_m2p_mfn0 = entries[0].mfn; + /* 32 bit toolstacks automatically get the compat m2p */ + ctx->x86.pv.compat_m2p_mfn0 = entries[0].mfn; #else - /* 64 bit toolstacks need to ask Xen specially for it */ - { + /* 64 bit toolstacks need to ask Xen specially for it */ struct xen_machphys_mfn_list xmml = { .max_extents = 1, .extent_start = { &ctx->x86.pv.compat_m2p_mfn0 }, @@ -168,8 +169,10 @@ int x86_pv_map_m2p(struct xc_sr_context *ctx) rc = -1; goto err; } - } #endif + } + else + ctx->x86.pv.compat_m2p_mfn0 = INVALID_MFN; /* All Done */ rc = 0; -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Tue Jun 08 10:33:11 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 08 Jun 2021 10:33:11 +0000 Received: from list by lists.xenproject.org with outflank-mailman.138401.256211 (Exim 4.92) (envelope-from ) id 1lqZ2g-00026p-Jv; Tue, 08 Jun 2021 10:33:06 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 138401.256211; Tue, 08 Jun 2021 10:33:06 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqZ2g-00026h-G9; Tue, 08 Jun 2021 10:33:06 +0000 Received: by outflank-mailman (input) for mailman id 138401; Tue, 08 Jun 2021 10:33:04 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqZ2e-00026b-E3 for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 10:33:04 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqZ2e-0005bK-DQ for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 10:33:04 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqZ2e-0003Xj-CR for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 10:33:04 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=47U0IcMaLLkig2PVttSKyZ4/BVS8tVPNs3KYjAu30qo=; b=YcTz+tMs1xkk4QPwaMbLNA3Tjh PcjFq/6sCsa0wZY3kpoJu31Hl5pPJqZkF/5Dd2olEibLnkjnpRHhGi7tqomRpzZ3ZmIPmnVOLTEbZ y4Yhtz+6ZY2nVjvrGOtVIaNwEnu/G3v6JDCi5KgAs7s7QsNhLnoxX8b5J9MnlsXgv3Sw=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] tools/firmware/ovmf: Use OvmfXen platform file is exist Message-Id: Date: Tue, 08 Jun 2021 10:33:04 +0000 commit aad7b5c11d51d57659978e04702ac970906894e8 Author: Anthony PERARD AuthorDate: Tue Jun 1 11:28:03 2021 +0100 Commit: Ian Jackson CommitDate: Tue Jun 8 11:31:10 2021 +0100 tools/firmware/ovmf: Use OvmfXen platform file is exist A platform introduced in EDK II named OvmfXen is now the one to use for Xen instead of OvmfX64. It comes with PVH support. Also, the Xen support in OvmfX64 is deprecated, "deprecation notice: *dynamic* multi-VMM (QEMU vs. Xen) support in OvmfPkg" https://edk2.groups.io/g/devel/message/75498 Signed-off-by: Anthony PERARD Acked-by: Ian Jackson --- tools/firmware/ovmf-makefile | 10 ++++++++-- 1 file changed, 8 insertions(+), 2 deletions(-) diff --git a/tools/firmware/ovmf-makefile b/tools/firmware/ovmf-makefile index 55f9992145..637ee509c3 100644 --- a/tools/firmware/ovmf-makefile +++ b/tools/firmware/ovmf-makefile @@ -17,8 +17,14 @@ all: build .PHONY: build build: if test -e .git ; then $(GIT) submodule update --init --recursive ; fi - OvmfPkg/build.sh -a X64 -b $(TARGET) -n 4 - cp Build/OvmfX64/$(TARGET)_GCC*/FV/OVMF.fd ovmf.bin + set -ex; \ + if test -e OvmfPkg/OvmfXen.dsc; then \ + OvmfPkg/build.sh -a X64 -b $(TARGET) -n 4 -p OvmfPkg/OvmfXen.dsc; \ + cp Build/OvmfXen/$(TARGET)_GCC*/FV/OVMF.fd ovmf.bin; \ + else \ + OvmfPkg/build.sh -a X64 -b $(TARGET) -n 4; \ + cp Build/OvmfX64/$(TARGET)_GCC*/FV/OVMF.fd ovmf.bin; \ + fi .PHONY: clean clean: -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Tue Jun 08 12:55:13 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 08 Jun 2021 12:55:13 +0000 Received: from list by lists.xenproject.org with outflank-mailman.138514.256394 (Exim 4.92) (envelope-from ) id 1lqbG6-0006Ix-RB; Tue, 08 Jun 2021 12:55:06 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 138514.256394; Tue, 08 Jun 2021 12:55:06 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqbG6-0006Ip-OF; Tue, 08 Jun 2021 12:55:06 +0000 Received: by outflank-mailman (input) for mailman id 138514; Tue, 08 Jun 2021 12:55:05 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqbG5-0006Ij-Nb for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 12:55:05 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqbG5-00086O-LJ for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 12:55:05 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqbG5-0004oT-KO for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 12:55:05 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=3JcFAau0JjKe6ogsCBVgbC2BOqR0mepVuSlHJq1RaYA=; b=btPv5gKm7cN44u+9aijUDiG1Ht p7Jw7dgN1/HvOzLzHxc2SdkKgm3Na+CuY0owlFp1gft++lTU8hzBzPcns4AkcTQ8I1mJHx6yg3cDo ok12yckACuS85pezyxsVFD0aLkOv7s5eF0LfbuYq/FrEbPQgiiBKYH6wHtLuPW4QLp34=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] evtchn: slightly defer lock acquire where possible Message-Id: Date: Tue, 08 Jun 2021 12:55:05 +0000 commit 411076f4d635f6cba5e74e8173d0ac4c576024e8 Author: Jan Beulich AuthorDate: Tue Jun 8 14:46:06 2021 +0200 Commit: Jan Beulich CommitDate: Tue Jun 8 14:46:06 2021 +0200 evtchn: slightly defer lock acquire where possible port_is_valid() and evtchn_from_port() are fine to use without holding any locks. Accordingly acquire the per-domain lock slightly later in evtchn_close() and evtchn_bind_vcpu(). Especially for the use by the former (but there are pre-existing uses) add a comment about port_is_valid()'s guarantees. Signed-off-by: Jan Beulich Acked-by: Roger Pau Monné Reviewed-by: Julien Grall --- xen/common/event_channel.c | 20 +++++++------------- xen/include/xen/event.h | 10 ++++++++++ 2 files changed, 17 insertions(+), 13 deletions(-) diff --git a/xen/common/event_channel.c b/xen/common/event_channel.c index 5479315aae..8a5f8bb9d4 100644 --- a/xen/common/event_channel.c +++ b/xen/common/event_channel.c @@ -606,17 +606,14 @@ int evtchn_close(struct domain *d1, int port1, bool guest) int port2; long rc = 0; - again: - spin_lock(&d1->event_lock); - if ( !port_is_valid(d1, port1) ) - { - rc = -EINVAL; - goto out; - } + return -EINVAL; chn1 = evtchn_from_port(d1, port1); + again: + spin_lock(&d1->event_lock); + /* Guest cannot close a Xen-attached event channel. */ if ( unlikely(consumer_is_xen(chn1)) && guest ) { @@ -1041,16 +1038,13 @@ long evtchn_bind_vcpu(unsigned int port, unsigned int vcpu_id) if ( (v = domain_vcpu(d, vcpu_id)) == NULL ) return -ENOENT; - spin_lock(&d->event_lock); - if ( !port_is_valid(d, port) ) - { - rc = -EINVAL; - goto out; - } + return -EINVAL; chn = evtchn_from_port(d, port); + spin_lock(&d->event_lock); + /* Guest cannot re-bind a Xen-attached event channel. */ if ( unlikely(consumer_is_xen(chn)) ) { diff --git a/xen/include/xen/event.h b/xen/include/xen/event.h index a0a85cdda8..a185a5b875 100644 --- a/xen/include/xen/event.h +++ b/xen/include/xen/event.h @@ -120,6 +120,16 @@ static inline void evtchn_read_unlock(struct evtchn *evtchn) read_unlock(&evtchn->lock); } +/* + * While calling the function is okay without holding a suitable lock yet + * (see the comment ahead of struct evtchn_port_ops for which ones those + * are), for a dying domain it may start returning false at any point - see + * evtchn_destroy(). This is not a fundamental problem though, as the + * struct evtchn instance won't disappear (and will continue to hold valid + * data) until final cleanup of the domain, at which point the domain itself + * cannot be looked up anymore and hence calls here can't occur any longer + * in the first place. + */ static inline bool_t port_is_valid(struct domain *d, unsigned int p) { if ( p >= read_atomic(&d->valid_evtchns) ) -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Tue Jun 08 12:55:16 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 08 Jun 2021 12:55:16 +0000 Received: from list by lists.xenproject.org with outflank-mailman.138515.256398 (Exim 4.92) (envelope-from ) id 1lqbGG-0006Kb-T1; Tue, 08 Jun 2021 12:55:16 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 138515.256398; Tue, 08 Jun 2021 12:55:16 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqbGG-0006KT-Px; Tue, 08 Jun 2021 12:55:16 +0000 Received: by outflank-mailman (input) for mailman id 138515; Tue, 08 Jun 2021 12:55:15 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqbGF-0006KJ-Q1 for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 12:55:15 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqbGF-00086S-P5 for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 12:55:15 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqbGF-0004pB-NZ for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 12:55:15 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=11/04vzWk9oNF80O3kjP7gd/qrTp/fguYE1+tPmEeIQ=; b=tGasX/tIlO38E0Ms9ch96D1/7P 5sJs0q7dXo+IsH86Y6RYGKSvZkXl2+P1GSfFH/Q1kbsg/8L+Uq7UHXBtIp/5XpVrbuRCWlBqbjQpW 4yUabBsFP+IKJzrJI6I7UGz3QzFvi091iU6uouBmcdr84KCSm/jJ+0UjSytHBWlsKq+Q=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] evtchn: add helper for port_is_valid() + evtchn_from_port() Message-Id: Date: Tue, 08 Jun 2021 12:55:15 +0000 commit 3270a2d45fbbcdf9b3dd44b4b4776c1fc4682717 Author: Jan Beulich AuthorDate: Tue Jun 8 14:47:14 2021 +0200 Commit: Jan Beulich CommitDate: Tue Jun 8 14:47:14 2021 +0200 evtchn: add helper for port_is_valid() + evtchn_from_port() The combination is pretty common, so adding a simple local helper seems worthwhile. Make it const- and type-correct, in turn requiring the two called function to also be const-correct (and at this occasion also make them type-correct). Signed-off-by: Jan Beulich Acked-by: Julien Grall --- xen/common/event_channel.c | 73 +++++++++++++++++++++------------------------- xen/include/xen/event.h | 5 ++-- 2 files changed, 36 insertions(+), 42 deletions(-) diff --git a/xen/common/event_channel.c b/xen/common/event_channel.c index 8a5f8bb9d4..886c2726d6 100644 --- a/xen/common/event_channel.c +++ b/xen/common/event_channel.c @@ -147,6 +147,12 @@ static bool virq_is_global(unsigned int virq) return true; } +static struct evtchn *_evtchn_from_port(const struct domain *d, + evtchn_port_t port) +{ + return port_is_valid(d, port) ? evtchn_from_port(d, port) : NULL; +} + static void free_evtchn_bucket(struct domain *d, struct evtchn *bucket) { if ( !bucket ) @@ -319,7 +325,6 @@ static long evtchn_alloc_unbound(evtchn_alloc_unbound_t *alloc) return rc; } - static void double_evtchn_lock(struct evtchn *lchn, struct evtchn *rchn) { ASSERT(lchn != rchn); @@ -365,9 +370,9 @@ static long evtchn_bind_interdomain(evtchn_bind_interdomain_t *bind) ERROR_EXIT(lport); lchn = evtchn_from_port(ld, lport); - if ( !port_is_valid(rd, rport) ) + rchn = _evtchn_from_port(rd, rport); + if ( !rchn ) ERROR_EXIT_DOM(-EINVAL, rd); - rchn = evtchn_from_port(rd, rport); if ( (rchn->state != ECS_UNBOUND) || (rchn->u.unbound.remote_domid != ld->domain_id) ) ERROR_EXIT_DOM(-EINVAL, rd); @@ -602,15 +607,12 @@ static long evtchn_bind_pirq(evtchn_bind_pirq_t *bind) int evtchn_close(struct domain *d1, int port1, bool guest) { struct domain *d2 = NULL; - struct evtchn *chn1, *chn2; - int port2; + struct evtchn *chn1 = _evtchn_from_port(d1, port1), *chn2; long rc = 0; - if ( !port_is_valid(d1, port1) ) + if ( !chn1 ) return -EINVAL; - chn1 = evtchn_from_port(d1, port1); - again: spin_lock(&d1->event_lock); @@ -698,10 +700,8 @@ int evtchn_close(struct domain *d1, int port1, bool guest) goto out; } - port2 = chn1->u.interdomain.remote_port; - BUG_ON(!port_is_valid(d2, port2)); - - chn2 = evtchn_from_port(d2, port2); + chn2 = _evtchn_from_port(d2, chn1->u.interdomain.remote_port); + BUG_ON(!chn2); BUG_ON(chn2->state != ECS_INTERDOMAIN); BUG_ON(chn2->u.interdomain.remote_dom != d1); @@ -739,15 +739,13 @@ int evtchn_close(struct domain *d1, int port1, bool guest) int evtchn_send(struct domain *ld, unsigned int lport) { - struct evtchn *lchn, *rchn; + struct evtchn *lchn = _evtchn_from_port(ld, lport), *rchn; struct domain *rd; int rport, ret = 0; - if ( !port_is_valid(ld, lport) ) + if ( !lchn ) return -EINVAL; - lchn = evtchn_from_port(ld, lport); - evtchn_read_lock(lchn); /* Guest cannot send via a Xen-attached event channel. */ @@ -967,15 +965,15 @@ int evtchn_status(evtchn_status_t *status) if ( d == NULL ) return -ESRCH; - spin_lock(&d->event_lock); - - if ( !port_is_valid(d, port) ) + chn = _evtchn_from_port(d, port); + if ( !chn ) { - rc = -EINVAL; - goto out; + rcu_unlock_domain(d); + return -EINVAL; } - chn = evtchn_from_port(d, port); + spin_lock(&d->event_lock); + if ( consumer_is_xen(chn) ) { rc = -EACCES; @@ -1038,11 +1036,10 @@ long evtchn_bind_vcpu(unsigned int port, unsigned int vcpu_id) if ( (v = domain_vcpu(d, vcpu_id)) == NULL ) return -ENOENT; - if ( !port_is_valid(d, port) ) + chn = _evtchn_from_port(d, port); + if ( !chn ) return -EINVAL; - chn = evtchn_from_port(d, port); - spin_lock(&d->event_lock); /* Guest cannot re-bind a Xen-attached event channel. */ @@ -1088,13 +1085,11 @@ long evtchn_bind_vcpu(unsigned int port, unsigned int vcpu_id) int evtchn_unmask(unsigned int port) { struct domain *d = current->domain; - struct evtchn *evtchn; + struct evtchn *evtchn = _evtchn_from_port(d, port); - if ( unlikely(!port_is_valid(d, port)) ) + if ( unlikely(!evtchn) ) return -EINVAL; - evtchn = evtchn_from_port(d, port); - evtchn_read_lock(evtchn); evtchn_port_unmask(d, evtchn); @@ -1177,14 +1172,12 @@ static long evtchn_set_priority(const struct evtchn_set_priority *set_priority) { struct domain *d = current->domain; unsigned int port = set_priority->port; - struct evtchn *chn; + struct evtchn *chn = _evtchn_from_port(d, port); long ret; - if ( !port_is_valid(d, port) ) + if ( !chn ) return -EINVAL; - chn = evtchn_from_port(d, port); - evtchn_read_lock(chn); ret = evtchn_port_set_priority(d, chn, set_priority->priority); @@ -1410,10 +1403,10 @@ void free_xen_event_channel(struct domain *d, int port) void notify_via_xen_event_channel(struct domain *ld, int lport) { - struct evtchn *lchn, *rchn; + struct evtchn *lchn = _evtchn_from_port(ld, lport), *rchn; struct domain *rd; - if ( !port_is_valid(ld, lport) ) + if ( !lchn ) { /* * Make sure ->is_dying is read /after/ ->valid_evtchns, pairing @@ -1424,8 +1417,6 @@ void notify_via_xen_event_channel(struct domain *ld, int lport) return; } - lchn = evtchn_from_port(ld, lport); - if ( !evtchn_read_trylock(lchn) ) return; @@ -1580,12 +1571,14 @@ static void domain_dump_evtchn_info(struct domain *d) spin_lock(&d->event_lock); - for ( port = 1; port_is_valid(d, port); ++port ) + for ( port = 1; ; ++port ) { - const struct evtchn *chn; + const struct evtchn *chn = _evtchn_from_port(d, port); char *ssid; - chn = evtchn_from_port(d, port); + if ( !chn ) + break; + if ( chn->state == ECS_FREE ) continue; diff --git a/xen/include/xen/event.h b/xen/include/xen/event.h index a185a5b875..4f4e97e9c7 100644 --- a/xen/include/xen/event.h +++ b/xen/include/xen/event.h @@ -130,7 +130,7 @@ static inline void evtchn_read_unlock(struct evtchn *evtchn) * cannot be looked up anymore and hence calls here can't occur any longer * in the first place. */ -static inline bool_t port_is_valid(struct domain *d, unsigned int p) +static inline bool port_is_valid(const struct domain *d, evtchn_port_t p) { if ( p >= read_atomic(&d->valid_evtchns) ) return false; @@ -145,7 +145,8 @@ static inline bool_t port_is_valid(struct domain *d, unsigned int p) return true; } -static inline struct evtchn *evtchn_from_port(struct domain *d, unsigned int p) +static inline struct evtchn *evtchn_from_port(const struct domain *d, + evtchn_port_t p) { if ( p < EVTCHNS_PER_BUCKET ) return &d->evtchn[array_index_nospec(p, EVTCHNS_PER_BUCKET)]; -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Tue Jun 08 12:55:27 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 08 Jun 2021 12:55:27 +0000 Received: from list by lists.xenproject.org with outflank-mailman.138516.256402 (Exim 4.92) (envelope-from ) id 1lqbGR-0006Nd-V1; Tue, 08 Jun 2021 12:55:27 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 138516.256402; Tue, 08 Jun 2021 12:55:27 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqbGR-0006NT-Ra; Tue, 08 Jun 2021 12:55:27 +0000 Received: by outflank-mailman (input) for mailman id 138516; Tue, 08 Jun 2021 12:55:25 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqbGP-0006N3-TH for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 12:55:25 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqbGP-00086j-SP for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 12:55:25 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqbGP-0004qR-RO for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 12:55:25 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=ojNPNGDNeM/BQg9t9/G+UwWeIZgQf13TnTQxPZjYkGI=; b=BtCS7gxCUY+ir5VTNt5pEIx7xW wC+g7iToxlJTtYR05FPHEMWWStvoimYr4JuEGsktlcmgvwXnF6uKOrb6FUwczgPyR3zzaxXJsyGp7 Jn+6xgIQ4SVPxC6XeMxiYAQIIIoI4AcoINWR9RvQ+7cp79kZfV6QH6e6jsDIX54pXWHU=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] evtchn: type adjustments Message-Id: Date: Tue, 08 Jun 2021 12:55:25 +0000 commit dd77e859686f458a5313786deae0a63683d3cba3 Author: Jan Beulich AuthorDate: Tue Jun 8 14:47:47 2021 +0200 Commit: Jan Beulich CommitDate: Tue Jun 8 14:47:47 2021 +0200 evtchn: type adjustments First of all avoid "long" when "int" suffices, i.e. in particular when merely conveying error codes. 32-bit values are slightly cheaper to deal with on x86, and their processing is at least no more expensive on Arm. Where possible use evtchn_port_t for port numbers and unsigned int for other unsigned quantities in adjacent code. In evtchn_set_priority() eliminate a local variable altogether instead of changing its type. Signed-off-by: Jan Beulich Acked-by: Roger Pau Monné --- xen/common/event_channel.c | 42 ++++++++++++++++++++---------------------- xen/include/xen/event.h | 2 +- 2 files changed, 21 insertions(+), 23 deletions(-) diff --git a/xen/common/event_channel.c b/xen/common/event_channel.c index 886c2726d6..da88ad141a 100644 --- a/xen/common/event_channel.c +++ b/xen/common/event_channel.c @@ -284,13 +284,12 @@ void evtchn_free(struct domain *d, struct evtchn *chn) xsm_evtchn_close_post(chn); } -static long evtchn_alloc_unbound(evtchn_alloc_unbound_t *alloc) +static int evtchn_alloc_unbound(evtchn_alloc_unbound_t *alloc) { struct evtchn *chn; struct domain *d; - int port; + int port, rc; domid_t dom = alloc->dom; - long rc; d = rcu_lock_domain_by_any_id(dom); if ( d == NULL ) @@ -342,13 +341,13 @@ static void double_evtchn_unlock(struct evtchn *lchn, struct evtchn *rchn) evtchn_write_unlock(rchn); } -static long evtchn_bind_interdomain(evtchn_bind_interdomain_t *bind) +static int evtchn_bind_interdomain(evtchn_bind_interdomain_t *bind) { struct evtchn *lchn, *rchn; struct domain *ld = current->domain, *rd; - int lport, rport = bind->remote_port; + int lport, rc; + evtchn_port_t rport = bind->remote_port; domid_t rdom = bind->remote_dom; - long rc; if ( (rd = rcu_lock_domain_by_any_id(rdom)) == NULL ) return -ESRCH; @@ -484,12 +483,12 @@ int evtchn_bind_virq(evtchn_bind_virq_t *bind, evtchn_port_t port) } -static long evtchn_bind_ipi(evtchn_bind_ipi_t *bind) +static int evtchn_bind_ipi(evtchn_bind_ipi_t *bind) { struct evtchn *chn; struct domain *d = current->domain; - int port, vcpu = bind->vcpu; - long rc = 0; + int port, rc = 0; + unsigned int vcpu = bind->vcpu; if ( domain_vcpu(d, vcpu) == NULL ) return -ENOENT; @@ -543,16 +542,16 @@ static void unlink_pirq_port(struct evtchn *chn, struct vcpu *v) } -static long evtchn_bind_pirq(evtchn_bind_pirq_t *bind) +static int evtchn_bind_pirq(evtchn_bind_pirq_t *bind) { struct evtchn *chn; struct domain *d = current->domain; struct vcpu *v = d->vcpu[0]; struct pirq *info; - int port = 0, pirq = bind->pirq; - long rc; + int port = 0, rc; + unsigned int pirq = bind->pirq; - if ( (pirq < 0) || (pirq >= d->nr_pirqs) ) + if ( pirq >= d->nr_pirqs ) return -EINVAL; if ( !is_hvm_domain(d) && !pirq_access_permitted(d, pirq) ) @@ -608,7 +607,7 @@ int evtchn_close(struct domain *d1, int port1, bool guest) { struct domain *d2 = NULL; struct evtchn *chn1 = _evtchn_from_port(d1, port1), *chn2; - long rc = 0; + int rc = 0; if ( !chn1 ) return -EINVAL; @@ -959,7 +958,7 @@ int evtchn_status(evtchn_status_t *status) domid_t dom = status->dom; int port = status->port; struct evtchn *chn; - long rc = 0; + int rc = 0; d = rcu_lock_domain_by_any_id(dom); if ( d == NULL ) @@ -1025,11 +1024,11 @@ int evtchn_status(evtchn_status_t *status) } -long evtchn_bind_vcpu(unsigned int port, unsigned int vcpu_id) +int evtchn_bind_vcpu(evtchn_port_t port, unsigned int vcpu_id) { struct domain *d = current->domain; struct evtchn *chn; - long rc = 0; + int rc = 0; struct vcpu *v; /* Use the vcpu info to prevent speculative out-of-bound accesses */ @@ -1168,12 +1167,11 @@ int evtchn_reset(struct domain *d, bool resuming) return rc; } -static long evtchn_set_priority(const struct evtchn_set_priority *set_priority) +static int evtchn_set_priority(const struct evtchn_set_priority *set_priority) { struct domain *d = current->domain; - unsigned int port = set_priority->port; - struct evtchn *chn = _evtchn_from_port(d, port); - long ret; + struct evtchn *chn = _evtchn_from_port(d, set_priority->port); + int ret; if ( !chn ) return -EINVAL; @@ -1189,7 +1187,7 @@ static long evtchn_set_priority(const struct evtchn_set_priority *set_priority) long do_event_channel_op(int cmd, XEN_GUEST_HANDLE_PARAM(void) arg) { - long rc; + int rc; switch ( cmd ) { diff --git a/xen/include/xen/event.h b/xen/include/xen/event.h index 4f4e97e9c7..21c95e14fd 100644 --- a/xen/include/xen/event.h +++ b/xen/include/xen/event.h @@ -54,7 +54,7 @@ void send_guest_pirq(struct domain *, const struct pirq *); int evtchn_send(struct domain *d, unsigned int lport); /* Bind a local event-channel port to the specified VCPU. */ -long evtchn_bind_vcpu(unsigned int port, unsigned int vcpu_id); +int evtchn_bind_vcpu(evtchn_port_t port, unsigned int vcpu_id); /* Bind a VIRQ. */ int evtchn_bind_virq(evtchn_bind_virq_t *bind, evtchn_port_t port); -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Tue Jun 08 16:33:12 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 08 Jun 2021 16:33:12 +0000 Received: from list by lists.xenproject.org with outflank-mailman.138594.256492 (Exim 4.92) (envelope-from ) id 1lqef4-00044Q-Bv; Tue, 08 Jun 2021 16:33:06 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 138594.256492; Tue, 08 Jun 2021 16:33:06 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqef4-00044I-93; Tue, 08 Jun 2021 16:33:06 +0000 Received: by outflank-mailman (input) for mailman id 138594; Tue, 08 Jun 2021 16:33:05 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqef3-00044C-3c for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 16:33:05 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqef3-000486-1f for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 16:33:05 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqef3-0006Ac-0v for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 16:33:05 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=Luw/iIxVtu9PSjNV6xB75NihrsIVUkCKQlW5jalMKPc=; b=Wm+LtTHH2PLXJitKh5t/OOYC3g iIMIT4eT34pbMbtWNYC/yC8xbvzZDC3bSJYddG4319CdgSakapoZTVbUz73xn00sP1JW3eNzABqoW uMnvR7KvacDgyWUvhuIp9ryZDc03sq9xyw6/WWUJAq1kraOZ1S2bp+0PxInRqXuEvCc8=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] x86/cpuid: Half revert "x86/cpuid: Drop special_features[]" Message-Id: Date: Tue, 08 Jun 2021 16:33:05 +0000 commit 0ba0663b1b32d9351890dfd02bdebb3d238897bd Author: Andrew Cooper AuthorDate: Tue Jun 8 17:13:59 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 17:27:06 2021 +0100 x86/cpuid: Half revert "x86/cpuid: Drop special_features[]" xen-cpuid does print out the list of special features, and this is helpful to keep. Fixes: ba6950fb070 ("x86/cpuid: Drop special_features[]") Reported-by: Jan Beulich Reported-by: Sander Eikelenboom Signed-off-by: Andrew Cooper Acked-by: Jan Beulich --- xen/tools/gen-cpuid.py | 3 +++ 1 file changed, 3 insertions(+) diff --git a/xen/tools/gen-cpuid.py b/xen/tools/gen-cpuid.py index c6b5056a8d..b953648b65 100755 --- a/xen/tools/gen-cpuid.py +++ b/xen/tools/gen-cpuid.py @@ -362,6 +362,8 @@ def write_results(state): #define INIT_KNOWN_FEATURES { \\\n%s\n} +#define INIT_SPECIAL_FEATURES { \\\n%s\n} + #define INIT_PV_DEF_FEATURES { \\\n%s\n} #define INIT_PV_MAX_FEATURES { \\\n%s\n} @@ -382,6 +384,7 @@ def write_results(state): """ % (state.nr_entries, next(featureset_to_uint32s(state.common_1d, 1)), format_uint32s(state, state.names.keys(), 4), + format_uint32s(state, state.raw['!'], 4), format_uint32s(state, state.pv_def, 4), format_uint32s(state, state.pv_max, 4), format_uint32s(state, state.hvm_shadow_def, 4), -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Tue Jun 08 17:11:10 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 08 Jun 2021 17:11:10 +0000 Received: from list by lists.xenproject.org with outflank-mailman.138687.256697 (Exim 4.92) (envelope-from ) id 1lqfFr-0001ig-4B; Tue, 08 Jun 2021 17:11:07 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 138687.256697; Tue, 08 Jun 2021 17:11:07 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfFq-0001iQ-Ux; Tue, 08 Jun 2021 17:11:06 +0000 Received: by outflank-mailman (input) for mailman id 138687; Tue, 08 Jun 2021 17:11:05 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfFp-0001iJ-IZ for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:11:05 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfFp-0004wO-FJ for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:11:05 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqfFp-0007LC-EH for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:11:05 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=XOc8dHMUlZjOJ709J7o3zF78Pr0y8DLs1vhiktAvy1k=; b=Y10RokB0876+qDI2Eq8Q4FA79A rTp9nWEoVQwFmH7St16/8lvjXqfl0vW6jAiLUKp9pDoDwPL6qo/jV+lif6600cw9JNSNTZE8cK9AE IydEoJxDn5mnfc08zSXhqLbz1uSEjEdBzg11w4AOTOY0L4Z/9jv0ot4VdK2045/fxf6Y=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] xen/arm: Create dom0less domUs earlier Message-Id: Date: Tue, 08 Jun 2021 17:11:05 +0000 commit 371347c5b64da699d9f5a0edda5dc496fd2b7a5c Author: Julien Grall AuthorDate: Mon May 17 17:47:13 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 17:43:06 2021 +0100 xen/arm: Create dom0less domUs earlier In a follow-up patch we will need to unallocate the boot modules before heap_init_late() is called. The modules will contain the domUs kernel and initramfs. Therefore Xen will need to create extra domUs (used by dom0less) before heap_init_late(). This has two consequences on dom0less: 1) Domains will not be unpaused as soon as they are created but once all have been created. However, Xen doesn't guarantee an order to unpause, so this is not something one could rely on. 2) The memory allocated for a domU will not be scrubbed anymore when an admin select bootscrub=on. This is not something we advertised, but if this is a concern we can introduce either force scrub for all domUs or a per-domain flag in the DT. The behavior for bootscrub=off and bootscrub=idle (default) has not changed. This is part of XSA-372 / CVE-2021-28693. Signed-off-by: Julien Grall Reviewed-by: Jan Beulich Reviewed-by: Stefano Stabellini Tested-by: Stefano Stabellini --- xen/arch/arm/domain_build.c | 6 +----- xen/arch/arm/setup.c | 14 +++++++------- xen/include/asm-arm/setup.h | 2 +- 3 files changed, 9 insertions(+), 13 deletions(-) diff --git a/xen/arch/arm/domain_build.c b/xen/arch/arm/domain_build.c index 282416e74d..6c86d52781 100644 --- a/xen/arch/arm/domain_build.c +++ b/xen/arch/arm/domain_build.c @@ -2521,8 +2521,6 @@ void __init create_domUs(void) if ( construct_domU(d, node) != 0 ) panic("Could not set up domain %s\n", dt_node_name(node)); - - domain_unpause_by_systemcontroller(d); } } @@ -2584,7 +2582,7 @@ static int __init construct_dom0(struct domain *d) return construct_domain(d, &kinfo); } -struct domain* __init create_dom0(void) +void __init create_dom0(void) { struct domain *dom0; struct xen_domctl_createdomain dom0_cfg = { @@ -2615,8 +2613,6 @@ struct domain* __init create_dom0(void) if ( construct_dom0(dom0) != 0) panic("Could not set up DOM0 guest OS\n"); - - return dom0; } /* diff --git a/xen/arch/arm/setup.c b/xen/arch/arm/setup.c index 00aad1c194..e17532c132 100644 --- a/xen/arch/arm/setup.c +++ b/xen/arch/arm/setup.c @@ -836,7 +836,7 @@ void __init start_xen(unsigned long boot_phys_offset, int cpus, i; const char *cmdline; struct bootmodule *xen_bootmodule; - struct domain *dom0 = NULL; + struct domain *d; int rc; dcache_line_bytes = read_dcache_line_bytes(); @@ -992,10 +992,13 @@ void __init start_xen(unsigned long boot_phys_offset, /* Create initial domain 0. */ if ( !is_dom0less_mode() ) - dom0 = create_dom0(); + create_dom0(); else printk(XENLOG_INFO "Xen dom0less mode detected\n"); + if ( acpi_disabled ) + create_domUs(); + heap_init_late(); init_trace_bufs(); @@ -1009,11 +1012,8 @@ void __init start_xen(unsigned long boot_phys_offset, system_state = SYS_STATE_active; - if ( acpi_disabled ) - create_domUs(); - - if ( dom0 ) - domain_unpause_by_systemcontroller(dom0); + for_each_domain( d ) + domain_unpause_by_systemcontroller(d); /* Switch on to the dynamically allocated stack for the idle vcpu * since the static one we're running on is about to be freed. */ diff --git a/xen/include/asm-arm/setup.h b/xen/include/asm-arm/setup.h index 5283244015..c4b6af6029 100644 --- a/xen/include/asm-arm/setup.h +++ b/xen/include/asm-arm/setup.h @@ -94,7 +94,7 @@ void acpi_create_efi_mmap_table(struct domain *d, int acpi_make_efi_nodes(void *fdt, struct membank tbl_add[]); void create_domUs(void); -struct domain* create_dom0(void); +void create_dom0(void); void discard_initial_modules(void); void fw_unreserved_regions(paddr_t s, paddr_t e, -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Tue Jun 08 17:11:17 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 08 Jun 2021 17:11:17 +0000 Received: from list by lists.xenproject.org with outflank-mailman.138688.256699 (Exim 4.92) (envelope-from ) id 1lqfG1-0001l4-59; Tue, 08 Jun 2021 17:11:17 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 138688.256699; Tue, 08 Jun 2021 17:11:17 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfG1-0001kw-2I; Tue, 08 Jun 2021 17:11:17 +0000 Received: by outflank-mailman (input) for mailman id 138688; Tue, 08 Jun 2021 17:11:15 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfFz-0001ki-JH for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:11:15 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfFz-0004wb-IQ for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:11:15 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqfFz-0007Lt-HT for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:11:15 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=wTaYyQoKUrYfh9I6UmPdyZwv0lE+k3p02CFZRklJVW0=; b=zhSqsjb8d149ZCEMwIMkULLU45 HvbUWN36Gtij08UlcHbMe/rDG325K7xXy5/NtpCwkKw8Ov7dIL0LnvkfK+tVGEJsQ8EsthElAVMSb QywfltO9QBQg2aF+zbjGGrNfmckltm7yuRo2R92bCLLuItndzhALtdxGB1Z/hnE/6tMY=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] xen/arm: Boot modules should always be scrubbed if bootscrub={on, idle} Message-Id: Date: Tue, 08 Jun 2021 17:11:15 +0000 commit fd5dc41ceaed9cfcfa011cdfd50f264c89277a90 Author: Julien Grall AuthorDate: Sat Apr 17 17:38:28 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 17:43:06 2021 +0100 xen/arm: Boot modules should always be scrubbed if bootscrub={on, idle} The function to initialize the pages (see init_heap_pages()) will request scrub when the admin request idle bootscrub (default) and state == SYS_STATE_active. When bootscrub=on, Xen will scrub any free pages in heap_init_late(). Currently, the boot modules (e.g. kernels, initramfs) will be discarded/ freed after heap_init_late() is called and system_state switched to SYS_STATE_active. This means the pages associated with the boot modules will not get scrubbed before getting re-purposed. If the memory is assigned to an untrusted domU, it may be able to retrieve secrets from the modules. This is part of XSA-372 / CVE-2021-28693. Fixes: 1774e9b1df27 ("xen/arm: introduce create_domUs") Signed-off-by: Julien Grall Reviewed-by: Jan Beulich Reviewed-by: Stefano Stabellini Tested-by: Stefano Stabellini --- xen/arch/arm/setup.c | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-) diff --git a/xen/arch/arm/setup.c b/xen/arch/arm/setup.c index e17532c132..63a908e325 100644 --- a/xen/arch/arm/setup.c +++ b/xen/arch/arm/setup.c @@ -71,8 +71,6 @@ domid_t __read_mostly max_init_domid; static __used void init_done(void) { - discard_initial_modules(); - /* Must be done past setting system_state. */ unregister_init_virtual_region(); @@ -999,6 +997,12 @@ void __init start_xen(unsigned long boot_phys_offset, if ( acpi_disabled ) create_domUs(); + /* + * This needs to be called **before** heap_init_late() so modules + * will be scrubbed (unless suppressed). + */ + discard_initial_modules(); + heap_init_late(); init_trace_bufs(); -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Tue Jun 08 17:11:27 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 08 Jun 2021 17:11:27 +0000 Received: from list by lists.xenproject.org with outflank-mailman.138689.256703 (Exim 4.92) (envelope-from ) id 1lqfGB-0001oI-6l; Tue, 08 Jun 2021 17:11:27 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 138689.256703; Tue, 08 Jun 2021 17:11:27 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfGB-0001oA-3j; Tue, 08 Jun 2021 17:11:27 +0000 Received: by outflank-mailman (input) for mailman id 138689; Tue, 08 Jun 2021 17:11:25 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfG9-0001nl-Md for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:11:25 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfG9-0004wo-Lt for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:11:25 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqfG9-0007Mi-Km for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:11:25 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=iCRipB43LDKNh68gWtjN3Z+x3HxXvfwx8rQqr6kQCJg=; b=ssWVb1lUK1gNLjnic0KzXfKLdj wqUrl2mB5Bjl+rCnnQ4bEYdvkHbkzRoVHj60GRgKPu11TVXw42zSpvMQVLtH71uhsNdmczUpOGIYH Hb48yb2xsl6hKInNbIIIotnE4NsOKwr8vJybxb/3aFvVLwDFQ8mq6cOP/9iC//9UfQ6I=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] VT-d: size qinval queue dynamically Message-Id: Date: Tue, 08 Jun 2021 17:11:25 +0000 commit cbfa62bb140c8f10a0ae96db3ce06e22b3b9d302 Author: Jan Beulich AuthorDate: Tue Jun 8 17:38:55 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 17:43:06 2021 +0100 VT-d: size qinval queue dynamically With the present synchronous model, we need two slots for every operation (the operation itself and a wait descriptor). There can be one such pair of requests pending per CPU. To ensure that under all normal circumstances a slot is always available when one is requested, size the queue ring according to the number of present CPUs. This is part of XSA-373 / CVE-2021-28692. Signed-off-by: Jan Beulich Reviewed-by: Paul Durrant --- xen/drivers/passthrough/vtd/iommu.h | 10 +------- xen/drivers/passthrough/vtd/qinval.c | 46 +++++++++++++++++++++++++++--------- 2 files changed, 36 insertions(+), 20 deletions(-) diff --git a/xen/drivers/passthrough/vtd/iommu.h b/xen/drivers/passthrough/vtd/iommu.h index 68b3c76ea9..cd516ac5f8 100644 --- a/xen/drivers/passthrough/vtd/iommu.h +++ b/xen/drivers/passthrough/vtd/iommu.h @@ -452,17 +452,9 @@ struct qinval_entry { }q; }; -/* Order of queue invalidation pages(max is 8) */ -#define QINVAL_PAGE_ORDER 2 - -#define QINVAL_ARCH_PAGE_ORDER (QINVAL_PAGE_ORDER + PAGE_SHIFT_4K - PAGE_SHIFT) -#define QINVAL_ARCH_PAGE_NR ( QINVAL_ARCH_PAGE_ORDER < 0 ? \ - 1 : \ - 1 << QINVAL_ARCH_PAGE_ORDER ) - /* Each entry is 16 bytes, so 2^8 entries per page */ #define QINVAL_ENTRY_ORDER ( PAGE_SHIFT - 4 ) -#define QINVAL_ENTRY_NR (1 << (QINVAL_PAGE_ORDER + 8)) +#define QINVAL_MAX_ENTRY_NR (1u << (7 + QINVAL_ENTRY_ORDER)) /* Status data flag */ #define QINVAL_STAT_INIT 0 diff --git a/xen/drivers/passthrough/vtd/qinval.c b/xen/drivers/passthrough/vtd/qinval.c index 0976def43d..69fef844bb 100644 --- a/xen/drivers/passthrough/vtd/qinval.c +++ b/xen/drivers/passthrough/vtd/qinval.c @@ -31,6 +31,9 @@ #define VTD_QI_TIMEOUT 1 +static unsigned int __read_mostly qi_pg_order; +static unsigned int __read_mostly qi_entry_nr; + static int __must_check invalidate_sync(struct vtd_iommu *iommu); static void print_qi_regs(const struct vtd_iommu *iommu) @@ -47,7 +50,7 @@ static unsigned int qinval_next_index(struct vtd_iommu *iommu) tail >>= QINVAL_INDEX_SHIFT; /* (tail+1 == head) indicates a full queue, wait for HW */ - while ( (tail + 1) % QINVAL_ENTRY_NR == + while ( ((tail + 1) & (qi_entry_nr - 1)) == (dmar_readl(iommu->reg, DMAR_IQH_REG) >> QINVAL_INDEX_SHIFT) ) cpu_relax(); @@ -60,7 +63,7 @@ static void qinval_update_qtail(struct vtd_iommu *iommu, unsigned int index) /* Need hold register lock when update tail */ ASSERT( spin_is_locked(&iommu->register_lock) ); - val = (index + 1) % QINVAL_ENTRY_NR; + val = (index + 1) & (qi_entry_nr - 1); dmar_writel(iommu->reg, DMAR_IQT_REG, val << QINVAL_INDEX_SHIFT); } @@ -395,8 +398,28 @@ int enable_qinval(struct vtd_iommu *iommu) if ( iommu->qinval_maddr == 0 ) { - iommu->qinval_maddr = alloc_pgtable_maddr(QINVAL_ARCH_PAGE_NR, - iommu->node); + if ( !qi_entry_nr ) + { + /* + * With the present synchronous model, we need two slots for every + * operation (the operation itself and a wait descriptor). There + * can be one such pair of requests pending per CPU. One extra + * entry is needed as the ring is considered full when there's + * only one entry left. + */ + BUILD_BUG_ON(CONFIG_NR_CPUS * 2 >= QINVAL_MAX_ENTRY_NR); + qi_pg_order = get_order_from_bytes((num_present_cpus() * 2 + 1) << + (PAGE_SHIFT - + QINVAL_ENTRY_ORDER)); + qi_entry_nr = 1u << (qi_pg_order + QINVAL_ENTRY_ORDER); + + dprintk(XENLOG_INFO VTDPREFIX, + "QI: using %u-entry ring(s)\n", qi_entry_nr); + } + + iommu->qinval_maddr = + alloc_pgtable_maddr(qi_entry_nr >> QINVAL_ENTRY_ORDER, + iommu->node); if ( iommu->qinval_maddr == 0 ) { dprintk(XENLOG_WARNING VTDPREFIX, @@ -410,15 +433,16 @@ int enable_qinval(struct vtd_iommu *iommu) spin_lock_irqsave(&iommu->register_lock, flags); - /* Setup Invalidation Queue Address(IQA) register with the - * address of the page we just allocated. QS field at - * bits[2:0] to indicate size of queue is one 4KB page. - * That's 256 entries. Queued Head (IQH) and Queue Tail (IQT) - * registers are automatically reset to 0 with write - * to IQA register. + /* + * Setup Invalidation Queue Address (IQA) register with the address of the + * pages we just allocated. The QS field at bits[2:0] indicates the size + * (page order) of the queue. + * + * Queued Head (IQH) and Queue Tail (IQT) registers are automatically + * reset to 0 with write to IQA register. */ dmar_writeq(iommu->reg, DMAR_IQA_REG, - iommu->qinval_maddr | QINVAL_PAGE_ORDER); + iommu->qinval_maddr | qi_pg_order); dmar_writeq(iommu->reg, DMAR_IQT_REG, 0); -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Tue Jun 08 17:11:36 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 08 Jun 2021 17:11:36 +0000 Received: from list by lists.xenproject.org with outflank-mailman.138690.256706 (Exim 4.92) (envelope-from ) id 1lqfGK-0001rS-8G; Tue, 08 Jun 2021 17:11:36 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 138690.256706; Tue, 08 Jun 2021 17:11:36 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfGK-0001rK-5E; Tue, 08 Jun 2021 17:11:36 +0000 Received: by outflank-mailman (input) for mailman id 138690; Tue, 08 Jun 2021 17:11:35 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfGJ-0001rB-RT for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:11:35 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfGJ-0004x2-Pc for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:11:35 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqfGJ-0007Ni-OM for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:11:35 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=pIwg6oEbvvA8nfyh3aJ1YgQ0dXffBe5DvMWBSjkez2M=; b=ZfichTabIsriQ9y4vrFBg8AkER V7V2TPXGIM6Lp8ynqRzwQyQV50srDXkX99gNY9fokclzF1DXu4HuSsF26pPBN5muYr0BIHX0MhxUM 9Ks66gPu3g8m8OCbcB/OcZxL3s8oNvc6pulwuTtr28zkyQ5qKPZjgnM4DHU06DH+4zVo=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] AMD/IOMMU: size command buffer dynamically Message-Id: Date: Tue, 08 Jun 2021 17:11:35 +0000 commit df242851ddc93ac0b0a3a20ecab34acc29e3b36b Author: Jan Beulich AuthorDate: Tue Jun 8 17:38:55 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 17:43:06 2021 +0100 AMD/IOMMU: size command buffer dynamically With the present synchronous model, we need two slots for every operation (the operation itself and a wait command). There can be one such pair of commands pending per CPU. To ensure that under all normal circumstances a slot is always available when one is requested, size the command ring according to the number of present CPUs. This is part of XSA-373 / CVE-2021-28692. Signed-off-by: Jan Beulich Reviewed-by: Paul Durrant --- xen/drivers/passthrough/amd/iommu-defs.h | 7 ++----- xen/drivers/passthrough/amd/iommu_cmd.c | 4 ++-- xen/drivers/passthrough/amd/iommu_init.c | 28 +++++++++++++++++++++++++--- 3 files changed, 29 insertions(+), 10 deletions(-) diff --git a/xen/drivers/passthrough/amd/iommu-defs.h b/xen/drivers/passthrough/amd/iommu-defs.h index 56f385f270..0c97db2a32 100644 --- a/xen/drivers/passthrough/amd/iommu-defs.h +++ b/xen/drivers/passthrough/amd/iommu-defs.h @@ -20,9 +20,6 @@ #ifndef AMD_IOMMU_DEFS_H #define AMD_IOMMU_DEFS_H -/* IOMMU Command Buffer entries: in power of 2 increments, minimum of 256 */ -#define IOMMU_CMD_BUFFER_DEFAULT_ENTRIES 512 - /* IOMMU Event Log entries: in power of 2 increments, minimum of 256 */ #define IOMMU_EVENT_LOG_DEFAULT_ENTRIES 512 @@ -164,8 +161,8 @@ struct amd_iommu_dte { #define IOMMU_CMD_BUFFER_LENGTH_MASK 0x0F000000 #define IOMMU_CMD_BUFFER_LENGTH_SHIFT 24 -#define IOMMU_CMD_BUFFER_ENTRY_SIZE 16 -#define IOMMU_CMD_BUFFER_POWER_OF2_ENTRIES_PER_PAGE 8 +#define IOMMU_CMD_BUFFER_ENTRY_ORDER 4 +#define IOMMU_CMD_BUFFER_MAX_ENTRIES (1u << 15) #define IOMMU_CMD_OPCODE_MASK 0xF0000000 #define IOMMU_CMD_OPCODE_SHIFT 28 diff --git a/xen/drivers/passthrough/amd/iommu_cmd.c b/xen/drivers/passthrough/amd/iommu_cmd.c index 6c0647c524..54d7528f96 100644 --- a/xen/drivers/passthrough/amd/iommu_cmd.c +++ b/xen/drivers/passthrough/amd/iommu_cmd.c @@ -24,7 +24,7 @@ static int queue_iommu_command(struct amd_iommu *iommu, u32 cmd[]) { uint32_t tail, head; - tail = iommu->cmd_buffer.tail + IOMMU_CMD_BUFFER_ENTRY_SIZE; + tail = iommu->cmd_buffer.tail + sizeof(cmd_entry_t); if ( tail == iommu->cmd_buffer.size ) tail = 0; @@ -33,7 +33,7 @@ static int queue_iommu_command(struct amd_iommu *iommu, u32 cmd[]) if ( head != tail ) { memcpy(iommu->cmd_buffer.buffer + iommu->cmd_buffer.tail, - cmd, IOMMU_CMD_BUFFER_ENTRY_SIZE); + cmd, sizeof(cmd_entry_t)); iommu->cmd_buffer.tail = tail; return 1; diff --git a/xen/drivers/passthrough/amd/iommu_init.c b/xen/drivers/passthrough/amd/iommu_init.c index 24d1dfec40..bb52c181f8 100644 --- a/xen/drivers/passthrough/amd/iommu_init.c +++ b/xen/drivers/passthrough/amd/iommu_init.c @@ -118,7 +118,7 @@ static void register_iommu_cmd_buffer_in_mmio_space(struct amd_iommu *iommu) writel(entry, iommu->mmio_base + IOMMU_CMD_BUFFER_BASE_LOW_OFFSET); power_of2_entries = get_order_from_bytes(iommu->cmd_buffer.size) + - IOMMU_CMD_BUFFER_POWER_OF2_ENTRIES_PER_PAGE; + PAGE_SHIFT - IOMMU_CMD_BUFFER_ENTRY_ORDER; entry = 0; iommu_set_addr_hi_to_reg(&entry, addr_hi); @@ -1018,9 +1018,31 @@ static void *__init allocate_ring_buffer(struct ring_buffer *ring_buf, static void * __init allocate_cmd_buffer(struct amd_iommu *iommu) { /* allocate 'command buffer' in power of 2 increments of 4K */ + static unsigned int __read_mostly nr_ents; + + if ( !nr_ents ) + { + unsigned int order; + + /* + * With the present synchronous model, we need two slots for every + * operation (the operation itself and a wait command). There can be + * one such pair of requests pending per CPU. One extra entry is + * needed as the ring is considered full when there's only one entry + * left. + */ + BUILD_BUG_ON(CONFIG_NR_CPUS * 2 >= IOMMU_CMD_BUFFER_MAX_ENTRIES); + order = get_order_from_bytes((num_present_cpus() * 2 + 1) << + IOMMU_CMD_BUFFER_ENTRY_ORDER); + nr_ents = 1u << (order + PAGE_SHIFT - IOMMU_CMD_BUFFER_ENTRY_ORDER); + + AMD_IOMMU_DEBUG("using %u-entry cmd ring(s)\n", nr_ents); + } + + BUILD_BUG_ON(sizeof(cmd_entry_t) != (1u << IOMMU_CMD_BUFFER_ENTRY_ORDER)); + return allocate_ring_buffer(&iommu->cmd_buffer, sizeof(cmd_entry_t), - IOMMU_CMD_BUFFER_DEFAULT_ENTRIES, - "Command Buffer", false); + nr_ents, "Command Buffer", false); } static void * __init allocate_event_log(struct amd_iommu *iommu) -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Tue Jun 08 17:11:46 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 08 Jun 2021 17:11:46 +0000 Received: from list by lists.xenproject.org with outflank-mailman.138691.256711 (Exim 4.92) (envelope-from ) id 1lqfGU-0001uZ-9a; Tue, 08 Jun 2021 17:11:46 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 138691.256711; Tue, 08 Jun 2021 17:11:46 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfGU-0001uR-6e; Tue, 08 Jun 2021 17:11:46 +0000 Received: by outflank-mailman (input) for mailman id 138691; Tue, 08 Jun 2021 17:11:45 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfGT-0001uL-Tj for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:11:45 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfGT-0004xS-Sw for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:11:45 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqfGT-0007P5-Ry for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:11:45 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=blz55LIuNw6+gOlo0COTV3n+I12ObnytkdPXkPELCgY=; b=GN7xjtIv5nnEKG/cKfAh46fY8E JeruClIn2cpkZ8wVYbxeOEEHIpz/4Rn3zNskecbzmNAbD8HK9XTJhO0qTthNCj70ac8R1DEdOv++0 AllAqjkqd43WL/PbKkqLUf9I4YqORMLJ2i4vFgDKgLeI1wKYKvUwsfdz8dTT229IYbhk=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] VT-d: eliminate flush related timeouts Message-Id: Date: Tue, 08 Jun 2021 17:11:45 +0000 commit 7c893abc4ee7e9af62297ba6fd5f27c89d0f33f5 Author: Jan Beulich AuthorDate: Tue Jun 8 17:38:55 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 17:43:06 2021 +0100 VT-d: eliminate flush related timeouts Leaving an in-progress operation pending when it appears to take too long is problematic: If e.g. a QI command completed later, the write to the "poll slot" may instead be understood to signal a subsequently started command's completion. Also our accounting of the timeout period was actually wrong: We included the time it took for the command to actually make it to the front of the queue, which could be heavily affected by guests other than the one for which the flush is being performed. Do away with all timeout detection on all flush related code paths. Log excessively long processing times (with a progressive threshold) to have some indication of problems in this area. Additionally log (once) if qinval_next_index() didn't immediately find an available slot. Together with the earlier change sizing the queue(s) dynamically, we should now have a guarantee that with our fully synchronous model any demand for slots can actually be satisfied. This is part of XSA-373 / CVE-2021-28692. Signed-off-by: Jan Beulich Reviewed-by: Paul Durrant --- xen/drivers/passthrough/vtd/dmar.h | 28 ++++++++++++++++++++++++++++ xen/drivers/passthrough/vtd/iommu.c | 12 ++++++------ xen/drivers/passthrough/vtd/qinval.c | 33 ++++++++++++++++++++++----------- 3 files changed, 56 insertions(+), 17 deletions(-) diff --git a/xen/drivers/passthrough/vtd/dmar.h b/xen/drivers/passthrough/vtd/dmar.h index 1a9c965e59..a44820a5e7 100644 --- a/xen/drivers/passthrough/vtd/dmar.h +++ b/xen/drivers/passthrough/vtd/dmar.h @@ -127,6 +127,34 @@ do { \ } \ } while (0) +#define IOMMU_FLUSH_WAIT(what, iommu, offset, op, cond, sts) \ +do { \ + static unsigned int __read_mostly threshold = 1; \ + s_time_t start = NOW(); \ + s_time_t timeout = start + DMAR_OPERATION_TIMEOUT * threshold; \ + \ + for ( ; ; ) \ + { \ + sts = op(iommu->reg, offset); \ + if ( cond ) \ + break; \ + if ( timeout && NOW() > timeout ) \ + { \ + threshold |= threshold << 1; \ + printk(XENLOG_WARNING VTDPREFIX \ + " IOMMU#%u: %s flush taking too long\n", \ + iommu->index, what); \ + timeout = 0; \ + } \ + cpu_relax(); \ + } \ + \ + if ( !timeout ) \ + printk(XENLOG_WARNING VTDPREFIX \ + " IOMMU#%u: %s flush took %lums\n", \ + iommu->index, what, (NOW() - start) / 10000000); \ +} while ( false ) + int vtd_hw_check(void); void disable_pmr(struct vtd_iommu *iommu); int is_igd_drhd(struct acpi_drhd_unit *drhd); diff --git a/xen/drivers/passthrough/vtd/iommu.c b/xen/drivers/passthrough/vtd/iommu.c index 82ae2f5e97..b7aa15e5ef 100644 --- a/xen/drivers/passthrough/vtd/iommu.c +++ b/xen/drivers/passthrough/vtd/iommu.c @@ -373,8 +373,8 @@ static void iommu_flush_write_buffer(struct vtd_iommu *iommu) dmar_writel(iommu->reg, DMAR_GCMD_REG, val | DMA_GCMD_WBF); /* Make sure hardware complete it */ - IOMMU_WAIT_OP(iommu, DMAR_GSTS_REG, dmar_readl, - !(val & DMA_GSTS_WBFS), val); + IOMMU_FLUSH_WAIT("write buffer", iommu, DMAR_GSTS_REG, dmar_readl, + !(val & DMA_GSTS_WBFS), val); spin_unlock_irqrestore(&iommu->register_lock, flags); } @@ -423,8 +423,8 @@ int vtd_flush_context_reg(struct vtd_iommu *iommu, uint16_t did, dmar_writeq(iommu->reg, DMAR_CCMD_REG, val); /* Make sure hardware complete it */ - IOMMU_WAIT_OP(iommu, DMAR_CCMD_REG, dmar_readq, - !(val & DMA_CCMD_ICC), val); + IOMMU_FLUSH_WAIT("context", iommu, DMAR_CCMD_REG, dmar_readq, + !(val & DMA_CCMD_ICC), val); spin_unlock_irqrestore(&iommu->register_lock, flags); /* flush context entry will implicitly flush write buffer */ @@ -501,8 +501,8 @@ int vtd_flush_iotlb_reg(struct vtd_iommu *iommu, uint16_t did, uint64_t addr, dmar_writeq(iommu->reg, tlb_offset + 8, val); /* Make sure hardware complete it */ - IOMMU_WAIT_OP(iommu, (tlb_offset + 8), dmar_readq, - !(val & DMA_TLB_IVT), val); + IOMMU_FLUSH_WAIT("iotlb", iommu, (tlb_offset + 8), dmar_readq, + !(val & DMA_TLB_IVT), val); spin_unlock_irqrestore(&iommu->register_lock, flags); /* check IOTLB invalidation granularity */ diff --git a/xen/drivers/passthrough/vtd/qinval.c b/xen/drivers/passthrough/vtd/qinval.c index 69fef844bb..311bf107ed 100644 --- a/xen/drivers/passthrough/vtd/qinval.c +++ b/xen/drivers/passthrough/vtd/qinval.c @@ -29,8 +29,6 @@ #include "extern.h" #include "../ats.h" -#define VTD_QI_TIMEOUT 1 - static unsigned int __read_mostly qi_pg_order; static unsigned int __read_mostly qi_entry_nr; @@ -52,7 +50,11 @@ static unsigned int qinval_next_index(struct vtd_iommu *iommu) /* (tail+1 == head) indicates a full queue, wait for HW */ while ( ((tail + 1) & (qi_entry_nr - 1)) == (dmar_readl(iommu->reg, DMAR_IQH_REG) >> QINVAL_INDEX_SHIFT) ) + { + printk_once(XENLOG_ERR VTDPREFIX " IOMMU#%u: no QI slot available\n", + iommu->index); cpu_relax(); + } return tail; } @@ -172,23 +174,32 @@ static int __must_check queue_invalidate_wait(struct vtd_iommu *iommu, /* Now we don't support interrupt method */ if ( sw ) { - s_time_t timeout; - - /* In case all wait descriptor writes to same addr with same data */ - timeout = NOW() + MILLISECS(flush_dev_iotlb ? - iommu_dev_iotlb_timeout : VTD_QI_TIMEOUT); + static unsigned int __read_mostly threshold = 1; + s_time_t start = NOW(); + s_time_t timeout = start + (flush_dev_iotlb + ? iommu_dev_iotlb_timeout + : 100) * MILLISECS(threshold); while ( ACCESS_ONCE(*this_poll_slot) != QINVAL_STAT_DONE ) { - if ( NOW() > timeout ) + if ( timeout && NOW() > timeout ) { - print_qi_regs(iommu); + threshold |= threshold << 1; printk(XENLOG_WARNING VTDPREFIX - " Queue invalidate wait descriptor timed out\n"); - return -ETIMEDOUT; + " IOMMU#%u: QI%s wait descriptor taking too long\n", + iommu->index, flush_dev_iotlb ? " dev" : ""); + print_qi_regs(iommu); + timeout = 0; } cpu_relax(); } + + if ( !timeout ) + printk(XENLOG_WARNING VTDPREFIX + " IOMMU#%u: QI%s wait descriptor took %lums\n", + iommu->index, flush_dev_iotlb ? " dev" : "", + (NOW() - start) / 10000000); + return 0; } -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Tue Jun 08 17:11:57 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 08 Jun 2021 17:11:57 +0000 Received: from list by lists.xenproject.org with outflank-mailman.138694.256715 (Exim 4.92) (envelope-from ) id 1lqfGf-0001xZ-C6; Tue, 08 Jun 2021 17:11:57 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 138694.256715; Tue, 08 Jun 2021 17:11:57 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfGf-0001xR-99; Tue, 08 Jun 2021 17:11:57 +0000 Received: by outflank-mailman (input) for mailman id 138694; Tue, 08 Jun 2021 17:11:56 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfGe-0001xG-1G for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:11:56 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfGe-0004xc-0V for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:11:56 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqfGd-0007Qg-Vo for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:11:55 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=p2h2n5Hy1/U0p5ywJj/K8RYj33AHohcGhnZQ9AoQiPY=; b=p9mYiXNjKxFpSrAzSvmjKgMix2 vBI9stIquzECvgA5YMtlEPzZvgYNuFvTh/5PmKwBmm02f+dl+yUJvWd03vskYkRzmM8DAGUz30gLv 3GFNsFaKjjWpcBEGHHsyaPD210pgKW7Or+k4ksvaGYAobB/hT4Y/WB+xrxEFJ164rGM8=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] x86/spec-ctrl: Protect against Speculative Code Store Bypass Message-Id: Date: Tue, 08 Jun 2021 17:11:55 +0000 commit 45f59ed8865318bb0356954bad067f329677ce9e Author: Andrew Cooper AuthorDate: Thu Mar 11 14:39:11 2021 +0000 Commit: Andrew Cooper CommitDate: Tue Jun 8 17:43:06 2021 +0100 x86/spec-ctrl: Protect against Speculative Code Store Bypass Modern x86 processors have far-better-than-architecturally-guaranteed self modifying code detection. Typically, when a write hits an instruction in flight, a Machine Clear occurs to flush stale content in the frontend and backend. For self modifying code, before a write which hits an instruction in flight retires, the frontend can speculatively decode and execute the old instruction stream. Speculation of this form can suffer from type confusion in registers, and potentially leak data. Furthermore, updates are typically byte-wise, rather than atomic. Depending on timing, speculation can race ahead multiple times between individual writes, and execute the transiently-malformed instruction stream. Xen has stubs which are used in certain cases for emulation purposes. Inhibit speculation between updating the stub and executing it. This is XSA-375 / CVE-2021-0089. Signed-off-by: Andrew Cooper Reviewed-by: Jan Beulich --- xen/arch/x86/pv/emul-priv-op.c | 2 ++ xen/arch/x86/x86_emulate/x86_emulate.c | 1 + 2 files changed, 3 insertions(+) diff --git a/xen/arch/x86/pv/emul-priv-op.c b/xen/arch/x86/pv/emul-priv-op.c index 8889509d2a..11467a1e3a 100644 --- a/xen/arch/x86/pv/emul-priv-op.c +++ b/xen/arch/x86/pv/emul-priv-op.c @@ -138,6 +138,8 @@ static io_emul_stub_t *io_emul_stub_setup(struct priv_op_ctxt *ctxt, u8 opcode, /* Runtime confirmation that we haven't clobbered an adjacent stub. */ BUG_ON(STUB_BUF_SIZE / 2 < (p - ctxt->io_emul_stub)); + block_speculation(); /* SCSB */ + /* Handy function-typed pointer to the stub. */ return (void *)stub_va; diff --git a/xen/arch/x86/x86_emulate/x86_emulate.c b/xen/arch/x86/x86_emulate/x86_emulate.c index 31fdec030c..45828b726b 100644 --- a/xen/arch/x86/x86_emulate/x86_emulate.c +++ b/xen/arch/x86/x86_emulate/x86_emulate.c @@ -1257,6 +1257,7 @@ static inline int mkec(uint8_t e, int32_t ec, ...) # define invoke_stub(pre, post, constraints...) do { \ stub_exn.info = (union stub_exception_token) { .raw = ~0 }; \ stub_exn.line = __LINE__; /* Utility outweighs livepatching cost */ \ + block_speculation(); /* SCSB */ \ asm volatile ( pre "\n\tINDIRECT_CALL %[stub]\n\t" post "\n" \ ".Lret%=:\n\t" \ ".pushsection .fixup,\"ax\"\n" \ -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Tue Jun 08 17:12:07 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 08 Jun 2021 17:12:07 +0000 Received: from list by lists.xenproject.org with outflank-mailman.138696.256719 (Exim 4.92) (envelope-from ) id 1lqfGp-00021K-F7; Tue, 08 Jun 2021 17:12:07 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 138696.256719; Tue, 08 Jun 2021 17:12:07 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfGp-00021C-C3; Tue, 08 Jun 2021 17:12:07 +0000 Received: by outflank-mailman (input) for mailman id 138696; Tue, 08 Jun 2021 17:12:06 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfGo-00020v-4E for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:12:06 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfGo-0004xw-3V for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:12:06 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqfGo-0007Rn-2d for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:12:06 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=Q3jTmjACiuXpcoHNFMTFN3YwwbYbG5feVSwKX5D6qis=; b=PPyFy45rgAaKDSYxxqTtHTONI2 C18nn2vJcpqg15MbVeaEOHmSqefsihQ4MssIn696uW5c83Xz18oP4gYwqDNWG4psdQweF1wsyI++b KXpJBuaR5wXI+5Al9E8S9fj8AR4oDzP2GTTUDIOwO1aJdItUIb5YTiAT51WqTNKLDfcg=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] x86/spec-ctrl: Mitigate TAA after S3 resume Message-Id: Date: Tue, 08 Jun 2021 17:12:06 +0000 commit 8cf276cb2e0b99b96333865873f56b0b31555ff1 Author: Andrew Cooper AuthorDate: Thu May 20 01:21:39 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 17:43:06 2021 +0100 x86/spec-ctrl: Mitigate TAA after S3 resume The user chosen setting for MSR_TSX_CTRL needs restoring after S3. All APs get the correct setting via start_secondary(), but the BSP was missed out. This is XSA-377 / CVE-2021-28690. Fixes: 8c4330818f6 ("x86/spec-ctrl: Mitigate the TSX Asynchronous Abort sidechannel") Signed-off-by: Andrew Cooper Reviewed-by: Jan Beulich --- xen/arch/x86/acpi/power.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/xen/arch/x86/acpi/power.c b/xen/arch/x86/acpi/power.c index 91a8c4d0bd..31a56f02d0 100644 --- a/xen/arch/x86/acpi/power.c +++ b/xen/arch/x86/acpi/power.c @@ -288,6 +288,8 @@ static int enter_state(u32 state) microcode_update_one(); + tsx_init(); /* Needs microcode. May change HLE/RTM feature bits. */ + if ( !recheck_cpu_features(0) ) panic("Missing previously available feature(s)\n"); -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Tue Jun 08 17:12:17 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 08 Jun 2021 17:12:17 +0000 Received: from list by lists.xenproject.org with outflank-mailman.138697.256723 (Exim 4.92) (envelope-from ) id 1lqfGz-00024U-GX; Tue, 08 Jun 2021 17:12:17 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 138697.256723; Tue, 08 Jun 2021 17:12:17 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfGz-00024M-Dc; Tue, 08 Jun 2021 17:12:17 +0000 Received: by outflank-mailman (input) for mailman id 138697; Tue, 08 Jun 2021 17:12:17 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfGz-00024G-1B for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:12:17 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfGz-0004y8-0T for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:12:17 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqfGy-0007Tw-Vu for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:12:16 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=Z0/E1JcC9GlpeSrH1CxEWIB94L5IczkxhPacQKR4P6Y=; b=sWSzyYyc3vStnIW7Gkw30JGIBW DntSgh+aUjJpzaRm1VCQfELh75LtYhGciO4BnX/gGD26wrn2oTLZ3QvGEamPQJxC1VU4kX2Lw1tqJ 1AkyOqpeXn+klxAjs1HTUgo42GrDr2yCzBLYR7naJ2dBfjKTBGDq5owwWyp+HVbp5I14=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.15] xen/arm: Create dom0less domUs earlier Message-Id: Date: Tue, 08 Jun 2021 17:12:16 +0000 commit 0a64b182db36b3b1c6f11c3b41d9b27fbbb8a5c6 Author: Julien Grall AuthorDate: Mon May 17 17:47:13 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 17:52:45 2021 +0100 xen/arm: Create dom0less domUs earlier In a follow-up patch we will need to unallocate the boot modules before heap_init_late() is called. The modules will contain the domUs kernel and initramfs. Therefore Xen will need to create extra domUs (used by dom0less) before heap_init_late(). This has two consequences on dom0less: 1) Domains will not be unpaused as soon as they are created but once all have been created. However, Xen doesn't guarantee an order to unpause, so this is not something one could rely on. 2) The memory allocated for a domU will not be scrubbed anymore when an admin select bootscrub=on. This is not something we advertised, but if this is a concern we can introduce either force scrub for all domUs or a per-domain flag in the DT. The behavior for bootscrub=off and bootscrub=idle (default) has not changed. This is part of XSA-372 / CVE-2021-28693. Signed-off-by: Julien Grall Reviewed-by: Jan Beulich Reviewed-by: Stefano Stabellini Tested-by: Stefano Stabellini (cherry picked from commit 371347c5b64da699d9f5a0edda5dc496fd2b7a5c) --- xen/arch/arm/domain_build.c | 2 -- xen/arch/arm/setup.c | 11 ++++++----- 2 files changed, 6 insertions(+), 7 deletions(-) diff --git a/xen/arch/arm/domain_build.c b/xen/arch/arm/domain_build.c index 374bf655ee..4203ddcca0 100644 --- a/xen/arch/arm/domain_build.c +++ b/xen/arch/arm/domain_build.c @@ -2515,8 +2515,6 @@ void __init create_domUs(void) if ( construct_domU(d, node) != 0 ) panic("Could not set up domain %s\n", dt_node_name(node)); - - domain_unpause_by_systemcontroller(d); } } diff --git a/xen/arch/arm/setup.c b/xen/arch/arm/setup.c index 2532ec9739..441e0e16e9 100644 --- a/xen/arch/arm/setup.c +++ b/xen/arch/arm/setup.c @@ -804,7 +804,7 @@ void __init start_xen(unsigned long boot_phys_offset, int cpus, i; const char *cmdline; struct bootmodule *xen_bootmodule; - struct domain *dom0; + struct domain *dom0, *d; struct xen_domctl_createdomain dom0_cfg = { .flags = XEN_DOMCTL_CDF_hvm | XEN_DOMCTL_CDF_hap, .max_evtchn_port = -1, @@ -987,6 +987,9 @@ void __init start_xen(unsigned long boot_phys_offset, if ( construct_dom0(dom0) != 0) panic("Could not set up DOM0 guest OS\n"); + if ( acpi_disabled ) + create_domUs(); + heap_init_late(); init_trace_bufs(); @@ -1000,10 +1003,8 @@ void __init start_xen(unsigned long boot_phys_offset, system_state = SYS_STATE_active; - if ( acpi_disabled ) - create_domUs(); - - domain_unpause_by_systemcontroller(dom0); + for_each_domain( d ) + domain_unpause_by_systemcontroller(d); /* Switch on to the dynamically allocated stack for the idle vcpu * since the static one we're running on is about to be freed. */ -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.15 From xen-changelog-bounces@lists.xenproject.org Tue Jun 08 17:12:28 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 08 Jun 2021 17:12:28 +0000 Received: from list by lists.xenproject.org with outflank-mailman.138698.256727 (Exim 4.92) (envelope-from ) id 1lqfHA-00027E-I7; Tue, 08 Jun 2021 17:12:28 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 138698.256727; Tue, 08 Jun 2021 17:12:28 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfHA-000276-F9; Tue, 08 Jun 2021 17:12:28 +0000 Received: by outflank-mailman (input) for mailman id 138698; Tue, 08 Jun 2021 17:12:27 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfH9-00026y-43 for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:12:27 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfH9-0004ya-3L for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:12:27 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqfH9-0007Ui-2h for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:12:27 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=F3jdniV2k8jRWarOMrpsgSoH1TYyR0EtpI7sXKsGYgY=; b=Tko4MjO4X5U6LFyJZ2F5f6/JXl 5uSch+0TqN+rcnaOP1L9UZy1siiNaR3JxzXp3M7JRldVVtUHWTiUGKHflG8rgmHjQtmM3iy9yQ0bc s30IXyQf5HNMoiWRZHxDSprct8yq9fR1mnHiA2mnq1a7OUQ2gV56o2raKeqEx6xoebAo=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.15] xen/arm: Boot modules should always be scrubbed if bootscrub={on, idle} Message-Id: Date: Tue, 08 Jun 2021 17:12:27 +0000 commit 70441846070ca5540e78b4a2d73a01458539ae90 Author: Julien Grall AuthorDate: Sat Apr 17 17:38:28 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 17:52:45 2021 +0100 xen/arm: Boot modules should always be scrubbed if bootscrub={on, idle} The function to initialize the pages (see init_heap_pages()) will request scrub when the admin request idle bootscrub (default) and state == SYS_STATE_active. When bootscrub=on, Xen will scrub any free pages in heap_init_late(). Currently, the boot modules (e.g. kernels, initramfs) will be discarded/ freed after heap_init_late() is called and system_state switched to SYS_STATE_active. This means the pages associated with the boot modules will not get scrubbed before getting re-purposed. If the memory is assigned to an untrusted domU, it may be able to retrieve secrets from the modules. This is part of XSA-372 / CVE-2021-28693. Fixes: 1774e9b1df27 ("xen/arm: introduce create_domUs") Signed-off-by: Julien Grall Reviewed-by: Jan Beulich Reviewed-by: Stefano Stabellini Tested-by: Stefano Stabellini (cherry picked from commit fd5dc41ceaed9cfcfa011cdfd50f264c89277a90) --- xen/arch/arm/setup.c | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-) diff --git a/xen/arch/arm/setup.c b/xen/arch/arm/setup.c index 441e0e16e9..8afb78f2c9 100644 --- a/xen/arch/arm/setup.c +++ b/xen/arch/arm/setup.c @@ -72,8 +72,6 @@ domid_t __read_mostly max_init_domid; static __used void init_done(void) { - discard_initial_modules(); - /* Must be done past setting system_state. */ unregister_init_virtual_region(); @@ -990,6 +988,12 @@ void __init start_xen(unsigned long boot_phys_offset, if ( acpi_disabled ) create_domUs(); + /* + * This needs to be called **before** heap_init_late() so modules + * will be scrubbed (unless suppressed). + */ + discard_initial_modules(); + heap_init_late(); init_trace_bufs(); -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.15 From xen-changelog-bounces@lists.xenproject.org Tue Jun 08 17:12:38 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 08 Jun 2021 17:12:38 +0000 Received: from list by lists.xenproject.org with outflank-mailman.138700.256731 (Exim 4.92) (envelope-from ) id 1lqfHK-0002AN-Jd; Tue, 08 Jun 2021 17:12:38 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 138700.256731; Tue, 08 Jun 2021 17:12:38 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfHK-0002AF-Gh; Tue, 08 Jun 2021 17:12:38 +0000 Received: by outflank-mailman (input) for mailman id 138700; Tue, 08 Jun 2021 17:12:37 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfHJ-0002A1-6u for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:12:37 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfHJ-0004yr-6C for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:12:37 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqfHJ-0007VL-5W for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:12:37 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=pYCK9ZTQCSHu25Di3LnH3PtJsSy5oj4ZZtWDz6Wa/mA=; b=ez73AHTlnL22DfH+K7o4+UgaLM o8qC+FylXA7SoljPBNrThJCCVmk/V50qBeFLoLmp4xQyIcbqb5CdcLviM6i0lnfnVW8BDEybZtu0N TK7vBABVlU/JD00B+za/pyWscTcvQAE8p0eSVHuZQz0QVhemBDOW+QKvWBC7QuqwXbJw=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.15] VT-d: size qinval queue dynamically Message-Id: Date: Tue, 08 Jun 2021 17:12:37 +0000 commit 8368f214f9446c3b1abc0045d01a2bc05926215a Author: Jan Beulich AuthorDate: Tue Jun 8 17:50:42 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 17:52:45 2021 +0100 VT-d: size qinval queue dynamically With the present synchronous model, we need two slots for every operation (the operation itself and a wait descriptor). There can be one such pair of requests pending per CPU. To ensure that under all normal circumstances a slot is always available when one is requested, size the queue ring according to the number of present CPUs. This is part of XSA-373 / CVE-2021-28692. Signed-off-by: Jan Beulich Reviewed-by: Paul Durrant (cherry picked from commit cbfa62bb140c8f10a0ae96db3ce06e22b3b9d302) --- xen/drivers/passthrough/vtd/iommu.h | 10 +------- xen/drivers/passthrough/vtd/qinval.c | 46 +++++++++++++++++++++++++++--------- 2 files changed, 36 insertions(+), 20 deletions(-) diff --git a/xen/drivers/passthrough/vtd/iommu.h b/xen/drivers/passthrough/vtd/iommu.h index 57737b1a4a..e002dc6f01 100644 --- a/xen/drivers/passthrough/vtd/iommu.h +++ b/xen/drivers/passthrough/vtd/iommu.h @@ -450,17 +450,9 @@ struct qinval_entry { }q; }; -/* Order of queue invalidation pages(max is 8) */ -#define QINVAL_PAGE_ORDER 2 - -#define QINVAL_ARCH_PAGE_ORDER (QINVAL_PAGE_ORDER + PAGE_SHIFT_4K - PAGE_SHIFT) -#define QINVAL_ARCH_PAGE_NR ( QINVAL_ARCH_PAGE_ORDER < 0 ? \ - 1 : \ - 1 << QINVAL_ARCH_PAGE_ORDER ) - /* Each entry is 16 bytes, so 2^8 entries per page */ #define QINVAL_ENTRY_ORDER ( PAGE_SHIFT - 4 ) -#define QINVAL_ENTRY_NR (1 << (QINVAL_PAGE_ORDER + 8)) +#define QINVAL_MAX_ENTRY_NR (1u << (7 + QINVAL_ENTRY_ORDER)) /* Status data flag */ #define QINVAL_STAT_INIT 0 diff --git a/xen/drivers/passthrough/vtd/qinval.c b/xen/drivers/passthrough/vtd/qinval.c index dfcbb66c8a..86a941ce27 100644 --- a/xen/drivers/passthrough/vtd/qinval.c +++ b/xen/drivers/passthrough/vtd/qinval.c @@ -31,6 +31,9 @@ #define VTD_QI_TIMEOUT 1 +static unsigned int __read_mostly qi_pg_order; +static unsigned int __read_mostly qi_entry_nr; + static int __must_check invalidate_sync(struct vtd_iommu *iommu); static void print_qi_regs(struct vtd_iommu *iommu) @@ -55,7 +58,7 @@ static unsigned int qinval_next_index(struct vtd_iommu *iommu) tail >>= QINVAL_INDEX_SHIFT; /* (tail+1 == head) indicates a full queue, wait for HW */ - while ( ( tail + 1 ) % QINVAL_ENTRY_NR == + while ( ((tail + 1) & (qi_entry_nr - 1)) == ( dmar_readq(iommu->reg, DMAR_IQH_REG) >> QINVAL_INDEX_SHIFT ) ) cpu_relax(); @@ -68,7 +71,7 @@ static void qinval_update_qtail(struct vtd_iommu *iommu, unsigned int index) /* Need hold register lock when update tail */ ASSERT( spin_is_locked(&iommu->register_lock) ); - val = (index + 1) % QINVAL_ENTRY_NR; + val = (index + 1) & (qi_entry_nr - 1); dmar_writeq(iommu->reg, DMAR_IQT_REG, (val << QINVAL_INDEX_SHIFT)); } @@ -403,8 +406,28 @@ int enable_qinval(struct vtd_iommu *iommu) if ( iommu->qinval_maddr == 0 ) { - iommu->qinval_maddr = alloc_pgtable_maddr(QINVAL_ARCH_PAGE_NR, - iommu->node); + if ( !qi_entry_nr ) + { + /* + * With the present synchronous model, we need two slots for every + * operation (the operation itself and a wait descriptor). There + * can be one such pair of requests pending per CPU. One extra + * entry is needed as the ring is considered full when there's + * only one entry left. + */ + BUILD_BUG_ON(CONFIG_NR_CPUS * 2 >= QINVAL_MAX_ENTRY_NR); + qi_pg_order = get_order_from_bytes((num_present_cpus() * 2 + 1) << + (PAGE_SHIFT - + QINVAL_ENTRY_ORDER)); + qi_entry_nr = 1u << (qi_pg_order + QINVAL_ENTRY_ORDER); + + dprintk(XENLOG_INFO VTDPREFIX, + "QI: using %u-entry ring(s)\n", qi_entry_nr); + } + + iommu->qinval_maddr = + alloc_pgtable_maddr(qi_entry_nr >> QINVAL_ENTRY_ORDER, + iommu->node); if ( iommu->qinval_maddr == 0 ) { dprintk(XENLOG_WARNING VTDPREFIX, @@ -418,15 +441,16 @@ int enable_qinval(struct vtd_iommu *iommu) spin_lock_irqsave(&iommu->register_lock, flags); - /* Setup Invalidation Queue Address(IQA) register with the - * address of the page we just allocated. QS field at - * bits[2:0] to indicate size of queue is one 4KB page. - * That's 256 entries. Queued Head (IQH) and Queue Tail (IQT) - * registers are automatically reset to 0 with write - * to IQA register. + /* + * Setup Invalidation Queue Address (IQA) register with the address of the + * pages we just allocated. The QS field at bits[2:0] indicates the size + * (page order) of the queue. + * + * Queued Head (IQH) and Queue Tail (IQT) registers are automatically + * reset to 0 with write to IQA register. */ dmar_writeq(iommu->reg, DMAR_IQA_REG, - iommu->qinval_maddr | QINVAL_PAGE_ORDER); + iommu->qinval_maddr | qi_pg_order); dmar_writeq(iommu->reg, DMAR_IQT_REG, 0); -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.15 From xen-changelog-bounces@lists.xenproject.org Tue Jun 08 17:12:48 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 08 Jun 2021 17:12:48 +0000 Received: from list by lists.xenproject.org with outflank-mailman.138701.256735 (Exim 4.92) (envelope-from ) id 1lqfHU-0002DG-L5; Tue, 08 Jun 2021 17:12:48 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 138701.256735; Tue, 08 Jun 2021 17:12:48 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfHU-0002D8-IB; Tue, 08 Jun 2021 17:12:48 +0000 Received: by outflank-mailman (input) for mailman id 138701; Tue, 08 Jun 2021 17:12:47 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfHT-0002Cv-9j for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:12:47 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfHT-0004zR-94 for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:12:47 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqfHT-0007W8-8N for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:12:47 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=6UknDr16qPzi0m3Df6T49Z1hlBeQ4q+dcFnSetfitT4=; b=jv7acAfNa6k+SSmlDr+vkuHjjf mEJ3yw89Prh+wwno9lUWvOuK+yrSA+8pvVoxbRrjwmqTw/wo1/bSya6Z29hD8Z1DY1u6q6EsAui9q ikA79Y5TGmhsR0O4at2D/rAsvspvIiTsRtfu5D7uc31L7CY+3l9Kz/hjpTYy7lD/62yo=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.15] AMD/IOMMU: size command buffer dynamically Message-Id: Date: Tue, 08 Jun 2021 17:12:47 +0000 commit 35b583686c8085ac78922eb58a2930222c9c142c Author: Jan Beulich AuthorDate: Tue Jun 8 17:50:45 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 17:52:45 2021 +0100 AMD/IOMMU: size command buffer dynamically With the present synchronous model, we need two slots for every operation (the operation itself and a wait command). There can be one such pair of commands pending per CPU. To ensure that under all normal circumstances a slot is always available when one is requested, size the command ring according to the number of present CPUs. This is part of XSA-373 / CVE-2021-28692. Signed-off-by: Jan Beulich Reviewed-by: Paul Durrant (cherry picked from commit df242851ddc93ac0b0a3a20ecab34acc29e3b36b) --- xen/drivers/passthrough/amd/iommu-defs.h | 7 ++----- xen/drivers/passthrough/amd/iommu_cmd.c | 4 ++-- xen/drivers/passthrough/amd/iommu_init.c | 28 +++++++++++++++++++++++++--- 3 files changed, 29 insertions(+), 10 deletions(-) diff --git a/xen/drivers/passthrough/amd/iommu-defs.h b/xen/drivers/passthrough/amd/iommu-defs.h index 56f385f270..0c97db2a32 100644 --- a/xen/drivers/passthrough/amd/iommu-defs.h +++ b/xen/drivers/passthrough/amd/iommu-defs.h @@ -20,9 +20,6 @@ #ifndef AMD_IOMMU_DEFS_H #define AMD_IOMMU_DEFS_H -/* IOMMU Command Buffer entries: in power of 2 increments, minimum of 256 */ -#define IOMMU_CMD_BUFFER_DEFAULT_ENTRIES 512 - /* IOMMU Event Log entries: in power of 2 increments, minimum of 256 */ #define IOMMU_EVENT_LOG_DEFAULT_ENTRIES 512 @@ -164,8 +161,8 @@ struct amd_iommu_dte { #define IOMMU_CMD_BUFFER_LENGTH_MASK 0x0F000000 #define IOMMU_CMD_BUFFER_LENGTH_SHIFT 24 -#define IOMMU_CMD_BUFFER_ENTRY_SIZE 16 -#define IOMMU_CMD_BUFFER_POWER_OF2_ENTRIES_PER_PAGE 8 +#define IOMMU_CMD_BUFFER_ENTRY_ORDER 4 +#define IOMMU_CMD_BUFFER_MAX_ENTRIES (1u << 15) #define IOMMU_CMD_OPCODE_MASK 0xF0000000 #define IOMMU_CMD_OPCODE_SHIFT 28 diff --git a/xen/drivers/passthrough/amd/iommu_cmd.c b/xen/drivers/passthrough/amd/iommu_cmd.c index 6c0647c524..54d7528f96 100644 --- a/xen/drivers/passthrough/amd/iommu_cmd.c +++ b/xen/drivers/passthrough/amd/iommu_cmd.c @@ -24,7 +24,7 @@ static int queue_iommu_command(struct amd_iommu *iommu, u32 cmd[]) { uint32_t tail, head; - tail = iommu->cmd_buffer.tail + IOMMU_CMD_BUFFER_ENTRY_SIZE; + tail = iommu->cmd_buffer.tail + sizeof(cmd_entry_t); if ( tail == iommu->cmd_buffer.size ) tail = 0; @@ -33,7 +33,7 @@ static int queue_iommu_command(struct amd_iommu *iommu, u32 cmd[]) if ( head != tail ) { memcpy(iommu->cmd_buffer.buffer + iommu->cmd_buffer.tail, - cmd, IOMMU_CMD_BUFFER_ENTRY_SIZE); + cmd, sizeof(cmd_entry_t)); iommu->cmd_buffer.tail = tail; return 1; diff --git a/xen/drivers/passthrough/amd/iommu_init.c b/xen/drivers/passthrough/amd/iommu_init.c index 24d1dfec40..bb52c181f8 100644 --- a/xen/drivers/passthrough/amd/iommu_init.c +++ b/xen/drivers/passthrough/amd/iommu_init.c @@ -118,7 +118,7 @@ static void register_iommu_cmd_buffer_in_mmio_space(struct amd_iommu *iommu) writel(entry, iommu->mmio_base + IOMMU_CMD_BUFFER_BASE_LOW_OFFSET); power_of2_entries = get_order_from_bytes(iommu->cmd_buffer.size) + - IOMMU_CMD_BUFFER_POWER_OF2_ENTRIES_PER_PAGE; + PAGE_SHIFT - IOMMU_CMD_BUFFER_ENTRY_ORDER; entry = 0; iommu_set_addr_hi_to_reg(&entry, addr_hi); @@ -1018,9 +1018,31 @@ static void *__init allocate_ring_buffer(struct ring_buffer *ring_buf, static void * __init allocate_cmd_buffer(struct amd_iommu *iommu) { /* allocate 'command buffer' in power of 2 increments of 4K */ + static unsigned int __read_mostly nr_ents; + + if ( !nr_ents ) + { + unsigned int order; + + /* + * With the present synchronous model, we need two slots for every + * operation (the operation itself and a wait command). There can be + * one such pair of requests pending per CPU. One extra entry is + * needed as the ring is considered full when there's only one entry + * left. + */ + BUILD_BUG_ON(CONFIG_NR_CPUS * 2 >= IOMMU_CMD_BUFFER_MAX_ENTRIES); + order = get_order_from_bytes((num_present_cpus() * 2 + 1) << + IOMMU_CMD_BUFFER_ENTRY_ORDER); + nr_ents = 1u << (order + PAGE_SHIFT - IOMMU_CMD_BUFFER_ENTRY_ORDER); + + AMD_IOMMU_DEBUG("using %u-entry cmd ring(s)\n", nr_ents); + } + + BUILD_BUG_ON(sizeof(cmd_entry_t) != (1u << IOMMU_CMD_BUFFER_ENTRY_ORDER)); + return allocate_ring_buffer(&iommu->cmd_buffer, sizeof(cmd_entry_t), - IOMMU_CMD_BUFFER_DEFAULT_ENTRIES, - "Command Buffer", false); + nr_ents, "Command Buffer", false); } static void * __init allocate_event_log(struct amd_iommu *iommu) -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.15 From xen-changelog-bounces@lists.xenproject.org Tue Jun 08 17:12:58 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 08 Jun 2021 17:12:58 +0000 Received: from list by lists.xenproject.org with outflank-mailman.138702.256739 (Exim 4.92) (envelope-from ) id 1lqfHe-0002GQ-Me; Tue, 08 Jun 2021 17:12:58 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 138702.256739; Tue, 08 Jun 2021 17:12:58 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfHe-0002GG-Jf; Tue, 08 Jun 2021 17:12:58 +0000 Received: by outflank-mailman (input) for mailman id 138702; Tue, 08 Jun 2021 17:12:57 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfHd-0002G0-Ch for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:12:57 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfHd-0004zd-By for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:12:57 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqfHd-0007Wn-BG for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:12:57 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=XqhbKMZITh8UqMBnud880SikXPCE46rdHGKkg5qm6zQ=; b=i6KM9S2Er1pJ/LzBivEZr4u5OG Rtlsqa4TZtlyWz+9wb/rS/9WXVB2VycvKUR3EFkPM5oNgta9awITLNa8kmyfq6Jve6eUbNRdLaAxA 70hvFILQwGsfLyR2JLeg5T46/lWLBq1VMVC4+dEzTLvOTmtpHBxwK0r4rbNgM/dIyTVA=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.15] VT-d: eliminate flush related timeouts Message-Id: Date: Tue, 08 Jun 2021 17:12:57 +0000 commit 12ebf0f819517096e49a395bd9c4288b5010526d Author: Jan Beulich AuthorDate: Tue Jun 8 17:50:47 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 17:52:45 2021 +0100 VT-d: eliminate flush related timeouts Leaving an in-progress operation pending when it appears to take too long is problematic: If e.g. a QI command completed later, the write to the "poll slot" may instead be understood to signal a subsequently started command's completion. Also our accounting of the timeout period was actually wrong: We included the time it took for the command to actually make it to the front of the queue, which could be heavily affected by guests other than the one for which the flush is being performed. Do away with all timeout detection on all flush related code paths. Log excessively long processing times (with a progressive threshold) to have some indication of problems in this area. Additionally log (once) if qinval_next_index() didn't immediately find an available slot. Together with the earlier change sizing the queue(s) dynamically, we should now have a guarantee that with our fully synchronous model any demand for slots can actually be satisfied. This is part of XSA-373 / CVE-2021-28692. Signed-off-by: Jan Beulich Reviewed-by: Paul Durrant (cherry picked from commit 7c893abc4ee7e9af62297ba6fd5f27c89d0f33f5) --- xen/drivers/passthrough/vtd/dmar.h | 28 ++++++++++++++++++++++++++++ xen/drivers/passthrough/vtd/iommu.c | 12 ++++++------ xen/drivers/passthrough/vtd/qinval.c | 33 ++++++++++++++++++++++----------- 3 files changed, 56 insertions(+), 17 deletions(-) diff --git a/xen/drivers/passthrough/vtd/dmar.h b/xen/drivers/passthrough/vtd/dmar.h index 1a9c965e59..a44820a5e7 100644 --- a/xen/drivers/passthrough/vtd/dmar.h +++ b/xen/drivers/passthrough/vtd/dmar.h @@ -127,6 +127,34 @@ do { \ } \ } while (0) +#define IOMMU_FLUSH_WAIT(what, iommu, offset, op, cond, sts) \ +do { \ + static unsigned int __read_mostly threshold = 1; \ + s_time_t start = NOW(); \ + s_time_t timeout = start + DMAR_OPERATION_TIMEOUT * threshold; \ + \ + for ( ; ; ) \ + { \ + sts = op(iommu->reg, offset); \ + if ( cond ) \ + break; \ + if ( timeout && NOW() > timeout ) \ + { \ + threshold |= threshold << 1; \ + printk(XENLOG_WARNING VTDPREFIX \ + " IOMMU#%u: %s flush taking too long\n", \ + iommu->index, what); \ + timeout = 0; \ + } \ + cpu_relax(); \ + } \ + \ + if ( !timeout ) \ + printk(XENLOG_WARNING VTDPREFIX \ + " IOMMU#%u: %s flush took %lums\n", \ + iommu->index, what, (NOW() - start) / 10000000); \ +} while ( false ) + int vtd_hw_check(void); void disable_pmr(struct vtd_iommu *iommu); int is_igd_drhd(struct acpi_drhd_unit *drhd); diff --git a/xen/drivers/passthrough/vtd/iommu.c b/xen/drivers/passthrough/vtd/iommu.c index d14fb43182..c160a70688 100644 --- a/xen/drivers/passthrough/vtd/iommu.c +++ b/xen/drivers/passthrough/vtd/iommu.c @@ -373,8 +373,8 @@ static void iommu_flush_write_buffer(struct vtd_iommu *iommu) dmar_writel(iommu->reg, DMAR_GCMD_REG, val | DMA_GCMD_WBF); /* Make sure hardware complete it */ - IOMMU_WAIT_OP(iommu, DMAR_GSTS_REG, dmar_readl, - !(val & DMA_GSTS_WBFS), val); + IOMMU_FLUSH_WAIT("write buffer", iommu, DMAR_GSTS_REG, dmar_readl, + !(val & DMA_GSTS_WBFS), val); spin_unlock_irqrestore(&iommu->register_lock, flags); } @@ -423,8 +423,8 @@ int vtd_flush_context_reg(struct vtd_iommu *iommu, uint16_t did, dmar_writeq(iommu->reg, DMAR_CCMD_REG, val); /* Make sure hardware complete it */ - IOMMU_WAIT_OP(iommu, DMAR_CCMD_REG, dmar_readq, - !(val & DMA_CCMD_ICC), val); + IOMMU_FLUSH_WAIT("context", iommu, DMAR_CCMD_REG, dmar_readq, + !(val & DMA_CCMD_ICC), val); spin_unlock_irqrestore(&iommu->register_lock, flags); /* flush context entry will implicitly flush write buffer */ @@ -501,8 +501,8 @@ int vtd_flush_iotlb_reg(struct vtd_iommu *iommu, uint16_t did, uint64_t addr, dmar_writeq(iommu->reg, tlb_offset + 8, val); /* Make sure hardware complete it */ - IOMMU_WAIT_OP(iommu, (tlb_offset + 8), dmar_readq, - !(val & DMA_TLB_IVT), val); + IOMMU_FLUSH_WAIT("iotlb", iommu, (tlb_offset + 8), dmar_readq, + !(val & DMA_TLB_IVT), val); spin_unlock_irqrestore(&iommu->register_lock, flags); /* check IOTLB invalidation granularity */ diff --git a/xen/drivers/passthrough/vtd/qinval.c b/xen/drivers/passthrough/vtd/qinval.c index 86a941ce27..1056ab1feb 100644 --- a/xen/drivers/passthrough/vtd/qinval.c +++ b/xen/drivers/passthrough/vtd/qinval.c @@ -29,8 +29,6 @@ #include "extern.h" #include "../ats.h" -#define VTD_QI_TIMEOUT 1 - static unsigned int __read_mostly qi_pg_order; static unsigned int __read_mostly qi_entry_nr; @@ -60,7 +58,11 @@ static unsigned int qinval_next_index(struct vtd_iommu *iommu) /* (tail+1 == head) indicates a full queue, wait for HW */ while ( ((tail + 1) & (qi_entry_nr - 1)) == ( dmar_readq(iommu->reg, DMAR_IQH_REG) >> QINVAL_INDEX_SHIFT ) ) + { + printk_once(XENLOG_ERR VTDPREFIX " IOMMU#%u: no QI slot available\n", + iommu->index); cpu_relax(); + } return tail; } @@ -180,23 +182,32 @@ static int __must_check queue_invalidate_wait(struct vtd_iommu *iommu, /* Now we don't support interrupt method */ if ( sw ) { - s_time_t timeout; - - /* In case all wait descriptor writes to same addr with same data */ - timeout = NOW() + MILLISECS(flush_dev_iotlb ? - iommu_dev_iotlb_timeout : VTD_QI_TIMEOUT); + static unsigned int __read_mostly threshold = 1; + s_time_t start = NOW(); + s_time_t timeout = start + (flush_dev_iotlb + ? iommu_dev_iotlb_timeout + : 100) * MILLISECS(threshold); while ( ACCESS_ONCE(*this_poll_slot) != QINVAL_STAT_DONE ) { - if ( NOW() > timeout ) + if ( timeout && NOW() > timeout ) { - print_qi_regs(iommu); + threshold |= threshold << 1; printk(XENLOG_WARNING VTDPREFIX - " Queue invalidate wait descriptor timed out\n"); - return -ETIMEDOUT; + " IOMMU#%u: QI%s wait descriptor taking too long\n", + iommu->index, flush_dev_iotlb ? " dev" : ""); + print_qi_regs(iommu); + timeout = 0; } cpu_relax(); } + + if ( !timeout ) + printk(XENLOG_WARNING VTDPREFIX + " IOMMU#%u: QI%s wait descriptor took %lums\n", + iommu->index, flush_dev_iotlb ? " dev" : "", + (NOW() - start) / 10000000); + return 0; } -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.15 From xen-changelog-bounces@lists.xenproject.org Tue Jun 08 17:13:08 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 08 Jun 2021 17:13:08 +0000 Received: from list by lists.xenproject.org with outflank-mailman.138703.256743 (Exim 4.92) (envelope-from ) id 1lqfHo-0002Jx-QB; Tue, 08 Jun 2021 17:13:08 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 138703.256743; Tue, 08 Jun 2021 17:13:08 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfHo-0002Jp-Ml; Tue, 08 Jun 2021 17:13:08 +0000 Received: by outflank-mailman (input) for mailman id 138703; Tue, 08 Jun 2021 17:13:07 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfHn-0002JX-GQ for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:13:07 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfHn-000504-Fo for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:13:07 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqfHn-0007Xb-Eu for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:13:07 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=7v5rEFhcJfEZi40LpZOD01NGc8SQN1nnJQaYrCweNBY=; b=roUzTRioaJ4dM43Zo/fHZ4+IR0 ljvwzG6pcPXJYU16TOWfhEn+QgKwVQB5JfCQQS5/P4Tz62neLaSEnTPGt/AB5erzkzg9LGmX4BnOM C72CZGcr19VBNBDqluUMHhCa8gyuIc3AHttH/x54JuxhTSIt9Kf7ds9jfPba1wRH/qkk=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.15] x86/spec-ctrl: Protect against Speculative Code Store Bypass Message-Id: Date: Tue, 08 Jun 2021 17:13:07 +0000 commit 894636d192368b130146311af2e539346c6fd650 Author: Andrew Cooper AuthorDate: Thu Mar 11 14:39:11 2021 +0000 Commit: Andrew Cooper CommitDate: Tue Jun 8 17:52:45 2021 +0100 x86/spec-ctrl: Protect against Speculative Code Store Bypass Modern x86 processors have far-better-than-architecturally-guaranteed self modifying code detection. Typically, when a write hits an instruction in flight, a Machine Clear occurs to flush stale content in the frontend and backend. For self modifying code, before a write which hits an instruction in flight retires, the frontend can speculatively decode and execute the old instruction stream. Speculation of this form can suffer from type confusion in registers, and potentially leak data. Furthermore, updates are typically byte-wise, rather than atomic. Depending on timing, speculation can race ahead multiple times between individual writes, and execute the transiently-malformed instruction stream. Xen has stubs which are used in certain cases for emulation purposes. Inhibit speculation between updating the stub and executing it. This is XSA-375 / CVE-2021-0089. Signed-off-by: Andrew Cooper Reviewed-by: Jan Beulich (cherry picked from commit 45f59ed8865318bb0356954bad067f329677ce9e) --- xen/arch/x86/pv/emul-priv-op.c | 2 ++ xen/arch/x86/x86_emulate/x86_emulate.c | 1 + 2 files changed, 3 insertions(+) diff --git a/xen/arch/x86/pv/emul-priv-op.c b/xen/arch/x86/pv/emul-priv-op.c index 8889509d2a..11467a1e3a 100644 --- a/xen/arch/x86/pv/emul-priv-op.c +++ b/xen/arch/x86/pv/emul-priv-op.c @@ -138,6 +138,8 @@ static io_emul_stub_t *io_emul_stub_setup(struct priv_op_ctxt *ctxt, u8 opcode, /* Runtime confirmation that we haven't clobbered an adjacent stub. */ BUG_ON(STUB_BUF_SIZE / 2 < (p - ctxt->io_emul_stub)); + block_speculation(); /* SCSB */ + /* Handy function-typed pointer to the stub. */ return (void *)stub_va; diff --git a/xen/arch/x86/x86_emulate/x86_emulate.c b/xen/arch/x86/x86_emulate/x86_emulate.c index a8000cbcbd..7bfb91dfff 100644 --- a/xen/arch/x86/x86_emulate/x86_emulate.c +++ b/xen/arch/x86/x86_emulate/x86_emulate.c @@ -1257,6 +1257,7 @@ static inline int mkec(uint8_t e, int32_t ec, ...) # define invoke_stub(pre, post, constraints...) do { \ stub_exn.info = (union stub_exception_token) { .raw = ~0 }; \ stub_exn.line = __LINE__; /* Utility outweighs livepatching cost */ \ + block_speculation(); /* SCSB */ \ asm volatile ( pre "\n\tINDIRECT_CALL %[stub]\n\t" post "\n" \ ".Lret%=:\n\t" \ ".pushsection .fixup,\"ax\"\n" \ -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.15 From xen-changelog-bounces@lists.xenproject.org Tue Jun 08 17:13:19 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 08 Jun 2021 17:13:19 +0000 Received: from list by lists.xenproject.org with outflank-mailman.138723.256760 (Exim 4.92) (envelope-from ) id 1lqfHz-0002hc-8s; Tue, 08 Jun 2021 17:13:19 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 138723.256760; Tue, 08 Jun 2021 17:13:19 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfHz-0002hP-4f; Tue, 08 Jun 2021 17:13:19 +0000 Received: by outflank-mailman (input) for mailman id 138723; Tue, 08 Jun 2021 17:13:17 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfHx-0002fc-JK for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:13:17 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfHx-00050c-Ik for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:13:17 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqfHx-0007YO-Hw for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:13:17 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=cQmqNZl4YqfeCcOlieB9KENtmLwPo7eqjSVd7U7PEFU=; b=XJ4EUHhbrf4KwbpQgg8DC5tTRR Cx61Rbg0fMWOAZra9VgkjkmPL1SjNKmgXqSe61ppJvnZG1su2Zbodd22NAsALJIun4e53C7Vp8yQr 5rSwBUcg1XQTGEJFElGp37YT3OlIYfAhBS8IrqdVVTDBRL+pDKG6wcKOa7no6YTewA48=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.15] x86/spec-ctrl: Mitigate TAA after S3 resume Message-Id: Date: Tue, 08 Jun 2021 17:13:17 +0000 commit f034c96e882b81738720472cd28e75e6d6eb66fe Author: Andrew Cooper AuthorDate: Thu May 20 01:21:39 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 17:52:45 2021 +0100 x86/spec-ctrl: Mitigate TAA after S3 resume The user chosen setting for MSR_TSX_CTRL needs restoring after S3. All APs get the correct setting via start_secondary(), but the BSP was missed out. This is XSA-377 / CVE-2021-28690. Fixes: 8c4330818f6 ("x86/spec-ctrl: Mitigate the TSX Asynchronous Abort sidechannel") Signed-off-by: Andrew Cooper Reviewed-by: Jan Beulich (cherry picked from commit 8cf276cb2e0b99b96333865873f56b0b31555ff1) --- xen/arch/x86/acpi/power.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/xen/arch/x86/acpi/power.c b/xen/arch/x86/acpi/power.c index 91a8c4d0bd..31a56f02d0 100644 --- a/xen/arch/x86/acpi/power.c +++ b/xen/arch/x86/acpi/power.c @@ -288,6 +288,8 @@ static int enter_state(u32 state) microcode_update_one(); + tsx_init(); /* Needs microcode. May change HLE/RTM feature bits. */ + if ( !recheck_cpu_features(0) ) panic("Missing previously available feature(s)\n"); -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.15 From xen-changelog-bounces@lists.xenproject.org Tue Jun 08 17:33:08 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 08 Jun 2021 17:33:08 +0000 Received: from list by lists.xenproject.org with outflank-mailman.138855.256835 (Exim 4.92) (envelope-from ) id 1lqfb7-0000g1-OK; Tue, 08 Jun 2021 17:33:05 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 138855.256835; Tue, 08 Jun 2021 17:33:05 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfb7-0000ft-LR; Tue, 08 Jun 2021 17:33:05 +0000 Received: by outflank-mailman (input) for mailman id 138855; Tue, 08 Jun 2021 17:33:05 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfb7-0000fn-0Q for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:33:05 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfb6-0005MR-Um for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:33:04 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqfb6-0000qJ-Tl for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:33:04 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=3ZHeKwRHNPNRZ039NQd0502J53XvzOxuSLxlqCxuKrA=; b=Wd/VT5/wu3XeiyYu466E//OHPd e0lxBJsjuxpIC38NHlsURGc8uFpopDJp6Cd/hpCQJ1cQF8ZiVOxiF5FuQ3Rr6g0MHoYwNkPrCmMhq iwi+H5ea8/Th0zkAs86k+h3OJeOKYqaMMHVPg9fbTHFIhJd64NYtmFzfx65gFLd7X4jk=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] AMD/IOMMU: wait for command slot to be available Message-Id: Date: Tue, 08 Jun 2021 17:33:04 +0000 commit c4e44343579e2c3304d676404d81b2b9bd893c5b Author: Jan Beulich AuthorDate: Tue Jun 8 18:29:40 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 18:29:40 2021 +0100 AMD/IOMMU: wait for command slot to be available No caller cared about send_iommu_command() indicating unavailability of a slot. Hence if a sufficient number prior commands timed out, we did blindly assume that the requested command was submitted to the IOMMU when really it wasn't. This could mean both a hanging system (waiting for a command to complete that was never seen by the IOMMU) or blindly propagating success back to callers, making them believe they're fine to e.g. free previously unmapped pages. Fold the three involved functions into one, add spin waiting for an available slot along the lines of VT-d's qinval_next_index(), and as a consequence drop all error indicator return types/values. This is part of XSA-373 / CVE-2021-28692. Signed-off-by: Jan Beulich Reviewed-by: Paul Durrant --- xen/drivers/passthrough/amd/iommu_cmd.c | 39 +++++++++++---------------------- 1 file changed, 13 insertions(+), 26 deletions(-) diff --git a/xen/drivers/passthrough/amd/iommu_cmd.c b/xen/drivers/passthrough/amd/iommu_cmd.c index 54d7528f96..135883aaf8 100644 --- a/xen/drivers/passthrough/amd/iommu_cmd.c +++ b/xen/drivers/passthrough/amd/iommu_cmd.c @@ -20,43 +20,30 @@ #include "iommu.h" #include "../ats.h" -static int queue_iommu_command(struct amd_iommu *iommu, u32 cmd[]) +static void send_iommu_command(struct amd_iommu *iommu, + const uint32_t cmd[4]) { - uint32_t tail, head; + uint32_t tail; tail = iommu->cmd_buffer.tail + sizeof(cmd_entry_t); if ( tail == iommu->cmd_buffer.size ) tail = 0; - head = readl(iommu->mmio_base + - IOMMU_CMD_BUFFER_HEAD_OFFSET) & IOMMU_RING_BUFFER_PTR_MASK; - if ( head != tail ) + while ( tail == (readl(iommu->mmio_base + + IOMMU_CMD_BUFFER_HEAD_OFFSET) & + IOMMU_RING_BUFFER_PTR_MASK) ) { - memcpy(iommu->cmd_buffer.buffer + iommu->cmd_buffer.tail, - cmd, sizeof(cmd_entry_t)); - - iommu->cmd_buffer.tail = tail; - return 1; + printk_once(XENLOG_ERR "AMD IOMMU %pp: no cmd slot available\n", + &PCI_SBDF2(iommu->seg, iommu->bdf)); + cpu_relax(); } - return 0; -} - -static void commit_iommu_command_buffer(struct amd_iommu *iommu) -{ - writel(iommu->cmd_buffer.tail, - iommu->mmio_base + IOMMU_CMD_BUFFER_TAIL_OFFSET); -} + memcpy(iommu->cmd_buffer.buffer + iommu->cmd_buffer.tail, + cmd, sizeof(cmd_entry_t)); -static int send_iommu_command(struct amd_iommu *iommu, u32 cmd[]) -{ - if ( queue_iommu_command(iommu, cmd) ) - { - commit_iommu_command_buffer(iommu); - return 1; - } + iommu->cmd_buffer.tail = tail; - return 0; + writel(tail, iommu->mmio_base + IOMMU_CMD_BUFFER_TAIL_OFFSET); } static void flush_command_buffer(struct amd_iommu *iommu) -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Tue Jun 08 17:33:16 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 08 Jun 2021 17:33:16 +0000 Received: from list by lists.xenproject.org with outflank-mailman.138856.256839 (Exim 4.92) (envelope-from ) id 1lqfbI-0000ih-Q2; Tue, 08 Jun 2021 17:33:16 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 138856.256839; Tue, 08 Jun 2021 17:33:16 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfbI-0000iZ-N2; Tue, 08 Jun 2021 17:33:16 +0000 Received: by outflank-mailman (input) for mailman id 138856; Tue, 08 Jun 2021 17:33:15 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfbH-0000iL-3R for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:33:15 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfbH-0005MZ-1q for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:33:15 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqfbH-0000rJ-0m for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:33:15 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=lK63rkYtNDLAanjcTVWq4TDPqyCP97JDcRwKEp3B4+Q=; b=jm7anrVWQg25MpmzEYulqlcqCN h+1Xz5TLd658SQuve82Z022C+Qdp/FSoHgp/k+BoZ7bIe0jjgDw+EWA4OTcCv5G26WZT8Ef71rzdA R/sONJFZqi3AY/ttb2Wga30DQmn61ElklzhKhbBl2VyZ4oP1zLlO+uetzA0RHfbUatF0=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] AMD/IOMMU: drop command completion timeout Message-Id: Date: Tue, 08 Jun 2021 17:33:15 +0000 commit e4fee66043120c954fc309bbb37813604c1c0eb7 Author: Jan Beulich AuthorDate: Tue Jun 8 18:29:42 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 18:29:42 2021 +0100 AMD/IOMMU: drop command completion timeout First and foremost - such timeouts were not signaled to callers, making them believe they're fine to e.g. free previously unmapped pages. Mirror VT-d's behavior: A fixed number of loop iterations is not a suitable way to detect timeouts in an environment (CPU and bus speeds) independent manner anyway. Furthermore, leaving an in-progress operation pending when it appears to take too long is problematic: If a command completed later, the signaling of its completion may instead be understood to signal a subsequently started command's completion. Log excessively long processing times (with a progressive threshold) to have some indication of problems in this area. Allow callers to specify a non-default timeout bias for this logging, using the same values as VT-d does, which in particular means a (by default) much larger value for device IO TLB invalidation. This is part of XSA-373 / CVE-2021-28692. Signed-off-by: Jan Beulich Reviewed-by: Paul Durrant --- xen/drivers/passthrough/amd/iommu_cmd.c | 56 +++++++++++++++++++-------------- 1 file changed, 33 insertions(+), 23 deletions(-) diff --git a/xen/drivers/passthrough/amd/iommu_cmd.c b/xen/drivers/passthrough/amd/iommu_cmd.c index 135883aaf8..92a688dc9d 100644 --- a/xen/drivers/passthrough/amd/iommu_cmd.c +++ b/xen/drivers/passthrough/amd/iommu_cmd.c @@ -46,10 +46,12 @@ static void send_iommu_command(struct amd_iommu *iommu, writel(tail, iommu->mmio_base + IOMMU_CMD_BUFFER_TAIL_OFFSET); } -static void flush_command_buffer(struct amd_iommu *iommu) +static void flush_command_buffer(struct amd_iommu *iommu, + unsigned int timeout_base) { - unsigned int cmd[4], status, loop_count; - bool comp_wait; + uint32_t cmd[4]; + s_time_t start, timeout; + static unsigned int __read_mostly threshold = 1; /* RW1C 'ComWaitInt' in status register */ writel(IOMMU_STATUS_COMP_WAIT_INT, @@ -65,22 +67,29 @@ static void flush_command_buffer(struct amd_iommu *iommu) IOMMU_COMP_WAIT_I_FLAG_SHIFT, &cmd[0]); send_iommu_command(iommu, cmd); - /* Make loop_count long enough for polling completion wait bit */ - loop_count = 1000; - do { - status = readl(iommu->mmio_base + IOMMU_STATUS_MMIO_OFFSET); - comp_wait = status & IOMMU_STATUS_COMP_WAIT_INT; - --loop_count; - } while ( !comp_wait && loop_count ); - - if ( comp_wait ) + start = NOW(); + timeout = start + (timeout_base ?: 100) * MILLISECS(threshold); + while ( !(readl(iommu->mmio_base + IOMMU_STATUS_MMIO_OFFSET) & + IOMMU_STATUS_COMP_WAIT_INT) ) { - /* RW1C 'ComWaitInt' in status register */ - writel(IOMMU_STATUS_COMP_WAIT_INT, - iommu->mmio_base + IOMMU_STATUS_MMIO_OFFSET); - return; + if ( timeout && NOW() > timeout ) + { + threshold |= threshold << 1; + printk(XENLOG_WARNING + "AMD IOMMU %pp: %scompletion wait taking too long\n", + &PCI_SBDF2(iommu->seg, iommu->bdf), + timeout_base ? "iotlb " : ""); + timeout = 0; + } + cpu_relax(); } - AMD_IOMMU_DEBUG("Warning: ComWaitInt bit did not assert!\n"); + + if ( !timeout ) + printk(XENLOG_WARNING + "AMD IOMMU %pp: %scompletion wait took %lums\n", + &PCI_SBDF2(iommu->seg, iommu->bdf), + timeout_base ? "iotlb " : "", + (NOW() - start) / 10000000); } /* Build low level iommu command messages */ @@ -291,7 +300,7 @@ void amd_iommu_flush_iotlb(u8 devfn, const struct pci_dev *pdev, /* send INVALIDATE_IOTLB_PAGES command */ spin_lock_irqsave(&iommu->lock, flags); invalidate_iotlb_pages(iommu, maxpend, 0, queueid, daddr, req_id, order); - flush_command_buffer(iommu); + flush_command_buffer(iommu, iommu_dev_iotlb_timeout); spin_unlock_irqrestore(&iommu->lock, flags); } @@ -328,7 +337,7 @@ static void _amd_iommu_flush_pages(struct domain *d, { spin_lock_irqsave(&iommu->lock, flags); invalidate_iommu_pages(iommu, daddr, dom_id, order); - flush_command_buffer(iommu); + flush_command_buffer(iommu, 0); spin_unlock_irqrestore(&iommu->lock, flags); } @@ -352,7 +361,7 @@ void amd_iommu_flush_device(struct amd_iommu *iommu, uint16_t bdf) ASSERT( spin_is_locked(&iommu->lock) ); invalidate_dev_table_entry(iommu, bdf); - flush_command_buffer(iommu); + flush_command_buffer(iommu, 0); } void amd_iommu_flush_intremap(struct amd_iommu *iommu, uint16_t bdf) @@ -360,7 +369,7 @@ void amd_iommu_flush_intremap(struct amd_iommu *iommu, uint16_t bdf) ASSERT( spin_is_locked(&iommu->lock) ); invalidate_interrupt_table(iommu, bdf); - flush_command_buffer(iommu); + flush_command_buffer(iommu, 0); } void amd_iommu_flush_all_caches(struct amd_iommu *iommu) @@ -368,7 +377,7 @@ void amd_iommu_flush_all_caches(struct amd_iommu *iommu) ASSERT( spin_is_locked(&iommu->lock) ); invalidate_iommu_all(iommu); - flush_command_buffer(iommu); + flush_command_buffer(iommu, 0); } void amd_iommu_send_guest_cmd(struct amd_iommu *iommu, u32 cmd[]) @@ -378,7 +387,8 @@ void amd_iommu_send_guest_cmd(struct amd_iommu *iommu, u32 cmd[]) spin_lock_irqsave(&iommu->lock, flags); send_iommu_command(iommu, cmd); - flush_command_buffer(iommu); + /* TBD: Timeout selection may require peeking into cmd[]. */ + flush_command_buffer(iommu, 0); spin_unlock_irqrestore(&iommu->lock, flags); } -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Tue Jun 08 17:44:10 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 08 Jun 2021 17:44:10 +0000 Received: from list by lists.xenproject.org with outflank-mailman.138863.256843 (Exim 4.92) (envelope-from ) id 1lqflm-0001uv-Jg; Tue, 08 Jun 2021 17:44:06 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 138863.256843; Tue, 08 Jun 2021 17:44:06 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqflm-0001uo-Fg; Tue, 08 Jun 2021 17:44:06 +0000 Received: by outflank-mailman (input) for mailman id 138863; Tue, 08 Jun 2021 17:44:05 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfll-0001tS-HE for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:44:05 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfll-0005YB-D3 for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:44:05 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqfll-0001ue-CA for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:44:05 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=0169D13mJ1QYuMJ/uI4arr7ZWRFvYj1fHqYEkJDO14k=; b=0JDExqqKfrqQbUHZULX+ZD8r38 Qki67oH1AGMg2OqPo5BRI8lvDCG/hYtCTlprQsPrJacCOhBeY6Hj9i8aGoMb0af3pOF+wDJNu8xoK Laqjfnx8n5Pf7NtFygXbGJt40bIapGPELPL2dKeagTy122G/yha5Y0wmFGQY6UXSy7lM=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.15] AMD/IOMMU: wait for command slot to be available Message-Id: Date: Tue, 08 Jun 2021 17:44:05 +0000 commit 874dac9b27c09a1fe976dccd62d4509c0389ffea Author: Jan Beulich AuthorDate: Tue Jun 8 18:33:49 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 18:34:05 2021 +0100 AMD/IOMMU: wait for command slot to be available No caller cared about send_iommu_command() indicating unavailability of a slot. Hence if a sufficient number prior commands timed out, we did blindly assume that the requested command was submitted to the IOMMU when really it wasn't. This could mean both a hanging system (waiting for a command to complete that was never seen by the IOMMU) or blindly propagating success back to callers, making them believe they're fine to e.g. free previously unmapped pages. Fold the three involved functions into one, add spin waiting for an available slot along the lines of VT-d's qinval_next_index(), and as a consequence drop all error indicator return types/values. This is part of XSA-373 / CVE-2021-28692. Signed-off-by: Jan Beulich Reviewed-by: Paul Durrant (cherry picked from commit c4e44343579e2c3304d676404d81b2b9bd893c5b) --- xen/drivers/passthrough/amd/iommu_cmd.c | 39 +++++++++++---------------------- 1 file changed, 13 insertions(+), 26 deletions(-) diff --git a/xen/drivers/passthrough/amd/iommu_cmd.c b/xen/drivers/passthrough/amd/iommu_cmd.c index 54d7528f96..135883aaf8 100644 --- a/xen/drivers/passthrough/amd/iommu_cmd.c +++ b/xen/drivers/passthrough/amd/iommu_cmd.c @@ -20,43 +20,30 @@ #include "iommu.h" #include "../ats.h" -static int queue_iommu_command(struct amd_iommu *iommu, u32 cmd[]) +static void send_iommu_command(struct amd_iommu *iommu, + const uint32_t cmd[4]) { - uint32_t tail, head; + uint32_t tail; tail = iommu->cmd_buffer.tail + sizeof(cmd_entry_t); if ( tail == iommu->cmd_buffer.size ) tail = 0; - head = readl(iommu->mmio_base + - IOMMU_CMD_BUFFER_HEAD_OFFSET) & IOMMU_RING_BUFFER_PTR_MASK; - if ( head != tail ) + while ( tail == (readl(iommu->mmio_base + + IOMMU_CMD_BUFFER_HEAD_OFFSET) & + IOMMU_RING_BUFFER_PTR_MASK) ) { - memcpy(iommu->cmd_buffer.buffer + iommu->cmd_buffer.tail, - cmd, sizeof(cmd_entry_t)); - - iommu->cmd_buffer.tail = tail; - return 1; + printk_once(XENLOG_ERR "AMD IOMMU %pp: no cmd slot available\n", + &PCI_SBDF2(iommu->seg, iommu->bdf)); + cpu_relax(); } - return 0; -} - -static void commit_iommu_command_buffer(struct amd_iommu *iommu) -{ - writel(iommu->cmd_buffer.tail, - iommu->mmio_base + IOMMU_CMD_BUFFER_TAIL_OFFSET); -} + memcpy(iommu->cmd_buffer.buffer + iommu->cmd_buffer.tail, + cmd, sizeof(cmd_entry_t)); -static int send_iommu_command(struct amd_iommu *iommu, u32 cmd[]) -{ - if ( queue_iommu_command(iommu, cmd) ) - { - commit_iommu_command_buffer(iommu); - return 1; - } + iommu->cmd_buffer.tail = tail; - return 0; + writel(tail, iommu->mmio_base + IOMMU_CMD_BUFFER_TAIL_OFFSET); } static void flush_command_buffer(struct amd_iommu *iommu) -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.15 From xen-changelog-bounces@lists.xenproject.org Tue Jun 08 17:44:16 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 08 Jun 2021 17:44:16 +0000 Received: from list by lists.xenproject.org with outflank-mailman.138864.256847 (Exim 4.92) (envelope-from ) id 1lqflw-0001xY-KI; Tue, 08 Jun 2021 17:44:16 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 138864.256847; Tue, 08 Jun 2021 17:44:16 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqflw-0001xQ-HN; Tue, 08 Jun 2021 17:44:16 +0000 Received: by outflank-mailman (input) for mailman id 138864; Tue, 08 Jun 2021 17:44:15 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqflv-0001xC-Gd for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:44:15 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqflv-0005YJ-Fs for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:44:15 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqflv-0001vW-FA for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:44:15 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=dH9ERq0c1GMDyGriNsDSlIq6mzRgKdFChn88tRzujew=; b=rChDKloxV2ZCSwhWC8Z2ii+zwG /Bzrq6b6wENQDF+b7Euu8WxNg6DYwForq9bX95KhjstO0KCz9feVwt1rwkzPepesyw2EmYv6r5QE4 nrsK+QJU9NmzFIWhFgi+655DrSN2F21S6e+pvz5jEnlDWWFss8pquJBF7k3HG9lAgEf4=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.15] AMD/IOMMU: drop command completion timeout Message-Id: Date: Tue, 08 Jun 2021 17:44:15 +0000 commit a339ceaa8f17e827ee5eb25f05ad6f52ba8d6b1c Author: Jan Beulich AuthorDate: Tue Jun 8 18:34:36 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 18:34:38 2021 +0100 AMD/IOMMU: drop command completion timeout First and foremost - such timeouts were not signaled to callers, making them believe they're fine to e.g. free previously unmapped pages. Mirror VT-d's behavior: A fixed number of loop iterations is not a suitable way to detect timeouts in an environment (CPU and bus speeds) independent manner anyway. Furthermore, leaving an in-progress operation pending when it appears to take too long is problematic: If a command completed later, the signaling of its completion may instead be understood to signal a subsequently started command's completion. Log excessively long processing times (with a progressive threshold) to have some indication of problems in this area. Allow callers to specify a non-default timeout bias for this logging, using the same values as VT-d does, which in particular means a (by default) much larger value for device IO TLB invalidation. This is part of XSA-373 / CVE-2021-28692. Signed-off-by: Jan Beulich Reviewed-by: Paul Durrant (cherry picked from commit e4fee66043120c954fc309bbb37813604c1c0eb7) --- xen/drivers/passthrough/amd/iommu_cmd.c | 56 +++++++++++++++++++-------------- 1 file changed, 33 insertions(+), 23 deletions(-) diff --git a/xen/drivers/passthrough/amd/iommu_cmd.c b/xen/drivers/passthrough/amd/iommu_cmd.c index 135883aaf8..92a688dc9d 100644 --- a/xen/drivers/passthrough/amd/iommu_cmd.c +++ b/xen/drivers/passthrough/amd/iommu_cmd.c @@ -46,10 +46,12 @@ static void send_iommu_command(struct amd_iommu *iommu, writel(tail, iommu->mmio_base + IOMMU_CMD_BUFFER_TAIL_OFFSET); } -static void flush_command_buffer(struct amd_iommu *iommu) +static void flush_command_buffer(struct amd_iommu *iommu, + unsigned int timeout_base) { - unsigned int cmd[4], status, loop_count; - bool comp_wait; + uint32_t cmd[4]; + s_time_t start, timeout; + static unsigned int __read_mostly threshold = 1; /* RW1C 'ComWaitInt' in status register */ writel(IOMMU_STATUS_COMP_WAIT_INT, @@ -65,22 +67,29 @@ static void flush_command_buffer(struct amd_iommu *iommu) IOMMU_COMP_WAIT_I_FLAG_SHIFT, &cmd[0]); send_iommu_command(iommu, cmd); - /* Make loop_count long enough for polling completion wait bit */ - loop_count = 1000; - do { - status = readl(iommu->mmio_base + IOMMU_STATUS_MMIO_OFFSET); - comp_wait = status & IOMMU_STATUS_COMP_WAIT_INT; - --loop_count; - } while ( !comp_wait && loop_count ); - - if ( comp_wait ) + start = NOW(); + timeout = start + (timeout_base ?: 100) * MILLISECS(threshold); + while ( !(readl(iommu->mmio_base + IOMMU_STATUS_MMIO_OFFSET) & + IOMMU_STATUS_COMP_WAIT_INT) ) { - /* RW1C 'ComWaitInt' in status register */ - writel(IOMMU_STATUS_COMP_WAIT_INT, - iommu->mmio_base + IOMMU_STATUS_MMIO_OFFSET); - return; + if ( timeout && NOW() > timeout ) + { + threshold |= threshold << 1; + printk(XENLOG_WARNING + "AMD IOMMU %pp: %scompletion wait taking too long\n", + &PCI_SBDF2(iommu->seg, iommu->bdf), + timeout_base ? "iotlb " : ""); + timeout = 0; + } + cpu_relax(); } - AMD_IOMMU_DEBUG("Warning: ComWaitInt bit did not assert!\n"); + + if ( !timeout ) + printk(XENLOG_WARNING + "AMD IOMMU %pp: %scompletion wait took %lums\n", + &PCI_SBDF2(iommu->seg, iommu->bdf), + timeout_base ? "iotlb " : "", + (NOW() - start) / 10000000); } /* Build low level iommu command messages */ @@ -291,7 +300,7 @@ void amd_iommu_flush_iotlb(u8 devfn, const struct pci_dev *pdev, /* send INVALIDATE_IOTLB_PAGES command */ spin_lock_irqsave(&iommu->lock, flags); invalidate_iotlb_pages(iommu, maxpend, 0, queueid, daddr, req_id, order); - flush_command_buffer(iommu); + flush_command_buffer(iommu, iommu_dev_iotlb_timeout); spin_unlock_irqrestore(&iommu->lock, flags); } @@ -328,7 +337,7 @@ static void _amd_iommu_flush_pages(struct domain *d, { spin_lock_irqsave(&iommu->lock, flags); invalidate_iommu_pages(iommu, daddr, dom_id, order); - flush_command_buffer(iommu); + flush_command_buffer(iommu, 0); spin_unlock_irqrestore(&iommu->lock, flags); } @@ -352,7 +361,7 @@ void amd_iommu_flush_device(struct amd_iommu *iommu, uint16_t bdf) ASSERT( spin_is_locked(&iommu->lock) ); invalidate_dev_table_entry(iommu, bdf); - flush_command_buffer(iommu); + flush_command_buffer(iommu, 0); } void amd_iommu_flush_intremap(struct amd_iommu *iommu, uint16_t bdf) @@ -360,7 +369,7 @@ void amd_iommu_flush_intremap(struct amd_iommu *iommu, uint16_t bdf) ASSERT( spin_is_locked(&iommu->lock) ); invalidate_interrupt_table(iommu, bdf); - flush_command_buffer(iommu); + flush_command_buffer(iommu, 0); } void amd_iommu_flush_all_caches(struct amd_iommu *iommu) @@ -368,7 +377,7 @@ void amd_iommu_flush_all_caches(struct amd_iommu *iommu) ASSERT( spin_is_locked(&iommu->lock) ); invalidate_iommu_all(iommu); - flush_command_buffer(iommu); + flush_command_buffer(iommu, 0); } void amd_iommu_send_guest_cmd(struct amd_iommu *iommu, u32 cmd[]) @@ -378,7 +387,8 @@ void amd_iommu_send_guest_cmd(struct amd_iommu *iommu, u32 cmd[]) spin_lock_irqsave(&iommu->lock, flags); send_iommu_command(iommu, cmd); - flush_command_buffer(iommu); + /* TBD: Timeout selection may require peeking into cmd[]. */ + flush_command_buffer(iommu, 0); spin_unlock_irqrestore(&iommu->lock, flags); } -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.15 From xen-changelog-bounces@lists.xenproject.org Tue Jun 08 17:55:10 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 08 Jun 2021 17:55:10 +0000 Received: from list by lists.xenproject.org with outflank-mailman.138865.256851 (Exim 4.92) (envelope-from ) id 1lqfwQ-00033U-Gd; Tue, 08 Jun 2021 17:55:06 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 138865.256851; Tue, 08 Jun 2021 17:55:06 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfwQ-00033M-DW; Tue, 08 Jun 2021 17:55:06 +0000 Received: by outflank-mailman (input) for mailman id 138865; Tue, 08 Jun 2021 17:55:05 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfwP-00033G-Il for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:55:05 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfwP-0005ia-E9 for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:55:05 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqfwP-0002nm-D5 for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:55:05 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=JzPWVW4Q/sMDvoWWVZeMbyokZ83n686aLkaP/qgLzRg=; b=1CyFdAxVGdtaVD2edY2Q1/tdOB 6UfsBHQxLeaYRfLrnoV1tn8lCdJtsMwzLpDi1V+z7DspktatT8b2ddhVLse8IHfvoJary3jwEWbh3 Sxm3T9pj95FfktYJ+eceNLi3DKADvJEryl7k0Kq1NjilQYSHiDTeefF6THamLZp6ncno=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.14] xen/arm: Create dom0less domUs earlier Message-Id: Date: Tue, 08 Jun 2021 17:55:05 +0000 commit 5caa690ea7e9dffa65a496424353c96af10f0308 Author: Julien Grall AuthorDate: Mon May 17 17:47:13 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 18:36:03 2021 +0100 xen/arm: Create dom0less domUs earlier In a follow-up patch we will need to unallocate the boot modules before heap_init_late() is called. The modules will contain the domUs kernel and initramfs. Therefore Xen will need to create extra domUs (used by dom0less) before heap_init_late(). This has two consequences on dom0less: 1) Domains will not be unpaused as soon as they are created but once all have been created. However, Xen doesn't guarantee an order to unpause, so this is not something one could rely on. 2) The memory allocated for a domU will not be scrubbed anymore when an admin select bootscrub=on. This is not something we advertised, but if this is a concern we can introduce either force scrub for all domUs or a per-domain flag in the DT. The behavior for bootscrub=off and bootscrub=idle (default) has not changed. This is part of XSA-372 / CVE-2021-28693. Signed-off-by: Julien Grall Reviewed-by: Jan Beulich Reviewed-by: Stefano Stabellini Tested-by: Stefano Stabellini (cherry picked from commit 371347c5b64da699d9f5a0edda5dc496fd2b7a5c) --- xen/arch/arm/domain_build.c | 2 -- xen/arch/arm/setup.c | 9 +++++---- 2 files changed, 5 insertions(+), 6 deletions(-) diff --git a/xen/arch/arm/domain_build.c b/xen/arch/arm/domain_build.c index e824ba34b0..b07461f5d3 100644 --- a/xen/arch/arm/domain_build.c +++ b/xen/arch/arm/domain_build.c @@ -2515,8 +2515,6 @@ void __init create_domUs(void) if ( construct_domU(d, node) != 0 ) panic("Could not set up domain %s\n", dt_node_name(node)); - - domain_unpause_by_systemcontroller(d); } } diff --git a/xen/arch/arm/setup.c b/xen/arch/arm/setup.c index 7968cee47d..1f26080b30 100644 --- a/xen/arch/arm/setup.c +++ b/xen/arch/arm/setup.c @@ -779,7 +779,7 @@ void __init start_xen(unsigned long boot_phys_offset, int cpus, i; const char *cmdline; struct bootmodule *xen_bootmodule; - struct domain *dom0; + struct domain *dom0, *d; struct xen_domctl_createdomain dom0_cfg = { .flags = XEN_DOMCTL_CDF_hvm | XEN_DOMCTL_CDF_hap, .max_evtchn_port = -1, @@ -962,6 +962,8 @@ void __init start_xen(unsigned long boot_phys_offset, if ( construct_dom0(dom0) != 0) panic("Could not set up DOM0 guest OS\n"); + create_domUs(); + heap_init_late(); init_trace_bufs(); @@ -975,9 +977,8 @@ void __init start_xen(unsigned long boot_phys_offset, system_state = SYS_STATE_active; - create_domUs(); - - domain_unpause_by_systemcontroller(dom0); + for_each_domain( d ) + domain_unpause_by_systemcontroller(d); /* Switch on to the dynamically allocated stack for the idle vcpu * since the static one we're running on is about to be freed. */ -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.14 From xen-changelog-bounces@lists.xenproject.org Tue Jun 08 17:55:16 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 08 Jun 2021 17:55:16 +0000 Received: from list by lists.xenproject.org with outflank-mailman.138866.256856 (Exim 4.92) (envelope-from ) id 1lqfwa-00035T-Is; Tue, 08 Jun 2021 17:55:16 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 138866.256856; Tue, 08 Jun 2021 17:55:16 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfwa-00035K-FN; Tue, 08 Jun 2021 17:55:16 +0000 Received: by outflank-mailman (input) for mailman id 138866; Tue, 08 Jun 2021 17:55:15 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfwZ-00035A-I9 for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:55:15 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfwZ-0005ii-HJ for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:55:15 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqfwZ-0002oi-GE for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:55:15 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=TKNAZ6QbAas92mrQWAjm4PDzhKcWzY9sTkOFBIk0jLM=; b=EcJbVFqAgQmRl9ZAnhhgj2dMjh j2309IjOHdzMS+tu2CLGuNq+7jqb4ysFz09B+d7TiaPFos5B9wtFZoY0/e2JqArSrxxV+NSwmitxi eKPEZIndO1XKVFf85dt0qC85PW93JzRezYqSL40QyBMuGVe9KE2Y8fn5vjdRLnxefJR4=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.14] xen/arm: Boot modules should always be scrubbed if bootscrub={on, idle} Message-Id: Date: Tue, 08 Jun 2021 17:55:15 +0000 commit 7053c8ea5dc88e0a3b6579d3bb61a242fcd5c499 Author: Julien Grall AuthorDate: Sat Apr 17 17:38:28 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 18:36:03 2021 +0100 xen/arm: Boot modules should always be scrubbed if bootscrub={on, idle} The function to initialize the pages (see init_heap_pages()) will request scrub when the admin request idle bootscrub (default) and state == SYS_STATE_active. When bootscrub=on, Xen will scrub any free pages in heap_init_late(). Currently, the boot modules (e.g. kernels, initramfs) will be discarded/ freed after heap_init_late() is called and system_state switched to SYS_STATE_active. This means the pages associated with the boot modules will not get scrubbed before getting re-purposed. If the memory is assigned to an untrusted domU, it may be able to retrieve secrets from the modules. This is part of XSA-372 / CVE-2021-28693. Fixes: 1774e9b1df27 ("xen/arm: introduce create_domUs") Signed-off-by: Julien Grall Reviewed-by: Jan Beulich Reviewed-by: Stefano Stabellini Tested-by: Stefano Stabellini (cherry picked from commit fd5dc41ceaed9cfcfa011cdfd50f264c89277a90) --- xen/arch/arm/setup.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-) diff --git a/xen/arch/arm/setup.c b/xen/arch/arm/setup.c index 1f26080b30..34b1c1a11e 100644 --- a/xen/arch/arm/setup.c +++ b/xen/arch/arm/setup.c @@ -75,7 +75,6 @@ static __used void init_done(void) /* Must be done past setting system_state. */ unregister_init_virtual_region(); - discard_initial_modules(); free_init_memory(); startup_cpu_idle_loop(); } @@ -964,6 +963,12 @@ void __init start_xen(unsigned long boot_phys_offset, create_domUs(); + /* + * This needs to be called **before** heap_init_late() so modules + * will be scrubbed (unless suppressed). + */ + discard_initial_modules(); + heap_init_late(); init_trace_bufs(); -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.14 From xen-changelog-bounces@lists.xenproject.org Tue Jun 08 17:55:26 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 08 Jun 2021 17:55:26 +0000 Received: from list by lists.xenproject.org with outflank-mailman.138867.256859 (Exim 4.92) (envelope-from ) id 1lqfwk-00038K-Jf; Tue, 08 Jun 2021 17:55:26 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 138867.256859; Tue, 08 Jun 2021 17:55:26 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfwk-00038D-Gx; Tue, 08 Jun 2021 17:55:26 +0000 Received: by outflank-mailman (input) for mailman id 138867; Tue, 08 Jun 2021 17:55:25 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfwj-000386-LG for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:55:25 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfwj-0005iw-KN for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:55:25 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqfwj-0002ph-Jd for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:55:25 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=9WduRZN2jkHPdaLsaxDKAxi/P4LADkvTyJL5PlvviC4=; b=N2ffkfKm5CQGml+3C1vloe7nKT opOjIRmKEJfgelJLCf1QkCWCFNCDO6D3hvgsL0w3wGurjssq5uRQYGm2bBnfZUI+ourPWKPLHi5kh 8JcZZ71zNeJPZNrz7tp4KTCiLUQ/uTVSdx78WhokVIwiXtxXg6c68Dw8nd4lqgTTjxm4=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.14] VT-d: size qinval queue dynamically Message-Id: Date: Tue, 08 Jun 2021 17:55:25 +0000 commit 2665d9785e748bf9cd87ec5783099e644ce2bac2 Author: Jan Beulich AuthorDate: Tue Jun 8 18:05:25 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 18:36:03 2021 +0100 VT-d: size qinval queue dynamically With the present synchronous model, we need two slots for every operation (the operation itself and a wait descriptor). There can be one such pair of requests pending per CPU. To ensure that under all normal circumstances a slot is always available when one is requested, size the queue ring according to the number of present CPUs. This is part of XSA-373 / CVE-2021-28692. Signed-off-by: Jan Beulich Reviewed-by: Paul Durrant (cherry picked from commit cbfa62bb140c8f10a0ae96db3ce06e22b3b9d302) --- xen/drivers/passthrough/vtd/iommu.h | 10 +------- xen/drivers/passthrough/vtd/qinval.c | 46 +++++++++++++++++++++++++++--------- 2 files changed, 36 insertions(+), 20 deletions(-) diff --git a/xen/drivers/passthrough/vtd/iommu.h b/xen/drivers/passthrough/vtd/iommu.h index 57737b1a4a..e002dc6f01 100644 --- a/xen/drivers/passthrough/vtd/iommu.h +++ b/xen/drivers/passthrough/vtd/iommu.h @@ -450,17 +450,9 @@ struct qinval_entry { }q; }; -/* Order of queue invalidation pages(max is 8) */ -#define QINVAL_PAGE_ORDER 2 - -#define QINVAL_ARCH_PAGE_ORDER (QINVAL_PAGE_ORDER + PAGE_SHIFT_4K - PAGE_SHIFT) -#define QINVAL_ARCH_PAGE_NR ( QINVAL_ARCH_PAGE_ORDER < 0 ? \ - 1 : \ - 1 << QINVAL_ARCH_PAGE_ORDER ) - /* Each entry is 16 bytes, so 2^8 entries per page */ #define QINVAL_ENTRY_ORDER ( PAGE_SHIFT - 4 ) -#define QINVAL_ENTRY_NR (1 << (QINVAL_PAGE_ORDER + 8)) +#define QINVAL_MAX_ENTRY_NR (1u << (7 + QINVAL_ENTRY_ORDER)) /* Status data flag */ #define QINVAL_STAT_INIT 0 diff --git a/xen/drivers/passthrough/vtd/qinval.c b/xen/drivers/passthrough/vtd/qinval.c index dfcbb66c8a..86a941ce27 100644 --- a/xen/drivers/passthrough/vtd/qinval.c +++ b/xen/drivers/passthrough/vtd/qinval.c @@ -31,6 +31,9 @@ #define VTD_QI_TIMEOUT 1 +static unsigned int __read_mostly qi_pg_order; +static unsigned int __read_mostly qi_entry_nr; + static int __must_check invalidate_sync(struct vtd_iommu *iommu); static void print_qi_regs(struct vtd_iommu *iommu) @@ -55,7 +58,7 @@ static unsigned int qinval_next_index(struct vtd_iommu *iommu) tail >>= QINVAL_INDEX_SHIFT; /* (tail+1 == head) indicates a full queue, wait for HW */ - while ( ( tail + 1 ) % QINVAL_ENTRY_NR == + while ( ((tail + 1) & (qi_entry_nr - 1)) == ( dmar_readq(iommu->reg, DMAR_IQH_REG) >> QINVAL_INDEX_SHIFT ) ) cpu_relax(); @@ -68,7 +71,7 @@ static void qinval_update_qtail(struct vtd_iommu *iommu, unsigned int index) /* Need hold register lock when update tail */ ASSERT( spin_is_locked(&iommu->register_lock) ); - val = (index + 1) % QINVAL_ENTRY_NR; + val = (index + 1) & (qi_entry_nr - 1); dmar_writeq(iommu->reg, DMAR_IQT_REG, (val << QINVAL_INDEX_SHIFT)); } @@ -403,8 +406,28 @@ int enable_qinval(struct vtd_iommu *iommu) if ( iommu->qinval_maddr == 0 ) { - iommu->qinval_maddr = alloc_pgtable_maddr(QINVAL_ARCH_PAGE_NR, - iommu->node); + if ( !qi_entry_nr ) + { + /* + * With the present synchronous model, we need two slots for every + * operation (the operation itself and a wait descriptor). There + * can be one such pair of requests pending per CPU. One extra + * entry is needed as the ring is considered full when there's + * only one entry left. + */ + BUILD_BUG_ON(CONFIG_NR_CPUS * 2 >= QINVAL_MAX_ENTRY_NR); + qi_pg_order = get_order_from_bytes((num_present_cpus() * 2 + 1) << + (PAGE_SHIFT - + QINVAL_ENTRY_ORDER)); + qi_entry_nr = 1u << (qi_pg_order + QINVAL_ENTRY_ORDER); + + dprintk(XENLOG_INFO VTDPREFIX, + "QI: using %u-entry ring(s)\n", qi_entry_nr); + } + + iommu->qinval_maddr = + alloc_pgtable_maddr(qi_entry_nr >> QINVAL_ENTRY_ORDER, + iommu->node); if ( iommu->qinval_maddr == 0 ) { dprintk(XENLOG_WARNING VTDPREFIX, @@ -418,15 +441,16 @@ int enable_qinval(struct vtd_iommu *iommu) spin_lock_irqsave(&iommu->register_lock, flags); - /* Setup Invalidation Queue Address(IQA) register with the - * address of the page we just allocated. QS field at - * bits[2:0] to indicate size of queue is one 4KB page. - * That's 256 entries. Queued Head (IQH) and Queue Tail (IQT) - * registers are automatically reset to 0 with write - * to IQA register. + /* + * Setup Invalidation Queue Address (IQA) register with the address of the + * pages we just allocated. The QS field at bits[2:0] indicates the size + * (page order) of the queue. + * + * Queued Head (IQH) and Queue Tail (IQT) registers are automatically + * reset to 0 with write to IQA register. */ dmar_writeq(iommu->reg, DMAR_IQA_REG, - iommu->qinval_maddr | QINVAL_PAGE_ORDER); + iommu->qinval_maddr | qi_pg_order); dmar_writeq(iommu->reg, DMAR_IQT_REG, 0); -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.14 From xen-changelog-bounces@lists.xenproject.org Tue Jun 08 17:55:37 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 08 Jun 2021 17:55:37 +0000 Received: from list by lists.xenproject.org with outflank-mailman.138868.256863 (Exim 4.92) (envelope-from ) id 1lqfwv-0003C6-Lf; Tue, 08 Jun 2021 17:55:37 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 138868.256863; Tue, 08 Jun 2021 17:55:37 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfwv-0003Bz-Ig; Tue, 08 Jun 2021 17:55:37 +0000 Received: by outflank-mailman (input) for mailman id 138868; Tue, 08 Jun 2021 17:55:35 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfwt-0003Bc-Oo for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:55:35 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfwt-0005j7-Nx for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:55:35 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqfwt-0002ql-Mx for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:55:35 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=uNd2ljCZBSIi5lAZcmlXQYEZpOWwje7nVzXroWRx4XM=; b=kbFJGpKqYpWMKSs0CquoTgxE56 pi1lvfNiJpE+ic1XnADagqPSsWMir4OWJFseKRA5bM+ozeaRhylrA97g2p2MgQxqJBFpTELeER7FR fLrhg6m7/FYqZGjc1tW7tLo0JZQgIb2a03wl36nl4Kp0kmKf8qReLTEKpY8vEY33llvs=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.14] AMD/IOMMU: size command buffer dynamically Message-Id: Date: Tue, 08 Jun 2021 17:55:35 +0000 commit 46ff2453de9e58ad49d79eda1be6ff21c22b3144 Author: Jan Beulich AuthorDate: Tue Jun 8 18:15:45 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 18:36:03 2021 +0100 AMD/IOMMU: size command buffer dynamically With the present synchronous model, we need two slots for every operation (the operation itself and a wait command). There can be one such pair of commands pending per CPU. To ensure that under all normal circumstances a slot is always available when one is requested, size the command ring according to the number of present CPUs. This is part of XSA-373 / CVE-2021-28692. Signed-off-by: Jan Beulich Reviewed-by: Paul Durrant (cherry picked from commit df242851ddc93ac0b0a3a20ecab34acc29e3b36b) --- xen/drivers/passthrough/amd/iommu-defs.h | 7 ++----- xen/drivers/passthrough/amd/iommu_cmd.c | 4 ++-- xen/drivers/passthrough/amd/iommu_init.c | 28 +++++++++++++++++++++++++--- 3 files changed, 29 insertions(+), 10 deletions(-) diff --git a/xen/drivers/passthrough/amd/iommu-defs.h b/xen/drivers/passthrough/amd/iommu-defs.h index 56f385f270..0c97db2a32 100644 --- a/xen/drivers/passthrough/amd/iommu-defs.h +++ b/xen/drivers/passthrough/amd/iommu-defs.h @@ -20,9 +20,6 @@ #ifndef AMD_IOMMU_DEFS_H #define AMD_IOMMU_DEFS_H -/* IOMMU Command Buffer entries: in power of 2 increments, minimum of 256 */ -#define IOMMU_CMD_BUFFER_DEFAULT_ENTRIES 512 - /* IOMMU Event Log entries: in power of 2 increments, minimum of 256 */ #define IOMMU_EVENT_LOG_DEFAULT_ENTRIES 512 @@ -164,8 +161,8 @@ struct amd_iommu_dte { #define IOMMU_CMD_BUFFER_LENGTH_MASK 0x0F000000 #define IOMMU_CMD_BUFFER_LENGTH_SHIFT 24 -#define IOMMU_CMD_BUFFER_ENTRY_SIZE 16 -#define IOMMU_CMD_BUFFER_POWER_OF2_ENTRIES_PER_PAGE 8 +#define IOMMU_CMD_BUFFER_ENTRY_ORDER 4 +#define IOMMU_CMD_BUFFER_MAX_ENTRIES (1u << 15) #define IOMMU_CMD_OPCODE_MASK 0xF0000000 #define IOMMU_CMD_OPCODE_SHIFT 28 diff --git a/xen/drivers/passthrough/amd/iommu_cmd.c b/xen/drivers/passthrough/amd/iommu_cmd.c index 249ed345a0..760babf20e 100644 --- a/xen/drivers/passthrough/amd/iommu_cmd.c +++ b/xen/drivers/passthrough/amd/iommu_cmd.c @@ -24,7 +24,7 @@ static int queue_iommu_command(struct amd_iommu *iommu, u32 cmd[]) { uint32_t tail, head; - tail = iommu->cmd_buffer.tail + IOMMU_CMD_BUFFER_ENTRY_SIZE; + tail = iommu->cmd_buffer.tail + sizeof(cmd_entry_t); if ( tail == iommu->cmd_buffer.size ) tail = 0; @@ -33,7 +33,7 @@ static int queue_iommu_command(struct amd_iommu *iommu, u32 cmd[]) if ( head != tail ) { memcpy(iommu->cmd_buffer.buffer + iommu->cmd_buffer.tail, - cmd, IOMMU_CMD_BUFFER_ENTRY_SIZE); + cmd, sizeof(cmd_entry_t)); iommu->cmd_buffer.tail = tail; return 1; diff --git a/xen/drivers/passthrough/amd/iommu_init.c b/xen/drivers/passthrough/amd/iommu_init.c index 034f3b9c2c..9b7e439eb7 100644 --- a/xen/drivers/passthrough/amd/iommu_init.c +++ b/xen/drivers/passthrough/amd/iommu_init.c @@ -118,7 +118,7 @@ static void register_iommu_cmd_buffer_in_mmio_space(struct amd_iommu *iommu) writel(entry, iommu->mmio_base + IOMMU_CMD_BUFFER_BASE_LOW_OFFSET); power_of2_entries = get_order_from_bytes(iommu->cmd_buffer.size) + - IOMMU_CMD_BUFFER_POWER_OF2_ENTRIES_PER_PAGE; + PAGE_SHIFT - IOMMU_CMD_BUFFER_ENTRY_ORDER; entry = 0; iommu_set_addr_hi_to_reg(&entry, addr_hi); @@ -1022,9 +1022,31 @@ static void *__init allocate_ring_buffer(struct ring_buffer *ring_buf, static void * __init allocate_cmd_buffer(struct amd_iommu *iommu) { /* allocate 'command buffer' in power of 2 increments of 4K */ + static unsigned int __read_mostly nr_ents; + + if ( !nr_ents ) + { + unsigned int order; + + /* + * With the present synchronous model, we need two slots for every + * operation (the operation itself and a wait command). There can be + * one such pair of requests pending per CPU. One extra entry is + * needed as the ring is considered full when there's only one entry + * left. + */ + BUILD_BUG_ON(CONFIG_NR_CPUS * 2 >= IOMMU_CMD_BUFFER_MAX_ENTRIES); + order = get_order_from_bytes((num_present_cpus() * 2 + 1) << + IOMMU_CMD_BUFFER_ENTRY_ORDER); + nr_ents = 1u << (order + PAGE_SHIFT - IOMMU_CMD_BUFFER_ENTRY_ORDER); + + AMD_IOMMU_DEBUG("using %u-entry cmd ring(s)\n", nr_ents); + } + + BUILD_BUG_ON(sizeof(cmd_entry_t) != (1u << IOMMU_CMD_BUFFER_ENTRY_ORDER)); + return allocate_ring_buffer(&iommu->cmd_buffer, sizeof(cmd_entry_t), - IOMMU_CMD_BUFFER_DEFAULT_ENTRIES, - "Command Buffer", false); + nr_ents, "Command Buffer", false); } static void * __init allocate_event_log(struct amd_iommu *iommu) -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.14 From xen-changelog-bounces@lists.xenproject.org Tue Jun 08 17:55:47 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 08 Jun 2021 17:55:47 +0000 Received: from list by lists.xenproject.org with outflank-mailman.138869.256868 (Exim 4.92) (envelope-from ) id 1lqfx5-0003Ev-Nw; Tue, 08 Jun 2021 17:55:47 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 138869.256868; Tue, 08 Jun 2021 17:55:47 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfx5-0003Eh-KB; Tue, 08 Jun 2021 17:55:47 +0000 Received: by outflank-mailman (input) for mailman id 138869; Tue, 08 Jun 2021 17:55:45 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfx3-0003EP-S8 for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:55:45 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfx3-0005ky-RG for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:55:45 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqfx3-0002rM-QH for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:55:45 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=PKmAmH0p83zjZEdRPx/WAybik6nJuY0FBnT4uO8FMuk=; b=UYfXtvbGeaHqKRR3pMPcjjz6HE QGbg9lTmzCG4MhkuPtTDi7O3diX9/bwqHumMeT72exuiUtlXhc4KwqssRh0mFziO9Zrd+H7fZSQP1 oCgSV3TxcsK2zzIpgXozImJuW37K1EP20LS9sFdHzkTYeCdNeKGSMk8SeHPL+UzpssUg=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.14] VT-d: eliminate flush related timeouts Message-Id: Date: Tue, 08 Jun 2021 17:55:45 +0000 commit e5bce3a74bacfe570233d61e57d31438d21fafd5 Author: Jan Beulich AuthorDate: Tue Jun 8 18:16:02 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 18:36:03 2021 +0100 VT-d: eliminate flush related timeouts Leaving an in-progress operation pending when it appears to take too long is problematic: If e.g. a QI command completed later, the write to the "poll slot" may instead be understood to signal a subsequently started command's completion. Also our accounting of the timeout period was actually wrong: We included the time it took for the command to actually make it to the front of the queue, which could be heavily affected by guests other than the one for which the flush is being performed. Do away with all timeout detection on all flush related code paths. Log excessively long processing times (with a progressive threshold) to have some indication of problems in this area. Additionally log (once) if qinval_next_index() didn't immediately find an available slot. Together with the earlier change sizing the queue(s) dynamically, we should now have a guarantee that with our fully synchronous model any demand for slots can actually be satisfied. This is part of XSA-373 / CVE-2021-28692. Signed-off-by: Jan Beulich Reviewed-by: Paul Durrant (cherry picked from commit 7c893abc4ee7e9af62297ba6fd5f27c89d0f33f5) --- xen/drivers/passthrough/vtd/dmar.h | 28 ++++++++++++++++++++++++++++ xen/drivers/passthrough/vtd/iommu.c | 12 ++++++------ xen/drivers/passthrough/vtd/qinval.c | 33 ++++++++++++++++++++++----------- 3 files changed, 56 insertions(+), 17 deletions(-) diff --git a/xen/drivers/passthrough/vtd/dmar.h b/xen/drivers/passthrough/vtd/dmar.h index 1a9c965e59..a44820a5e7 100644 --- a/xen/drivers/passthrough/vtd/dmar.h +++ b/xen/drivers/passthrough/vtd/dmar.h @@ -127,6 +127,34 @@ do { \ } \ } while (0) +#define IOMMU_FLUSH_WAIT(what, iommu, offset, op, cond, sts) \ +do { \ + static unsigned int __read_mostly threshold = 1; \ + s_time_t start = NOW(); \ + s_time_t timeout = start + DMAR_OPERATION_TIMEOUT * threshold; \ + \ + for ( ; ; ) \ + { \ + sts = op(iommu->reg, offset); \ + if ( cond ) \ + break; \ + if ( timeout && NOW() > timeout ) \ + { \ + threshold |= threshold << 1; \ + printk(XENLOG_WARNING VTDPREFIX \ + " IOMMU#%u: %s flush taking too long\n", \ + iommu->index, what); \ + timeout = 0; \ + } \ + cpu_relax(); \ + } \ + \ + if ( !timeout ) \ + printk(XENLOG_WARNING VTDPREFIX \ + " IOMMU#%u: %s flush took %lums\n", \ + iommu->index, what, (NOW() - start) / 10000000); \ +} while ( false ) + int vtd_hw_check(void); void disable_pmr(struct vtd_iommu *iommu); int is_igd_drhd(struct acpi_drhd_unit *drhd); diff --git a/xen/drivers/passthrough/vtd/iommu.c b/xen/drivers/passthrough/vtd/iommu.c index 32ff60bfe3..bef3d594c9 100644 --- a/xen/drivers/passthrough/vtd/iommu.c +++ b/xen/drivers/passthrough/vtd/iommu.c @@ -326,8 +326,8 @@ static void iommu_flush_write_buffer(struct vtd_iommu *iommu) dmar_writel(iommu->reg, DMAR_GCMD_REG, val | DMA_GCMD_WBF); /* Make sure hardware complete it */ - IOMMU_WAIT_OP(iommu, DMAR_GSTS_REG, dmar_readl, - !(val & DMA_GSTS_WBFS), val); + IOMMU_FLUSH_WAIT("write buffer", iommu, DMAR_GSTS_REG, dmar_readl, + !(val & DMA_GSTS_WBFS), val); spin_unlock_irqrestore(&iommu->register_lock, flags); } @@ -376,8 +376,8 @@ int vtd_flush_context_reg(struct vtd_iommu *iommu, uint16_t did, dmar_writeq(iommu->reg, DMAR_CCMD_REG, val); /* Make sure hardware complete it */ - IOMMU_WAIT_OP(iommu, DMAR_CCMD_REG, dmar_readq, - !(val & DMA_CCMD_ICC), val); + IOMMU_FLUSH_WAIT("context", iommu, DMAR_CCMD_REG, dmar_readq, + !(val & DMA_CCMD_ICC), val); spin_unlock_irqrestore(&iommu->register_lock, flags); /* flush context entry will implicitly flush write buffer */ @@ -454,8 +454,8 @@ int vtd_flush_iotlb_reg(struct vtd_iommu *iommu, uint16_t did, uint64_t addr, dmar_writeq(iommu->reg, tlb_offset + 8, val); /* Make sure hardware complete it */ - IOMMU_WAIT_OP(iommu, (tlb_offset + 8), dmar_readq, - !(val & DMA_TLB_IVT), val); + IOMMU_FLUSH_WAIT("iotlb", iommu, (tlb_offset + 8), dmar_readq, + !(val & DMA_TLB_IVT), val); spin_unlock_irqrestore(&iommu->register_lock, flags); /* check IOTLB invalidation granularity */ diff --git a/xen/drivers/passthrough/vtd/qinval.c b/xen/drivers/passthrough/vtd/qinval.c index 86a941ce27..1056ab1feb 100644 --- a/xen/drivers/passthrough/vtd/qinval.c +++ b/xen/drivers/passthrough/vtd/qinval.c @@ -29,8 +29,6 @@ #include "extern.h" #include "../ats.h" -#define VTD_QI_TIMEOUT 1 - static unsigned int __read_mostly qi_pg_order; static unsigned int __read_mostly qi_entry_nr; @@ -60,7 +58,11 @@ static unsigned int qinval_next_index(struct vtd_iommu *iommu) /* (tail+1 == head) indicates a full queue, wait for HW */ while ( ((tail + 1) & (qi_entry_nr - 1)) == ( dmar_readq(iommu->reg, DMAR_IQH_REG) >> QINVAL_INDEX_SHIFT ) ) + { + printk_once(XENLOG_ERR VTDPREFIX " IOMMU#%u: no QI slot available\n", + iommu->index); cpu_relax(); + } return tail; } @@ -180,23 +182,32 @@ static int __must_check queue_invalidate_wait(struct vtd_iommu *iommu, /* Now we don't support interrupt method */ if ( sw ) { - s_time_t timeout; - - /* In case all wait descriptor writes to same addr with same data */ - timeout = NOW() + MILLISECS(flush_dev_iotlb ? - iommu_dev_iotlb_timeout : VTD_QI_TIMEOUT); + static unsigned int __read_mostly threshold = 1; + s_time_t start = NOW(); + s_time_t timeout = start + (flush_dev_iotlb + ? iommu_dev_iotlb_timeout + : 100) * MILLISECS(threshold); while ( ACCESS_ONCE(*this_poll_slot) != QINVAL_STAT_DONE ) { - if ( NOW() > timeout ) + if ( timeout && NOW() > timeout ) { - print_qi_regs(iommu); + threshold |= threshold << 1; printk(XENLOG_WARNING VTDPREFIX - " Queue invalidate wait descriptor timed out\n"); - return -ETIMEDOUT; + " IOMMU#%u: QI%s wait descriptor taking too long\n", + iommu->index, flush_dev_iotlb ? " dev" : ""); + print_qi_regs(iommu); + timeout = 0; } cpu_relax(); } + + if ( !timeout ) + printk(XENLOG_WARNING VTDPREFIX + " IOMMU#%u: QI%s wait descriptor took %lums\n", + iommu->index, flush_dev_iotlb ? " dev" : "", + (NOW() - start) / 10000000); + return 0; } -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.14 From xen-changelog-bounces@lists.xenproject.org Tue Jun 08 17:55:57 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 08 Jun 2021 17:55:57 +0000 Received: from list by lists.xenproject.org with outflank-mailman.138870.256871 (Exim 4.92) (envelope-from ) id 1lqfxF-0003IB-Qo; Tue, 08 Jun 2021 17:55:57 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 138870.256871; Tue, 08 Jun 2021 17:55:57 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfxF-0003Hz-Na; Tue, 08 Jun 2021 17:55:57 +0000 Received: by outflank-mailman (input) for mailman id 138870; Tue, 08 Jun 2021 17:55:56 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfxD-0003Hj-Ux for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:55:55 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfxD-0005lH-UH for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:55:55 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqfxD-0002s3-TL for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:55:55 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=ptf7IvCM5vWmPex1xJjysuUUOYCfHIyHerWKPpWT5/Y=; b=buJgcmBj/dbnnw1TJWeVRQx2QI L8GatHKuZ/0CO2VchnIqrVMte8MyQPZzFI12Cu3wudCaozmATr7lOHt3EIUwmdjMxopulgNlDM3W2 XFHbdeZrDE3sRZlkMQ7AXDHKRYZIPBWumAmhuTYWdbh2YttVsa0iTWuXzH1VuPRVR3I0=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.14] AMD/IOMMU: wait for command slot to be available Message-Id: Date: Tue, 08 Jun 2021 17:55:55 +0000 commit 766b1f447314dd5e8f14c540440cbfc080f3b998 Author: Jan Beulich AuthorDate: Tue Jun 8 18:16:15 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 18:36:03 2021 +0100 AMD/IOMMU: wait for command slot to be available No caller cared about send_iommu_command() indicating unavailability of a slot. Hence if a sufficient number prior commands timed out, we did blindly assume that the requested command was submitted to the IOMMU when really it wasn't. This could mean both a hanging system (waiting for a command to complete that was never seen by the IOMMU) or blindly propagating success back to callers, making them believe they're fine to e.g. free previously unmapped pages. Fold the three involved functions into one, add spin waiting for an available slot along the lines of VT-d's qinval_next_index(), and as a consequence drop all error indicator return types/values. This is part of XSA-373 / CVE-2021-28692. Signed-off-by: Jan Beulich Reviewed-by: Paul Durrant (cherry picked from commit c4e44343579e2c3304d676404d81b2b9bd893c5b) --- xen/drivers/passthrough/amd/iommu_cmd.c | 41 ++++++++++++--------------------- 1 file changed, 15 insertions(+), 26 deletions(-) diff --git a/xen/drivers/passthrough/amd/iommu_cmd.c b/xen/drivers/passthrough/amd/iommu_cmd.c index 760babf20e..fa9603ee45 100644 --- a/xen/drivers/passthrough/amd/iommu_cmd.c +++ b/xen/drivers/passthrough/amd/iommu_cmd.c @@ -20,43 +20,32 @@ #include "iommu.h" #include "../ats.h" -static int queue_iommu_command(struct amd_iommu *iommu, u32 cmd[]) +static void send_iommu_command(struct amd_iommu *iommu, + const uint32_t cmd[4]) { - uint32_t tail, head; + uint32_t tail; tail = iommu->cmd_buffer.tail + sizeof(cmd_entry_t); if ( tail == iommu->cmd_buffer.size ) tail = 0; - head = readl(iommu->mmio_base + - IOMMU_CMD_BUFFER_HEAD_OFFSET) & IOMMU_RING_BUFFER_PTR_MASK; - if ( head != tail ) + while ( tail == (readl(iommu->mmio_base + + IOMMU_CMD_BUFFER_HEAD_OFFSET) & + IOMMU_RING_BUFFER_PTR_MASK) ) { - memcpy(iommu->cmd_buffer.buffer + iommu->cmd_buffer.tail, - cmd, sizeof(cmd_entry_t)); - - iommu->cmd_buffer.tail = tail; - return 1; + printk_once(XENLOG_ERR + "AMD IOMMU %04x:%02x:%02x.%u: no cmd slot available\n", + iommu->seg, PCI_BUS(iommu->bdf), + PCI_SLOT(iommu->bdf), PCI_FUNC(iommu->bdf)); + cpu_relax(); } - return 0; -} - -static void commit_iommu_command_buffer(struct amd_iommu *iommu) -{ - writel(iommu->cmd_buffer.tail, - iommu->mmio_base + IOMMU_CMD_BUFFER_TAIL_OFFSET); -} + memcpy(iommu->cmd_buffer.buffer + iommu->cmd_buffer.tail, + cmd, sizeof(cmd_entry_t)); -static int send_iommu_command(struct amd_iommu *iommu, u32 cmd[]) -{ - if ( queue_iommu_command(iommu, cmd) ) - { - commit_iommu_command_buffer(iommu); - return 1; - } + iommu->cmd_buffer.tail = tail; - return 0; + writel(tail, iommu->mmio_base + IOMMU_CMD_BUFFER_TAIL_OFFSET); } static void flush_command_buffer(struct amd_iommu *iommu) -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.14 From xen-changelog-bounces@lists.xenproject.org Tue Jun 08 17:56:07 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 08 Jun 2021 17:56:07 +0000 Received: from list by lists.xenproject.org with outflank-mailman.138871.256875 (Exim 4.92) (envelope-from ) id 1lqfxP-0003LQ-SD; Tue, 08 Jun 2021 17:56:07 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 138871.256875; Tue, 08 Jun 2021 17:56:07 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfxP-0003LI-P7; Tue, 08 Jun 2021 17:56:07 +0000 Received: by outflank-mailman (input) for mailman id 138871; Tue, 08 Jun 2021 17:56:06 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfxO-0003L0-1Z for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:56:06 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfxO-0005le-0t for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:56:06 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqfxO-0002tH-08 for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:56:06 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=gvRI5hoQY5cBimsOD2a5mzUGm1FeSCnlkGcyFKRWUT8=; b=Goz26QL+xAEQshjjGI3ggzInbL z/Q07WHZPWRRj/F41XTXLwbd4qx9vfPAphYNN5WJHIhHrvLVzOT58+bJ0lvEw0cW4TTmreRw2/sui mi5Mp0bGLaD5KIOcGH8kYffyVqETvfXRtI6Ys7LtOdgTp/GCW8rLdvmKlHt6rJMIYqO0=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.14] AMD/IOMMU: drop command completion timeout Message-Id: Date: Tue, 08 Jun 2021 17:56:06 +0000 commit 51278ce831a8fcc4ffb4c841e3442de3fd4ea76b Author: Jan Beulich AuthorDate: Tue Jun 8 18:19:29 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 18:36:03 2021 +0100 AMD/IOMMU: drop command completion timeout First and foremost - such timeouts were not signaled to callers, making them believe they're fine to e.g. free previously unmapped pages. Mirror VT-d's behavior: A fixed number of loop iterations is not a suitable way to detect timeouts in an environment (CPU and bus speeds) independent manner anyway. Furthermore, leaving an in-progress operation pending when it appears to take too long is problematic: If a command completed later, the signaling of its completion may instead be understood to signal a subsequently started command's completion. Log excessively long processing times (with a progressive threshold) to have some indication of problems in this area. Allow callers to specify a non-default timeout bias for this logging, using the same values as VT-d does, which in particular means a (by default) much larger value for device IO TLB invalidation. This is part of XSA-373 / CVE-2021-28692. Signed-off-by: Jan Beulich Reviewed-by: Paul Durrant (cherry picked from commit e4fee66043120c954fc309bbb37813604c1c0eb7) --- xen/drivers/passthrough/amd/iommu_cmd.c | 58 ++++++++++++++++++++------------- 1 file changed, 35 insertions(+), 23 deletions(-) diff --git a/xen/drivers/passthrough/amd/iommu_cmd.c b/xen/drivers/passthrough/amd/iommu_cmd.c index fa9603ee45..fc5377f5f2 100644 --- a/xen/drivers/passthrough/amd/iommu_cmd.c +++ b/xen/drivers/passthrough/amd/iommu_cmd.c @@ -48,10 +48,12 @@ static void send_iommu_command(struct amd_iommu *iommu, writel(tail, iommu->mmio_base + IOMMU_CMD_BUFFER_TAIL_OFFSET); } -static void flush_command_buffer(struct amd_iommu *iommu) +static void flush_command_buffer(struct amd_iommu *iommu, + unsigned int timeout_base) { - unsigned int cmd[4], status, loop_count; - bool comp_wait; + uint32_t cmd[4]; + s_time_t start, timeout; + static unsigned int __read_mostly threshold = 1; /* RW1C 'ComWaitInt' in status register */ writel(IOMMU_STATUS_COMP_WAIT_INT, @@ -67,22 +69,31 @@ static void flush_command_buffer(struct amd_iommu *iommu) IOMMU_COMP_WAIT_I_FLAG_SHIFT, &cmd[0]); send_iommu_command(iommu, cmd); - /* Make loop_count long enough for polling completion wait bit */ - loop_count = 1000; - do { - status = readl(iommu->mmio_base + IOMMU_STATUS_MMIO_OFFSET); - comp_wait = status & IOMMU_STATUS_COMP_WAIT_INT; - --loop_count; - } while ( !comp_wait && loop_count ); - - if ( comp_wait ) + start = NOW(); + timeout = start + (timeout_base ?: 100) * MILLISECS(threshold); + while ( !(readl(iommu->mmio_base + IOMMU_STATUS_MMIO_OFFSET) & + IOMMU_STATUS_COMP_WAIT_INT) ) { - /* RW1C 'ComWaitInt' in status register */ - writel(IOMMU_STATUS_COMP_WAIT_INT, - iommu->mmio_base + IOMMU_STATUS_MMIO_OFFSET); - return; + if ( timeout && NOW() > timeout ) + { + threshold |= threshold << 1; + printk(XENLOG_WARNING + "AMD IOMMU %04x:%02x:%02x.%u: %scompletion wait taking too long\n", + iommu->seg, PCI_BUS(iommu->bdf), + PCI_SLOT(iommu->bdf), PCI_FUNC(iommu->bdf), + timeout_base ? "iotlb " : ""); + timeout = 0; + } + cpu_relax(); } - AMD_IOMMU_DEBUG("Warning: ComWaitInt bit did not assert!\n"); + + if ( !timeout ) + printk(XENLOG_WARNING + "AMD IOMMU %04x:%02x:%02x.%u: %scompletion wait took %lums\n", + iommu->seg, PCI_BUS(iommu->bdf), + PCI_SLOT(iommu->bdf), PCI_FUNC(iommu->bdf), + timeout_base ? "iotlb " : "", + (NOW() - start) / 10000000); } /* Build low level iommu command messages */ @@ -294,7 +305,7 @@ void amd_iommu_flush_iotlb(u8 devfn, const struct pci_dev *pdev, /* send INVALIDATE_IOTLB_PAGES command */ spin_lock_irqsave(&iommu->lock, flags); invalidate_iotlb_pages(iommu, maxpend, 0, queueid, daddr, req_id, order); - flush_command_buffer(iommu); + flush_command_buffer(iommu, iommu_dev_iotlb_timeout); spin_unlock_irqrestore(&iommu->lock, flags); } @@ -331,7 +342,7 @@ static void _amd_iommu_flush_pages(struct domain *d, { spin_lock_irqsave(&iommu->lock, flags); invalidate_iommu_pages(iommu, daddr, dom_id, order); - flush_command_buffer(iommu); + flush_command_buffer(iommu, 0); spin_unlock_irqrestore(&iommu->lock, flags); } @@ -355,7 +366,7 @@ void amd_iommu_flush_device(struct amd_iommu *iommu, uint16_t bdf) ASSERT( spin_is_locked(&iommu->lock) ); invalidate_dev_table_entry(iommu, bdf); - flush_command_buffer(iommu); + flush_command_buffer(iommu, 0); } void amd_iommu_flush_intremap(struct amd_iommu *iommu, uint16_t bdf) @@ -363,7 +374,7 @@ void amd_iommu_flush_intremap(struct amd_iommu *iommu, uint16_t bdf) ASSERT( spin_is_locked(&iommu->lock) ); invalidate_interrupt_table(iommu, bdf); - flush_command_buffer(iommu); + flush_command_buffer(iommu, 0); } void amd_iommu_flush_all_caches(struct amd_iommu *iommu) @@ -371,7 +382,7 @@ void amd_iommu_flush_all_caches(struct amd_iommu *iommu) ASSERT( spin_is_locked(&iommu->lock) ); invalidate_iommu_all(iommu); - flush_command_buffer(iommu); + flush_command_buffer(iommu, 0); } void amd_iommu_send_guest_cmd(struct amd_iommu *iommu, u32 cmd[]) @@ -381,7 +392,8 @@ void amd_iommu_send_guest_cmd(struct amd_iommu *iommu, u32 cmd[]) spin_lock_irqsave(&iommu->lock, flags); send_iommu_command(iommu, cmd); - flush_command_buffer(iommu); + /* TBD: Timeout selection may require peeking into cmd[]. */ + flush_command_buffer(iommu, 0); spin_unlock_irqrestore(&iommu->lock, flags); } -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.14 From xen-changelog-bounces@lists.xenproject.org Tue Jun 08 17:56:17 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 08 Jun 2021 17:56:17 +0000 Received: from list by lists.xenproject.org with outflank-mailman.138872.256879 (Exim 4.92) (envelope-from ) id 1lqfxZ-0003ON-TT; Tue, 08 Jun 2021 17:56:17 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 138872.256879; Tue, 08 Jun 2021 17:56:17 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfxZ-0003OF-Qe; Tue, 08 Jun 2021 17:56:17 +0000 Received: by outflank-mailman (input) for mailman id 138872; Tue, 08 Jun 2021 17:56:16 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfxY-0003Nx-4z for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:56:16 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfxY-0005lt-4J for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:56:16 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqfxY-0002uJ-3Z for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:56:16 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=CNXEH2WEcrnp0y1wy4EULqv9A87UZuhSkS3rjK4bA+Q=; b=RpOitAn5lqJU76hu3lpv4twK1b yrBlsqHinKcDQBjJhTXca4pnHxz1Fv4fBBcr/EWVcWIpJk7NMf07chRw0NHu4KgF0xNcpP0F7MPMP 8N2fpZOz5A3Ttk7nnkGSPXgZ7Y8GOCkdmMTCgo1ETPWt2gBp69mr2b+gU98QaqAM2dj0=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.14] x86/spec-ctrl: Protect against Speculative Code Store Bypass Message-Id: Date: Tue, 08 Jun 2021 17:56:16 +0000 commit fcf98eff7321a5ed8d9b4fccf44cde8ad7e2ec7e Author: Andrew Cooper AuthorDate: Thu Mar 11 14:39:11 2021 +0000 Commit: Andrew Cooper CommitDate: Tue Jun 8 18:36:03 2021 +0100 x86/spec-ctrl: Protect against Speculative Code Store Bypass Modern x86 processors have far-better-than-architecturally-guaranteed self modifying code detection. Typically, when a write hits an instruction in flight, a Machine Clear occurs to flush stale content in the frontend and backend. For self modifying code, before a write which hits an instruction in flight retires, the frontend can speculatively decode and execute the old instruction stream. Speculation of this form can suffer from type confusion in registers, and potentially leak data. Furthermore, updates are typically byte-wise, rather than atomic. Depending on timing, speculation can race ahead multiple times between individual writes, and execute the transiently-malformed instruction stream. Xen has stubs which are used in certain cases for emulation purposes. Inhibit speculation between updating the stub and executing it. This is XSA-375 / CVE-2021-0089. Signed-off-by: Andrew Cooper Reviewed-by: Jan Beulich (cherry picked from commit 45f59ed8865318bb0356954bad067f329677ce9e) --- xen/arch/x86/pv/emul-priv-op.c | 2 ++ xen/arch/x86/x86_emulate/x86_emulate.c | 1 + 2 files changed, 3 insertions(+) diff --git a/xen/arch/x86/pv/emul-priv-op.c b/xen/arch/x86/pv/emul-priv-op.c index 5c7b9117ae..5399fe382f 100644 --- a/xen/arch/x86/pv/emul-priv-op.c +++ b/xen/arch/x86/pv/emul-priv-op.c @@ -138,6 +138,8 @@ static io_emul_stub_t *io_emul_stub_setup(struct priv_op_ctxt *ctxt, u8 opcode, /* Runtime confirmation that we haven't clobbered an adjacent stub. */ BUG_ON(STUB_BUF_SIZE / 2 < (p - ctxt->io_emul_stub)); + block_speculation(); /* SCSB */ + /* Handy function-typed pointer to the stub. */ return (void *)stub_va; diff --git a/xen/arch/x86/x86_emulate/x86_emulate.c b/xen/arch/x86/x86_emulate/x86_emulate.c index 16d164904c..f186ae65fd 100644 --- a/xen/arch/x86/x86_emulate/x86_emulate.c +++ b/xen/arch/x86/x86_emulate/x86_emulate.c @@ -1256,6 +1256,7 @@ static inline int mkec(uint8_t e, int32_t ec, ...) # define invoke_stub(pre, post, constraints...) do { \ stub_exn.info = (union stub_exception_token) { .raw = ~0 }; \ stub_exn.line = __LINE__; /* Utility outweighs livepatching cost */ \ + block_speculation(); /* SCSB */ \ asm volatile ( pre "\n\tINDIRECT_CALL %[stub]\n\t" post "\n" \ ".Lret%=:\n\t" \ ".pushsection .fixup,\"ax\"\n" \ -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.14 From xen-changelog-bounces@lists.xenproject.org Tue Jun 08 17:56:28 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 08 Jun 2021 17:56:28 +0000 Received: from list by lists.xenproject.org with outflank-mailman.138873.256883 (Exim 4.92) (envelope-from ) id 1lqfxj-0003RC-V6; Tue, 08 Jun 2021 17:56:27 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 138873.256883; Tue, 08 Jun 2021 17:56:27 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfxj-0003R2-SB; Tue, 08 Jun 2021 17:56:27 +0000 Received: by outflank-mailman (input) for mailman id 138873; Tue, 08 Jun 2021 17:56:26 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfxi-0003Qn-7y for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:56:26 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqfxi-0005m7-7G for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:56:26 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqfxi-0002v2-6U for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 17:56:26 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=4Cp0fjxBO+wDqvB2DVOMQ90KYKl1u5Jl2qJ5PjeoDxk=; b=p9QZ+dIDHE6yyajEAe8kDfhO9j vVhIMEsTYhyAOqFV0eQl9MYzycNjWdv8KjgAk3kVOra/074B4ZhCZDLTW0BY6clXLLw4IrkMUFxvD fS8o7xzxVzrHVuJf/0BwEapa7PLwS/FRGILoy0/xSaLyCBUwlpzrdp4QGk+7zBtQZcq8=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.14] x86/spec-ctrl: Mitigate TAA after S3 resume Message-Id: Date: Tue, 08 Jun 2021 17:56:26 +0000 commit 0ff7f9c5aa02cd2469a8fc03f1ed262f18933721 Author: Andrew Cooper AuthorDate: Thu May 20 01:21:39 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 18:36:03 2021 +0100 x86/spec-ctrl: Mitigate TAA after S3 resume The user chosen setting for MSR_TSX_CTRL needs restoring after S3. All APs get the correct setting via start_secondary(), but the BSP was missed out. This is XSA-377 / CVE-2021-28690. Fixes: 8c4330818f6 ("x86/spec-ctrl: Mitigate the TSX Asynchronous Abort sidechannel") Signed-off-by: Andrew Cooper Reviewed-by: Jan Beulich (cherry picked from commit 8cf276cb2e0b99b96333865873f56b0b31555ff1) --- xen/arch/x86/acpi/power.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/xen/arch/x86/acpi/power.c b/xen/arch/x86/acpi/power.c index 7f162a4df9..604cb7e222 100644 --- a/xen/arch/x86/acpi/power.c +++ b/xen/arch/x86/acpi/power.c @@ -286,6 +286,8 @@ static int enter_state(u32 state) microcode_update_one(); + tsx_init(); /* Needs microcode. May change HLE/RTM feature bits. */ + if ( !recheck_cpu_features(0) ) panic("Missing previously available feature(s)\n"); -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.14 From xen-changelog-bounces@lists.xenproject.org Tue Jun 08 18:11:09 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 08 Jun 2021 18:11:09 +0000 Received: from list by lists.xenproject.org with outflank-mailman.138883.256902 (Exim 4.92) (envelope-from ) id 1lqgBu-0005nO-B7; Tue, 08 Jun 2021 18:11:06 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 138883.256902; Tue, 08 Jun 2021 18:11:06 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqgBu-0005nG-85; Tue, 08 Jun 2021 18:11:06 +0000 Received: by outflank-mailman (input) for mailman id 138883; Tue, 08 Jun 2021 18:11:05 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqgBt-0005nA-Ln for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 18:11:05 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqgBt-00068c-L0 for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 18:11:05 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqgBt-0004RR-K2 for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 18:11:05 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=uBtoO11ztDGTihjh29tjNqgUiHIQ2DmplQHDLk+QAyw=; b=h9LPmL/hYO+wad2M02jngCsfAh n8SyPweG8TQ27R5QFRZOT3k4NLPddiyT328FU+7Ji0B5m1ZSG+WYC5bsbf3otugBJ+bJP9Z7rXEMN dW9pFQf8kQvzwpP8N31Kd3DcDhYt9f3EGGZTP/xmfp3Wal6PDK7BtSzfNLBDo0nLCsOI=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.13] xen/arm: Create dom0less domUs earlier Message-Id: Date: Tue, 08 Jun 2021 18:11:05 +0000 commit 95197d46f5e291b0bf12cf617e602bbc67fd633f Author: Julien Grall AuthorDate: Mon May 17 17:47:13 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 18:53:03 2021 +0100 xen/arm: Create dom0less domUs earlier In a follow-up patch we will need to unallocate the boot modules before heap_init_late() is called. The modules will contain the domUs kernel and initramfs. Therefore Xen will need to create extra domUs (used by dom0less) before heap_init_late(). This has two consequences on dom0less: 1) Domains will not be unpaused as soon as they are created but once all have been created. However, Xen doesn't guarantee an order to unpause, so this is not something one could rely on. 2) The memory allocated for a domU will not be scrubbed anymore when an admin select bootscrub=on. This is not something we advertised, but if this is a concern we can introduce either force scrub for all domUs or a per-domain flag in the DT. The behavior for bootscrub=off and bootscrub=idle (default) has not changed. This is part of XSA-372 / CVE-2021-28693. Signed-off-by: Julien Grall Reviewed-by: Jan Beulich Reviewed-by: Stefano Stabellini Tested-by: Stefano Stabellini (cherry picked from commit 371347c5b64da699d9f5a0edda5dc496fd2b7a5c) --- xen/arch/arm/domain_build.c | 2 -- xen/arch/arm/setup.c | 9 +++++---- 2 files changed, 5 insertions(+), 6 deletions(-) diff --git a/xen/arch/arm/domain_build.c b/xen/arch/arm/domain_build.c index 8c5ff11cdc..2b37ea019d 100644 --- a/xen/arch/arm/domain_build.c +++ b/xen/arch/arm/domain_build.c @@ -2509,8 +2509,6 @@ void __init create_domUs(void) if ( construct_domU(d, node) != 0 ) panic("Could not set up domain %s\n", dt_node_name(node)); - - domain_unpause_by_systemcontroller(d); } } diff --git a/xen/arch/arm/setup.c b/xen/arch/arm/setup.c index 3c899cd4a0..9a5aaeed3b 100644 --- a/xen/arch/arm/setup.c +++ b/xen/arch/arm/setup.c @@ -784,7 +784,7 @@ void __init start_xen(unsigned long boot_phys_offset, int cpus, i; const char *cmdline; struct bootmodule *xen_bootmodule; - struct domain *dom0; + struct domain *dom0, *d; struct xen_domctl_createdomain dom0_cfg = { .flags = XEN_DOMCTL_CDF_hvm | XEN_DOMCTL_CDF_hap, .max_evtchn_port = -1, @@ -972,6 +972,8 @@ void __init start_xen(unsigned long boot_phys_offset, if ( construct_dom0(dom0) != 0) panic("Could not set up DOM0 guest OS\n"); + create_domUs(); + heap_init_late(); init_trace_bufs(); @@ -985,9 +987,8 @@ void __init start_xen(unsigned long boot_phys_offset, system_state = SYS_STATE_active; - create_domUs(); - - domain_unpause_by_systemcontroller(dom0); + for_each_domain( d ) + domain_unpause_by_systemcontroller(d); /* Switch on to the dynamically allocated stack for the idle vcpu * since the static one we're running on is about to be freed. */ -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.13 From xen-changelog-bounces@lists.xenproject.org Tue Jun 08 18:11:16 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 08 Jun 2021 18:11:16 +0000 Received: from list by lists.xenproject.org with outflank-mailman.138884.256906 (Exim 4.92) (envelope-from ) id 1lqgC4-0005pS-Cf; Tue, 08 Jun 2021 18:11:16 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 138884.256906; Tue, 08 Jun 2021 18:11:16 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqgC4-0005pK-9e; Tue, 08 Jun 2021 18:11:16 +0000 Received: by outflank-mailman (input) for mailman id 138884; Tue, 08 Jun 2021 18:11:15 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqgC3-0005pE-Ot for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 18:11:15 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqgC3-00068p-OA for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 18:11:15 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqgC3-0004SS-ND for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 18:11:15 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=etMsUJhCtqtP1ojcNqzqZJDih9ZU8flpAp7pWTR6aXE=; b=E0l0HdAoRD+tSuQYpPou3Mnb7F Lt6GhGvXtgH3J4tvBopuU0kRnr8ssZsVME1mVWb7gZLHvjKmAzuqDADq2FsPUJ3omM3f64f7F+dgk zUrTzw+I03eNCs6AxxkcOTGdNddC2efbZn2xB6/q8RLpe4/nNH0cD87L07HwXPqrRT5M=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.13] xen/arm: Boot modules should always be scrubbed if bootscrub={on, idle} Message-Id: Date: Tue, 08 Jun 2021 18:11:15 +0000 commit def4352113d21acf4d49579c4ed3b86c6f2c922e Author: Julien Grall AuthorDate: Sat Apr 17 17:38:28 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 18:53:04 2021 +0100 xen/arm: Boot modules should always be scrubbed if bootscrub={on, idle} The function to initialize the pages (see init_heap_pages()) will request scrub when the admin request idle bootscrub (default) and state == SYS_STATE_active. When bootscrub=on, Xen will scrub any free pages in heap_init_late(). Currently, the boot modules (e.g. kernels, initramfs) will be discarded/ freed after heap_init_late() is called and system_state switched to SYS_STATE_active. This means the pages associated with the boot modules will not get scrubbed before getting re-purposed. If the memory is assigned to an untrusted domU, it may be able to retrieve secrets from the modules. This is part of XSA-372 / CVE-2021-28693. Fixes: 1774e9b1df27 ("xen/arm: introduce create_domUs") Signed-off-by: Julien Grall Reviewed-by: Jan Beulich Reviewed-by: Stefano Stabellini Tested-by: Stefano Stabellini (cherry picked from commit fd5dc41ceaed9cfcfa011cdfd50f264c89277a90) --- xen/arch/arm/setup.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-) diff --git a/xen/arch/arm/setup.c b/xen/arch/arm/setup.c index 9a5aaeed3b..de21fbce0e 100644 --- a/xen/arch/arm/setup.c +++ b/xen/arch/arm/setup.c @@ -74,7 +74,6 @@ static __used void init_done(void) /* Must be done past setting system_state. */ unregister_init_virtual_region(); - discard_initial_modules(); free_init_memory(); startup_cpu_idle_loop(); } @@ -974,6 +973,12 @@ void __init start_xen(unsigned long boot_phys_offset, create_domUs(); + /* + * This needs to be called **before** heap_init_late() so modules + * will be scrubbed (unless suppressed). + */ + discard_initial_modules(); + heap_init_late(); init_trace_bufs(); -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.13 From xen-changelog-bounces@lists.xenproject.org Tue Jun 08 18:11:26 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 08 Jun 2021 18:11:26 +0000 Received: from list by lists.xenproject.org with outflank-mailman.138885.256910 (Exim 4.92) (envelope-from ) id 1lqgCE-0005tE-FX; Tue, 08 Jun 2021 18:11:26 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 138885.256910; Tue, 08 Jun 2021 18:11:26 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqgCE-0005t4-Cf; Tue, 08 Jun 2021 18:11:26 +0000 Received: by outflank-mailman (input) for mailman id 138885; Tue, 08 Jun 2021 18:11:25 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqgCD-0005sw-S6 for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 18:11:25 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqgCD-000699-RH for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 18:11:25 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqgCD-0004TM-QQ for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 18:11:25 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=EZoCoaPk25ZSZrdX1qO79p0Aqg8suPp/zCq1XwsiDZg=; b=JIm+D8yDxYf6drgNtIBCmIMM0a DisgBiKcX5WCuRocb1igpHFVGSWgTpYlLzyuLU9dGv/raMqDph1cPZPJeg8pZtxmnPjhf0cAKtQSm lEdugfzqN87zykRaF/ZdoI3PGXI9slXwf0H1jykAbP0KQgFKV45yuke6gEf5noFaTcHU=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.13] VT-d: size qinval queue dynamically Message-Id: Date: Tue, 08 Jun 2021 18:11:25 +0000 commit 9e3c8b1448d50a3154882a686719f8acc0dad6e7 Author: Jan Beulich AuthorDate: Tue Jun 8 18:55:50 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 18:55:58 2021 +0100 VT-d: size qinval queue dynamically With the present synchronous model, we need two slots for every operation (the operation itself and a wait descriptor). There can be one such pair of requests pending per CPU. To ensure that under all normal circumstances a slot is always available when one is requested, size the queue ring according to the number of present CPUs. This is part of XSA-373 / CVE-2021-28692. Signed-off-by: Jan Beulich Reviewed-by: Paul Durrant (cherry picked from commit cbfa62bb140c8f10a0ae96db3ce06e22b3b9d302) --- xen/drivers/passthrough/vtd/iommu.h | 10 +------- xen/drivers/passthrough/vtd/qinval.c | 46 +++++++++++++++++++++++++++--------- 2 files changed, 36 insertions(+), 20 deletions(-) diff --git a/xen/drivers/passthrough/vtd/iommu.h b/xen/drivers/passthrough/vtd/iommu.h index 57737b1a4a..e002dc6f01 100644 --- a/xen/drivers/passthrough/vtd/iommu.h +++ b/xen/drivers/passthrough/vtd/iommu.h @@ -450,17 +450,9 @@ struct qinval_entry { }q; }; -/* Order of queue invalidation pages(max is 8) */ -#define QINVAL_PAGE_ORDER 2 - -#define QINVAL_ARCH_PAGE_ORDER (QINVAL_PAGE_ORDER + PAGE_SHIFT_4K - PAGE_SHIFT) -#define QINVAL_ARCH_PAGE_NR ( QINVAL_ARCH_PAGE_ORDER < 0 ? \ - 1 : \ - 1 << QINVAL_ARCH_PAGE_ORDER ) - /* Each entry is 16 bytes, so 2^8 entries per page */ #define QINVAL_ENTRY_ORDER ( PAGE_SHIFT - 4 ) -#define QINVAL_ENTRY_NR (1 << (QINVAL_PAGE_ORDER + 8)) +#define QINVAL_MAX_ENTRY_NR (1u << (7 + QINVAL_ENTRY_ORDER)) /* Status data flag */ #define QINVAL_STAT_INIT 0 diff --git a/xen/drivers/passthrough/vtd/qinval.c b/xen/drivers/passthrough/vtd/qinval.c index dfcbb66c8a..86a941ce27 100644 --- a/xen/drivers/passthrough/vtd/qinval.c +++ b/xen/drivers/passthrough/vtd/qinval.c @@ -31,6 +31,9 @@ #define VTD_QI_TIMEOUT 1 +static unsigned int __read_mostly qi_pg_order; +static unsigned int __read_mostly qi_entry_nr; + static int __must_check invalidate_sync(struct vtd_iommu *iommu); static void print_qi_regs(struct vtd_iommu *iommu) @@ -55,7 +58,7 @@ static unsigned int qinval_next_index(struct vtd_iommu *iommu) tail >>= QINVAL_INDEX_SHIFT; /* (tail+1 == head) indicates a full queue, wait for HW */ - while ( ( tail + 1 ) % QINVAL_ENTRY_NR == + while ( ((tail + 1) & (qi_entry_nr - 1)) == ( dmar_readq(iommu->reg, DMAR_IQH_REG) >> QINVAL_INDEX_SHIFT ) ) cpu_relax(); @@ -68,7 +71,7 @@ static void qinval_update_qtail(struct vtd_iommu *iommu, unsigned int index) /* Need hold register lock when update tail */ ASSERT( spin_is_locked(&iommu->register_lock) ); - val = (index + 1) % QINVAL_ENTRY_NR; + val = (index + 1) & (qi_entry_nr - 1); dmar_writeq(iommu->reg, DMAR_IQT_REG, (val << QINVAL_INDEX_SHIFT)); } @@ -403,8 +406,28 @@ int enable_qinval(struct vtd_iommu *iommu) if ( iommu->qinval_maddr == 0 ) { - iommu->qinval_maddr = alloc_pgtable_maddr(QINVAL_ARCH_PAGE_NR, - iommu->node); + if ( !qi_entry_nr ) + { + /* + * With the present synchronous model, we need two slots for every + * operation (the operation itself and a wait descriptor). There + * can be one such pair of requests pending per CPU. One extra + * entry is needed as the ring is considered full when there's + * only one entry left. + */ + BUILD_BUG_ON(CONFIG_NR_CPUS * 2 >= QINVAL_MAX_ENTRY_NR); + qi_pg_order = get_order_from_bytes((num_present_cpus() * 2 + 1) << + (PAGE_SHIFT - + QINVAL_ENTRY_ORDER)); + qi_entry_nr = 1u << (qi_pg_order + QINVAL_ENTRY_ORDER); + + dprintk(XENLOG_INFO VTDPREFIX, + "QI: using %u-entry ring(s)\n", qi_entry_nr); + } + + iommu->qinval_maddr = + alloc_pgtable_maddr(qi_entry_nr >> QINVAL_ENTRY_ORDER, + iommu->node); if ( iommu->qinval_maddr == 0 ) { dprintk(XENLOG_WARNING VTDPREFIX, @@ -418,15 +441,16 @@ int enable_qinval(struct vtd_iommu *iommu) spin_lock_irqsave(&iommu->register_lock, flags); - /* Setup Invalidation Queue Address(IQA) register with the - * address of the page we just allocated. QS field at - * bits[2:0] to indicate size of queue is one 4KB page. - * That's 256 entries. Queued Head (IQH) and Queue Tail (IQT) - * registers are automatically reset to 0 with write - * to IQA register. + /* + * Setup Invalidation Queue Address (IQA) register with the address of the + * pages we just allocated. The QS field at bits[2:0] indicates the size + * (page order) of the queue. + * + * Queued Head (IQH) and Queue Tail (IQT) registers are automatically + * reset to 0 with write to IQA register. */ dmar_writeq(iommu->reg, DMAR_IQA_REG, - iommu->qinval_maddr | QINVAL_PAGE_ORDER); + iommu->qinval_maddr | qi_pg_order); dmar_writeq(iommu->reg, DMAR_IQT_REG, 0); -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.13 From xen-changelog-bounces@lists.xenproject.org Tue Jun 08 18:11:37 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 08 Jun 2021 18:11:37 +0000 Received: from list by lists.xenproject.org with outflank-mailman.138886.256914 (Exim 4.92) (envelope-from ) id 1lqgCP-0005wN-Hp; Tue, 08 Jun 2021 18:11:37 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 138886.256914; Tue, 08 Jun 2021 18:11:37 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqgCP-0005wG-EG; Tue, 08 Jun 2021 18:11:37 +0000 Received: by outflank-mailman (input) for mailman id 138886; Tue, 08 Jun 2021 18:11:36 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqgCO-0005w2-3d for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 18:11:36 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqgCN-00069T-Vc for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 18:11:35 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqgCN-0004UF-US for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 18:11:35 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=B7igF+Nj5qyH8N1rTmYkBqQF2b1kVjknG4OG8tFDtEA=; b=SZi2bqKRz99Hr24bnXrjJGuU8A 7NDOAioQCRkimbRi/8OHVNAkgHCzMEnVTUzHffJEOIlu6T5jEp81in1ApDGyd7Aef6/BOsob6Dkfl n2TJvmCqWMEBBDuYO/xZoHTM0xfDSbgD6ZeoVdCAC7jiRDq1jGXVmt3LJjw2ALwPWZdQ=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.13] AMD/IOMMU: size command buffer dynamically Message-Id: Date: Tue, 08 Jun 2021 18:11:35 +0000 commit 83c0f6bbb25769b9ac711ea6b218d5ee858a360e Author: Jan Beulich AuthorDate: Tue Jun 8 18:56:24 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 18:56:26 2021 +0100 AMD/IOMMU: size command buffer dynamically With the present synchronous model, we need two slots for every operation (the operation itself and a wait command). There can be one such pair of commands pending per CPU. To ensure that under all normal circumstances a slot is always available when one is requested, size the command ring according to the number of present CPUs. This is part of XSA-373 / CVE-2021-28692. Signed-off-by: Jan Beulich Reviewed-by: Paul Durrant (cherry picked from commit df242851ddc93ac0b0a3a20ecab34acc29e3b36b) --- xen/drivers/passthrough/amd/iommu_cmd.c | 4 ++-- xen/drivers/passthrough/amd/iommu_init.c | 28 +++++++++++++++++++++++++--- xen/include/asm-x86/hvm/svm/amd-iommu-defs.h | 7 ++----- 3 files changed, 29 insertions(+), 10 deletions(-) diff --git a/xen/drivers/passthrough/amd/iommu_cmd.c b/xen/drivers/passthrough/amd/iommu_cmd.c index af3a1fb865..c9c7d7b290 100644 --- a/xen/drivers/passthrough/amd/iommu_cmd.c +++ b/xen/drivers/passthrough/amd/iommu_cmd.c @@ -35,8 +35,8 @@ static int queue_iommu_command(struct amd_iommu *iommu, u32 cmd[]) if ( head != tail ) { memcpy(iommu->cmd_buffer.buffer + - (iommu->cmd_buffer.tail * IOMMU_CMD_BUFFER_ENTRY_SIZE), - cmd, IOMMU_CMD_BUFFER_ENTRY_SIZE); + (iommu->cmd_buffer.tail * sizeof(cmd_entry_t)), + cmd, sizeof(cmd_entry_t)); iommu->cmd_buffer.tail = tail; return 1; diff --git a/xen/drivers/passthrough/amd/iommu_init.c b/xen/drivers/passthrough/amd/iommu_init.c index 566e6defa1..2209eba066 100644 --- a/xen/drivers/passthrough/amd/iommu_init.c +++ b/xen/drivers/passthrough/amd/iommu_init.c @@ -125,7 +125,7 @@ static void register_iommu_cmd_buffer_in_mmio_space(struct amd_iommu *iommu) writel(entry, iommu->mmio_base + IOMMU_CMD_BUFFER_BASE_LOW_OFFSET); power_of2_entries = get_order_from_bytes(iommu->cmd_buffer.alloc_size) + - IOMMU_CMD_BUFFER_POWER_OF2_ENTRIES_PER_PAGE; + PAGE_SHIFT - IOMMU_CMD_BUFFER_ENTRY_ORDER; entry = 0; iommu_set_addr_hi_to_reg(&entry, addr_hi); @@ -1050,9 +1050,31 @@ static void *__init allocate_ring_buffer(struct ring_buffer *ring_buf, static void * __init allocate_cmd_buffer(struct amd_iommu *iommu) { /* allocate 'command buffer' in power of 2 increments of 4K */ + static unsigned int __read_mostly nr_ents; + + if ( !nr_ents ) + { + unsigned int order; + + /* + * With the present synchronous model, we need two slots for every + * operation (the operation itself and a wait command). There can be + * one such pair of requests pending per CPU. One extra entry is + * needed as the ring is considered full when there's only one entry + * left. + */ + BUILD_BUG_ON(CONFIG_NR_CPUS * 2 >= IOMMU_CMD_BUFFER_MAX_ENTRIES); + order = get_order_from_bytes((num_present_cpus() * 2 + 1) << + IOMMU_CMD_BUFFER_ENTRY_ORDER); + nr_ents = 1u << (order + PAGE_SHIFT - IOMMU_CMD_BUFFER_ENTRY_ORDER); + + AMD_IOMMU_DEBUG("using %u-entry cmd ring(s)\n", nr_ents); + } + + BUILD_BUG_ON(sizeof(cmd_entry_t) != (1u << IOMMU_CMD_BUFFER_ENTRY_ORDER)); + return allocate_ring_buffer(&iommu->cmd_buffer, sizeof(cmd_entry_t), - IOMMU_CMD_BUFFER_DEFAULT_ENTRIES, - "Command Buffer", false); + nr_ents, "Command Buffer", false); } static void * __init allocate_event_log(struct amd_iommu *iommu) diff --git a/xen/include/asm-x86/hvm/svm/amd-iommu-defs.h b/xen/include/asm-x86/hvm/svm/amd-iommu-defs.h index 6049507a30..a54d6e9fc6 100644 --- a/xen/include/asm-x86/hvm/svm/amd-iommu-defs.h +++ b/xen/include/asm-x86/hvm/svm/amd-iommu-defs.h @@ -20,9 +20,6 @@ #ifndef _ASM_X86_64_AMD_IOMMU_DEFS_H #define _ASM_X86_64_AMD_IOMMU_DEFS_H -/* IOMMU Command Buffer entries: in power of 2 increments, minimum of 256 */ -#define IOMMU_CMD_BUFFER_DEFAULT_ENTRIES 512 - /* IOMMU Event Log entries: in power of 2 increments, minimum of 256 */ #define IOMMU_EVENT_LOG_DEFAULT_ENTRIES 512 @@ -168,8 +165,8 @@ struct amd_iommu_dte { #define IOMMU_CMD_BUFFER_LENGTH_MASK 0x0F000000 #define IOMMU_CMD_BUFFER_LENGTH_SHIFT 24 -#define IOMMU_CMD_BUFFER_ENTRY_SIZE 16 -#define IOMMU_CMD_BUFFER_POWER_OF2_ENTRIES_PER_PAGE 8 +#define IOMMU_CMD_BUFFER_ENTRY_ORDER 4 +#define IOMMU_CMD_BUFFER_MAX_ENTRIES (1u << 15) #define IOMMU_CMD_OPCODE_MASK 0xF0000000 #define IOMMU_CMD_OPCODE_SHIFT 28 -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.13 From xen-changelog-bounces@lists.xenproject.org Tue Jun 08 18:11:47 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 08 Jun 2021 18:11:47 +0000 Received: from list by lists.xenproject.org with outflank-mailman.138887.256918 (Exim 4.92) (envelope-from ) id 1lqgCZ-0005za-Ir; Tue, 08 Jun 2021 18:11:47 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 138887.256918; Tue, 08 Jun 2021 18:11:47 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqgCZ-0005zS-Fw; Tue, 08 Jun 2021 18:11:47 +0000 Received: by outflank-mailman (input) for mailman id 138887; Tue, 08 Jun 2021 18:11:46 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqgCY-0005zF-3c for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 18:11:46 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqgCY-0006A2-2m for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 18:11:46 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqgCY-0004VN-1o for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 18:11:46 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=CtGGmWtMZNbcsLZpzrDKUA+FGeU7q8c7Z9VTtfx60zU=; b=iYxSGJWk6GfrSrrrBxHbk2Q+cJ GiMjdJAcNJLLbyrWM66nLnqX51DXZDUQXWDBcvap8HANI5yybHg6x4apWaJKM6w92xEVRWTv5k9bF I/2iZJu1K5r3suV7Pr565hzyDLJHpcdtFo4i3ainlEb9MxSW4fpwj9jKIqDdajz05x3c=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.13] VT-d: eliminate flush related timeouts Message-Id: Date: Tue, 08 Jun 2021 18:11:46 +0000 commit d3d8a298f7fd27371414fc771786c2a0f1564152 Author: Jan Beulich AuthorDate: Tue Jun 8 18:56:39 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 18:56:42 2021 +0100 VT-d: eliminate flush related timeouts Leaving an in-progress operation pending when it appears to take too long is problematic: If e.g. a QI command completed later, the write to the "poll slot" may instead be understood to signal a subsequently started command's completion. Also our accounting of the timeout period was actually wrong: We included the time it took for the command to actually make it to the front of the queue, which could be heavily affected by guests other than the one for which the flush is being performed. Do away with all timeout detection on all flush related code paths. Log excessively long processing times (with a progressive threshold) to have some indication of problems in this area. Additionally log (once) if qinval_next_index() didn't immediately find an available slot. Together with the earlier change sizing the queue(s) dynamically, we should now have a guarantee that with our fully synchronous model any demand for slots can actually be satisfied. This is part of XSA-373 / CVE-2021-28692. Signed-off-by: Jan Beulich Reviewed-by: Paul Durrant (cherry picked from commit 7c893abc4ee7e9af62297ba6fd5f27c89d0f33f5) --- xen/drivers/passthrough/vtd/dmar.h | 28 ++++++++++++++++++++++++++++ xen/drivers/passthrough/vtd/iommu.c | 12 ++++++------ xen/drivers/passthrough/vtd/qinval.c | 33 ++++++++++++++++++++++----------- 3 files changed, 56 insertions(+), 17 deletions(-) diff --git a/xen/drivers/passthrough/vtd/dmar.h b/xen/drivers/passthrough/vtd/dmar.h index 1a9c965e59..a44820a5e7 100644 --- a/xen/drivers/passthrough/vtd/dmar.h +++ b/xen/drivers/passthrough/vtd/dmar.h @@ -127,6 +127,34 @@ do { \ } \ } while (0) +#define IOMMU_FLUSH_WAIT(what, iommu, offset, op, cond, sts) \ +do { \ + static unsigned int __read_mostly threshold = 1; \ + s_time_t start = NOW(); \ + s_time_t timeout = start + DMAR_OPERATION_TIMEOUT * threshold; \ + \ + for ( ; ; ) \ + { \ + sts = op(iommu->reg, offset); \ + if ( cond ) \ + break; \ + if ( timeout && NOW() > timeout ) \ + { \ + threshold |= threshold << 1; \ + printk(XENLOG_WARNING VTDPREFIX \ + " IOMMU#%u: %s flush taking too long\n", \ + iommu->index, what); \ + timeout = 0; \ + } \ + cpu_relax(); \ + } \ + \ + if ( !timeout ) \ + printk(XENLOG_WARNING VTDPREFIX \ + " IOMMU#%u: %s flush took %lums\n", \ + iommu->index, what, (NOW() - start) / 10000000); \ +} while ( false ) + int vtd_hw_check(void); void disable_pmr(struct vtd_iommu *iommu); int is_igd_drhd(struct acpi_drhd_unit *drhd); diff --git a/xen/drivers/passthrough/vtd/iommu.c b/xen/drivers/passthrough/vtd/iommu.c index 34ee5d22f1..dfd867d40b 100644 --- a/xen/drivers/passthrough/vtd/iommu.c +++ b/xen/drivers/passthrough/vtd/iommu.c @@ -320,8 +320,8 @@ static void iommu_flush_write_buffer(struct vtd_iommu *iommu) dmar_writel(iommu->reg, DMAR_GCMD_REG, val | DMA_GCMD_WBF); /* Make sure hardware complete it */ - IOMMU_WAIT_OP(iommu, DMAR_GSTS_REG, dmar_readl, - !(val & DMA_GSTS_WBFS), val); + IOMMU_FLUSH_WAIT("write buffer", iommu, DMAR_GSTS_REG, dmar_readl, + !(val & DMA_GSTS_WBFS), val); spin_unlock_irqrestore(&iommu->register_lock, flags); } @@ -370,8 +370,8 @@ int vtd_flush_context_reg(struct vtd_iommu *iommu, uint16_t did, dmar_writeq(iommu->reg, DMAR_CCMD_REG, val); /* Make sure hardware complete it */ - IOMMU_WAIT_OP(iommu, DMAR_CCMD_REG, dmar_readq, - !(val & DMA_CCMD_ICC), val); + IOMMU_FLUSH_WAIT("context", iommu, DMAR_CCMD_REG, dmar_readq, + !(val & DMA_CCMD_ICC), val); spin_unlock_irqrestore(&iommu->register_lock, flags); /* flush context entry will implicitly flush write buffer */ @@ -448,8 +448,8 @@ int vtd_flush_iotlb_reg(struct vtd_iommu *iommu, uint16_t did, uint64_t addr, dmar_writeq(iommu->reg, tlb_offset + 8, val); /* Make sure hardware complete it */ - IOMMU_WAIT_OP(iommu, (tlb_offset + 8), dmar_readq, - !(val & DMA_TLB_IVT), val); + IOMMU_FLUSH_WAIT("iotlb", iommu, (tlb_offset + 8), dmar_readq, + !(val & DMA_TLB_IVT), val); spin_unlock_irqrestore(&iommu->register_lock, flags); /* check IOTLB invalidation granularity */ diff --git a/xen/drivers/passthrough/vtd/qinval.c b/xen/drivers/passthrough/vtd/qinval.c index 86a941ce27..1056ab1feb 100644 --- a/xen/drivers/passthrough/vtd/qinval.c +++ b/xen/drivers/passthrough/vtd/qinval.c @@ -29,8 +29,6 @@ #include "extern.h" #include "../ats.h" -#define VTD_QI_TIMEOUT 1 - static unsigned int __read_mostly qi_pg_order; static unsigned int __read_mostly qi_entry_nr; @@ -60,7 +58,11 @@ static unsigned int qinval_next_index(struct vtd_iommu *iommu) /* (tail+1 == head) indicates a full queue, wait for HW */ while ( ((tail + 1) & (qi_entry_nr - 1)) == ( dmar_readq(iommu->reg, DMAR_IQH_REG) >> QINVAL_INDEX_SHIFT ) ) + { + printk_once(XENLOG_ERR VTDPREFIX " IOMMU#%u: no QI slot available\n", + iommu->index); cpu_relax(); + } return tail; } @@ -180,23 +182,32 @@ static int __must_check queue_invalidate_wait(struct vtd_iommu *iommu, /* Now we don't support interrupt method */ if ( sw ) { - s_time_t timeout; - - /* In case all wait descriptor writes to same addr with same data */ - timeout = NOW() + MILLISECS(flush_dev_iotlb ? - iommu_dev_iotlb_timeout : VTD_QI_TIMEOUT); + static unsigned int __read_mostly threshold = 1; + s_time_t start = NOW(); + s_time_t timeout = start + (flush_dev_iotlb + ? iommu_dev_iotlb_timeout + : 100) * MILLISECS(threshold); while ( ACCESS_ONCE(*this_poll_slot) != QINVAL_STAT_DONE ) { - if ( NOW() > timeout ) + if ( timeout && NOW() > timeout ) { - print_qi_regs(iommu); + threshold |= threshold << 1; printk(XENLOG_WARNING VTDPREFIX - " Queue invalidate wait descriptor timed out\n"); - return -ETIMEDOUT; + " IOMMU#%u: QI%s wait descriptor taking too long\n", + iommu->index, flush_dev_iotlb ? " dev" : ""); + print_qi_regs(iommu); + timeout = 0; } cpu_relax(); } + + if ( !timeout ) + printk(XENLOG_WARNING VTDPREFIX + " IOMMU#%u: QI%s wait descriptor took %lums\n", + iommu->index, flush_dev_iotlb ? " dev" : "", + (NOW() - start) / 10000000); + return 0; } -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.13 From xen-changelog-bounces@lists.xenproject.org Tue Jun 08 18:11:57 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 08 Jun 2021 18:11:57 +0000 Received: from list by lists.xenproject.org with outflank-mailman.138888.256922 (Exim 4.92) (envelope-from ) id 1lqgCj-00062N-Kh; Tue, 08 Jun 2021 18:11:57 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 138888.256922; Tue, 08 Jun 2021 18:11:57 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqgCj-00062F-Ha; Tue, 08 Jun 2021 18:11:57 +0000 Received: by outflank-mailman (input) for mailman id 138888; Tue, 08 Jun 2021 18:11:56 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqgCi-000625-6p for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 18:11:56 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqgCi-0006AF-64 for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 18:11:56 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqgCi-0004WF-58 for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 18:11:56 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=30ieUQSZo1y8HGfMh7ld+vjkP4Fg8ANE4bN2uRu0bG8=; b=4SdpfoNJaczVdyfIAwM4uAX+/8 kb11P4hsaEbyXF3k9gVKskNX/Cp5Mm6hzeJSQi1fLisexsUNOFluLKVTRwrzTMEzBMjAbEx0H6u1M f9FCbWNABWA3EssNKmmrBL+AMblIms/DTuioIPFuXfvGEm1usfuGTWCZU/kH/NVDlZo8=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.13] AMD/IOMMU: wait for command slot to be available Message-Id: Date: Tue, 08 Jun 2021 18:11:56 +0000 commit 0a3eb9bbc2188f36692a570c85816c6d3d0da697 Author: Jan Beulich AuthorDate: Tue Jun 8 18:58:05 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 18:58:12 2021 +0100 AMD/IOMMU: wait for command slot to be available No caller cared about send_iommu_command() indicating unavailability of a slot. Hence if a sufficient number prior commands timed out, we did blindly assume that the requested command was submitted to the IOMMU when really it wasn't. This could mean both a hanging system (waiting for a command to complete that was never seen by the IOMMU) or blindly propagating success back to callers, making them believe they're fine to e.g. free previously unmapped pages. Fold the three involved functions into one, add spin waiting for an available slot along the lines of VT-d's qinval_next_index(), and as a consequence drop all error indicator return types/values. This is part of XSA-373 / CVE-2021-28692. Signed-off-by: Jan Beulich Reviewed-by: Paul Durrant (cherry picked from commit c4e44343579e2c3304d676404d81b2b9bd893c5b) --- xen/drivers/passthrough/amd/iommu_cmd.c | 42 ++++++++++++--------------------- 1 file changed, 15 insertions(+), 27 deletions(-) diff --git a/xen/drivers/passthrough/amd/iommu_cmd.c b/xen/drivers/passthrough/amd/iommu_cmd.c index c9c7d7b290..3e48fc3733 100644 --- a/xen/drivers/passthrough/amd/iommu_cmd.c +++ b/xen/drivers/passthrough/amd/iommu_cmd.c @@ -22,48 +22,36 @@ #include #include "../ats.h" -static int queue_iommu_command(struct amd_iommu *iommu, u32 cmd[]) +static void send_iommu_command(struct amd_iommu *iommu, + const uint32_t cmd[4]) { - uint32_t tail, head; + uint32_t tail; tail = iommu->cmd_buffer.tail; if ( ++tail == iommu->cmd_buffer.entries ) tail = 0; - head = iommu_get_rb_pointer(readl(iommu->mmio_base + - IOMMU_CMD_BUFFER_HEAD_OFFSET)); - if ( head != tail ) + while ( tail == iommu_get_rb_pointer(readl(iommu->mmio_base + + IOMMU_CMD_BUFFER_HEAD_OFFSET)) ) { - memcpy(iommu->cmd_buffer.buffer + - (iommu->cmd_buffer.tail * sizeof(cmd_entry_t)), - cmd, sizeof(cmd_entry_t)); - - iommu->cmd_buffer.tail = tail; - return 1; + printk_once(XENLOG_ERR + "AMD IOMMU %04x:%02x:%02x.%u: no cmd slot available\n", + iommu->seg, PCI_BUS(iommu->bdf), + PCI_SLOT(iommu->bdf), PCI_FUNC(iommu->bdf)); + cpu_relax(); } - return 0; -} + memcpy(iommu->cmd_buffer.buffer + + (iommu->cmd_buffer.tail * sizeof(cmd_entry_t)), + cmd, sizeof(cmd_entry_t)); -static void commit_iommu_command_buffer(struct amd_iommu *iommu) -{ - u32 tail = 0; + iommu->cmd_buffer.tail = tail; + tail = 0; iommu_set_rb_pointer(&tail, iommu->cmd_buffer.tail); writel(tail, iommu->mmio_base+IOMMU_CMD_BUFFER_TAIL_OFFSET); } -int send_iommu_command(struct amd_iommu *iommu, u32 cmd[]) -{ - if ( queue_iommu_command(iommu, cmd) ) - { - commit_iommu_command_buffer(iommu); - return 1; - } - - return 0; -} - static void flush_command_buffer(struct amd_iommu *iommu) { u32 cmd[4], status; -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.13 From xen-changelog-bounces@lists.xenproject.org Tue Jun 08 18:12:07 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 08 Jun 2021 18:12:07 +0000 Received: from list by lists.xenproject.org with outflank-mailman.138889.256926 (Exim 4.92) (envelope-from ) id 1lqgCt-000665-NM; Tue, 08 Jun 2021 18:12:07 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 138889.256926; Tue, 08 Jun 2021 18:12:07 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqgCt-00065x-KP; Tue, 08 Jun 2021 18:12:07 +0000 Received: by outflank-mailman (input) for mailman id 138889; Tue, 08 Jun 2021 18:12:06 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqgCs-00065j-9o for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 18:12:06 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqgCs-0006Ad-97 for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 18:12:06 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqgCs-0004XG-8C for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 18:12:06 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=bSGFlmBE3vBO65BmDTEUHf74TYEPiKc1qqNYwpX5U+M=; b=q/ndRaUu9+uYXpaMRgpQeXloQF 4hTSpNKQ1lmant5wEBFMNY+wewJDRaPznZL/sCjI21fheA668dcBm1BG6q46rf/L34uuzLuna83NG mLR8iMVEXhqOlFby+JU4KzjfPb+nkbAuz3F3e4b0jlB6MENPi4MqLQOK5nym/Gm0vvzc=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.13] AMD/IOMMU: drop command completion timeout Message-Id: Date: Tue, 08 Jun 2021 18:12:06 +0000 commit f7993297b3469f4396af8ed14ba376982cabd64b Author: Jan Beulich AuthorDate: Tue Jun 8 18:58:24 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 18:58:26 2021 +0100 AMD/IOMMU: drop command completion timeout First and foremost - such timeouts were not signaled to callers, making them believe they're fine to e.g. free previously unmapped pages. Mirror VT-d's behavior: A fixed number of loop iterations is not a suitable way to detect timeouts in an environment (CPU and bus speeds) independent manner anyway. Furthermore, leaving an in-progress operation pending when it appears to take too long is problematic: If a command completed later, the signaling of its completion may instead be understood to signal a subsequently started command's completion. Log excessively long processing times (with a progressive threshold) to have some indication of problems in this area. Allow callers to specify a non-default timeout bias for this logging, using the same values as VT-d does, which in particular means a (by default) much larger value for device IO TLB invalidation. This is part of XSA-373 / CVE-2021-28692. Signed-off-by: Jan Beulich Reviewed-by: Paul Durrant (cherry picked from commit e4fee66043120c954fc309bbb37813604c1c0eb7) --- xen/drivers/passthrough/amd/iommu_cmd.c | 60 +++++++++++++++++++-------------- 1 file changed, 35 insertions(+), 25 deletions(-) diff --git a/xen/drivers/passthrough/amd/iommu_cmd.c b/xen/drivers/passthrough/amd/iommu_cmd.c index 3e48fc3733..801b27af58 100644 --- a/xen/drivers/passthrough/amd/iommu_cmd.c +++ b/xen/drivers/passthrough/amd/iommu_cmd.c @@ -52,10 +52,12 @@ static void send_iommu_command(struct amd_iommu *iommu, writel(tail, iommu->mmio_base+IOMMU_CMD_BUFFER_TAIL_OFFSET); } -static void flush_command_buffer(struct amd_iommu *iommu) +static void flush_command_buffer(struct amd_iommu *iommu, + unsigned int timeout_base) { - u32 cmd[4], status; - int loop_count, comp_wait; + uint32_t cmd[4]; + s_time_t start, timeout; + static unsigned int __read_mostly threshold = 1; /* RW1C 'ComWaitInt' in status register */ writel(IOMMU_STATUS_COMP_WAIT_INT_MASK, @@ -71,24 +73,31 @@ static void flush_command_buffer(struct amd_iommu *iommu) IOMMU_COMP_WAIT_I_FLAG_SHIFT, &cmd[0]); send_iommu_command(iommu, cmd); - /* Make loop_count long enough for polling completion wait bit */ - loop_count = 1000; - do { - status = readl(iommu->mmio_base + IOMMU_STATUS_MMIO_OFFSET); - comp_wait = get_field_from_reg_u32(status, - IOMMU_STATUS_COMP_WAIT_INT_MASK, - IOMMU_STATUS_COMP_WAIT_INT_SHIFT); - --loop_count; - } while ( !comp_wait && loop_count ); - - if ( comp_wait ) + start = NOW(); + timeout = start + (timeout_base ?: 100) * MILLISECS(threshold); + while ( !(readl(iommu->mmio_base + IOMMU_STATUS_MMIO_OFFSET) & + IOMMU_STATUS_COMP_WAIT_INT_MASK) ) { - /* RW1C 'ComWaitInt' in status register */ - writel(IOMMU_STATUS_COMP_WAIT_INT_MASK, - iommu->mmio_base + IOMMU_STATUS_MMIO_OFFSET); - return; + if ( timeout && NOW() > timeout ) + { + threshold |= threshold << 1; + printk(XENLOG_WARNING + "AMD IOMMU %04x:%02x:%02x.%u: %scompletion wait taking too long\n", + iommu->seg, PCI_BUS(iommu->bdf), + PCI_SLOT(iommu->bdf), PCI_FUNC(iommu->bdf), + timeout_base ? "iotlb " : ""); + timeout = 0; + } + cpu_relax(); } - AMD_IOMMU_DEBUG("Warning: ComWaitInt bit did not assert!\n"); + + if ( !timeout ) + printk(XENLOG_WARNING + "AMD IOMMU %04x:%02x:%02x.%u: %scompletion wait took %lums\n", + iommu->seg, PCI_BUS(iommu->bdf), + PCI_SLOT(iommu->bdf), PCI_FUNC(iommu->bdf), + timeout_base ? "iotlb " : "", + (NOW() - start) / 10000000); } /* Build low level iommu command messages */ @@ -300,7 +309,7 @@ void amd_iommu_flush_iotlb(u8 devfn, const struct pci_dev *pdev, /* send INVALIDATE_IOTLB_PAGES command */ spin_lock_irqsave(&iommu->lock, flags); invalidate_iotlb_pages(iommu, maxpend, 0, queueid, daddr, req_id, order); - flush_command_buffer(iommu); + flush_command_buffer(iommu, iommu_dev_iotlb_timeout); spin_unlock_irqrestore(&iommu->lock, flags); } @@ -337,7 +346,7 @@ static void _amd_iommu_flush_pages(struct domain *d, { spin_lock_irqsave(&iommu->lock, flags); invalidate_iommu_pages(iommu, daddr, dom_id, order); - flush_command_buffer(iommu); + flush_command_buffer(iommu, 0); spin_unlock_irqrestore(&iommu->lock, flags); } @@ -361,7 +370,7 @@ void amd_iommu_flush_device(struct amd_iommu *iommu, uint16_t bdf) ASSERT( spin_is_locked(&iommu->lock) ); invalidate_dev_table_entry(iommu, bdf); - flush_command_buffer(iommu); + flush_command_buffer(iommu, 0); } void amd_iommu_flush_intremap(struct amd_iommu *iommu, uint16_t bdf) @@ -369,7 +378,7 @@ void amd_iommu_flush_intremap(struct amd_iommu *iommu, uint16_t bdf) ASSERT( spin_is_locked(&iommu->lock) ); invalidate_interrupt_table(iommu, bdf); - flush_command_buffer(iommu); + flush_command_buffer(iommu, 0); } void amd_iommu_flush_all_caches(struct amd_iommu *iommu) @@ -377,7 +386,7 @@ void amd_iommu_flush_all_caches(struct amd_iommu *iommu) ASSERT( spin_is_locked(&iommu->lock) ); invalidate_iommu_all(iommu); - flush_command_buffer(iommu); + flush_command_buffer(iommu, 0); } void amd_iommu_send_guest_cmd(struct amd_iommu *iommu, u32 cmd[]) @@ -387,7 +396,8 @@ void amd_iommu_send_guest_cmd(struct amd_iommu *iommu, u32 cmd[]) spin_lock_irqsave(&iommu->lock, flags); send_iommu_command(iommu, cmd); - flush_command_buffer(iommu); + /* TBD: Timeout selection may require peeking into cmd[]. */ + flush_command_buffer(iommu, 0); spin_unlock_irqrestore(&iommu->lock, flags); } -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.13 From xen-changelog-bounces@lists.xenproject.org Tue Jun 08 18:12:17 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 08 Jun 2021 18:12:17 +0000 Received: from list by lists.xenproject.org with outflank-mailman.138890.256930 (Exim 4.92) (envelope-from ) id 1lqgD3-00068u-P4; Tue, 08 Jun 2021 18:12:17 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 138890.256930; Tue, 08 Jun 2021 18:12:17 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqgD3-00068m-Ly; Tue, 08 Jun 2021 18:12:17 +0000 Received: by outflank-mailman (input) for mailman id 138890; Tue, 08 Jun 2021 18:12:16 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqgD2-00068W-DK for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 18:12:16 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqgD2-0006Ao-Ch for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 18:12:16 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqgD2-0004YA-Bv for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 18:12:16 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=SFUWI9jGl3UUV1h1RaS4593maBVv44iIr54OelipdGI=; b=AzW6sZAy9xhar7NXrI+pXNHUY2 O2gbrw6KEm8662XKg7NPZXP28Jg9bDx9kZkdwaF/dtHEOy7jDSY56qWAV13Y0wD/7qPJgaBOoINYr TMkTl6Bdj9+OmoQkRhvIyd1E4j1jq0iyLpUCyM6R7AdqvSKkfn+8MsplZU+g7CH069Jk=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.13] x86/spec-ctrl: Protect against Speculative Code Store Bypass Message-Id: Date: Tue, 08 Jun 2021 18:12:16 +0000 commit 97af34f9c36cb77132d91dfbc0cf8d502185764d Author: Andrew Cooper AuthorDate: Thu Mar 11 14:39:11 2021 +0000 Commit: Andrew Cooper CommitDate: Tue Jun 8 18:58:34 2021 +0100 x86/spec-ctrl: Protect against Speculative Code Store Bypass Modern x86 processors have far-better-than-architecturally-guaranteed self modifying code detection. Typically, when a write hits an instruction in flight, a Machine Clear occurs to flush stale content in the frontend and backend. For self modifying code, before a write which hits an instruction in flight retires, the frontend can speculatively decode and execute the old instruction stream. Speculation of this form can suffer from type confusion in registers, and potentially leak data. Furthermore, updates are typically byte-wise, rather than atomic. Depending on timing, speculation can race ahead multiple times between individual writes, and execute the transiently-malformed instruction stream. Xen has stubs which are used in certain cases for emulation purposes. Inhibit speculation between updating the stub and executing it. This is XSA-375 / CVE-2021-0089. Signed-off-by: Andrew Cooper Reviewed-by: Jan Beulich (cherry picked from commit 45f59ed8865318bb0356954bad067f329677ce9e) --- xen/arch/x86/pv/emul-priv-op.c | 2 ++ xen/arch/x86/x86_emulate/x86_emulate.c | 1 + 2 files changed, 3 insertions(+) diff --git a/xen/arch/x86/pv/emul-priv-op.c b/xen/arch/x86/pv/emul-priv-op.c index 6dc4f92a84..59c15ca0e7 100644 --- a/xen/arch/x86/pv/emul-priv-op.c +++ b/xen/arch/x86/pv/emul-priv-op.c @@ -97,6 +97,8 @@ static io_emul_stub_t *io_emul_stub_setup(struct priv_op_ctxt *ctxt, u8 opcode, BUILD_BUG_ON(STUB_BUF_SIZE / 2 < MAX(9, /* Default emul stub */ 5 + IOEMUL_QUIRK_STUB_BYTES)); + block_speculation(); /* SCSB */ + /* Handy function-typed pointer to the stub. */ return (void *)stub_va; } diff --git a/xen/arch/x86/x86_emulate/x86_emulate.c b/xen/arch/x86/x86_emulate/x86_emulate.c index 2dee940adb..195924076c 100644 --- a/xen/arch/x86/x86_emulate/x86_emulate.c +++ b/xen/arch/x86/x86_emulate/x86_emulate.c @@ -1172,6 +1172,7 @@ static inline int mkec(uint8_t e, int32_t ec, ...) # define invoke_stub(pre, post, constraints...) do { \ stub_exn.info = (union stub_exception_token) { .raw = ~0 }; \ stub_exn.line = __LINE__; /* Utility outweighs livepatching cost */ \ + block_speculation(); /* SCSB */ \ asm volatile ( pre "\n\tINDIRECT_CALL %[stub]\n\t" post "\n" \ ".Lret%=:\n\t" \ ".pushsection .fixup,\"ax\"\n" \ -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.13 From xen-changelog-bounces@lists.xenproject.org Tue Jun 08 18:12:27 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 08 Jun 2021 18:12:27 +0000 Received: from list by lists.xenproject.org with outflank-mailman.138891.256934 (Exim 4.92) (envelope-from ) id 1lqgDD-0006BP-Qe; Tue, 08 Jun 2021 18:12:27 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 138891.256934; Tue, 08 Jun 2021 18:12:27 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqgDD-0006BE-Nb; Tue, 08 Jun 2021 18:12:27 +0000 Received: by outflank-mailman (input) for mailman id 138891; Tue, 08 Jun 2021 18:12:26 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqgDC-0006B3-GZ for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 18:12:26 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqgDC-0006B4-Fq for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 18:12:26 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqgDC-0004ZF-Ev for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 18:12:26 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=mvWkglMnEXsQ0QqxYaj++ZtncC8oUkpj1UGhlQC41Ks=; b=tVdqIYsoAJe6lI3HGL0SWUfa0h CcLd3LuPmgS97dD6TItX+cFfGR7+PHYmP5ylrk1jGp4Wvus0I3YG5RJnBFInDn/WJwUxxS9u0XbUy eXKoXQ6HJZmLkzUJ2scIphZjqLesp8HixCoL31mnsgX/8+aC2t4Uox4XSam5K66JVQjw=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.13] x86/spec-ctrl: Mitigate TAA after S3 resume Message-Id: Date: Tue, 08 Jun 2021 18:12:26 +0000 commit 9bd6416528f9080cead0f8c22441f5568dbd0bf3 Author: Andrew Cooper AuthorDate: Thu May 20 01:21:39 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 18:58:34 2021 +0100 x86/spec-ctrl: Mitigate TAA after S3 resume The user chosen setting for MSR_TSX_CTRL needs restoring after S3. All APs get the correct setting via start_secondary(), but the BSP was missed out. This is XSA-377 / CVE-2021-28690. Fixes: 8c4330818f6 ("x86/spec-ctrl: Mitigate the TSX Asynchronous Abort sidechannel") Signed-off-by: Andrew Cooper Reviewed-by: Jan Beulich (cherry picked from commit 8cf276cb2e0b99b96333865873f56b0b31555ff1) --- xen/arch/x86/acpi/power.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/xen/arch/x86/acpi/power.c b/xen/arch/x86/acpi/power.c index ec679c4d3a..19b04aed22 100644 --- a/xen/arch/x86/acpi/power.c +++ b/xen/arch/x86/acpi/power.c @@ -288,6 +288,8 @@ static int enter_state(u32 state) microcode_update_one(); + tsx_init(); /* Needs microcode. May change HLE/RTM feature bits. */ + if ( !recheck_cpu_features(0) ) panic("Missing previously available feature(s)\n"); -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.13 From xen-changelog-bounces@lists.xenproject.org Tue Jun 08 18:22:09 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 08 Jun 2021 18:22:09 +0000 Received: from list by lists.xenproject.org with outflank-mailman.138893.256938 (Exim 4.92) (envelope-from ) id 1lqgMY-00077B-HW; Tue, 08 Jun 2021 18:22:06 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 138893.256938; Tue, 08 Jun 2021 18:22:06 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqgMY-000773-EW; Tue, 08 Jun 2021 18:22:06 +0000 Received: by outflank-mailman (input) for mailman id 138893; Tue, 08 Jun 2021 18:22:05 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqgMX-00076x-Hw for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 18:22:05 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqgMX-0006Lk-EV for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 18:22:05 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqgMX-0005Tx-DT for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 18:22:05 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=VDdSV06G+Y2IqC+aaMx9TW3Nz1IbkJgXcohGFhahEOA=; b=lJeEU3K2ujLjrKT18es4OT6SW8 R42XDHxO8BhQ55gS9lgGnpw8hBauPyR7pg0qa7BXpQFfkn+Gs4gwwF/f6V1cSWztytyXa0rDEgJRk 1NcRawNqIo+XmdK9C0ops/Kahb6Fda4Luv2vxLJbBfptF1RhWQ0SVuJC9A9+itCquZMA=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.12] xen/arm: Create dom0less domUs earlier Message-Id: Date: Tue, 08 Jun 2021 18:22:05 +0000 commit 2c395704e6fedcb1d2dee71c59ce37259aaad3b9 Author: Julien Grall AuthorDate: Mon May 17 17:47:13 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 19:14:03 2021 +0100 xen/arm: Create dom0less domUs earlier In a follow-up patch we will need to unallocate the boot modules before heap_init_late() is called. The modules will contain the domUs kernel and initramfs. Therefore Xen will need to create extra domUs (used by dom0less) before heap_init_late(). This has two consequences on dom0less: 1) Domains will not be unpaused as soon as they are created but once all have been created. However, Xen doesn't guarantee an order to unpause, so this is not something one could rely on. 2) The memory allocated for a domU will not be scrubbed anymore when an admin select bootscrub=on. This is not something we advertised, but if this is a concern we can introduce either force scrub for all domUs or a per-domain flag in the DT. The behavior for bootscrub=off and bootscrub=idle (default) has not changed. This is part of XSA-372 / CVE-2021-28693. Signed-off-by: Julien Grall Reviewed-by: Jan Beulich Reviewed-by: Stefano Stabellini Tested-by: Stefano Stabellini (cherry picked from commit 371347c5b64da699d9f5a0edda5dc496fd2b7a5c) --- xen/arch/arm/domain_build.c | 2 -- xen/arch/arm/setup.c | 9 +++++---- 2 files changed, 5 insertions(+), 6 deletions(-) diff --git a/xen/arch/arm/domain_build.c b/xen/arch/arm/domain_build.c index d9836779d1..6c5a6db144 100644 --- a/xen/arch/arm/domain_build.c +++ b/xen/arch/arm/domain_build.c @@ -2092,8 +2092,6 @@ void __init create_domUs(void) if ( construct_domU(d, node) != 0 ) panic("Could not set up domain %s\n", dt_node_name(node)); - - domain_unpause_by_systemcontroller(d); } } diff --git a/xen/arch/arm/setup.c b/xen/arch/arm/setup.c index d986f84f8d..0e54e9c73e 100644 --- a/xen/arch/arm/setup.c +++ b/xen/arch/arm/setup.c @@ -736,7 +736,7 @@ void __init start_xen(unsigned long boot_phys_offset, int cpus, i; const char *cmdline; struct bootmodule *xen_bootmodule; - struct domain *dom0; + struct domain *dom0, *d; struct xen_domctl_createdomain dom0_cfg = { .flags = XEN_DOMCTL_CDF_hvm_guest | XEN_DOMCTL_CDF_hap, .max_evtchn_port = -1, @@ -902,6 +902,8 @@ void __init start_xen(unsigned long boot_phys_offset, if ( construct_dom0(dom0) != 0) panic("Could not set up DOM0 guest OS\n"); + create_domUs(); + heap_init_late(); init_trace_bufs(); @@ -915,9 +917,8 @@ void __init start_xen(unsigned long boot_phys_offset, system_state = SYS_STATE_active; - create_domUs(); - - domain_unpause_by_systemcontroller(dom0); + for_each_domain( d ) + domain_unpause_by_systemcontroller(d); /* Switch on to the dynamically allocated stack for the idle vcpu * since the static one we're running on is about to be freed. */ -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.12 From xen-changelog-bounces@lists.xenproject.org Tue Jun 08 18:22:16 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 08 Jun 2021 18:22:16 +0000 Received: from list by lists.xenproject.org with outflank-mailman.138894.256942 (Exim 4.92) (envelope-from ) id 1lqgMi-00079E-JG; Tue, 08 Jun 2021 18:22:16 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 138894.256942; Tue, 08 Jun 2021 18:22:16 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqgMi-000796-G6; Tue, 08 Jun 2021 18:22:16 +0000 Received: by outflank-mailman (input) for mailman id 138894; Tue, 08 Jun 2021 18:22:15 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqgMh-00078x-Ie for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 18:22:15 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqgMh-0006Ls-Ho for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 18:22:15 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqgMh-0005VE-Gn for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 18:22:15 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=XOLLIqX4lG8mR1NeREicZKMAcefRWFJE6SpZnokGTp8=; b=fUDdtVtWxrJCpMka4G2yonBvSr 50pyKTgDDq+lG/YP4etwgX7vtkfSuPKJdBrpTDKr1GJesCKKYc1u9ajtxCQG2AqGYoBZBP+Wi5ALo i4OlkJdczuxS5GsP/Zv2ZS0EhBloVIc6IHlNM+oCj2AgKj8hRPwMkbm7bRQinElYKhp8=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.12] xen/arm: Boot modules should always be scrubbed if bootscrub={on, idle} Message-Id: Date: Tue, 08 Jun 2021 18:22:15 +0000 commit aa8866c1e0a92342166c2786ed4ae2455eb7ce61 Author: Julien Grall AuthorDate: Sat Apr 17 17:38:28 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 19:14:03 2021 +0100 xen/arm: Boot modules should always be scrubbed if bootscrub={on, idle} The function to initialize the pages (see init_heap_pages()) will request scrub when the admin request idle bootscrub (default) and state == SYS_STATE_active. When bootscrub=on, Xen will scrub any free pages in heap_init_late(). Currently, the boot modules (e.g. kernels, initramfs) will be discarded/ freed after heap_init_late() is called and system_state switched to SYS_STATE_active. This means the pages associated with the boot modules will not get scrubbed before getting re-purposed. If the memory is assigned to an untrusted domU, it may be able to retrieve secrets from the modules. This is part of XSA-372 / CVE-2021-28693. Fixes: 1774e9b1df27 ("xen/arm: introduce create_domUs") Signed-off-by: Julien Grall Reviewed-by: Jan Beulich Reviewed-by: Stefano Stabellini Tested-by: Stefano Stabellini (cherry picked from commit fd5dc41ceaed9cfcfa011cdfd50f264c89277a90) --- xen/arch/arm/setup.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-) diff --git a/xen/arch/arm/setup.c b/xen/arch/arm/setup.c index 0e54e9c73e..ba95c06d89 100644 --- a/xen/arch/arm/setup.c +++ b/xen/arch/arm/setup.c @@ -73,7 +73,6 @@ static __used void init_done(void) /* Must be done past setting system_state. */ unregister_init_virtual_region(); - discard_initial_modules(); free_init_memory(); startup_cpu_idle_loop(); } @@ -904,6 +903,12 @@ void __init start_xen(unsigned long boot_phys_offset, create_domUs(); + /* + * This needs to be called **before** heap_init_late() so modules + * will be scrubbed (unless suppressed). + */ + discard_initial_modules(); + heap_init_late(); init_trace_bufs(); -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.12 From xen-changelog-bounces@lists.xenproject.org Tue Jun 08 18:22:26 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 08 Jun 2021 18:22:26 +0000 Received: from list by lists.xenproject.org with outflank-mailman.138895.256946 (Exim 4.92) (envelope-from ) id 1lqgMs-0007CB-L6; Tue, 08 Jun 2021 18:22:26 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 138895.256946; Tue, 08 Jun 2021 18:22:26 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqgMs-0007C3-He; Tue, 08 Jun 2021 18:22:26 +0000 Received: by outflank-mailman (input) for mailman id 138895; Tue, 08 Jun 2021 18:22:25 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqgMr-0007Bn-Lp for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 18:22:25 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqgMr-0006M4-L2 for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 18:22:25 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqgMr-0005Wi-K2 for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 18:22:25 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=eCnPludtaDA6uJJH2BPoR30ylkMwqMraOXtFozC0WKQ=; b=CHPaBGvx8kOb+1P0Jy2d/cVlKZ uqiQc6VDqtV7J1M4MxaT7fxI3kgSHk1Li1bYWv+rgNHpIIqX3Tfs3wAgQW7bBgoxcJgGQurwvScR9 uKuKHjEI3+UMNQaMwPndIlA+E3CfXhUEIrVyqrgNaGwzNskxW7s/le3wKksKNZB8D1Z8=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.12] VT-d: size qinval queue dynamically Message-Id: Date: Tue, 08 Jun 2021 18:22:25 +0000 commit 95d23c7563636e6304931a559b6610387ef2bf4b Author: Jan Beulich AuthorDate: Tue Jun 8 19:10:30 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 19:14:03 2021 +0100 VT-d: size qinval queue dynamically With the present synchronous model, we need two slots for every operation (the operation itself and a wait descriptor). There can be one such pair of requests pending per CPU. To ensure that under all normal circumstances a slot is always available when one is requested, size the queue ring according to the number of present CPUs. This is part of XSA-373 / CVE-2021-28692. Signed-off-by: Jan Beulich Reviewed-by: Paul Durrant (cherry picked from commit cbfa62bb140c8f10a0ae96db3ce06e22b3b9d302) --- xen/drivers/passthrough/vtd/iommu.h | 10 +------- xen/drivers/passthrough/vtd/qinval.c | 44 ++++++++++++++++++++++++++++-------- 2 files changed, 35 insertions(+), 19 deletions(-) diff --git a/xen/drivers/passthrough/vtd/iommu.h b/xen/drivers/passthrough/vtd/iommu.h index c9290a3996..3fa6a4ff13 100644 --- a/xen/drivers/passthrough/vtd/iommu.h +++ b/xen/drivers/passthrough/vtd/iommu.h @@ -450,17 +450,9 @@ struct qinval_entry { }q; }; -/* Order of queue invalidation pages(max is 8) */ -#define QINVAL_PAGE_ORDER 2 - -#define QINVAL_ARCH_PAGE_ORDER (QINVAL_PAGE_ORDER + PAGE_SHIFT_4K - PAGE_SHIFT) -#define QINVAL_ARCH_PAGE_NR ( QINVAL_ARCH_PAGE_ORDER < 0 ? \ - 1 : \ - 1 << QINVAL_ARCH_PAGE_ORDER ) - /* Each entry is 16 bytes, so 2^8 entries per page */ #define QINVAL_ENTRY_ORDER ( PAGE_SHIFT - 4 ) -#define QINVAL_ENTRY_NR (1 << (QINVAL_PAGE_ORDER + 8)) +#define QINVAL_MAX_ENTRY_NR (1u << (7 + QINVAL_ENTRY_ORDER)) /* Status data flag */ #define QINVAL_STAT_INIT 0 diff --git a/xen/drivers/passthrough/vtd/qinval.c b/xen/drivers/passthrough/vtd/qinval.c index 09cbd36ebb..103afc15fe 100644 --- a/xen/drivers/passthrough/vtd/qinval.c +++ b/xen/drivers/passthrough/vtd/qinval.c @@ -31,6 +31,9 @@ #define VTD_QI_TIMEOUT 1 +static unsigned int __read_mostly qi_pg_order; +static unsigned int __read_mostly qi_entry_nr; + static int __must_check invalidate_sync(struct iommu *iommu); static void print_qi_regs(struct iommu *iommu) @@ -55,7 +58,7 @@ static unsigned int qinval_next_index(struct iommu *iommu) tail >>= QINVAL_INDEX_SHIFT; /* (tail+1 == head) indicates a full queue, wait for HW */ - while ( ( tail + 1 ) % QINVAL_ENTRY_NR == + while ( ((tail + 1) & (qi_entry_nr - 1)) == ( dmar_readq(iommu->reg, DMAR_IQH_REG) >> QINVAL_INDEX_SHIFT ) ) cpu_relax(); @@ -68,7 +71,7 @@ static void qinval_update_qtail(struct iommu *iommu, unsigned int index) /* Need hold register lock when update tail */ ASSERT( spin_is_locked(&iommu->register_lock) ); - val = (index + 1) % QINVAL_ENTRY_NR; + val = (index + 1) & (qi_entry_nr - 1); dmar_writeq(iommu->reg, DMAR_IQT_REG, (val << QINVAL_INDEX_SHIFT)); } @@ -417,7 +420,27 @@ int enable_qinval(struct iommu *iommu) if ( qi_ctrl->qinval_maddr == 0 ) { drhd = iommu_to_drhd(iommu); - qi_ctrl->qinval_maddr = alloc_pgtable_maddr(drhd, QINVAL_ARCH_PAGE_NR); + if ( !qi_entry_nr ) + { + /* + * With the present synchronous model, we need two slots for every + * operation (the operation itself and a wait descriptor). There + * can be one such pair of requests pending per CPU. One extra + * entry is needed as the ring is considered full when there's + * only one entry left. + */ + BUILD_BUG_ON(CONFIG_NR_CPUS * 2 >= QINVAL_MAX_ENTRY_NR); + qi_pg_order = get_order_from_bytes((num_present_cpus() * 2 + 1) << + (PAGE_SHIFT - + QINVAL_ENTRY_ORDER)); + qi_entry_nr = 1u << (qi_pg_order + QINVAL_ENTRY_ORDER); + + dprintk(XENLOG_INFO VTDPREFIX, + "QI: using %u-entry ring(s)\n", qi_entry_nr); + } + + qi_ctrl->qinval_maddr = + alloc_pgtable_maddr(drhd, qi_entry_nr >> QINVAL_ENTRY_ORDER); if ( qi_ctrl->qinval_maddr == 0 ) { dprintk(XENLOG_WARNING VTDPREFIX, @@ -431,15 +454,16 @@ int enable_qinval(struct iommu *iommu) spin_lock_irqsave(&iommu->register_lock, flags); - /* Setup Invalidation Queue Address(IQA) register with the - * address of the page we just allocated. QS field at - * bits[2:0] to indicate size of queue is one 4KB page. - * That's 256 entries. Queued Head (IQH) and Queue Tail (IQT) - * registers are automatically reset to 0 with write - * to IQA register. + /* + * Setup Invalidation Queue Address (IQA) register with the address of the + * pages we just allocated. The QS field at bits[2:0] indicates the size + * (page order) of the queue. + * + * Queued Head (IQH) and Queue Tail (IQT) registers are automatically + * reset to 0 with write to IQA register. */ dmar_writeq(iommu->reg, DMAR_IQA_REG, - qi_ctrl->qinval_maddr | QINVAL_PAGE_ORDER); + qi_ctrl->qinval_maddr | qi_pg_order); dmar_writeq(iommu->reg, DMAR_IQT_REG, 0); -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.12 From xen-changelog-bounces@lists.xenproject.org Tue Jun 08 18:22:37 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 08 Jun 2021 18:22:37 +0000 Received: from list by lists.xenproject.org with outflank-mailman.138896.256950 (Exim 4.92) (envelope-from ) id 1lqgN3-0007GJ-O4; Tue, 08 Jun 2021 18:22:37 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 138896.256950; Tue, 08 Jun 2021 18:22:37 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqgN3-0007GB-L8; Tue, 08 Jun 2021 18:22:37 +0000 Received: by outflank-mailman (input) for mailman id 138896; Tue, 08 Jun 2021 18:22:35 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqgN1-0007Fp-SV for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 18:22:35 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqgN1-0006MG-Rc for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 18:22:35 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqgN1-0005Xi-QN for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 18:22:35 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=m2jwL4wUwGIfee+UHkTjwO6FN96aQdFnB4e8vDebKIQ=; b=Iys4hla/T3qI4z8ntSW4dXiD9o zzA2KPuS7lcsNvTZChdG6q1qhjei4AyUX84PO9SPlZOLDymtrhaYk9vEx8vvs7D72rZ1TlDd7863z ARdjlidwDmsDQkgBVj2eRZeFg+kgr3o6keHpiixrlFORhqrU2/F3vC7szxE5kC6AJw2M=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.12] AMD/IOMMU: size command buffer dynamically Message-Id: Date: Tue, 08 Jun 2021 18:22:35 +0000 commit ced413bdfddbfbaf72caf79333962fe199bea6d9 Author: Jan Beulich AuthorDate: Tue Jun 8 19:10:44 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 19:14:03 2021 +0100 AMD/IOMMU: size command buffer dynamically With the present synchronous model, we need two slots for every operation (the operation itself and a wait command). There can be one such pair of commands pending per CPU. To ensure that under all normal circumstances a slot is always available when one is requested, size the command ring according to the number of present CPUs. This is part of XSA-373 / CVE-2021-28692. Signed-off-by: Jan Beulich Reviewed-by: Paul Durrant (cherry picked from commit df242851ddc93ac0b0a3a20ecab34acc29e3b36b) --- xen/drivers/passthrough/amd/iommu_cmd.c | 4 ++-- xen/drivers/passthrough/amd/iommu_init.c | 28 +++++++++++++++++++++++++--- xen/include/asm-x86/hvm/svm/amd-iommu-defs.h | 7 ++----- 3 files changed, 29 insertions(+), 10 deletions(-) diff --git a/xen/drivers/passthrough/amd/iommu_cmd.c b/xen/drivers/passthrough/amd/iommu_cmd.c index af3a1fb865..c9c7d7b290 100644 --- a/xen/drivers/passthrough/amd/iommu_cmd.c +++ b/xen/drivers/passthrough/amd/iommu_cmd.c @@ -35,8 +35,8 @@ static int queue_iommu_command(struct amd_iommu *iommu, u32 cmd[]) if ( head != tail ) { memcpy(iommu->cmd_buffer.buffer + - (iommu->cmd_buffer.tail * IOMMU_CMD_BUFFER_ENTRY_SIZE), - cmd, IOMMU_CMD_BUFFER_ENTRY_SIZE); + (iommu->cmd_buffer.tail * sizeof(cmd_entry_t)), + cmd, sizeof(cmd_entry_t)); iommu->cmd_buffer.tail = tail; return 1; diff --git a/xen/drivers/passthrough/amd/iommu_init.c b/xen/drivers/passthrough/amd/iommu_init.c index 534d6bb889..98298f0bf0 100644 --- a/xen/drivers/passthrough/amd/iommu_init.c +++ b/xen/drivers/passthrough/amd/iommu_init.c @@ -136,7 +136,7 @@ static void register_iommu_cmd_buffer_in_mmio_space(struct amd_iommu *iommu) writel(entry, iommu->mmio_base + IOMMU_CMD_BUFFER_BASE_LOW_OFFSET); power_of2_entries = get_order_from_bytes(iommu->cmd_buffer.alloc_size) + - IOMMU_CMD_BUFFER_POWER_OF2_ENTRIES_PER_PAGE; + PAGE_SHIFT - IOMMU_CMD_BUFFER_ENTRY_ORDER; entry = 0; iommu_set_addr_hi_to_reg(&entry, addr_hi); @@ -1000,9 +1000,31 @@ static void * __init allocate_ring_buffer(struct ring_buffer *ring_buf, static void * __init allocate_cmd_buffer(struct amd_iommu *iommu) { /* allocate 'command buffer' in power of 2 increments of 4K */ + static unsigned int __read_mostly nr_ents; + + if ( !nr_ents ) + { + unsigned int order; + + /* + * With the present synchronous model, we need two slots for every + * operation (the operation itself and a wait command). There can be + * one such pair of requests pending per CPU. One extra entry is + * needed as the ring is considered full when there's only one entry + * left. + */ + BUILD_BUG_ON(CONFIG_NR_CPUS * 2 >= IOMMU_CMD_BUFFER_MAX_ENTRIES); + order = get_order_from_bytes((num_present_cpus() * 2 + 1) << + IOMMU_CMD_BUFFER_ENTRY_ORDER); + nr_ents = 1u << (order + PAGE_SHIFT - IOMMU_CMD_BUFFER_ENTRY_ORDER); + + AMD_IOMMU_DEBUG("using %u-entry cmd ring(s)\n", nr_ents); + } + + BUILD_BUG_ON(sizeof(cmd_entry_t) != (1u << IOMMU_CMD_BUFFER_ENTRY_ORDER)); + return allocate_ring_buffer(&iommu->cmd_buffer, sizeof(cmd_entry_t), - IOMMU_CMD_BUFFER_DEFAULT_ENTRIES, - "Command Buffer"); + nr_ents, "Command Buffer"); } static void * __init allocate_event_log(struct amd_iommu *iommu) diff --git a/xen/include/asm-x86/hvm/svm/amd-iommu-defs.h b/xen/include/asm-x86/hvm/svm/amd-iommu-defs.h index a217245249..d9ec20cc90 100644 --- a/xen/include/asm-x86/hvm/svm/amd-iommu-defs.h +++ b/xen/include/asm-x86/hvm/svm/amd-iommu-defs.h @@ -20,9 +20,6 @@ #ifndef _ASM_X86_64_AMD_IOMMU_DEFS_H #define _ASM_X86_64_AMD_IOMMU_DEFS_H -/* IOMMU Command Buffer entries: in power of 2 increments, minimum of 256 */ -#define IOMMU_CMD_BUFFER_DEFAULT_ENTRIES 512 - /* IOMMU Event Log entries: in power of 2 increments, minimum of 256 */ #define IOMMU_EVENT_LOG_DEFAULT_ENTRIES 512 @@ -184,8 +181,8 @@ #define IOMMU_CMD_BUFFER_LENGTH_MASK 0x0F000000 #define IOMMU_CMD_BUFFER_LENGTH_SHIFT 24 -#define IOMMU_CMD_BUFFER_ENTRY_SIZE 16 -#define IOMMU_CMD_BUFFER_POWER_OF2_ENTRIES_PER_PAGE 8 +#define IOMMU_CMD_BUFFER_ENTRY_ORDER 4 +#define IOMMU_CMD_BUFFER_MAX_ENTRIES (1u << 15) #define IOMMU_CMD_OPCODE_MASK 0xF0000000 #define IOMMU_CMD_OPCODE_SHIFT 28 -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.12 From xen-changelog-bounces@lists.xenproject.org Tue Jun 08 18:22:47 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 08 Jun 2021 18:22:47 +0000 Received: from list by lists.xenproject.org with outflank-mailman.138897.256954 (Exim 4.92) (envelope-from ) id 1lqgND-0007J0-Pg; Tue, 08 Jun 2021 18:22:47 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 138897.256954; Tue, 08 Jun 2021 18:22:47 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqgND-0007Is-Mj; Tue, 08 Jun 2021 18:22:47 +0000 Received: by outflank-mailman (input) for mailman id 138897; Tue, 08 Jun 2021 18:22:46 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqgNB-0007If-WE for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 18:22:46 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqgNB-0006Mr-VK for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 18:22:45 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqgNB-0005YS-UK for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 18:22:45 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=ogeoUuK2Tena2MN4WqHczQ0Yw7j1QEL5e7MME6sJc+o=; b=f3W0gxGz2ZR1U88VRqHvkkPkym tpDT7imE/69GvI0luI9XCoJzOShV7kbtfGjI2l/TyTgw5sEn/FnP0jNGWB/CGR0UCyw++8ahknqrc APXjLFPPDnt/zkcnggUwr0cM0kENiz7ApXxpule4y5GB/c/0wOB+bsi+O6zo0BHYbTyo=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.12] VT-d: eliminate flush related timeouts Message-Id: Date: Tue, 08 Jun 2021 18:22:45 +0000 commit e318c1330fbfb3388bdb8e5501a0ab89d6aa5a5a Author: Jan Beulich AuthorDate: Tue Jun 8 19:10:58 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 19:14:03 2021 +0100 VT-d: eliminate flush related timeouts Leaving an in-progress operation pending when it appears to take too long is problematic: If e.g. a QI command completed later, the write to the "poll slot" may instead be understood to signal a subsequently started command's completion. Also our accounting of the timeout period was actually wrong: We included the time it took for the command to actually make it to the front of the queue, which could be heavily affected by guests other than the one for which the flush is being performed. Do away with all timeout detection on all flush related code paths. Log excessively long processing times (with a progressive threshold) to have some indication of problems in this area. Additionally log (once) if qinval_next_index() didn't immediately find an available slot. Together with the earlier change sizing the queue(s) dynamically, we should now have a guarantee that with our fully synchronous model any demand for slots can actually be satisfied. This is part of XSA-373 / CVE-2021-28692. Signed-off-by: Jan Beulich Reviewed-by: Paul Durrant (cherry picked from commit 7c893abc4ee7e9af62297ba6fd5f27c89d0f33f5) --- xen/drivers/passthrough/vtd/dmar.h | 28 ++++++++++++++++++++++++++++ xen/drivers/passthrough/vtd/iommu.c | 12 ++++++------ xen/drivers/passthrough/vtd/qinval.c | 33 ++++++++++++++++++++++----------- 3 files changed, 56 insertions(+), 17 deletions(-) diff --git a/xen/drivers/passthrough/vtd/dmar.h b/xen/drivers/passthrough/vtd/dmar.h index 95bb132471..8438595556 100644 --- a/xen/drivers/passthrough/vtd/dmar.h +++ b/xen/drivers/passthrough/vtd/dmar.h @@ -127,6 +127,34 @@ do { \ } \ } while (0) +#define IOMMU_FLUSH_WAIT(what, iommu, offset, op, cond, sts) \ +do { \ + static unsigned int __read_mostly threshold = 1; \ + s_time_t start = NOW(); \ + s_time_t timeout = start + DMAR_OPERATION_TIMEOUT * threshold; \ + \ + for ( ; ; ) \ + { \ + sts = op(iommu->reg, offset); \ + if ( cond ) \ + break; \ + if ( timeout && NOW() > timeout ) \ + { \ + threshold |= threshold << 1; \ + printk(XENLOG_WARNING VTDPREFIX \ + " IOMMU#%u: %s flush taking too long\n", \ + iommu->index, what); \ + timeout = 0; \ + } \ + cpu_relax(); \ + } \ + \ + if ( !timeout ) \ + printk(XENLOG_WARNING VTDPREFIX \ + " IOMMU#%u: %s flush took %lums\n", \ + iommu->index, what, (NOW() - start) / 10000000); \ +} while ( false ) + int vtd_hw_check(void); void disable_pmr(struct iommu *iommu); int is_igd_drhd(struct acpi_drhd_unit *drhd); diff --git a/xen/drivers/passthrough/vtd/iommu.c b/xen/drivers/passthrough/vtd/iommu.c index f36fded6fe..89e0cab531 100644 --- a/xen/drivers/passthrough/vtd/iommu.c +++ b/xen/drivers/passthrough/vtd/iommu.c @@ -357,8 +357,8 @@ static void iommu_flush_write_buffer(struct iommu *iommu) dmar_writel(iommu->reg, DMAR_GCMD_REG, val | DMA_GCMD_WBF); /* Make sure hardware complete it */ - IOMMU_WAIT_OP(iommu, DMAR_GSTS_REG, dmar_readl, - !(val & DMA_GSTS_WBFS), val); + IOMMU_FLUSH_WAIT("write buffer", iommu, DMAR_GSTS_REG, dmar_readl, + !(val & DMA_GSTS_WBFS), val); spin_unlock_irqrestore(&iommu->register_lock, flags); } @@ -408,8 +408,8 @@ static int __must_check flush_context_reg(void *_iommu, u16 did, u16 source_id, dmar_writeq(iommu->reg, DMAR_CCMD_REG, val); /* Make sure hardware complete it */ - IOMMU_WAIT_OP(iommu, DMAR_CCMD_REG, dmar_readq, - !(val & DMA_CCMD_ICC), val); + IOMMU_FLUSH_WAIT("context", iommu, DMAR_CCMD_REG, dmar_readq, + !(val & DMA_CCMD_ICC), val); spin_unlock_irqrestore(&iommu->register_lock, flags); /* flush context entry will implicitly flush write buffer */ @@ -491,8 +491,8 @@ static int __must_check flush_iotlb_reg(void *_iommu, u16 did, u64 addr, dmar_writeq(iommu->reg, tlb_offset + 8, val); /* Make sure hardware complete it */ - IOMMU_WAIT_OP(iommu, (tlb_offset + 8), dmar_readq, - !(val & DMA_TLB_IVT), val); + IOMMU_FLUSH_WAIT("iotlb", iommu, (tlb_offset + 8), dmar_readq, + !(val & DMA_TLB_IVT), val); spin_unlock_irqrestore(&iommu->register_lock, flags); /* check IOTLB invalidation granularity */ diff --git a/xen/drivers/passthrough/vtd/qinval.c b/xen/drivers/passthrough/vtd/qinval.c index 103afc15fe..17c4235597 100644 --- a/xen/drivers/passthrough/vtd/qinval.c +++ b/xen/drivers/passthrough/vtd/qinval.c @@ -29,8 +29,6 @@ #include "extern.h" #include "../ats.h" -#define VTD_QI_TIMEOUT 1 - static unsigned int __read_mostly qi_pg_order; static unsigned int __read_mostly qi_entry_nr; @@ -60,7 +58,11 @@ static unsigned int qinval_next_index(struct iommu *iommu) /* (tail+1 == head) indicates a full queue, wait for HW */ while ( ((tail + 1) & (qi_entry_nr - 1)) == ( dmar_readq(iommu->reg, DMAR_IQH_REG) >> QINVAL_INDEX_SHIFT ) ) + { + printk_once(XENLOG_ERR VTDPREFIX " IOMMU#%u: no QI slot available\n", + iommu->index); cpu_relax(); + } return tail; } @@ -180,23 +182,32 @@ static int __must_check queue_invalidate_wait(struct iommu *iommu, /* Now we don't support interrupt method */ if ( sw ) { - s_time_t timeout; - - /* In case all wait descriptor writes to same addr with same data */ - timeout = NOW() + MILLISECS(flush_dev_iotlb ? - iommu_dev_iotlb_timeout : VTD_QI_TIMEOUT); + static unsigned int __read_mostly threshold = 1; + s_time_t start = NOW(); + s_time_t timeout = start + (flush_dev_iotlb + ? iommu_dev_iotlb_timeout + : 100) * MILLISECS(threshold); while ( ACCESS_ONCE(*this_poll_slot) != QINVAL_STAT_DONE ) { - if ( NOW() > timeout ) + if ( timeout && NOW() > timeout ) { - print_qi_regs(iommu); + threshold |= threshold << 1; printk(XENLOG_WARNING VTDPREFIX - " Queue invalidate wait descriptor timed out\n"); - return -ETIMEDOUT; + " IOMMU#%u: QI%s wait descriptor taking too long\n", + iommu->index, flush_dev_iotlb ? " dev" : ""); + print_qi_regs(iommu); + timeout = 0; } cpu_relax(); } + + if ( !timeout ) + printk(XENLOG_WARNING VTDPREFIX + " IOMMU#%u: QI%s wait descriptor took %lums\n", + iommu->index, flush_dev_iotlb ? " dev" : "", + (NOW() - start) / 10000000); + return 0; } -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.12 From xen-changelog-bounces@lists.xenproject.org Tue Jun 08 18:22:57 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 08 Jun 2021 18:22:57 +0000 Received: from list by lists.xenproject.org with outflank-mailman.138898.256957 (Exim 4.92) (envelope-from ) id 1lqgNN-0007M9-RI; Tue, 08 Jun 2021 18:22:57 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 138898.256957; Tue, 08 Jun 2021 18:22:57 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqgNN-0007Lz-OJ; Tue, 08 Jun 2021 18:22:57 +0000 Received: by outflank-mailman (input) for mailman id 138898; Tue, 08 Jun 2021 18:22:56 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqgNM-0007Lj-2j for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 18:22:56 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqgNM-0006N6-1y for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 18:22:56 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqgNM-0005ZV-1H for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 18:22:56 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=8ENBmDe6mANuUwyeSfxEBd98xW+ElA0TwBqtu0BV2Y0=; b=pAEEwrCsdrhbriEz/N7MfXTe6O 07OMfFQlJSo/1S+VQIPk7NPZ4NDRvMpqjcBodLcrrsIoCUfouC51aPbpLM9LQpux6jHcLm8XfdZfm oNnQGioZAsQURHfVQIAHzoKuFfZ/uEvbaZBqWhhENNFtFs4LnhyO5y/PD7UQsxesguLg=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.12] AMD/IOMMU: wait for command slot to be available Message-Id: Date: Tue, 08 Jun 2021 18:22:56 +0000 commit e5f3be9d6ef53d4dfef0f7a550ce6cd93b9629e2 Author: Jan Beulich AuthorDate: Tue Jun 8 19:11:18 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 19:14:03 2021 +0100 AMD/IOMMU: wait for command slot to be available No caller cared about send_iommu_command() indicating unavailability of a slot. Hence if a sufficient number prior commands timed out, we did blindly assume that the requested command was submitted to the IOMMU when really it wasn't. This could mean both a hanging system (waiting for a command to complete that was never seen by the IOMMU) or blindly propagating success back to callers, making them believe they're fine to e.g. free previously unmapped pages. Fold the three involved functions into one, add spin waiting for an available slot along the lines of VT-d's qinval_next_index(), and as a consequence drop all error indicator return types/values. This is part of XSA-373 / CVE-2021-28692. Signed-off-by: Jan Beulich Reviewed-by: Paul Durrant (cherry picked from commit c4e44343579e2c3304d676404d81b2b9bd893c5b) --- xen/drivers/passthrough/amd/iommu_cmd.c | 42 ++++++++++++--------------------- 1 file changed, 15 insertions(+), 27 deletions(-) diff --git a/xen/drivers/passthrough/amd/iommu_cmd.c b/xen/drivers/passthrough/amd/iommu_cmd.c index c9c7d7b290..3e48fc3733 100644 --- a/xen/drivers/passthrough/amd/iommu_cmd.c +++ b/xen/drivers/passthrough/amd/iommu_cmd.c @@ -22,48 +22,36 @@ #include #include "../ats.h" -static int queue_iommu_command(struct amd_iommu *iommu, u32 cmd[]) +static void send_iommu_command(struct amd_iommu *iommu, + const uint32_t cmd[4]) { - uint32_t tail, head; + uint32_t tail; tail = iommu->cmd_buffer.tail; if ( ++tail == iommu->cmd_buffer.entries ) tail = 0; - head = iommu_get_rb_pointer(readl(iommu->mmio_base + - IOMMU_CMD_BUFFER_HEAD_OFFSET)); - if ( head != tail ) + while ( tail == iommu_get_rb_pointer(readl(iommu->mmio_base + + IOMMU_CMD_BUFFER_HEAD_OFFSET)) ) { - memcpy(iommu->cmd_buffer.buffer + - (iommu->cmd_buffer.tail * sizeof(cmd_entry_t)), - cmd, sizeof(cmd_entry_t)); - - iommu->cmd_buffer.tail = tail; - return 1; + printk_once(XENLOG_ERR + "AMD IOMMU %04x:%02x:%02x.%u: no cmd slot available\n", + iommu->seg, PCI_BUS(iommu->bdf), + PCI_SLOT(iommu->bdf), PCI_FUNC(iommu->bdf)); + cpu_relax(); } - return 0; -} + memcpy(iommu->cmd_buffer.buffer + + (iommu->cmd_buffer.tail * sizeof(cmd_entry_t)), + cmd, sizeof(cmd_entry_t)); -static void commit_iommu_command_buffer(struct amd_iommu *iommu) -{ - u32 tail = 0; + iommu->cmd_buffer.tail = tail; + tail = 0; iommu_set_rb_pointer(&tail, iommu->cmd_buffer.tail); writel(tail, iommu->mmio_base+IOMMU_CMD_BUFFER_TAIL_OFFSET); } -int send_iommu_command(struct amd_iommu *iommu, u32 cmd[]) -{ - if ( queue_iommu_command(iommu, cmd) ) - { - commit_iommu_command_buffer(iommu); - return 1; - } - - return 0; -} - static void flush_command_buffer(struct amd_iommu *iommu) { u32 cmd[4], status; -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.12 From xen-changelog-bounces@lists.xenproject.org Tue Jun 08 18:23:07 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 08 Jun 2021 18:23:07 +0000 Received: from list by lists.xenproject.org with outflank-mailman.138899.256963 (Exim 4.92) (envelope-from ) id 1lqgNX-0007PP-U0; Tue, 08 Jun 2021 18:23:07 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 138899.256963; Tue, 08 Jun 2021 18:23:07 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqgNX-0007PH-Pv; Tue, 08 Jun 2021 18:23:07 +0000 Received: by outflank-mailman (input) for mailman id 138899; Tue, 08 Jun 2021 18:23:06 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqgNW-0007P5-5c for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 18:23:06 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqgNW-0006NZ-4p for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 18:23:06 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqgNW-0005b0-44 for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 18:23:06 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=Ak1YRI/mBLMiPVDF4qTGWxUQanoS1z5zfNEvEbOtrZ8=; b=w+i8ojH3xf1ZesuC7aIFFdhW8p 5mKbd2jPqEKsB+KslUzLbWUet1CcwkyoBHaXr3zJhbrs23i64tlMp/HFCNX9odIZITOFMGETt4d0K Zk9aadjJZdwKDhKTBq1De2zGqEwlzROeBi1D6X2SozNr4jzTUt8s+CUsLoaFGPxGSmCA=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.12] AMD/IOMMU: drop command completion timeout Message-Id: Date: Tue, 08 Jun 2021 18:23:06 +0000 commit d024fe144d072ec1e3c7e03a7992a147cfdbd35e Author: Jan Beulich AuthorDate: Tue Jun 8 19:11:49 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 19:14:03 2021 +0100 AMD/IOMMU: drop command completion timeout First and foremost - such timeouts were not signaled to callers, making them believe they're fine to e.g. free previously unmapped pages. Mirror VT-d's behavior: A fixed number of loop iterations is not a suitable way to detect timeouts in an environment (CPU and bus speeds) independent manner anyway. Furthermore, leaving an in-progress operation pending when it appears to take too long is problematic: If a command completed later, the signaling of its completion may instead be understood to signal a subsequently started command's completion. Log excessively long processing times (with a progressive threshold) to have some indication of problems in this area. Allow callers to specify a non-default timeout bias for this logging, using the same values as VT-d does, which in particular means a (by default) much larger value for device IO TLB invalidation. This is part of XSA-373 / CVE-2021-28692. Signed-off-by: Jan Beulich Reviewed-by: Paul Durrant (cherry picked from commit e4fee66043120c954fc309bbb37813604c1c0eb7) --- xen/drivers/passthrough/amd/iommu_cmd.c | 60 +++++++++++++++++++-------------- 1 file changed, 35 insertions(+), 25 deletions(-) diff --git a/xen/drivers/passthrough/amd/iommu_cmd.c b/xen/drivers/passthrough/amd/iommu_cmd.c index 3e48fc3733..801b27af58 100644 --- a/xen/drivers/passthrough/amd/iommu_cmd.c +++ b/xen/drivers/passthrough/amd/iommu_cmd.c @@ -52,10 +52,12 @@ static void send_iommu_command(struct amd_iommu *iommu, writel(tail, iommu->mmio_base+IOMMU_CMD_BUFFER_TAIL_OFFSET); } -static void flush_command_buffer(struct amd_iommu *iommu) +static void flush_command_buffer(struct amd_iommu *iommu, + unsigned int timeout_base) { - u32 cmd[4], status; - int loop_count, comp_wait; + uint32_t cmd[4]; + s_time_t start, timeout; + static unsigned int __read_mostly threshold = 1; /* RW1C 'ComWaitInt' in status register */ writel(IOMMU_STATUS_COMP_WAIT_INT_MASK, @@ -71,24 +73,31 @@ static void flush_command_buffer(struct amd_iommu *iommu) IOMMU_COMP_WAIT_I_FLAG_SHIFT, &cmd[0]); send_iommu_command(iommu, cmd); - /* Make loop_count long enough for polling completion wait bit */ - loop_count = 1000; - do { - status = readl(iommu->mmio_base + IOMMU_STATUS_MMIO_OFFSET); - comp_wait = get_field_from_reg_u32(status, - IOMMU_STATUS_COMP_WAIT_INT_MASK, - IOMMU_STATUS_COMP_WAIT_INT_SHIFT); - --loop_count; - } while ( !comp_wait && loop_count ); - - if ( comp_wait ) + start = NOW(); + timeout = start + (timeout_base ?: 100) * MILLISECS(threshold); + while ( !(readl(iommu->mmio_base + IOMMU_STATUS_MMIO_OFFSET) & + IOMMU_STATUS_COMP_WAIT_INT_MASK) ) { - /* RW1C 'ComWaitInt' in status register */ - writel(IOMMU_STATUS_COMP_WAIT_INT_MASK, - iommu->mmio_base + IOMMU_STATUS_MMIO_OFFSET); - return; + if ( timeout && NOW() > timeout ) + { + threshold |= threshold << 1; + printk(XENLOG_WARNING + "AMD IOMMU %04x:%02x:%02x.%u: %scompletion wait taking too long\n", + iommu->seg, PCI_BUS(iommu->bdf), + PCI_SLOT(iommu->bdf), PCI_FUNC(iommu->bdf), + timeout_base ? "iotlb " : ""); + timeout = 0; + } + cpu_relax(); } - AMD_IOMMU_DEBUG("Warning: ComWaitInt bit did not assert!\n"); + + if ( !timeout ) + printk(XENLOG_WARNING + "AMD IOMMU %04x:%02x:%02x.%u: %scompletion wait took %lums\n", + iommu->seg, PCI_BUS(iommu->bdf), + PCI_SLOT(iommu->bdf), PCI_FUNC(iommu->bdf), + timeout_base ? "iotlb " : "", + (NOW() - start) / 10000000); } /* Build low level iommu command messages */ @@ -300,7 +309,7 @@ void amd_iommu_flush_iotlb(u8 devfn, const struct pci_dev *pdev, /* send INVALIDATE_IOTLB_PAGES command */ spin_lock_irqsave(&iommu->lock, flags); invalidate_iotlb_pages(iommu, maxpend, 0, queueid, daddr, req_id, order); - flush_command_buffer(iommu); + flush_command_buffer(iommu, iommu_dev_iotlb_timeout); spin_unlock_irqrestore(&iommu->lock, flags); } @@ -337,7 +346,7 @@ static void _amd_iommu_flush_pages(struct domain *d, { spin_lock_irqsave(&iommu->lock, flags); invalidate_iommu_pages(iommu, daddr, dom_id, order); - flush_command_buffer(iommu); + flush_command_buffer(iommu, 0); spin_unlock_irqrestore(&iommu->lock, flags); } @@ -361,7 +370,7 @@ void amd_iommu_flush_device(struct amd_iommu *iommu, uint16_t bdf) ASSERT( spin_is_locked(&iommu->lock) ); invalidate_dev_table_entry(iommu, bdf); - flush_command_buffer(iommu); + flush_command_buffer(iommu, 0); } void amd_iommu_flush_intremap(struct amd_iommu *iommu, uint16_t bdf) @@ -369,7 +378,7 @@ void amd_iommu_flush_intremap(struct amd_iommu *iommu, uint16_t bdf) ASSERT( spin_is_locked(&iommu->lock) ); invalidate_interrupt_table(iommu, bdf); - flush_command_buffer(iommu); + flush_command_buffer(iommu, 0); } void amd_iommu_flush_all_caches(struct amd_iommu *iommu) @@ -377,7 +386,7 @@ void amd_iommu_flush_all_caches(struct amd_iommu *iommu) ASSERT( spin_is_locked(&iommu->lock) ); invalidate_iommu_all(iommu); - flush_command_buffer(iommu); + flush_command_buffer(iommu, 0); } void amd_iommu_send_guest_cmd(struct amd_iommu *iommu, u32 cmd[]) @@ -387,7 +396,8 @@ void amd_iommu_send_guest_cmd(struct amd_iommu *iommu, u32 cmd[]) spin_lock_irqsave(&iommu->lock, flags); send_iommu_command(iommu, cmd); - flush_command_buffer(iommu); + /* TBD: Timeout selection may require peeking into cmd[]. */ + flush_command_buffer(iommu, 0); spin_unlock_irqrestore(&iommu->lock, flags); } -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.12 From xen-changelog-bounces@lists.xenproject.org Tue Jun 08 18:23:18 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 08 Jun 2021 18:23:18 +0000 Received: from list by lists.xenproject.org with outflank-mailman.138900.256967 (Exim 4.92) (envelope-from ) id 1lqgNi-0007Sq-0R; Tue, 08 Jun 2021 18:23:18 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 138900.256967; Tue, 08 Jun 2021 18:23:17 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqgNh-0007Sg-TR; Tue, 08 Jun 2021 18:23:17 +0000 Received: by outflank-mailman (input) for mailman id 138900; Tue, 08 Jun 2021 18:23:16 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqgNg-0007SE-9r for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 18:23:16 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqgNg-0006Nk-90 for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 18:23:16 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqgNg-0005bn-7t for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 18:23:16 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=hnUpw4IK5rnAql8cff5G6jUk+HxquxSk8azeqOEB2g4=; b=eeSdbQbq6Vnh6HrrFpfqCJe6ZT L4sJzUFaznoxEEzUUw8jiHL5zS/n4oO6rWE7+O0csLuZ3rTjY/xwBm98dtoAudwVCHKtPD0mwI00p qo1kRiJQKDE+r02JwK1d3GaUsADLVFXW0tj/Io7/GqrWl4WxYs+KuVVn4SpaQg3q8uzE=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.12] x86/spec-ctrl: Protect against Speculative Code Store Bypass Message-Id: Date: Tue, 08 Jun 2021 18:23:16 +0000 commit 99f2c4672368481307a52c24ff1a630a7451ba5a Author: Andrew Cooper AuthorDate: Thu Mar 11 14:39:11 2021 +0000 Commit: Andrew Cooper CommitDate: Tue Jun 8 19:14:03 2021 +0100 x86/spec-ctrl: Protect against Speculative Code Store Bypass Modern x86 processors have far-better-than-architecturally-guaranteed self modifying code detection. Typically, when a write hits an instruction in flight, a Machine Clear occurs to flush stale content in the frontend and backend. For self modifying code, before a write which hits an instruction in flight retires, the frontend can speculatively decode and execute the old instruction stream. Speculation of this form can suffer from type confusion in registers, and potentially leak data. Furthermore, updates are typically byte-wise, rather than atomic. Depending on timing, speculation can race ahead multiple times between individual writes, and execute the transiently-malformed instruction stream. Xen has stubs which are used in certain cases for emulation purposes. Inhibit speculation between updating the stub and executing it. This is XSA-375 / CVE-2021-0089. Signed-off-by: Andrew Cooper Reviewed-by: Jan Beulich (cherry picked from commit 45f59ed8865318bb0356954bad067f329677ce9e) --- xen/arch/x86/pv/emul-priv-op.c | 2 ++ xen/arch/x86/x86_emulate/x86_emulate.c | 1 + 2 files changed, 3 insertions(+) diff --git a/xen/arch/x86/pv/emul-priv-op.c b/xen/arch/x86/pv/emul-priv-op.c index 933036ea34..e9147e24c7 100644 --- a/xen/arch/x86/pv/emul-priv-op.c +++ b/xen/arch/x86/pv/emul-priv-op.c @@ -97,6 +97,8 @@ static io_emul_stub_t *io_emul_stub_setup(struct priv_op_ctxt *ctxt, u8 opcode, BUILD_BUG_ON(STUB_BUF_SIZE / 2 < MAX(9, /* Default emul stub */ 5 + IOEMUL_QUIRK_STUB_BYTES)); + asm volatile ( "lfence" ::: "memory" ); /* SCSB */ + /* Handy function-typed pointer to the stub. */ return (void *)stub_va; } diff --git a/xen/arch/x86/x86_emulate/x86_emulate.c b/xen/arch/x86/x86_emulate/x86_emulate.c index 09de6ee15d..c7c55fa572 100644 --- a/xen/arch/x86/x86_emulate/x86_emulate.c +++ b/xen/arch/x86/x86_emulate/x86_emulate.c @@ -1093,6 +1093,7 @@ static inline int mkec(uint8_t e, int32_t ec, ...) # define invoke_stub(pre, post, constraints...) do { \ stub_exn.info = (union stub_exception_token) { .raw = ~0 }; \ stub_exn.line = __LINE__; /* Utility outweighs livepatching cost */ \ + asm volatile ( "lfence" ::: "memory" ); /* SCSB */ \ asm volatile ( pre "\n\tINDIRECT_CALL %[stub]\n\t" post "\n" \ ".Lret%=:\n\t" \ ".pushsection .fixup,\"ax\"\n" \ -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.12 From xen-changelog-bounces@lists.xenproject.org Tue Jun 08 18:23:28 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 08 Jun 2021 18:23:28 +0000 Received: from list by lists.xenproject.org with outflank-mailman.138901.256970 (Exim 4.92) (envelope-from ) id 1lqgNs-0007Xm-1d; Tue, 08 Jun 2021 18:23:28 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 138901.256970; Tue, 08 Jun 2021 18:23:28 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqgNr-0007Xe-Uw; Tue, 08 Jun 2021 18:23:27 +0000 Received: by outflank-mailman (input) for mailman id 138901; Tue, 08 Jun 2021 18:23:26 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqgNq-0007XI-Cn for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 18:23:26 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqgNq-0006Ny-C4 for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 18:23:26 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqgNq-0005e7-BJ for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 18:23:26 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=NbiMKSM3/M0U3OQrQ3fgTnP6jGth+b9qIz2kDEY3GlQ=; b=2/fMMhAlXPJ+I7NJPrCMIdIL2K mXUCNDmxbr6Rmpjt2TpsKGegGtE6CIcgw9eJ+e8RKcT79J7sivd2Vgh23Z0agb5G920Ilhu/S+t+x m75lLtYgkkBDqanuTUOFc3HpBbKLJR2i3RQyVqNsmoxmu9A+UYowEGcoJLZPwPpBDDuQ=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.12] x86/spec-ctrl: Mitigate TAA after S3 resume Message-Id: Date: Tue, 08 Jun 2021 18:23:26 +0000 commit ea20eee97e9e0861127a8070cc7b9ae3557b09fb Author: Andrew Cooper AuthorDate: Thu May 20 01:21:39 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 19:14:03 2021 +0100 x86/spec-ctrl: Mitigate TAA after S3 resume The user chosen setting for MSR_TSX_CTRL needs restoring after S3. All APs get the correct setting via start_secondary(), but the BSP was missed out. This is XSA-377 / CVE-2021-28690. Fixes: 8c4330818f6 ("x86/spec-ctrl: Mitigate the TSX Asynchronous Abort sidechannel") Signed-off-by: Andrew Cooper Reviewed-by: Jan Beulich (cherry picked from commit 8cf276cb2e0b99b96333865873f56b0b31555ff1) --- xen/arch/x86/acpi/power.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/xen/arch/x86/acpi/power.c b/xen/arch/x86/acpi/power.c index a07aa3b9ed..660363a3df 100644 --- a/xen/arch/x86/acpi/power.c +++ b/xen/arch/x86/acpi/power.c @@ -259,6 +259,8 @@ static int enter_state(u32 state) microcode_resume_cpu(0); + tsx_init(); /* Needs microcode. May change HLE/RTM feature bits. */ + if ( !recheck_cpu_features(0) ) panic("Missing previously available feature(s)\n"); -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.12 From xen-changelog-bounces@lists.xenproject.org Tue Jun 08 18:23:39 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 08 Jun 2021 18:23:39 +0000 Received: from list by lists.xenproject.org with outflank-mailman.138902.256974 (Exim 4.92) (envelope-from ) id 1lqgO3-0007b5-3E; Tue, 08 Jun 2021 18:23:39 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 138902.256974; Tue, 08 Jun 2021 18:23:39 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqgO3-0007aw-0I; Tue, 08 Jun 2021 18:23:39 +0000 Received: by outflank-mailman (input) for mailman id 138902; Tue, 08 Jun 2021 18:23:37 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqgO1-0007ah-Au for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 18:23:37 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqgO1-0006OF-AD for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 18:23:37 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqgO1-0005g8-9R for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 18:23:37 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=wBaSaf3yJfzhUWWKxeUy1vF9JSFIujapVIjAv8g84uE=; b=DrDtUExj0ucjELIIV3RnTkCwtt beLWqV7VrZKTB3Fpj6wn+upaWnMZzuZTvj/K9wjmYlPlR3+UMb+XIJ1MV6BOXbuZW4E7KJYs6cOmv 7WqYRNRy4J7xJd2WbK0/+sN4lkU98AWGsQuI8BF86nE92sM55Q2zx81dLk6rSg2JCEfM=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.11] VT-d: size qinval queue dynamically Message-Id: Date: Tue, 08 Jun 2021 18:23:37 +0000 commit 37518c21cf97b2d333403f994a772cbbbe5bedfc Author: Jan Beulich AuthorDate: Tue Jun 8 19:15:06 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 19:16:38 2021 +0100 VT-d: size qinval queue dynamically With the present synchronous model, we need two slots for every operation (the operation itself and a wait descriptor). There can be one such pair of requests pending per CPU. To ensure that under all normal circumstances a slot is always available when one is requested, size the queue ring according to the number of present CPUs. This is part of XSA-373 / CVE-2021-28692. Signed-off-by: Jan Beulich Reviewed-by: Paul Durrant (cherry picked from commit cbfa62bb140c8f10a0ae96db3ce06e22b3b9d302) --- xen/drivers/passthrough/vtd/iommu.h | 10 +------- xen/drivers/passthrough/vtd/qinval.c | 44 ++++++++++++++++++++++++++++-------- 2 files changed, 35 insertions(+), 19 deletions(-) diff --git a/xen/drivers/passthrough/vtd/iommu.h b/xen/drivers/passthrough/vtd/iommu.h index e76b261c81..fc52a3249e 100644 --- a/xen/drivers/passthrough/vtd/iommu.h +++ b/xen/drivers/passthrough/vtd/iommu.h @@ -447,17 +447,9 @@ struct qinval_entry { }q; }; -/* Order of queue invalidation pages(max is 8) */ -#define QINVAL_PAGE_ORDER 2 - -#define QINVAL_ARCH_PAGE_ORDER (QINVAL_PAGE_ORDER + PAGE_SHIFT_4K - PAGE_SHIFT) -#define QINVAL_ARCH_PAGE_NR ( QINVAL_ARCH_PAGE_ORDER < 0 ? \ - 1 : \ - 1 << QINVAL_ARCH_PAGE_ORDER ) - /* Each entry is 16 bytes, so 2^8 entries per page */ #define QINVAL_ENTRY_ORDER ( PAGE_SHIFT - 4 ) -#define QINVAL_ENTRY_NR (1 << (QINVAL_PAGE_ORDER + 8)) +#define QINVAL_MAX_ENTRY_NR (1u << (7 + QINVAL_ENTRY_ORDER)) /* Status data flag */ #define QINVAL_STAT_INIT 0 diff --git a/xen/drivers/passthrough/vtd/qinval.c b/xen/drivers/passthrough/vtd/qinval.c index 09cbd36ebb..103afc15fe 100644 --- a/xen/drivers/passthrough/vtd/qinval.c +++ b/xen/drivers/passthrough/vtd/qinval.c @@ -31,6 +31,9 @@ #define VTD_QI_TIMEOUT 1 +static unsigned int __read_mostly qi_pg_order; +static unsigned int __read_mostly qi_entry_nr; + static int __must_check invalidate_sync(struct iommu *iommu); static void print_qi_regs(struct iommu *iommu) @@ -55,7 +58,7 @@ static unsigned int qinval_next_index(struct iommu *iommu) tail >>= QINVAL_INDEX_SHIFT; /* (tail+1 == head) indicates a full queue, wait for HW */ - while ( ( tail + 1 ) % QINVAL_ENTRY_NR == + while ( ((tail + 1) & (qi_entry_nr - 1)) == ( dmar_readq(iommu->reg, DMAR_IQH_REG) >> QINVAL_INDEX_SHIFT ) ) cpu_relax(); @@ -68,7 +71,7 @@ static void qinval_update_qtail(struct iommu *iommu, unsigned int index) /* Need hold register lock when update tail */ ASSERT( spin_is_locked(&iommu->register_lock) ); - val = (index + 1) % QINVAL_ENTRY_NR; + val = (index + 1) & (qi_entry_nr - 1); dmar_writeq(iommu->reg, DMAR_IQT_REG, (val << QINVAL_INDEX_SHIFT)); } @@ -417,7 +420,27 @@ int enable_qinval(struct iommu *iommu) if ( qi_ctrl->qinval_maddr == 0 ) { drhd = iommu_to_drhd(iommu); - qi_ctrl->qinval_maddr = alloc_pgtable_maddr(drhd, QINVAL_ARCH_PAGE_NR); + if ( !qi_entry_nr ) + { + /* + * With the present synchronous model, we need two slots for every + * operation (the operation itself and a wait descriptor). There + * can be one such pair of requests pending per CPU. One extra + * entry is needed as the ring is considered full when there's + * only one entry left. + */ + BUILD_BUG_ON(CONFIG_NR_CPUS * 2 >= QINVAL_MAX_ENTRY_NR); + qi_pg_order = get_order_from_bytes((num_present_cpus() * 2 + 1) << + (PAGE_SHIFT - + QINVAL_ENTRY_ORDER)); + qi_entry_nr = 1u << (qi_pg_order + QINVAL_ENTRY_ORDER); + + dprintk(XENLOG_INFO VTDPREFIX, + "QI: using %u-entry ring(s)\n", qi_entry_nr); + } + + qi_ctrl->qinval_maddr = + alloc_pgtable_maddr(drhd, qi_entry_nr >> QINVAL_ENTRY_ORDER); if ( qi_ctrl->qinval_maddr == 0 ) { dprintk(XENLOG_WARNING VTDPREFIX, @@ -431,15 +454,16 @@ int enable_qinval(struct iommu *iommu) spin_lock_irqsave(&iommu->register_lock, flags); - /* Setup Invalidation Queue Address(IQA) register with the - * address of the page we just allocated. QS field at - * bits[2:0] to indicate size of queue is one 4KB page. - * That's 256 entries. Queued Head (IQH) and Queue Tail (IQT) - * registers are automatically reset to 0 with write - * to IQA register. + /* + * Setup Invalidation Queue Address (IQA) register with the address of the + * pages we just allocated. The QS field at bits[2:0] indicates the size + * (page order) of the queue. + * + * Queued Head (IQH) and Queue Tail (IQT) registers are automatically + * reset to 0 with write to IQA register. */ dmar_writeq(iommu->reg, DMAR_IQA_REG, - qi_ctrl->qinval_maddr | QINVAL_PAGE_ORDER); + qi_ctrl->qinval_maddr | qi_pg_order); dmar_writeq(iommu->reg, DMAR_IQT_REG, 0); -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.11 From xen-changelog-bounces@lists.xenproject.org Tue Jun 08 18:23:49 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 08 Jun 2021 18:23:49 +0000 Received: from list by lists.xenproject.org with outflank-mailman.138903.256978 (Exim 4.92) (envelope-from ) id 1lqgOD-0007dm-4m; Tue, 08 Jun 2021 18:23:49 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 138903.256978; Tue, 08 Jun 2021 18:23:49 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqgOD-0007de-1k; Tue, 08 Jun 2021 18:23:49 +0000 Received: by outflank-mailman (input) for mailman id 138903; Tue, 08 Jun 2021 18:23:47 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqgOB-0007dJ-E7 for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 18:23:47 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqgOB-0006Od-DP for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 18:23:47 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqgOB-0005hF-CX for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 18:23:47 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=jftHxmn84EeMDeaAZO/bjC9JbuyGhg8nGT7iO7KoXaw=; b=j4RWzc3qmJTc3pa9Kzj6D/YT2a ImDbTw83G88+cNiM0O225xVtNo1xkCP5iqE5dt8Du1beA+20PcDNCwfHTPP5ryTIIlMJfGphdq6bl WgVDG+YXcQ/93RIdz7h4TTlHQLVEsn3tG3lD1Tx/EVV4XfighdMGcIdgNp55w4gLqi2M=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.11] AMD/IOMMU: size command buffer dynamically Message-Id: Date: Tue, 08 Jun 2021 18:23:47 +0000 commit 89d3cc62a50475a63a388487b8d4b574dfd426a5 Author: Jan Beulich AuthorDate: Tue Jun 8 19:15:18 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 19:16:38 2021 +0100 AMD/IOMMU: size command buffer dynamically With the present synchronous model, we need two slots for every operation (the operation itself and a wait command). There can be one such pair of commands pending per CPU. To ensure that under all normal circumstances a slot is always available when one is requested, size the command ring according to the number of present CPUs. This is part of XSA-373 / CVE-2021-28692. Signed-off-by: Jan Beulich Reviewed-by: Paul Durrant (cherry picked from commit df242851ddc93ac0b0a3a20ecab34acc29e3b36b) --- xen/drivers/passthrough/amd/iommu_cmd.c | 12 ++++-------- xen/drivers/passthrough/amd/iommu_init.c | 28 +++++++++++++++++++++++++--- xen/include/asm-x86/hvm/svm/amd-iommu-defs.h | 8 ++------ 3 files changed, 31 insertions(+), 17 deletions(-) diff --git a/xen/drivers/passthrough/amd/iommu_cmd.c b/xen/drivers/passthrough/amd/iommu_cmd.c index 08247fa354..2c464fcb58 100644 --- a/xen/drivers/passthrough/amd/iommu_cmd.c +++ b/xen/drivers/passthrough/amd/iommu_cmd.c @@ -24,8 +24,7 @@ static int queue_iommu_command(struct amd_iommu *iommu, u32 cmd[]) { - u32 tail, head, *cmd_buffer; - int i; + uint32_t tail, head; tail = iommu->cmd_buffer.tail; if ( ++tail == iommu->cmd_buffer.entries ) @@ -35,12 +34,9 @@ static int queue_iommu_command(struct amd_iommu *iommu, u32 cmd[]) IOMMU_CMD_BUFFER_HEAD_OFFSET)); if ( head != tail ) { - cmd_buffer = (u32 *)(iommu->cmd_buffer.buffer + - (iommu->cmd_buffer.tail * - IOMMU_CMD_BUFFER_ENTRY_SIZE)); - - for ( i = 0; i < IOMMU_CMD_BUFFER_U32_PER_ENTRY; i++ ) - cmd_buffer[i] = cmd[i]; + memcpy(iommu->cmd_buffer.buffer + + (iommu->cmd_buffer.tail * sizeof(cmd_entry_t)), + cmd, sizeof(cmd_entry_t)); iommu->cmd_buffer.tail = tail; return 1; diff --git a/xen/drivers/passthrough/amd/iommu_init.c b/xen/drivers/passthrough/amd/iommu_init.c index 7a089b4c60..cf792ef77f 100644 --- a/xen/drivers/passthrough/amd/iommu_init.c +++ b/xen/drivers/passthrough/amd/iommu_init.c @@ -136,7 +136,7 @@ static void register_iommu_cmd_buffer_in_mmio_space(struct amd_iommu *iommu) writel(entry, iommu->mmio_base + IOMMU_CMD_BUFFER_BASE_LOW_OFFSET); power_of2_entries = get_order_from_bytes(iommu->cmd_buffer.alloc_size) + - IOMMU_CMD_BUFFER_POWER_OF2_ENTRIES_PER_PAGE; + PAGE_SHIFT - IOMMU_CMD_BUFFER_ENTRY_ORDER; entry = 0; iommu_set_addr_hi_to_reg(&entry, addr_hi); @@ -1000,9 +1000,31 @@ static void * __init allocate_ring_buffer(struct ring_buffer *ring_buf, static void * __init allocate_cmd_buffer(struct amd_iommu *iommu) { /* allocate 'command buffer' in power of 2 increments of 4K */ + static unsigned int __read_mostly nr_ents; + + if ( !nr_ents ) + { + unsigned int order; + + /* + * With the present synchronous model, we need two slots for every + * operation (the operation itself and a wait command). There can be + * one such pair of requests pending per CPU. One extra entry is + * needed as the ring is considered full when there's only one entry + * left. + */ + BUILD_BUG_ON(CONFIG_NR_CPUS * 2 >= IOMMU_CMD_BUFFER_MAX_ENTRIES); + order = get_order_from_bytes((num_present_cpus() * 2 + 1) << + IOMMU_CMD_BUFFER_ENTRY_ORDER); + nr_ents = 1u << (order + PAGE_SHIFT - IOMMU_CMD_BUFFER_ENTRY_ORDER); + + AMD_IOMMU_DEBUG("using %u-entry cmd ring(s)\n", nr_ents); + } + + BUILD_BUG_ON(sizeof(cmd_entry_t) != (1u << IOMMU_CMD_BUFFER_ENTRY_ORDER)); + return allocate_ring_buffer(&iommu->cmd_buffer, sizeof(cmd_entry_t), - IOMMU_CMD_BUFFER_DEFAULT_ENTRIES, - "Command Buffer"); + nr_ents, "Command Buffer"); } static void * __init allocate_event_log(struct amd_iommu *iommu) diff --git a/xen/include/asm-x86/hvm/svm/amd-iommu-defs.h b/xen/include/asm-x86/hvm/svm/amd-iommu-defs.h index c479f0bb02..d2e285f03c 100644 --- a/xen/include/asm-x86/hvm/svm/amd-iommu-defs.h +++ b/xen/include/asm-x86/hvm/svm/amd-iommu-defs.h @@ -20,9 +20,6 @@ #ifndef _ASM_X86_64_AMD_IOMMU_DEFS_H #define _ASM_X86_64_AMD_IOMMU_DEFS_H -/* IOMMU Command Buffer entries: in power of 2 increments, minimum of 256 */ -#define IOMMU_CMD_BUFFER_DEFAULT_ENTRIES 512 - /* IOMMU Event Log entries: in power of 2 increments, minimum of 256 */ #define IOMMU_EVENT_LOG_DEFAULT_ENTRIES 512 @@ -185,9 +182,8 @@ #define IOMMU_CMD_BUFFER_LENGTH_MASK 0x0F000000 #define IOMMU_CMD_BUFFER_LENGTH_SHIFT 24 -#define IOMMU_CMD_BUFFER_ENTRY_SIZE 16 -#define IOMMU_CMD_BUFFER_POWER_OF2_ENTRIES_PER_PAGE 8 -#define IOMMU_CMD_BUFFER_U32_PER_ENTRY (IOMMU_CMD_BUFFER_ENTRY_SIZE / 4) +#define IOMMU_CMD_BUFFER_ENTRY_ORDER 4 +#define IOMMU_CMD_BUFFER_MAX_ENTRIES (1u << 15) #define IOMMU_CMD_OPCODE_MASK 0xF0000000 #define IOMMU_CMD_OPCODE_SHIFT 28 -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.11 From xen-changelog-bounces@lists.xenproject.org Tue Jun 08 18:23:59 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 08 Jun 2021 18:23:59 +0000 Received: from list by lists.xenproject.org with outflank-mailman.138904.256982 (Exim 4.92) (envelope-from ) id 1lqgON-0007gg-6i; Tue, 08 Jun 2021 18:23:59 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 138904.256982; Tue, 08 Jun 2021 18:23:59 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqgON-0007gY-3M; Tue, 08 Jun 2021 18:23:59 +0000 Received: by outflank-mailman (input) for mailman id 138904; Tue, 08 Jun 2021 18:23:57 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqgOL-0007gC-I0 for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 18:23:57 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqgOL-0006On-HB for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 18:23:57 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqgOL-0005i2-Fm for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 18:23:57 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=WHI13fusLR9OGakBnQwmz0diaQ/iWZDraG9Qa03wBmI=; b=EFuHhJvdVNHri+z65VSxSRQkeH m2wQJKgJOIDQq13VFmuseUTf1lQWsrd4dyaG+Zb3WM5PpD4myuCyLlbyF0Kjqmu4qR/1bv06TK3vR 6XbgfyrrkmMTnb7lSFqehXt0g3aOqTBBmjJaR15FcmcdXTJQayZpjIw53sB4z6KYqwlQ=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.11] VT-d: eliminate flush related timeouts Message-Id: Date: Tue, 08 Jun 2021 18:23:57 +0000 commit dc3e5441867de26b921c4e10241fb8ad1c9d3fad Author: Jan Beulich AuthorDate: Tue Jun 8 19:15:39 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 19:16:38 2021 +0100 VT-d: eliminate flush related timeouts Leaving an in-progress operation pending when it appears to take too long is problematic: If e.g. a QI command completed later, the write to the "poll slot" may instead be understood to signal a subsequently started command's completion. Also our accounting of the timeout period was actually wrong: We included the time it took for the command to actually make it to the front of the queue, which could be heavily affected by guests other than the one for which the flush is being performed. Do away with all timeout detection on all flush related code paths. Log excessively long processing times (with a progressive threshold) to have some indication of problems in this area. Additionally log (once) if qinval_next_index() didn't immediately find an available slot. Together with the earlier change sizing the queue(s) dynamically, we should now have a guarantee that with our fully synchronous model any demand for slots can actually be satisfied. This is part of XSA-373 / CVE-2021-28692. Signed-off-by: Jan Beulich Reviewed-by: Paul Durrant (cherry picked from commit 7c893abc4ee7e9af62297ba6fd5f27c89d0f33f5) --- xen/drivers/passthrough/vtd/dmar.h | 28 ++++++++++++++++++++++++++++ xen/drivers/passthrough/vtd/iommu.c | 12 ++++++------ xen/drivers/passthrough/vtd/qinval.c | 33 ++++++++++++++++++++++----------- 3 files changed, 56 insertions(+), 17 deletions(-) diff --git a/xen/drivers/passthrough/vtd/dmar.h b/xen/drivers/passthrough/vtd/dmar.h index 8f1b18ebb2..f67cba1386 100644 --- a/xen/drivers/passthrough/vtd/dmar.h +++ b/xen/drivers/passthrough/vtd/dmar.h @@ -127,6 +127,34 @@ do { \ } \ } while (0) +#define IOMMU_FLUSH_WAIT(what, iommu, offset, op, cond, sts) \ +do { \ + static unsigned int __read_mostly threshold = 1; \ + s_time_t start = NOW(); \ + s_time_t timeout = start + DMAR_OPERATION_TIMEOUT * threshold; \ + \ + for ( ; ; ) \ + { \ + sts = op(iommu->reg, offset); \ + if ( cond ) \ + break; \ + if ( timeout && NOW() > timeout ) \ + { \ + threshold |= threshold << 1; \ + printk(XENLOG_WARNING VTDPREFIX \ + " IOMMU#%u: %s flush taking too long\n", \ + iommu->index, what); \ + timeout = 0; \ + } \ + cpu_relax(); \ + } \ + \ + if ( !timeout ) \ + printk(XENLOG_WARNING VTDPREFIX \ + " IOMMU#%u: %s flush took %lums\n", \ + iommu->index, what, (NOW() - start) / 10000000); \ +} while ( false ) + int vtd_hw_check(void); void disable_pmr(struct iommu *iommu); int is_igd_drhd(struct acpi_drhd_unit *drhd); diff --git a/xen/drivers/passthrough/vtd/iommu.c b/xen/drivers/passthrough/vtd/iommu.c index f499c6f430..698b176a81 100644 --- a/xen/drivers/passthrough/vtd/iommu.c +++ b/xen/drivers/passthrough/vtd/iommu.c @@ -357,8 +357,8 @@ static void iommu_flush_write_buffer(struct iommu *iommu) dmar_writel(iommu->reg, DMAR_GCMD_REG, val | DMA_GCMD_WBF); /* Make sure hardware complete it */ - IOMMU_WAIT_OP(iommu, DMAR_GSTS_REG, dmar_readl, - !(val & DMA_GSTS_WBFS), val); + IOMMU_FLUSH_WAIT("write buffer", iommu, DMAR_GSTS_REG, dmar_readl, + !(val & DMA_GSTS_WBFS), val); spin_unlock_irqrestore(&iommu->register_lock, flags); } @@ -408,8 +408,8 @@ static int __must_check flush_context_reg(void *_iommu, u16 did, u16 source_id, dmar_writeq(iommu->reg, DMAR_CCMD_REG, val); /* Make sure hardware complete it */ - IOMMU_WAIT_OP(iommu, DMAR_CCMD_REG, dmar_readq, - !(val & DMA_CCMD_ICC), val); + IOMMU_FLUSH_WAIT("context", iommu, DMAR_CCMD_REG, dmar_readq, + !(val & DMA_CCMD_ICC), val); spin_unlock_irqrestore(&iommu->register_lock, flags); /* flush context entry will implicitly flush write buffer */ @@ -491,8 +491,8 @@ static int __must_check flush_iotlb_reg(void *_iommu, u16 did, u64 addr, dmar_writeq(iommu->reg, tlb_offset + 8, val); /* Make sure hardware complete it */ - IOMMU_WAIT_OP(iommu, (tlb_offset + 8), dmar_readq, - !(val & DMA_TLB_IVT), val); + IOMMU_FLUSH_WAIT("iotlb", iommu, (tlb_offset + 8), dmar_readq, + !(val & DMA_TLB_IVT), val); spin_unlock_irqrestore(&iommu->register_lock, flags); /* check IOTLB invalidation granularity */ diff --git a/xen/drivers/passthrough/vtd/qinval.c b/xen/drivers/passthrough/vtd/qinval.c index 103afc15fe..17c4235597 100644 --- a/xen/drivers/passthrough/vtd/qinval.c +++ b/xen/drivers/passthrough/vtd/qinval.c @@ -29,8 +29,6 @@ #include "extern.h" #include "../ats.h" -#define VTD_QI_TIMEOUT 1 - static unsigned int __read_mostly qi_pg_order; static unsigned int __read_mostly qi_entry_nr; @@ -60,7 +58,11 @@ static unsigned int qinval_next_index(struct iommu *iommu) /* (tail+1 == head) indicates a full queue, wait for HW */ while ( ((tail + 1) & (qi_entry_nr - 1)) == ( dmar_readq(iommu->reg, DMAR_IQH_REG) >> QINVAL_INDEX_SHIFT ) ) + { + printk_once(XENLOG_ERR VTDPREFIX " IOMMU#%u: no QI slot available\n", + iommu->index); cpu_relax(); + } return tail; } @@ -180,23 +182,32 @@ static int __must_check queue_invalidate_wait(struct iommu *iommu, /* Now we don't support interrupt method */ if ( sw ) { - s_time_t timeout; - - /* In case all wait descriptor writes to same addr with same data */ - timeout = NOW() + MILLISECS(flush_dev_iotlb ? - iommu_dev_iotlb_timeout : VTD_QI_TIMEOUT); + static unsigned int __read_mostly threshold = 1; + s_time_t start = NOW(); + s_time_t timeout = start + (flush_dev_iotlb + ? iommu_dev_iotlb_timeout + : 100) * MILLISECS(threshold); while ( ACCESS_ONCE(*this_poll_slot) != QINVAL_STAT_DONE ) { - if ( NOW() > timeout ) + if ( timeout && NOW() > timeout ) { - print_qi_regs(iommu); + threshold |= threshold << 1; printk(XENLOG_WARNING VTDPREFIX - " Queue invalidate wait descriptor timed out\n"); - return -ETIMEDOUT; + " IOMMU#%u: QI%s wait descriptor taking too long\n", + iommu->index, flush_dev_iotlb ? " dev" : ""); + print_qi_regs(iommu); + timeout = 0; } cpu_relax(); } + + if ( !timeout ) + printk(XENLOG_WARNING VTDPREFIX + " IOMMU#%u: QI%s wait descriptor took %lums\n", + iommu->index, flush_dev_iotlb ? " dev" : "", + (NOW() - start) / 10000000); + return 0; } -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.11 From xen-changelog-bounces@lists.xenproject.org Tue Jun 08 18:24:09 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 08 Jun 2021 18:24:09 +0000 Received: from list by lists.xenproject.org with outflank-mailman.138905.256986 (Exim 4.92) (envelope-from ) id 1lqgOX-0007m6-AW; Tue, 08 Jun 2021 18:24:09 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 138905.256986; Tue, 08 Jun 2021 18:24:09 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqgOX-0007lx-73; Tue, 08 Jun 2021 18:24:09 +0000 Received: by outflank-mailman (input) for mailman id 138905; Tue, 08 Jun 2021 18:24:07 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqgOV-0007lj-Kf for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 18:24:07 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqgOV-0006PB-K1 for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 18:24:07 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqgOV-0005in-JD for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 18:24:07 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=RYVKSa9X1oqEkHiwW7cIks4unOOY59VzjgWI0nj38cE=; b=0XVPqC/8ptpmITKlKiSztIJcaz moB/8BS1P1lQ+hY2JdCXWnwWLFqGNwE9VPE3g9XxKAytqHONs0Hdbzv5NtORJYXp0qC4dUxrpOUHd vvXgT+oOIrAmyJrwUHGgB426TpOwBtuhz+i650nHWeUOmPY6aWBOPjee/48yPlFZeNdM=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.11] AMD/IOMMU: wait for command slot to be available Message-Id: Date: Tue, 08 Jun 2021 18:24:07 +0000 commit 9f8bf2a7ddab2a5c874072cfa430ffae9772de1d Author: Jan Beulich AuthorDate: Tue Jun 8 19:15:57 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 19:16:38 2021 +0100 AMD/IOMMU: wait for command slot to be available No caller cared about send_iommu_command() indicating unavailability of a slot. Hence if a sufficient number prior commands timed out, we did blindly assume that the requested command was submitted to the IOMMU when really it wasn't. This could mean both a hanging system (waiting for a command to complete that was never seen by the IOMMU) or blindly propagating success back to callers, making them believe they're fine to e.g. free previously unmapped pages. Fold the three involved functions into one, add spin waiting for an available slot along the lines of VT-d's qinval_next_index(), and as a consequence drop all error indicator return types/values. This is part of XSA-373 / CVE-2021-28692. Signed-off-by: Jan Beulich Reviewed-by: Paul Durrant (cherry picked from commit c4e44343579e2c3304d676404d81b2b9bd893c5b) --- xen/drivers/passthrough/amd/iommu_cmd.c | 42 ++++++++++++--------------------- 1 file changed, 15 insertions(+), 27 deletions(-) diff --git a/xen/drivers/passthrough/amd/iommu_cmd.c b/xen/drivers/passthrough/amd/iommu_cmd.c index 2c464fcb58..6625ee8518 100644 --- a/xen/drivers/passthrough/amd/iommu_cmd.c +++ b/xen/drivers/passthrough/amd/iommu_cmd.c @@ -22,48 +22,36 @@ #include #include "../ats.h" -static int queue_iommu_command(struct amd_iommu *iommu, u32 cmd[]) +static void send_iommu_command(struct amd_iommu *iommu, + const uint32_t cmd[4]) { - uint32_t tail, head; + uint32_t tail; tail = iommu->cmd_buffer.tail; if ( ++tail == iommu->cmd_buffer.entries ) tail = 0; - head = iommu_get_rb_pointer(readl(iommu->mmio_base + - IOMMU_CMD_BUFFER_HEAD_OFFSET)); - if ( head != tail ) + while ( tail == iommu_get_rb_pointer(readl(iommu->mmio_base + + IOMMU_CMD_BUFFER_HEAD_OFFSET)) ) { - memcpy(iommu->cmd_buffer.buffer + - (iommu->cmd_buffer.tail * sizeof(cmd_entry_t)), - cmd, sizeof(cmd_entry_t)); - - iommu->cmd_buffer.tail = tail; - return 1; + printk_once(XENLOG_ERR + "AMD IOMMU %04x:%02x:%02x.%u: no cmd slot available\n", + iommu->seg, PCI_BUS(iommu->bdf), + PCI_SLOT(iommu->bdf), PCI_FUNC(iommu->bdf)); + cpu_relax(); } - return 0; -} + memcpy(iommu->cmd_buffer.buffer + + (iommu->cmd_buffer.tail * sizeof(cmd_entry_t)), + cmd, sizeof(cmd_entry_t)); -static void commit_iommu_command_buffer(struct amd_iommu *iommu) -{ - u32 tail = 0; + iommu->cmd_buffer.tail = tail; + tail = 0; iommu_set_rb_pointer(&tail, iommu->cmd_buffer.tail); writel(tail, iommu->mmio_base+IOMMU_CMD_BUFFER_TAIL_OFFSET); } -int send_iommu_command(struct amd_iommu *iommu, u32 cmd[]) -{ - if ( queue_iommu_command(iommu, cmd) ) - { - commit_iommu_command_buffer(iommu); - return 1; - } - - return 0; -} - static void flush_command_buffer(struct amd_iommu *iommu) { u32 cmd[4], status; -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.11 From xen-changelog-bounces@lists.xenproject.org Tue Jun 08 18:24:19 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 08 Jun 2021 18:24:19 +0000 Received: from list by lists.xenproject.org with outflank-mailman.138906.256990 (Exim 4.92) (envelope-from ) id 1lqgOh-0007ov-Bt; Tue, 08 Jun 2021 18:24:19 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 138906.256990; Tue, 08 Jun 2021 18:24:19 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqgOh-0007on-8l; Tue, 08 Jun 2021 18:24:19 +0000 Received: by outflank-mailman (input) for mailman id 138906; Tue, 08 Jun 2021 18:24:17 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqgOf-0007oZ-OV for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 18:24:17 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqgOf-0006PO-Ni for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 18:24:17 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqgOf-0005k7-Ma for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 18:24:17 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=/+Wcuh3IumyyewQkARWSlEkVm8vPFxeZQB1GmZxQyCM=; b=IcU8lThqTWxzu+ruyRAl0ws2UW 2c1zx/GERlgNMn+raz5qEZ4NVp7TVAhl6xolzX+Gg3DetJKoxPRFOGSR4sFk+myC2w0/vViZFT3Z7 BJbMsHPVngi1VaigabqeluCTyx8A/vktk8O94cePF8SCP4OVEKrWuYsh9eHRyLS1HaRA=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.11] AMD/IOMMU: drop command completion timeout Message-Id: Date: Tue, 08 Jun 2021 18:24:17 +0000 commit 944da2f5b8530433aeaa46493db48f3ddb0aa302 Author: Jan Beulich AuthorDate: Tue Jun 8 19:16:07 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 19:16:38 2021 +0100 AMD/IOMMU: drop command completion timeout First and foremost - such timeouts were not signaled to callers, making them believe they're fine to e.g. free previously unmapped pages. Mirror VT-d's behavior: A fixed number of loop iterations is not a suitable way to detect timeouts in an environment (CPU and bus speeds) independent manner anyway. Furthermore, leaving an in-progress operation pending when it appears to take too long is problematic: If a command completed later, the signaling of its completion may instead be understood to signal a subsequently started command's completion. Log excessively long processing times (with a progressive threshold) to have some indication of problems in this area. Allow callers to specify a non-default timeout bias for this logging, using the same values as VT-d does, which in particular means a (by default) much larger value for device IO TLB invalidation. This is part of XSA-373 / CVE-2021-28692. Signed-off-by: Jan Beulich Reviewed-by: Paul Durrant (cherry picked from commit e4fee66043120c954fc309bbb37813604c1c0eb7) --- xen/drivers/passthrough/amd/iommu_cmd.c | 60 +++++++++++++++++++-------------- 1 file changed, 35 insertions(+), 25 deletions(-) diff --git a/xen/drivers/passthrough/amd/iommu_cmd.c b/xen/drivers/passthrough/amd/iommu_cmd.c index 6625ee8518..ebdd0f5529 100644 --- a/xen/drivers/passthrough/amd/iommu_cmd.c +++ b/xen/drivers/passthrough/amd/iommu_cmd.c @@ -52,10 +52,12 @@ static void send_iommu_command(struct amd_iommu *iommu, writel(tail, iommu->mmio_base+IOMMU_CMD_BUFFER_TAIL_OFFSET); } -static void flush_command_buffer(struct amd_iommu *iommu) +static void flush_command_buffer(struct amd_iommu *iommu, + unsigned int timeout_base) { - u32 cmd[4], status; - int loop_count, comp_wait; + uint32_t cmd[4]; + s_time_t start, timeout; + static unsigned int __read_mostly threshold = 1; /* RW1C 'ComWaitInt' in status register */ writel(IOMMU_STATUS_COMP_WAIT_INT_MASK, @@ -71,24 +73,31 @@ static void flush_command_buffer(struct amd_iommu *iommu) IOMMU_COMP_WAIT_I_FLAG_SHIFT, &cmd[0]); send_iommu_command(iommu, cmd); - /* Make loop_count long enough for polling completion wait bit */ - loop_count = 1000; - do { - status = readl(iommu->mmio_base + IOMMU_STATUS_MMIO_OFFSET); - comp_wait = get_field_from_reg_u32(status, - IOMMU_STATUS_COMP_WAIT_INT_MASK, - IOMMU_STATUS_COMP_WAIT_INT_SHIFT); - --loop_count; - } while ( !comp_wait && loop_count ); - - if ( comp_wait ) + start = NOW(); + timeout = start + (timeout_base ?: 100) * MILLISECS(threshold); + while ( !(readl(iommu->mmio_base + IOMMU_STATUS_MMIO_OFFSET) & + IOMMU_STATUS_COMP_WAIT_INT_MASK) ) { - /* RW1C 'ComWaitInt' in status register */ - writel(IOMMU_STATUS_COMP_WAIT_INT_MASK, - iommu->mmio_base + IOMMU_STATUS_MMIO_OFFSET); - return; + if ( timeout && NOW() > timeout ) + { + threshold |= threshold << 1; + printk(XENLOG_WARNING + "AMD IOMMU %04x:%02x:%02x.%u: %scompletion wait taking too long\n", + iommu->seg, PCI_BUS(iommu->bdf), + PCI_SLOT(iommu->bdf), PCI_FUNC(iommu->bdf), + timeout_base ? "iotlb " : ""); + timeout = 0; + } + cpu_relax(); } - AMD_IOMMU_DEBUG("Warning: ComWaitInt bit did not assert!\n"); + + if ( !timeout ) + printk(XENLOG_WARNING + "AMD IOMMU %04x:%02x:%02x.%u: %scompletion wait took %lums\n", + iommu->seg, PCI_BUS(iommu->bdf), + PCI_SLOT(iommu->bdf), PCI_FUNC(iommu->bdf), + timeout_base ? "iotlb " : "", + (NOW() - start) / 10000000); } /* Build low level iommu command messages */ @@ -300,7 +309,7 @@ void amd_iommu_flush_iotlb(u8 devfn, const struct pci_dev *pdev, /* send INVALIDATE_IOTLB_PAGES command */ spin_lock_irqsave(&iommu->lock, flags); invalidate_iotlb_pages(iommu, maxpend, 0, queueid, gaddr, req_id, order); - flush_command_buffer(iommu); + flush_command_buffer(iommu, iommu_dev_iotlb_timeout); spin_unlock_irqrestore(&iommu->lock, flags); } @@ -337,7 +346,7 @@ static void _amd_iommu_flush_pages(struct domain *d, { spin_lock_irqsave(&iommu->lock, flags); invalidate_iommu_pages(iommu, gaddr, dom_id, order); - flush_command_buffer(iommu); + flush_command_buffer(iommu, 0); spin_unlock_irqrestore(&iommu->lock, flags); } @@ -361,7 +370,7 @@ void amd_iommu_flush_device(struct amd_iommu *iommu, uint16_t bdf) ASSERT( spin_is_locked(&iommu->lock) ); invalidate_dev_table_entry(iommu, bdf); - flush_command_buffer(iommu); + flush_command_buffer(iommu, 0); } void amd_iommu_flush_intremap(struct amd_iommu *iommu, uint16_t bdf) @@ -369,7 +378,7 @@ void amd_iommu_flush_intremap(struct amd_iommu *iommu, uint16_t bdf) ASSERT( spin_is_locked(&iommu->lock) ); invalidate_interrupt_table(iommu, bdf); - flush_command_buffer(iommu); + flush_command_buffer(iommu, 0); } void amd_iommu_flush_all_caches(struct amd_iommu *iommu) @@ -377,7 +386,7 @@ void amd_iommu_flush_all_caches(struct amd_iommu *iommu) ASSERT( spin_is_locked(&iommu->lock) ); invalidate_iommu_all(iommu); - flush_command_buffer(iommu); + flush_command_buffer(iommu, 0); } void amd_iommu_send_guest_cmd(struct amd_iommu *iommu, u32 cmd[]) @@ -387,7 +396,8 @@ void amd_iommu_send_guest_cmd(struct amd_iommu *iommu, u32 cmd[]) spin_lock_irqsave(&iommu->lock, flags); send_iommu_command(iommu, cmd); - flush_command_buffer(iommu); + /* TBD: Timeout selection may require peeking into cmd[]. */ + flush_command_buffer(iommu, 0); spin_unlock_irqrestore(&iommu->lock, flags); } -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.11 From xen-changelog-bounces@lists.xenproject.org Tue Jun 08 18:24:28 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 08 Jun 2021 18:24:28 +0000 Received: from list by lists.xenproject.org with outflank-mailman.138907.256996 (Exim 4.92) (envelope-from ) id 1lqgOq-0007rk-E2; Tue, 08 Jun 2021 18:24:28 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 138907.256996; Tue, 08 Jun 2021 18:24:28 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqgOq-0007rb-AM; Tue, 08 Jun 2021 18:24:28 +0000 Received: by outflank-mailman (input) for mailman id 138907; Tue, 08 Jun 2021 18:24:27 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqgOp-0007rM-Ro for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 18:24:27 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqgOp-0006PZ-RB for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 18:24:27 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqgOp-0005kq-QQ for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 18:24:27 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=aedNczdzhnYGDOxzfj9/51/54W/1LUmuZKneY5qy3jA=; b=VP4+x6+4qnceFYdc+UomuzExlL ZwZYNt+M3BdQv7K3cgldS4x6YhL4ox9oeClucUlHAf4sIrNgPiBZdZ0w/azGkQ3p6SuTJjICxDdq2 hBUN0sLaYkjMO8jBkdRUB2OYOpaeA2glrhyd8zxQ3u5YodeKz+6sSzaqsKsHQUd0/A2A=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.11] x86/spec-ctrl: Protect against Speculative Code Store Bypass Message-Id: Date: Tue, 08 Jun 2021 18:24:27 +0000 commit 7937627baf95e3f142639df7cb750658ca1871f3 Author: Andrew Cooper AuthorDate: Thu Mar 11 14:39:11 2021 +0000 Commit: Andrew Cooper CommitDate: Tue Jun 8 19:16:38 2021 +0100 x86/spec-ctrl: Protect against Speculative Code Store Bypass Modern x86 processors have far-better-than-architecturally-guaranteed self modifying code detection. Typically, when a write hits an instruction in flight, a Machine Clear occurs to flush stale content in the frontend and backend. For self modifying code, before a write which hits an instruction in flight retires, the frontend can speculatively decode and execute the old instruction stream. Speculation of this form can suffer from type confusion in registers, and potentially leak data. Furthermore, updates are typically byte-wise, rather than atomic. Depending on timing, speculation can race ahead multiple times between individual writes, and execute the transiently-malformed instruction stream. Xen has stubs which are used in certain cases for emulation purposes. Inhibit speculation between updating the stub and executing it. This is XSA-375 / CVE-2021-0089. Signed-off-by: Andrew Cooper Reviewed-by: Jan Beulich (cherry picked from commit 45f59ed8865318bb0356954bad067f329677ce9e) --- xen/arch/x86/pv/emul-priv-op.c | 2 ++ xen/arch/x86/x86_emulate/x86_emulate.c | 1 + 2 files changed, 3 insertions(+) diff --git a/xen/arch/x86/pv/emul-priv-op.c b/xen/arch/x86/pv/emul-priv-op.c index 755f00db33..fdba01ce96 100644 --- a/xen/arch/x86/pv/emul-priv-op.c +++ b/xen/arch/x86/pv/emul-priv-op.c @@ -101,6 +101,8 @@ static io_emul_stub_t *io_emul_stub_setup(struct priv_op_ctxt *ctxt, u8 opcode, BUILD_BUG_ON(STUB_BUF_SIZE / 2 < MAX(9, /* Default emul stub */ 5 + IOEMUL_QUIRK_STUB_BYTES)); + asm volatile ( "lfence" ::: "memory" ); /* SCSB */ + /* Handy function-typed pointer to the stub. */ return (void *)stub_va; } diff --git a/xen/arch/x86/x86_emulate/x86_emulate.c b/xen/arch/x86/x86_emulate/x86_emulate.c index 1f0c208fc5..19594ab45d 100644 --- a/xen/arch/x86/x86_emulate/x86_emulate.c +++ b/xen/arch/x86/x86_emulate/x86_emulate.c @@ -1008,6 +1008,7 @@ static inline int mkec(uint8_t e, int32_t ec, ...) # define invoke_stub(pre, post, constraints...) do { \ stub_exn.info = (union stub_exception_token) { .raw = ~0 }; \ stub_exn.line = __LINE__; /* Utility outweighs livepatching cost */ \ + asm volatile ( "lfence" ::: "memory" ); /* SCSB */ \ asm volatile ( pre "\n\tINDIRECT_CALL %[stub]\n\t" post "\n" \ ".Lret%=:\n\t" \ ".pushsection .fixup,\"ax\"\n" \ -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.11 From xen-changelog-bounces@lists.xenproject.org Tue Jun 08 18:24:38 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 08 Jun 2021 18:24:38 +0000 Received: from list by lists.xenproject.org with outflank-mailman.138908.256998 (Exim 4.92) (envelope-from ) id 1lqgP0-0007uv-Ey; Tue, 08 Jun 2021 18:24:38 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 138908.256998; Tue, 08 Jun 2021 18:24:38 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqgP0-0007un-Bw; Tue, 08 Jun 2021 18:24:38 +0000 Received: by outflank-mailman (input) for mailman id 138908; Tue, 08 Jun 2021 18:24:38 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqgOz-0007uh-Uc for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 18:24:37 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqgOz-0006Pl-Tv for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 18:24:37 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqgOz-0005lc-TG for xen-changelog@lists.xenproject.org; Tue, 08 Jun 2021 18:24:37 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=aWnw0wSMQdsqKFtHJeaWQsIwwukiB/NOlfYmnzo4VbQ=; b=ERUKl3wj8GuHWKcLABUI95f+5p jnxgkNSIokfw1ZYEKVVZ1CWxZ93XyTzf+kulceKc7xt6DNFONt5o8L3AVzazdu+zZzOjsUukDQVnf ZKcC3hhaYGxJK5UQrEidk7MzAaaplXyRfxYy+sI9WrQq0J64iawsyYeHr1aHDeM+d/B8=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.11] x86/spec-ctrl: Mitigate TAA after S3 resume Message-Id: Date: Tue, 08 Jun 2021 18:24:37 +0000 commit ef32c7afa2731b758226d6e10a1e489b1a15fc41 Author: Andrew Cooper AuthorDate: Thu May 20 01:21:39 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 19:16:38 2021 +0100 x86/spec-ctrl: Mitigate TAA after S3 resume The user chosen setting for MSR_TSX_CTRL needs restoring after S3. All APs get the correct setting via start_secondary(), but the BSP was missed out. This is XSA-377 / CVE-2021-28690. Fixes: 8c4330818f6 ("x86/spec-ctrl: Mitigate the TSX Asynchronous Abort sidechannel") Signed-off-by: Andrew Cooper Reviewed-by: Jan Beulich (cherry picked from commit 8cf276cb2e0b99b96333865873f56b0b31555ff1) --- xen/arch/x86/acpi/power.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/xen/arch/x86/acpi/power.c b/xen/arch/x86/acpi/power.c index 30e1bd5cd3..451cba622c 100644 --- a/xen/arch/x86/acpi/power.c +++ b/xen/arch/x86/acpi/power.c @@ -259,6 +259,8 @@ static int enter_state(u32 state) microcode_resume_cpu(0); + tsx_init(); /* Needs microcode. May change HLE/RTM feature bits. */ + if ( !recheck_cpu_features(0) ) panic("Missing previously available feature(s)."); -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.11 From xen-changelog-bounces@lists.xenproject.org Wed Jun 09 02:22:09 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Wed, 09 Jun 2021 02:22:09 +0000 Received: from list by lists.xenproject.org with outflank-mailman.139013.257128 (Exim 4.92) (envelope-from ) id 1lqnr1-0001YL-Sc; Wed, 09 Jun 2021 02:22:03 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 139013.257128; Wed, 09 Jun 2021 02:22:03 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqnr1-0001YC-PB; Wed, 09 Jun 2021 02:22:03 +0000 Received: by outflank-mailman (input) for mailman id 139013; Wed, 09 Jun 2021 02:22:02 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqnr0-0001Y6-Nm for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 02:22:02 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqnr0-0003tK-N1 for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 02:22:02 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqnr0-0005KF-Lw for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 02:22:02 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=9GvvUhyala6vqZlcIPtbTgTC5PGoBuQgTWLcgN7IJFM=; b=EsMbOxJRey5FvHJEUfyLiMnqZT 7CMj4NJCB+/drJ9QePdtgKf0OH92oAeL3tzLjB/Bg7a8xXO/iTdhIIrPdR5ix8qUTYxjnA9p70gO9 z+Sg1u92ZQqmKWzojdfbIw6HZ1wx1l0W6BgSlaloAQBOW6LOlb2nHGxZ8XgQKQgFyidE=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.15] xen/arm: Create dom0less domUs earlier Message-Id: Date: Wed, 09 Jun 2021 02:22:02 +0000 commit 0a64b182db36b3b1c6f11c3b41d9b27fbbb8a5c6 Author: Julien Grall AuthorDate: Mon May 17 17:47:13 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 17:52:45 2021 +0100 xen/arm: Create dom0less domUs earlier In a follow-up patch we will need to unallocate the boot modules before heap_init_late() is called. The modules will contain the domUs kernel and initramfs. Therefore Xen will need to create extra domUs (used by dom0less) before heap_init_late(). This has two consequences on dom0less: 1) Domains will not be unpaused as soon as they are created but once all have been created. However, Xen doesn't guarantee an order to unpause, so this is not something one could rely on. 2) The memory allocated for a domU will not be scrubbed anymore when an admin select bootscrub=on. This is not something we advertised, but if this is a concern we can introduce either force scrub for all domUs or a per-domain flag in the DT. The behavior for bootscrub=off and bootscrub=idle (default) has not changed. This is part of XSA-372 / CVE-2021-28693. Signed-off-by: Julien Grall Reviewed-by: Jan Beulich Reviewed-by: Stefano Stabellini Tested-by: Stefano Stabellini (cherry picked from commit 371347c5b64da699d9f5a0edda5dc496fd2b7a5c) --- xen/arch/arm/domain_build.c | 2 -- xen/arch/arm/setup.c | 11 ++++++----- 2 files changed, 6 insertions(+), 7 deletions(-) diff --git a/xen/arch/arm/domain_build.c b/xen/arch/arm/domain_build.c index 374bf655ee..4203ddcca0 100644 --- a/xen/arch/arm/domain_build.c +++ b/xen/arch/arm/domain_build.c @@ -2515,8 +2515,6 @@ void __init create_domUs(void) if ( construct_domU(d, node) != 0 ) panic("Could not set up domain %s\n", dt_node_name(node)); - - domain_unpause_by_systemcontroller(d); } } diff --git a/xen/arch/arm/setup.c b/xen/arch/arm/setup.c index 2532ec9739..441e0e16e9 100644 --- a/xen/arch/arm/setup.c +++ b/xen/arch/arm/setup.c @@ -804,7 +804,7 @@ void __init start_xen(unsigned long boot_phys_offset, int cpus, i; const char *cmdline; struct bootmodule *xen_bootmodule; - struct domain *dom0; + struct domain *dom0, *d; struct xen_domctl_createdomain dom0_cfg = { .flags = XEN_DOMCTL_CDF_hvm | XEN_DOMCTL_CDF_hap, .max_evtchn_port = -1, @@ -987,6 +987,9 @@ void __init start_xen(unsigned long boot_phys_offset, if ( construct_dom0(dom0) != 0) panic("Could not set up DOM0 guest OS\n"); + if ( acpi_disabled ) + create_domUs(); + heap_init_late(); init_trace_bufs(); @@ -1000,10 +1003,8 @@ void __init start_xen(unsigned long boot_phys_offset, system_state = SYS_STATE_active; - if ( acpi_disabled ) - create_domUs(); - - domain_unpause_by_systemcontroller(dom0); + for_each_domain( d ) + domain_unpause_by_systemcontroller(d); /* Switch on to the dynamically allocated stack for the idle vcpu * since the static one we're running on is about to be freed. */ -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.15 From xen-changelog-bounces@lists.xenproject.org Wed Jun 09 02:22:13 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Wed, 09 Jun 2021 02:22:13 +0000 Received: from list by lists.xenproject.org with outflank-mailman.139014.257132 (Exim 4.92) (envelope-from ) id 1lqnrB-0001ab-Uc; Wed, 09 Jun 2021 02:22:13 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 139014.257132; Wed, 09 Jun 2021 02:22:13 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqnrB-0001aQ-Qo; Wed, 09 Jun 2021 02:22:13 +0000 Received: by outflank-mailman (input) for mailman id 139014; Wed, 09 Jun 2021 02:22:12 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqnrA-0001aG-Rc for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 02:22:12 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqnrA-0003tQ-Qv for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 02:22:12 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqnrA-0005Lo-PT for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 02:22:12 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=PcpG8vwXgyqe/Yq8SmdvJUc95bGyDPvHJc3YLOKh+lc=; b=FwFR7fo2BCyBJS7659wSBBMYfd Kkr1GZMBCH5hBfh4q/hvDYdIY0h8nybyIvrbcTHC6CHC2RpGv579RzbvIfX4GFg9DxBFhpXFIbvne 1hK0Ey3f2ZBVOq7geNVS8Bgm/ahEnJ7ujAOgMawel7znl85RWZ0tWhvyj35byE/YhS/g=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.15] xen/arm: Boot modules should always be scrubbed if bootscrub={on, idle} Message-Id: Date: Wed, 09 Jun 2021 02:22:12 +0000 commit 70441846070ca5540e78b4a2d73a01458539ae90 Author: Julien Grall AuthorDate: Sat Apr 17 17:38:28 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 17:52:45 2021 +0100 xen/arm: Boot modules should always be scrubbed if bootscrub={on, idle} The function to initialize the pages (see init_heap_pages()) will request scrub when the admin request idle bootscrub (default) and state == SYS_STATE_active. When bootscrub=on, Xen will scrub any free pages in heap_init_late(). Currently, the boot modules (e.g. kernels, initramfs) will be discarded/ freed after heap_init_late() is called and system_state switched to SYS_STATE_active. This means the pages associated with the boot modules will not get scrubbed before getting re-purposed. If the memory is assigned to an untrusted domU, it may be able to retrieve secrets from the modules. This is part of XSA-372 / CVE-2021-28693. Fixes: 1774e9b1df27 ("xen/arm: introduce create_domUs") Signed-off-by: Julien Grall Reviewed-by: Jan Beulich Reviewed-by: Stefano Stabellini Tested-by: Stefano Stabellini (cherry picked from commit fd5dc41ceaed9cfcfa011cdfd50f264c89277a90) --- xen/arch/arm/setup.c | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-) diff --git a/xen/arch/arm/setup.c b/xen/arch/arm/setup.c index 441e0e16e9..8afb78f2c9 100644 --- a/xen/arch/arm/setup.c +++ b/xen/arch/arm/setup.c @@ -72,8 +72,6 @@ domid_t __read_mostly max_init_domid; static __used void init_done(void) { - discard_initial_modules(); - /* Must be done past setting system_state. */ unregister_init_virtual_region(); @@ -990,6 +988,12 @@ void __init start_xen(unsigned long boot_phys_offset, if ( acpi_disabled ) create_domUs(); + /* + * This needs to be called **before** heap_init_late() so modules + * will be scrubbed (unless suppressed). + */ + discard_initial_modules(); + heap_init_late(); init_trace_bufs(); -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.15 From xen-changelog-bounces@lists.xenproject.org Wed Jun 09 02:22:25 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Wed, 09 Jun 2021 02:22:25 +0000 Received: from list by lists.xenproject.org with outflank-mailman.139015.257136 (Exim 4.92) (envelope-from ) id 1lqnrN-0001ei-0v; Wed, 09 Jun 2021 02:22:25 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 139015.257136; Wed, 09 Jun 2021 02:22:24 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqnrM-0001ea-UC; Wed, 09 Jun 2021 02:22:24 +0000 Received: by outflank-mailman (input) for mailman id 139015; Wed, 09 Jun 2021 02:22:23 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqnrK-0001dx-Un for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 02:22:22 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqnrK-0003tc-U4 for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 02:22:22 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqnrK-0005NN-T1 for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 02:22:22 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=P2+mPQpcQQOMD5Dn9QiRVe6H8tZMFjg9NUAmnS80WWA=; b=mkQFjrvpkvkNUzYUSZa7ewjkEq rYxQQOj9k33Gvs6IFm5fDP0DU+Fww+s7SYI44N/I2dKre4u/lQzHWKUF/pjIUIKZVuli3+mnZAPjS L6GzpExnY6+bf7tv4DSBeG8YmHzio580Wl94VXfIZZYKV3Hx2h2mrnt3BhK1Mjctw7V0=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.15] VT-d: size qinval queue dynamically Message-Id: Date: Wed, 09 Jun 2021 02:22:22 +0000 commit 8368f214f9446c3b1abc0045d01a2bc05926215a Author: Jan Beulich AuthorDate: Tue Jun 8 17:50:42 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 17:52:45 2021 +0100 VT-d: size qinval queue dynamically With the present synchronous model, we need two slots for every operation (the operation itself and a wait descriptor). There can be one such pair of requests pending per CPU. To ensure that under all normal circumstances a slot is always available when one is requested, size the queue ring according to the number of present CPUs. This is part of XSA-373 / CVE-2021-28692. Signed-off-by: Jan Beulich Reviewed-by: Paul Durrant (cherry picked from commit cbfa62bb140c8f10a0ae96db3ce06e22b3b9d302) --- xen/drivers/passthrough/vtd/iommu.h | 10 +------- xen/drivers/passthrough/vtd/qinval.c | 46 +++++++++++++++++++++++++++--------- 2 files changed, 36 insertions(+), 20 deletions(-) diff --git a/xen/drivers/passthrough/vtd/iommu.h b/xen/drivers/passthrough/vtd/iommu.h index 57737b1a4a..e002dc6f01 100644 --- a/xen/drivers/passthrough/vtd/iommu.h +++ b/xen/drivers/passthrough/vtd/iommu.h @@ -450,17 +450,9 @@ struct qinval_entry { }q; }; -/* Order of queue invalidation pages(max is 8) */ -#define QINVAL_PAGE_ORDER 2 - -#define QINVAL_ARCH_PAGE_ORDER (QINVAL_PAGE_ORDER + PAGE_SHIFT_4K - PAGE_SHIFT) -#define QINVAL_ARCH_PAGE_NR ( QINVAL_ARCH_PAGE_ORDER < 0 ? \ - 1 : \ - 1 << QINVAL_ARCH_PAGE_ORDER ) - /* Each entry is 16 bytes, so 2^8 entries per page */ #define QINVAL_ENTRY_ORDER ( PAGE_SHIFT - 4 ) -#define QINVAL_ENTRY_NR (1 << (QINVAL_PAGE_ORDER + 8)) +#define QINVAL_MAX_ENTRY_NR (1u << (7 + QINVAL_ENTRY_ORDER)) /* Status data flag */ #define QINVAL_STAT_INIT 0 diff --git a/xen/drivers/passthrough/vtd/qinval.c b/xen/drivers/passthrough/vtd/qinval.c index dfcbb66c8a..86a941ce27 100644 --- a/xen/drivers/passthrough/vtd/qinval.c +++ b/xen/drivers/passthrough/vtd/qinval.c @@ -31,6 +31,9 @@ #define VTD_QI_TIMEOUT 1 +static unsigned int __read_mostly qi_pg_order; +static unsigned int __read_mostly qi_entry_nr; + static int __must_check invalidate_sync(struct vtd_iommu *iommu); static void print_qi_regs(struct vtd_iommu *iommu) @@ -55,7 +58,7 @@ static unsigned int qinval_next_index(struct vtd_iommu *iommu) tail >>= QINVAL_INDEX_SHIFT; /* (tail+1 == head) indicates a full queue, wait for HW */ - while ( ( tail + 1 ) % QINVAL_ENTRY_NR == + while ( ((tail + 1) & (qi_entry_nr - 1)) == ( dmar_readq(iommu->reg, DMAR_IQH_REG) >> QINVAL_INDEX_SHIFT ) ) cpu_relax(); @@ -68,7 +71,7 @@ static void qinval_update_qtail(struct vtd_iommu *iommu, unsigned int index) /* Need hold register lock when update tail */ ASSERT( spin_is_locked(&iommu->register_lock) ); - val = (index + 1) % QINVAL_ENTRY_NR; + val = (index + 1) & (qi_entry_nr - 1); dmar_writeq(iommu->reg, DMAR_IQT_REG, (val << QINVAL_INDEX_SHIFT)); } @@ -403,8 +406,28 @@ int enable_qinval(struct vtd_iommu *iommu) if ( iommu->qinval_maddr == 0 ) { - iommu->qinval_maddr = alloc_pgtable_maddr(QINVAL_ARCH_PAGE_NR, - iommu->node); + if ( !qi_entry_nr ) + { + /* + * With the present synchronous model, we need two slots for every + * operation (the operation itself and a wait descriptor). There + * can be one such pair of requests pending per CPU. One extra + * entry is needed as the ring is considered full when there's + * only one entry left. + */ + BUILD_BUG_ON(CONFIG_NR_CPUS * 2 >= QINVAL_MAX_ENTRY_NR); + qi_pg_order = get_order_from_bytes((num_present_cpus() * 2 + 1) << + (PAGE_SHIFT - + QINVAL_ENTRY_ORDER)); + qi_entry_nr = 1u << (qi_pg_order + QINVAL_ENTRY_ORDER); + + dprintk(XENLOG_INFO VTDPREFIX, + "QI: using %u-entry ring(s)\n", qi_entry_nr); + } + + iommu->qinval_maddr = + alloc_pgtable_maddr(qi_entry_nr >> QINVAL_ENTRY_ORDER, + iommu->node); if ( iommu->qinval_maddr == 0 ) { dprintk(XENLOG_WARNING VTDPREFIX, @@ -418,15 +441,16 @@ int enable_qinval(struct vtd_iommu *iommu) spin_lock_irqsave(&iommu->register_lock, flags); - /* Setup Invalidation Queue Address(IQA) register with the - * address of the page we just allocated. QS field at - * bits[2:0] to indicate size of queue is one 4KB page. - * That's 256 entries. Queued Head (IQH) and Queue Tail (IQT) - * registers are automatically reset to 0 with write - * to IQA register. + /* + * Setup Invalidation Queue Address (IQA) register with the address of the + * pages we just allocated. The QS field at bits[2:0] indicates the size + * (page order) of the queue. + * + * Queued Head (IQH) and Queue Tail (IQT) registers are automatically + * reset to 0 with write to IQA register. */ dmar_writeq(iommu->reg, DMAR_IQA_REG, - iommu->qinval_maddr | QINVAL_PAGE_ORDER); + iommu->qinval_maddr | qi_pg_order); dmar_writeq(iommu->reg, DMAR_IQT_REG, 0); -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.15 From xen-changelog-bounces@lists.xenproject.org Wed Jun 09 02:22:35 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Wed, 09 Jun 2021 02:22:35 +0000 Received: from list by lists.xenproject.org with outflank-mailman.139016.257140 (Exim 4.92) (envelope-from ) id 1lqnrX-0001hw-2R; Wed, 09 Jun 2021 02:22:35 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 139016.257140; Wed, 09 Jun 2021 02:22:35 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqnrW-0001ho-Vp; Wed, 09 Jun 2021 02:22:34 +0000 Received: by outflank-mailman (input) for mailman id 139016; Wed, 09 Jun 2021 02:22:33 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqnrV-0001hV-1i for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 02:22:33 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqnrV-0003tr-0u for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 02:22:33 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqnrV-0005OR-08 for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 02:22:33 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=27onMpl19ihLZ7+mHDvEZq4vjCe8AH6l1J0GlX0PW1M=; b=D4BZQNEpdYYd18FXtNJOkkqYO9 ou1wqu3Z6T58DAVS0dsFOLui2cDDSweF+yHgXsskpqBrz+2dp/37Dqa3D1GX4V0YsBpgQJ9n+qbPv mnyM5Z87+iEArT8mBQSF0ck6fu02MCeT4WLi+l2iB8VXejnL0fYr4lkoR1uUlJYRu6pY=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.15] AMD/IOMMU: size command buffer dynamically Message-Id: Date: Wed, 09 Jun 2021 02:22:33 +0000 commit 35b583686c8085ac78922eb58a2930222c9c142c Author: Jan Beulich AuthorDate: Tue Jun 8 17:50:45 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 17:52:45 2021 +0100 AMD/IOMMU: size command buffer dynamically With the present synchronous model, we need two slots for every operation (the operation itself and a wait command). There can be one such pair of commands pending per CPU. To ensure that under all normal circumstances a slot is always available when one is requested, size the command ring according to the number of present CPUs. This is part of XSA-373 / CVE-2021-28692. Signed-off-by: Jan Beulich Reviewed-by: Paul Durrant (cherry picked from commit df242851ddc93ac0b0a3a20ecab34acc29e3b36b) --- xen/drivers/passthrough/amd/iommu-defs.h | 7 ++----- xen/drivers/passthrough/amd/iommu_cmd.c | 4 ++-- xen/drivers/passthrough/amd/iommu_init.c | 28 +++++++++++++++++++++++++--- 3 files changed, 29 insertions(+), 10 deletions(-) diff --git a/xen/drivers/passthrough/amd/iommu-defs.h b/xen/drivers/passthrough/amd/iommu-defs.h index 56f385f270..0c97db2a32 100644 --- a/xen/drivers/passthrough/amd/iommu-defs.h +++ b/xen/drivers/passthrough/amd/iommu-defs.h @@ -20,9 +20,6 @@ #ifndef AMD_IOMMU_DEFS_H #define AMD_IOMMU_DEFS_H -/* IOMMU Command Buffer entries: in power of 2 increments, minimum of 256 */ -#define IOMMU_CMD_BUFFER_DEFAULT_ENTRIES 512 - /* IOMMU Event Log entries: in power of 2 increments, minimum of 256 */ #define IOMMU_EVENT_LOG_DEFAULT_ENTRIES 512 @@ -164,8 +161,8 @@ struct amd_iommu_dte { #define IOMMU_CMD_BUFFER_LENGTH_MASK 0x0F000000 #define IOMMU_CMD_BUFFER_LENGTH_SHIFT 24 -#define IOMMU_CMD_BUFFER_ENTRY_SIZE 16 -#define IOMMU_CMD_BUFFER_POWER_OF2_ENTRIES_PER_PAGE 8 +#define IOMMU_CMD_BUFFER_ENTRY_ORDER 4 +#define IOMMU_CMD_BUFFER_MAX_ENTRIES (1u << 15) #define IOMMU_CMD_OPCODE_MASK 0xF0000000 #define IOMMU_CMD_OPCODE_SHIFT 28 diff --git a/xen/drivers/passthrough/amd/iommu_cmd.c b/xen/drivers/passthrough/amd/iommu_cmd.c index 6c0647c524..54d7528f96 100644 --- a/xen/drivers/passthrough/amd/iommu_cmd.c +++ b/xen/drivers/passthrough/amd/iommu_cmd.c @@ -24,7 +24,7 @@ static int queue_iommu_command(struct amd_iommu *iommu, u32 cmd[]) { uint32_t tail, head; - tail = iommu->cmd_buffer.tail + IOMMU_CMD_BUFFER_ENTRY_SIZE; + tail = iommu->cmd_buffer.tail + sizeof(cmd_entry_t); if ( tail == iommu->cmd_buffer.size ) tail = 0; @@ -33,7 +33,7 @@ static int queue_iommu_command(struct amd_iommu *iommu, u32 cmd[]) if ( head != tail ) { memcpy(iommu->cmd_buffer.buffer + iommu->cmd_buffer.tail, - cmd, IOMMU_CMD_BUFFER_ENTRY_SIZE); + cmd, sizeof(cmd_entry_t)); iommu->cmd_buffer.tail = tail; return 1; diff --git a/xen/drivers/passthrough/amd/iommu_init.c b/xen/drivers/passthrough/amd/iommu_init.c index 24d1dfec40..bb52c181f8 100644 --- a/xen/drivers/passthrough/amd/iommu_init.c +++ b/xen/drivers/passthrough/amd/iommu_init.c @@ -118,7 +118,7 @@ static void register_iommu_cmd_buffer_in_mmio_space(struct amd_iommu *iommu) writel(entry, iommu->mmio_base + IOMMU_CMD_BUFFER_BASE_LOW_OFFSET); power_of2_entries = get_order_from_bytes(iommu->cmd_buffer.size) + - IOMMU_CMD_BUFFER_POWER_OF2_ENTRIES_PER_PAGE; + PAGE_SHIFT - IOMMU_CMD_BUFFER_ENTRY_ORDER; entry = 0; iommu_set_addr_hi_to_reg(&entry, addr_hi); @@ -1018,9 +1018,31 @@ static void *__init allocate_ring_buffer(struct ring_buffer *ring_buf, static void * __init allocate_cmd_buffer(struct amd_iommu *iommu) { /* allocate 'command buffer' in power of 2 increments of 4K */ + static unsigned int __read_mostly nr_ents; + + if ( !nr_ents ) + { + unsigned int order; + + /* + * With the present synchronous model, we need two slots for every + * operation (the operation itself and a wait command). There can be + * one such pair of requests pending per CPU. One extra entry is + * needed as the ring is considered full when there's only one entry + * left. + */ + BUILD_BUG_ON(CONFIG_NR_CPUS * 2 >= IOMMU_CMD_BUFFER_MAX_ENTRIES); + order = get_order_from_bytes((num_present_cpus() * 2 + 1) << + IOMMU_CMD_BUFFER_ENTRY_ORDER); + nr_ents = 1u << (order + PAGE_SHIFT - IOMMU_CMD_BUFFER_ENTRY_ORDER); + + AMD_IOMMU_DEBUG("using %u-entry cmd ring(s)\n", nr_ents); + } + + BUILD_BUG_ON(sizeof(cmd_entry_t) != (1u << IOMMU_CMD_BUFFER_ENTRY_ORDER)); + return allocate_ring_buffer(&iommu->cmd_buffer, sizeof(cmd_entry_t), - IOMMU_CMD_BUFFER_DEFAULT_ENTRIES, - "Command Buffer", false); + nr_ents, "Command Buffer", false); } static void * __init allocate_event_log(struct amd_iommu *iommu) -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.15 From xen-changelog-bounces@lists.xenproject.org Wed Jun 09 02:22:45 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Wed, 09 Jun 2021 02:22:45 +0000 Received: from list by lists.xenproject.org with outflank-mailman.139017.257144 (Exim 4.92) (envelope-from ) id 1lqnrh-0001lA-40; Wed, 09 Jun 2021 02:22:45 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 139017.257144; Wed, 09 Jun 2021 02:22:45 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqnrh-0001l0-14; Wed, 09 Jun 2021 02:22:45 +0000 Received: by outflank-mailman (input) for mailman id 139017; Wed, 09 Jun 2021 02:22:43 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqnrf-0001kl-4z for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 02:22:43 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqnrf-0003u1-49 for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 02:22:43 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqnrf-0005PW-3C for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 02:22:43 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=62IkZZGgIDfVUmEVg7KrtvM6Y1d1slBX+t76gb8rnEU=; b=upzOWGMjcfArxAPLcotLCYdaex wJmibWYx/uoDV7K14Ft0Dd97heRrDny21ZFmlN1MwpHBYpHa2zlXt3Z3i5YGHC0jl876tR4XlrjOW opPSn1qzRG8ROZsqagXO82cV++1nljuoH24sUl6+VXAcpjMzCJ9o/SjIHaBbydpOU8L0=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.15] VT-d: eliminate flush related timeouts Message-Id: Date: Wed, 09 Jun 2021 02:22:43 +0000 commit 12ebf0f819517096e49a395bd9c4288b5010526d Author: Jan Beulich AuthorDate: Tue Jun 8 17:50:47 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 17:52:45 2021 +0100 VT-d: eliminate flush related timeouts Leaving an in-progress operation pending when it appears to take too long is problematic: If e.g. a QI command completed later, the write to the "poll slot" may instead be understood to signal a subsequently started command's completion. Also our accounting of the timeout period was actually wrong: We included the time it took for the command to actually make it to the front of the queue, which could be heavily affected by guests other than the one for which the flush is being performed. Do away with all timeout detection on all flush related code paths. Log excessively long processing times (with a progressive threshold) to have some indication of problems in this area. Additionally log (once) if qinval_next_index() didn't immediately find an available slot. Together with the earlier change sizing the queue(s) dynamically, we should now have a guarantee that with our fully synchronous model any demand for slots can actually be satisfied. This is part of XSA-373 / CVE-2021-28692. Signed-off-by: Jan Beulich Reviewed-by: Paul Durrant (cherry picked from commit 7c893abc4ee7e9af62297ba6fd5f27c89d0f33f5) --- xen/drivers/passthrough/vtd/dmar.h | 28 ++++++++++++++++++++++++++++ xen/drivers/passthrough/vtd/iommu.c | 12 ++++++------ xen/drivers/passthrough/vtd/qinval.c | 33 ++++++++++++++++++++++----------- 3 files changed, 56 insertions(+), 17 deletions(-) diff --git a/xen/drivers/passthrough/vtd/dmar.h b/xen/drivers/passthrough/vtd/dmar.h index 1a9c965e59..a44820a5e7 100644 --- a/xen/drivers/passthrough/vtd/dmar.h +++ b/xen/drivers/passthrough/vtd/dmar.h @@ -127,6 +127,34 @@ do { \ } \ } while (0) +#define IOMMU_FLUSH_WAIT(what, iommu, offset, op, cond, sts) \ +do { \ + static unsigned int __read_mostly threshold = 1; \ + s_time_t start = NOW(); \ + s_time_t timeout = start + DMAR_OPERATION_TIMEOUT * threshold; \ + \ + for ( ; ; ) \ + { \ + sts = op(iommu->reg, offset); \ + if ( cond ) \ + break; \ + if ( timeout && NOW() > timeout ) \ + { \ + threshold |= threshold << 1; \ + printk(XENLOG_WARNING VTDPREFIX \ + " IOMMU#%u: %s flush taking too long\n", \ + iommu->index, what); \ + timeout = 0; \ + } \ + cpu_relax(); \ + } \ + \ + if ( !timeout ) \ + printk(XENLOG_WARNING VTDPREFIX \ + " IOMMU#%u: %s flush took %lums\n", \ + iommu->index, what, (NOW() - start) / 10000000); \ +} while ( false ) + int vtd_hw_check(void); void disable_pmr(struct vtd_iommu *iommu); int is_igd_drhd(struct acpi_drhd_unit *drhd); diff --git a/xen/drivers/passthrough/vtd/iommu.c b/xen/drivers/passthrough/vtd/iommu.c index d14fb43182..c160a70688 100644 --- a/xen/drivers/passthrough/vtd/iommu.c +++ b/xen/drivers/passthrough/vtd/iommu.c @@ -373,8 +373,8 @@ static void iommu_flush_write_buffer(struct vtd_iommu *iommu) dmar_writel(iommu->reg, DMAR_GCMD_REG, val | DMA_GCMD_WBF); /* Make sure hardware complete it */ - IOMMU_WAIT_OP(iommu, DMAR_GSTS_REG, dmar_readl, - !(val & DMA_GSTS_WBFS), val); + IOMMU_FLUSH_WAIT("write buffer", iommu, DMAR_GSTS_REG, dmar_readl, + !(val & DMA_GSTS_WBFS), val); spin_unlock_irqrestore(&iommu->register_lock, flags); } @@ -423,8 +423,8 @@ int vtd_flush_context_reg(struct vtd_iommu *iommu, uint16_t did, dmar_writeq(iommu->reg, DMAR_CCMD_REG, val); /* Make sure hardware complete it */ - IOMMU_WAIT_OP(iommu, DMAR_CCMD_REG, dmar_readq, - !(val & DMA_CCMD_ICC), val); + IOMMU_FLUSH_WAIT("context", iommu, DMAR_CCMD_REG, dmar_readq, + !(val & DMA_CCMD_ICC), val); spin_unlock_irqrestore(&iommu->register_lock, flags); /* flush context entry will implicitly flush write buffer */ @@ -501,8 +501,8 @@ int vtd_flush_iotlb_reg(struct vtd_iommu *iommu, uint16_t did, uint64_t addr, dmar_writeq(iommu->reg, tlb_offset + 8, val); /* Make sure hardware complete it */ - IOMMU_WAIT_OP(iommu, (tlb_offset + 8), dmar_readq, - !(val & DMA_TLB_IVT), val); + IOMMU_FLUSH_WAIT("iotlb", iommu, (tlb_offset + 8), dmar_readq, + !(val & DMA_TLB_IVT), val); spin_unlock_irqrestore(&iommu->register_lock, flags); /* check IOTLB invalidation granularity */ diff --git a/xen/drivers/passthrough/vtd/qinval.c b/xen/drivers/passthrough/vtd/qinval.c index 86a941ce27..1056ab1feb 100644 --- a/xen/drivers/passthrough/vtd/qinval.c +++ b/xen/drivers/passthrough/vtd/qinval.c @@ -29,8 +29,6 @@ #include "extern.h" #include "../ats.h" -#define VTD_QI_TIMEOUT 1 - static unsigned int __read_mostly qi_pg_order; static unsigned int __read_mostly qi_entry_nr; @@ -60,7 +58,11 @@ static unsigned int qinval_next_index(struct vtd_iommu *iommu) /* (tail+1 == head) indicates a full queue, wait for HW */ while ( ((tail + 1) & (qi_entry_nr - 1)) == ( dmar_readq(iommu->reg, DMAR_IQH_REG) >> QINVAL_INDEX_SHIFT ) ) + { + printk_once(XENLOG_ERR VTDPREFIX " IOMMU#%u: no QI slot available\n", + iommu->index); cpu_relax(); + } return tail; } @@ -180,23 +182,32 @@ static int __must_check queue_invalidate_wait(struct vtd_iommu *iommu, /* Now we don't support interrupt method */ if ( sw ) { - s_time_t timeout; - - /* In case all wait descriptor writes to same addr with same data */ - timeout = NOW() + MILLISECS(flush_dev_iotlb ? - iommu_dev_iotlb_timeout : VTD_QI_TIMEOUT); + static unsigned int __read_mostly threshold = 1; + s_time_t start = NOW(); + s_time_t timeout = start + (flush_dev_iotlb + ? iommu_dev_iotlb_timeout + : 100) * MILLISECS(threshold); while ( ACCESS_ONCE(*this_poll_slot) != QINVAL_STAT_DONE ) { - if ( NOW() > timeout ) + if ( timeout && NOW() > timeout ) { - print_qi_regs(iommu); + threshold |= threshold << 1; printk(XENLOG_WARNING VTDPREFIX - " Queue invalidate wait descriptor timed out\n"); - return -ETIMEDOUT; + " IOMMU#%u: QI%s wait descriptor taking too long\n", + iommu->index, flush_dev_iotlb ? " dev" : ""); + print_qi_regs(iommu); + timeout = 0; } cpu_relax(); } + + if ( !timeout ) + printk(XENLOG_WARNING VTDPREFIX + " IOMMU#%u: QI%s wait descriptor took %lums\n", + iommu->index, flush_dev_iotlb ? " dev" : "", + (NOW() - start) / 10000000); + return 0; } -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.15 From xen-changelog-bounces@lists.xenproject.org Wed Jun 09 02:22:55 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Wed, 09 Jun 2021 02:22:55 +0000 Received: from list by lists.xenproject.org with outflank-mailman.139018.257148 (Exim 4.92) (envelope-from ) id 1lqnrr-0001nn-5U; Wed, 09 Jun 2021 02:22:55 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 139018.257148; Wed, 09 Jun 2021 02:22:55 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqnrr-0001nf-2f; Wed, 09 Jun 2021 02:22:55 +0000 Received: by outflank-mailman (input) for mailman id 139018; Wed, 09 Jun 2021 02:22:53 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqnrp-0001nK-8a for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 02:22:53 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqnrp-0003ub-7h for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 02:22:53 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqnrp-0005QR-6p for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 02:22:53 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=ITrPQTV6CZERanKMd5HnUiT4SNyq/Xlnv2vvI3mOyz0=; b=OZrlgAUGNToI359AU4maDU0AVR a4aVoK4dX9NE34vIBjhV5sbkhOkHnpTk1WBA8xN02nRwahCdgWh1FRy3O95F1hwvTkxfrgpV/w/FP WZj9P+KHK8W0bo2jhMaSXYjU9PUl8j0ILlviC0V/egVz5pqMMNNKBX2fGKDEawPoEIB8=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.15] x86/spec-ctrl: Protect against Speculative Code Store Bypass Message-Id: Date: Wed, 09 Jun 2021 02:22:53 +0000 commit 894636d192368b130146311af2e539346c6fd650 Author: Andrew Cooper AuthorDate: Thu Mar 11 14:39:11 2021 +0000 Commit: Andrew Cooper CommitDate: Tue Jun 8 17:52:45 2021 +0100 x86/spec-ctrl: Protect against Speculative Code Store Bypass Modern x86 processors have far-better-than-architecturally-guaranteed self modifying code detection. Typically, when a write hits an instruction in flight, a Machine Clear occurs to flush stale content in the frontend and backend. For self modifying code, before a write which hits an instruction in flight retires, the frontend can speculatively decode and execute the old instruction stream. Speculation of this form can suffer from type confusion in registers, and potentially leak data. Furthermore, updates are typically byte-wise, rather than atomic. Depending on timing, speculation can race ahead multiple times between individual writes, and execute the transiently-malformed instruction stream. Xen has stubs which are used in certain cases for emulation purposes. Inhibit speculation between updating the stub and executing it. This is XSA-375 / CVE-2021-0089. Signed-off-by: Andrew Cooper Reviewed-by: Jan Beulich (cherry picked from commit 45f59ed8865318bb0356954bad067f329677ce9e) --- xen/arch/x86/pv/emul-priv-op.c | 2 ++ xen/arch/x86/x86_emulate/x86_emulate.c | 1 + 2 files changed, 3 insertions(+) diff --git a/xen/arch/x86/pv/emul-priv-op.c b/xen/arch/x86/pv/emul-priv-op.c index 8889509d2a..11467a1e3a 100644 --- a/xen/arch/x86/pv/emul-priv-op.c +++ b/xen/arch/x86/pv/emul-priv-op.c @@ -138,6 +138,8 @@ static io_emul_stub_t *io_emul_stub_setup(struct priv_op_ctxt *ctxt, u8 opcode, /* Runtime confirmation that we haven't clobbered an adjacent stub. */ BUG_ON(STUB_BUF_SIZE / 2 < (p - ctxt->io_emul_stub)); + block_speculation(); /* SCSB */ + /* Handy function-typed pointer to the stub. */ return (void *)stub_va; diff --git a/xen/arch/x86/x86_emulate/x86_emulate.c b/xen/arch/x86/x86_emulate/x86_emulate.c index a8000cbcbd..7bfb91dfff 100644 --- a/xen/arch/x86/x86_emulate/x86_emulate.c +++ b/xen/arch/x86/x86_emulate/x86_emulate.c @@ -1257,6 +1257,7 @@ static inline int mkec(uint8_t e, int32_t ec, ...) # define invoke_stub(pre, post, constraints...) do { \ stub_exn.info = (union stub_exception_token) { .raw = ~0 }; \ stub_exn.line = __LINE__; /* Utility outweighs livepatching cost */ \ + block_speculation(); /* SCSB */ \ asm volatile ( pre "\n\tINDIRECT_CALL %[stub]\n\t" post "\n" \ ".Lret%=:\n\t" \ ".pushsection .fixup,\"ax\"\n" \ -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.15 From xen-changelog-bounces@lists.xenproject.org Wed Jun 09 02:23:05 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Wed, 09 Jun 2021 02:23:05 +0000 Received: from list by lists.xenproject.org with outflank-mailman.139019.257151 (Exim 4.92) (envelope-from ) id 1lqns1-0001qn-7P; Wed, 09 Jun 2021 02:23:05 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 139019.257151; Wed, 09 Jun 2021 02:23:05 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqns1-0001qg-46; Wed, 09 Jun 2021 02:23:05 +0000 Received: by outflank-mailman (input) for mailman id 139019; Wed, 09 Jun 2021 02:23:03 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqnrz-0001qQ-BY for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 02:23:03 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqnrz-0003vF-Aj for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 02:23:03 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqnrz-0005Rf-9u for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 02:23:03 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=le8R+OIKtRN/cFF7ly9FtOdyCJqjElKSmmVG3mYzJzk=; b=XPNq+oZV81FHEsgwfShrZueq95 1l0YyH46Sshgl4+ETvnenaEt4bmIkgI8fvOnju2+WNSj6RHvjJ+0h8lAZtIjcED9STKd5yRWDf7N0 OIPNciwYoReoZOf3mT58IYgbiG3AthePxQW5wMZV3ztimFaVFmB+I+Lp15nvQ5Yx1zLA=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.15] x86/spec-ctrl: Mitigate TAA after S3 resume Message-Id: Date: Wed, 09 Jun 2021 02:23:03 +0000 commit f034c96e882b81738720472cd28e75e6d6eb66fe Author: Andrew Cooper AuthorDate: Thu May 20 01:21:39 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 17:52:45 2021 +0100 x86/spec-ctrl: Mitigate TAA after S3 resume The user chosen setting for MSR_TSX_CTRL needs restoring after S3. All APs get the correct setting via start_secondary(), but the BSP was missed out. This is XSA-377 / CVE-2021-28690. Fixes: 8c4330818f6 ("x86/spec-ctrl: Mitigate the TSX Asynchronous Abort sidechannel") Signed-off-by: Andrew Cooper Reviewed-by: Jan Beulich (cherry picked from commit 8cf276cb2e0b99b96333865873f56b0b31555ff1) --- xen/arch/x86/acpi/power.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/xen/arch/x86/acpi/power.c b/xen/arch/x86/acpi/power.c index 91a8c4d0bd..31a56f02d0 100644 --- a/xen/arch/x86/acpi/power.c +++ b/xen/arch/x86/acpi/power.c @@ -288,6 +288,8 @@ static int enter_state(u32 state) microcode_update_one(); + tsx_init(); /* Needs microcode. May change HLE/RTM feature bits. */ + if ( !recheck_cpu_features(0) ) panic("Missing previously available feature(s)\n"); -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.15 From xen-changelog-bounces@lists.xenproject.org Wed Jun 09 06:11:16 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Wed, 09 Jun 2021 06:11:16 +0000 Received: from list by lists.xenproject.org with outflank-mailman.139029.257172 (Exim 4.92) (envelope-from ) id 1lqrQo-00059D-EI; Wed, 09 Jun 2021 06:11:14 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 139029.257172; Wed, 09 Jun 2021 06:11:14 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqrQo-000594-BF; Wed, 09 Jun 2021 06:11:14 +0000 Received: by outflank-mailman (input) for mailman id 139029; Wed, 09 Jun 2021 06:11:12 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqrQm-00058f-SP for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 06:11:12 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqrQm-0008FP-RX for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 06:11:12 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqrQm-0001M3-QV for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 06:11:12 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=TYQokdWZqRehgd1haoPrRwKOsYwFgHNEoWCBrISyUXQ=; b=JFq097IyJJhahoAJgnhMGIuymA K5tUetfX2NeL8BtbwLVGgiCkTdIILlc2iuy9kAa6CQeJY0+gxUSVE1h5PHCns3wrxdu9hNTOqgPK6 VQiyQZ2xqeEYb5g5mylfZN+kdTREZcwu/sRnz00AiVL07HrTFDrJ9ImkwSgEoY072DsY=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.14] xen/arm: Boot modules should always be scrubbed if bootscrub={on, idle} Message-Id: Date: Wed, 09 Jun 2021 06:11:12 +0000 commit 7053c8ea5dc88e0a3b6579d3bb61a242fcd5c499 Author: Julien Grall AuthorDate: Sat Apr 17 17:38:28 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 18:36:03 2021 +0100 xen/arm: Boot modules should always be scrubbed if bootscrub={on, idle} The function to initialize the pages (see init_heap_pages()) will request scrub when the admin request idle bootscrub (default) and state == SYS_STATE_active. When bootscrub=on, Xen will scrub any free pages in heap_init_late(). Currently, the boot modules (e.g. kernels, initramfs) will be discarded/ freed after heap_init_late() is called and system_state switched to SYS_STATE_active. This means the pages associated with the boot modules will not get scrubbed before getting re-purposed. If the memory is assigned to an untrusted domU, it may be able to retrieve secrets from the modules. This is part of XSA-372 / CVE-2021-28693. Fixes: 1774e9b1df27 ("xen/arm: introduce create_domUs") Signed-off-by: Julien Grall Reviewed-by: Jan Beulich Reviewed-by: Stefano Stabellini Tested-by: Stefano Stabellini (cherry picked from commit fd5dc41ceaed9cfcfa011cdfd50f264c89277a90) --- xen/arch/arm/setup.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-) diff --git a/xen/arch/arm/setup.c b/xen/arch/arm/setup.c index 1f26080b30..34b1c1a11e 100644 --- a/xen/arch/arm/setup.c +++ b/xen/arch/arm/setup.c @@ -75,7 +75,6 @@ static __used void init_done(void) /* Must be done past setting system_state. */ unregister_init_virtual_region(); - discard_initial_modules(); free_init_memory(); startup_cpu_idle_loop(); } @@ -964,6 +963,12 @@ void __init start_xen(unsigned long boot_phys_offset, create_domUs(); + /* + * This needs to be called **before** heap_init_late() so modules + * will be scrubbed (unless suppressed). + */ + discard_initial_modules(); + heap_init_late(); init_trace_bufs(); -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.14 From xen-changelog-bounces@lists.xenproject.org Wed Jun 09 06:11:16 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Wed, 09 Jun 2021 06:11:16 +0000 Received: from list by lists.xenproject.org with outflank-mailman.139028.257168 (Exim 4.92) (envelope-from ) id 1lqrQe-00057a-Ca; Wed, 09 Jun 2021 06:11:04 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 139028.257168; Wed, 09 Jun 2021 06:11:04 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqrQe-00057S-9k; Wed, 09 Jun 2021 06:11:04 +0000 Received: by outflank-mailman (input) for mailman id 139028; Wed, 09 Jun 2021 06:11:02 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqrQc-00057M-P8 for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 06:11:02 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqrQc-0008FG-OK for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 06:11:02 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqrQc-0001Kc-NE for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 06:11:02 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=IdMTTR0wyVIAxHapYYpt94uvywSyN/A7Q6OEzue5u08=; b=hsiIymdXHoq92JvopKtSAJWWOk y0yjfHa0jRVmBtP7awibmfRtkIQN0Crd8X98WsF9tU2GWAczXU/B9/R5aRLf7TPZGlj6hmD32ZVu6 6gftYDnSP9ZGaKJNkql7T0Kp+3gxqTz2ky0/hC2g9TNHXHZvdP4zIhbMTq/Cji53Irfs=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.14] xen/arm: Create dom0less domUs earlier Message-Id: Date: Wed, 09 Jun 2021 06:11:02 +0000 commit 5caa690ea7e9dffa65a496424353c96af10f0308 Author: Julien Grall AuthorDate: Mon May 17 17:47:13 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 18:36:03 2021 +0100 xen/arm: Create dom0less domUs earlier In a follow-up patch we will need to unallocate the boot modules before heap_init_late() is called. The modules will contain the domUs kernel and initramfs. Therefore Xen will need to create extra domUs (used by dom0less) before heap_init_late(). This has two consequences on dom0less: 1) Domains will not be unpaused as soon as they are created but once all have been created. However, Xen doesn't guarantee an order to unpause, so this is not something one could rely on. 2) The memory allocated for a domU will not be scrubbed anymore when an admin select bootscrub=on. This is not something we advertised, but if this is a concern we can introduce either force scrub for all domUs or a per-domain flag in the DT. The behavior for bootscrub=off and bootscrub=idle (default) has not changed. This is part of XSA-372 / CVE-2021-28693. Signed-off-by: Julien Grall Reviewed-by: Jan Beulich Reviewed-by: Stefano Stabellini Tested-by: Stefano Stabellini (cherry picked from commit 371347c5b64da699d9f5a0edda5dc496fd2b7a5c) --- xen/arch/arm/domain_build.c | 2 -- xen/arch/arm/setup.c | 9 +++++---- 2 files changed, 5 insertions(+), 6 deletions(-) diff --git a/xen/arch/arm/domain_build.c b/xen/arch/arm/domain_build.c index e824ba34b0..b07461f5d3 100644 --- a/xen/arch/arm/domain_build.c +++ b/xen/arch/arm/domain_build.c @@ -2515,8 +2515,6 @@ void __init create_domUs(void) if ( construct_domU(d, node) != 0 ) panic("Could not set up domain %s\n", dt_node_name(node)); - - domain_unpause_by_systemcontroller(d); } } diff --git a/xen/arch/arm/setup.c b/xen/arch/arm/setup.c index 7968cee47d..1f26080b30 100644 --- a/xen/arch/arm/setup.c +++ b/xen/arch/arm/setup.c @@ -779,7 +779,7 @@ void __init start_xen(unsigned long boot_phys_offset, int cpus, i; const char *cmdline; struct bootmodule *xen_bootmodule; - struct domain *dom0; + struct domain *dom0, *d; struct xen_domctl_createdomain dom0_cfg = { .flags = XEN_DOMCTL_CDF_hvm | XEN_DOMCTL_CDF_hap, .max_evtchn_port = -1, @@ -962,6 +962,8 @@ void __init start_xen(unsigned long boot_phys_offset, if ( construct_dom0(dom0) != 0) panic("Could not set up DOM0 guest OS\n"); + create_domUs(); + heap_init_late(); init_trace_bufs(); @@ -975,9 +977,8 @@ void __init start_xen(unsigned long boot_phys_offset, system_state = SYS_STATE_active; - create_domUs(); - - domain_unpause_by_systemcontroller(dom0); + for_each_domain( d ) + domain_unpause_by_systemcontroller(d); /* Switch on to the dynamically allocated stack for the idle vcpu * since the static one we're running on is about to be freed. */ -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.14 From xen-changelog-bounces@lists.xenproject.org Wed Jun 09 06:11:24 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Wed, 09 Jun 2021 06:11:24 +0000 Received: from list by lists.xenproject.org with outflank-mailman.139030.257176 (Exim 4.92) (envelope-from ) id 1lqrQy-0005CI-Fz; Wed, 09 Jun 2021 06:11:24 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 139030.257176; Wed, 09 Jun 2021 06:11:24 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqrQy-0005CA-Cz; Wed, 09 Jun 2021 06:11:24 +0000 Received: by outflank-mailman (input) for mailman id 139030; Wed, 09 Jun 2021 06:11:23 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqrQw-0005Bw-W3 for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 06:11:22 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqrQw-0008FX-VC for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 06:11:22 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqrQw-0001N9-U0 for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 06:11:22 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=UT3AXJhnJqCSj2FeRrtEn86nen1FPbnGecJ4HM31bq4=; b=wBXl/HkS7dGwP7eii2hS9qg+wb EZTpIgEzNd1zcVmk82xxvmsZ1lJDIQJ1sT/5HOZ7T1Tt/H4rOzMLNEsK0xDR95B97+EKgG11gnY2/ HUo9Cs1x4ew0Fy4aDxzWG0TsPgRWXA4HP9Kx35MPrx9YPyqV3PT0kUtTl9Qr0mSo2Sk4=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.14] VT-d: size qinval queue dynamically Message-Id: Date: Wed, 09 Jun 2021 06:11:22 +0000 commit 2665d9785e748bf9cd87ec5783099e644ce2bac2 Author: Jan Beulich AuthorDate: Tue Jun 8 18:05:25 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 18:36:03 2021 +0100 VT-d: size qinval queue dynamically With the present synchronous model, we need two slots for every operation (the operation itself and a wait descriptor). There can be one such pair of requests pending per CPU. To ensure that under all normal circumstances a slot is always available when one is requested, size the queue ring according to the number of present CPUs. This is part of XSA-373 / CVE-2021-28692. Signed-off-by: Jan Beulich Reviewed-by: Paul Durrant (cherry picked from commit cbfa62bb140c8f10a0ae96db3ce06e22b3b9d302) --- xen/drivers/passthrough/vtd/iommu.h | 10 +------- xen/drivers/passthrough/vtd/qinval.c | 46 +++++++++++++++++++++++++++--------- 2 files changed, 36 insertions(+), 20 deletions(-) diff --git a/xen/drivers/passthrough/vtd/iommu.h b/xen/drivers/passthrough/vtd/iommu.h index 57737b1a4a..e002dc6f01 100644 --- a/xen/drivers/passthrough/vtd/iommu.h +++ b/xen/drivers/passthrough/vtd/iommu.h @@ -450,17 +450,9 @@ struct qinval_entry { }q; }; -/* Order of queue invalidation pages(max is 8) */ -#define QINVAL_PAGE_ORDER 2 - -#define QINVAL_ARCH_PAGE_ORDER (QINVAL_PAGE_ORDER + PAGE_SHIFT_4K - PAGE_SHIFT) -#define QINVAL_ARCH_PAGE_NR ( QINVAL_ARCH_PAGE_ORDER < 0 ? \ - 1 : \ - 1 << QINVAL_ARCH_PAGE_ORDER ) - /* Each entry is 16 bytes, so 2^8 entries per page */ #define QINVAL_ENTRY_ORDER ( PAGE_SHIFT - 4 ) -#define QINVAL_ENTRY_NR (1 << (QINVAL_PAGE_ORDER + 8)) +#define QINVAL_MAX_ENTRY_NR (1u << (7 + QINVAL_ENTRY_ORDER)) /* Status data flag */ #define QINVAL_STAT_INIT 0 diff --git a/xen/drivers/passthrough/vtd/qinval.c b/xen/drivers/passthrough/vtd/qinval.c index dfcbb66c8a..86a941ce27 100644 --- a/xen/drivers/passthrough/vtd/qinval.c +++ b/xen/drivers/passthrough/vtd/qinval.c @@ -31,6 +31,9 @@ #define VTD_QI_TIMEOUT 1 +static unsigned int __read_mostly qi_pg_order; +static unsigned int __read_mostly qi_entry_nr; + static int __must_check invalidate_sync(struct vtd_iommu *iommu); static void print_qi_regs(struct vtd_iommu *iommu) @@ -55,7 +58,7 @@ static unsigned int qinval_next_index(struct vtd_iommu *iommu) tail >>= QINVAL_INDEX_SHIFT; /* (tail+1 == head) indicates a full queue, wait for HW */ - while ( ( tail + 1 ) % QINVAL_ENTRY_NR == + while ( ((tail + 1) & (qi_entry_nr - 1)) == ( dmar_readq(iommu->reg, DMAR_IQH_REG) >> QINVAL_INDEX_SHIFT ) ) cpu_relax(); @@ -68,7 +71,7 @@ static void qinval_update_qtail(struct vtd_iommu *iommu, unsigned int index) /* Need hold register lock when update tail */ ASSERT( spin_is_locked(&iommu->register_lock) ); - val = (index + 1) % QINVAL_ENTRY_NR; + val = (index + 1) & (qi_entry_nr - 1); dmar_writeq(iommu->reg, DMAR_IQT_REG, (val << QINVAL_INDEX_SHIFT)); } @@ -403,8 +406,28 @@ int enable_qinval(struct vtd_iommu *iommu) if ( iommu->qinval_maddr == 0 ) { - iommu->qinval_maddr = alloc_pgtable_maddr(QINVAL_ARCH_PAGE_NR, - iommu->node); + if ( !qi_entry_nr ) + { + /* + * With the present synchronous model, we need two slots for every + * operation (the operation itself and a wait descriptor). There + * can be one such pair of requests pending per CPU. One extra + * entry is needed as the ring is considered full when there's + * only one entry left. + */ + BUILD_BUG_ON(CONFIG_NR_CPUS * 2 >= QINVAL_MAX_ENTRY_NR); + qi_pg_order = get_order_from_bytes((num_present_cpus() * 2 + 1) << + (PAGE_SHIFT - + QINVAL_ENTRY_ORDER)); + qi_entry_nr = 1u << (qi_pg_order + QINVAL_ENTRY_ORDER); + + dprintk(XENLOG_INFO VTDPREFIX, + "QI: using %u-entry ring(s)\n", qi_entry_nr); + } + + iommu->qinval_maddr = + alloc_pgtable_maddr(qi_entry_nr >> QINVAL_ENTRY_ORDER, + iommu->node); if ( iommu->qinval_maddr == 0 ) { dprintk(XENLOG_WARNING VTDPREFIX, @@ -418,15 +441,16 @@ int enable_qinval(struct vtd_iommu *iommu) spin_lock_irqsave(&iommu->register_lock, flags); - /* Setup Invalidation Queue Address(IQA) register with the - * address of the page we just allocated. QS field at - * bits[2:0] to indicate size of queue is one 4KB page. - * That's 256 entries. Queued Head (IQH) and Queue Tail (IQT) - * registers are automatically reset to 0 with write - * to IQA register. + /* + * Setup Invalidation Queue Address (IQA) register with the address of the + * pages we just allocated. The QS field at bits[2:0] indicates the size + * (page order) of the queue. + * + * Queued Head (IQH) and Queue Tail (IQT) registers are automatically + * reset to 0 with write to IQA register. */ dmar_writeq(iommu->reg, DMAR_IQA_REG, - iommu->qinval_maddr | QINVAL_PAGE_ORDER); + iommu->qinval_maddr | qi_pg_order); dmar_writeq(iommu->reg, DMAR_IQT_REG, 0); -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.14 From xen-changelog-bounces@lists.xenproject.org Wed Jun 09 06:11:34 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Wed, 09 Jun 2021 06:11:34 +0000 Received: from list by lists.xenproject.org with outflank-mailman.139031.257180 (Exim 4.92) (envelope-from ) id 1lqrR8-0005Fy-HR; Wed, 09 Jun 2021 06:11:34 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 139031.257180; Wed, 09 Jun 2021 06:11:34 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqrR8-0005Fr-EU; Wed, 09 Jun 2021 06:11:34 +0000 Received: by outflank-mailman (input) for mailman id 139031; Wed, 09 Jun 2021 06:11:33 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqrR7-0005FO-3c for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 06:11:33 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqrR7-0008Fu-2l for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 06:11:33 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqrR7-0001OI-1c for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 06:11:33 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=9NpUEz0ocJUhTxUAfP86WxVIifG5Z4+BXR38nwhgWCY=; b=RHKdV/BGgK9W1LAFxnoftLoBMl mSn5oSTKRBJPyooLXKRUUNO2qk8Q74TQSARqGbg4OnM1cZTSloF4IePxGZnvQuqDZF2O0ARLzifz7 GfIsAYyNQ3mtPRPZcDdo2ZmTZvdxePjcfuc+4ZqwjJopaWl7rWHPnB5m+tn6Vh5le1/M=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.14] AMD/IOMMU: size command buffer dynamically Message-Id: Date: Wed, 09 Jun 2021 06:11:33 +0000 commit 46ff2453de9e58ad49d79eda1be6ff21c22b3144 Author: Jan Beulich AuthorDate: Tue Jun 8 18:15:45 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 18:36:03 2021 +0100 AMD/IOMMU: size command buffer dynamically With the present synchronous model, we need two slots for every operation (the operation itself and a wait command). There can be one such pair of commands pending per CPU. To ensure that under all normal circumstances a slot is always available when one is requested, size the command ring according to the number of present CPUs. This is part of XSA-373 / CVE-2021-28692. Signed-off-by: Jan Beulich Reviewed-by: Paul Durrant (cherry picked from commit df242851ddc93ac0b0a3a20ecab34acc29e3b36b) --- xen/drivers/passthrough/amd/iommu-defs.h | 7 ++----- xen/drivers/passthrough/amd/iommu_cmd.c | 4 ++-- xen/drivers/passthrough/amd/iommu_init.c | 28 +++++++++++++++++++++++++--- 3 files changed, 29 insertions(+), 10 deletions(-) diff --git a/xen/drivers/passthrough/amd/iommu-defs.h b/xen/drivers/passthrough/amd/iommu-defs.h index 56f385f270..0c97db2a32 100644 --- a/xen/drivers/passthrough/amd/iommu-defs.h +++ b/xen/drivers/passthrough/amd/iommu-defs.h @@ -20,9 +20,6 @@ #ifndef AMD_IOMMU_DEFS_H #define AMD_IOMMU_DEFS_H -/* IOMMU Command Buffer entries: in power of 2 increments, minimum of 256 */ -#define IOMMU_CMD_BUFFER_DEFAULT_ENTRIES 512 - /* IOMMU Event Log entries: in power of 2 increments, minimum of 256 */ #define IOMMU_EVENT_LOG_DEFAULT_ENTRIES 512 @@ -164,8 +161,8 @@ struct amd_iommu_dte { #define IOMMU_CMD_BUFFER_LENGTH_MASK 0x0F000000 #define IOMMU_CMD_BUFFER_LENGTH_SHIFT 24 -#define IOMMU_CMD_BUFFER_ENTRY_SIZE 16 -#define IOMMU_CMD_BUFFER_POWER_OF2_ENTRIES_PER_PAGE 8 +#define IOMMU_CMD_BUFFER_ENTRY_ORDER 4 +#define IOMMU_CMD_BUFFER_MAX_ENTRIES (1u << 15) #define IOMMU_CMD_OPCODE_MASK 0xF0000000 #define IOMMU_CMD_OPCODE_SHIFT 28 diff --git a/xen/drivers/passthrough/amd/iommu_cmd.c b/xen/drivers/passthrough/amd/iommu_cmd.c index 249ed345a0..760babf20e 100644 --- a/xen/drivers/passthrough/amd/iommu_cmd.c +++ b/xen/drivers/passthrough/amd/iommu_cmd.c @@ -24,7 +24,7 @@ static int queue_iommu_command(struct amd_iommu *iommu, u32 cmd[]) { uint32_t tail, head; - tail = iommu->cmd_buffer.tail + IOMMU_CMD_BUFFER_ENTRY_SIZE; + tail = iommu->cmd_buffer.tail + sizeof(cmd_entry_t); if ( tail == iommu->cmd_buffer.size ) tail = 0; @@ -33,7 +33,7 @@ static int queue_iommu_command(struct amd_iommu *iommu, u32 cmd[]) if ( head != tail ) { memcpy(iommu->cmd_buffer.buffer + iommu->cmd_buffer.tail, - cmd, IOMMU_CMD_BUFFER_ENTRY_SIZE); + cmd, sizeof(cmd_entry_t)); iommu->cmd_buffer.tail = tail; return 1; diff --git a/xen/drivers/passthrough/amd/iommu_init.c b/xen/drivers/passthrough/amd/iommu_init.c index 034f3b9c2c..9b7e439eb7 100644 --- a/xen/drivers/passthrough/amd/iommu_init.c +++ b/xen/drivers/passthrough/amd/iommu_init.c @@ -118,7 +118,7 @@ static void register_iommu_cmd_buffer_in_mmio_space(struct amd_iommu *iommu) writel(entry, iommu->mmio_base + IOMMU_CMD_BUFFER_BASE_LOW_OFFSET); power_of2_entries = get_order_from_bytes(iommu->cmd_buffer.size) + - IOMMU_CMD_BUFFER_POWER_OF2_ENTRIES_PER_PAGE; + PAGE_SHIFT - IOMMU_CMD_BUFFER_ENTRY_ORDER; entry = 0; iommu_set_addr_hi_to_reg(&entry, addr_hi); @@ -1022,9 +1022,31 @@ static void *__init allocate_ring_buffer(struct ring_buffer *ring_buf, static void * __init allocate_cmd_buffer(struct amd_iommu *iommu) { /* allocate 'command buffer' in power of 2 increments of 4K */ + static unsigned int __read_mostly nr_ents; + + if ( !nr_ents ) + { + unsigned int order; + + /* + * With the present synchronous model, we need two slots for every + * operation (the operation itself and a wait command). There can be + * one such pair of requests pending per CPU. One extra entry is + * needed as the ring is considered full when there's only one entry + * left. + */ + BUILD_BUG_ON(CONFIG_NR_CPUS * 2 >= IOMMU_CMD_BUFFER_MAX_ENTRIES); + order = get_order_from_bytes((num_present_cpus() * 2 + 1) << + IOMMU_CMD_BUFFER_ENTRY_ORDER); + nr_ents = 1u << (order + PAGE_SHIFT - IOMMU_CMD_BUFFER_ENTRY_ORDER); + + AMD_IOMMU_DEBUG("using %u-entry cmd ring(s)\n", nr_ents); + } + + BUILD_BUG_ON(sizeof(cmd_entry_t) != (1u << IOMMU_CMD_BUFFER_ENTRY_ORDER)); + return allocate_ring_buffer(&iommu->cmd_buffer, sizeof(cmd_entry_t), - IOMMU_CMD_BUFFER_DEFAULT_ENTRIES, - "Command Buffer", false); + nr_ents, "Command Buffer", false); } static void * __init allocate_event_log(struct amd_iommu *iommu) -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.14 From xen-changelog-bounces@lists.xenproject.org Wed Jun 09 06:11:44 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Wed, 09 Jun 2021 06:11:44 +0000 Received: from list by lists.xenproject.org with outflank-mailman.139032.257184 (Exim 4.92) (envelope-from ) id 1lqrRI-0005JN-JL; Wed, 09 Jun 2021 06:11:44 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 139032.257184; Wed, 09 Jun 2021 06:11:44 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqrRI-0005JF-GN; Wed, 09 Jun 2021 06:11:44 +0000 Received: by outflank-mailman (input) for mailman id 139032; Wed, 09 Jun 2021 06:11:43 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqrRH-0005J2-7s for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 06:11:43 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqrRH-0008G7-5w for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 06:11:43 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqrRH-0001PL-5B for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 06:11:43 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=0IYAkCq7/Yg4XtyGnMALU3tEdoVygoXRQKl30f9dRBA=; b=6ho9xx3MJVqDEWyF0ShjqSICtf 41CTzLlzDus4hDh1z8NOIIydtDxMyUIfXQbhM/n4CeCDVBqb2sTOhxODhVy1RjYhXtX4iNM1DGgas 58KRIOGG4ZpQ0kEFvINPOPXFkY4DY/d+SqCtMLYjgmfqkd/B/BTbv+xoLyeI51vopS+4=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.14] VT-d: eliminate flush related timeouts Message-Id: Date: Wed, 09 Jun 2021 06:11:43 +0000 commit e5bce3a74bacfe570233d61e57d31438d21fafd5 Author: Jan Beulich AuthorDate: Tue Jun 8 18:16:02 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 18:36:03 2021 +0100 VT-d: eliminate flush related timeouts Leaving an in-progress operation pending when it appears to take too long is problematic: If e.g. a QI command completed later, the write to the "poll slot" may instead be understood to signal a subsequently started command's completion. Also our accounting of the timeout period was actually wrong: We included the time it took for the command to actually make it to the front of the queue, which could be heavily affected by guests other than the one for which the flush is being performed. Do away with all timeout detection on all flush related code paths. Log excessively long processing times (with a progressive threshold) to have some indication of problems in this area. Additionally log (once) if qinval_next_index() didn't immediately find an available slot. Together with the earlier change sizing the queue(s) dynamically, we should now have a guarantee that with our fully synchronous model any demand for slots can actually be satisfied. This is part of XSA-373 / CVE-2021-28692. Signed-off-by: Jan Beulich Reviewed-by: Paul Durrant (cherry picked from commit 7c893abc4ee7e9af62297ba6fd5f27c89d0f33f5) --- xen/drivers/passthrough/vtd/dmar.h | 28 ++++++++++++++++++++++++++++ xen/drivers/passthrough/vtd/iommu.c | 12 ++++++------ xen/drivers/passthrough/vtd/qinval.c | 33 ++++++++++++++++++++++----------- 3 files changed, 56 insertions(+), 17 deletions(-) diff --git a/xen/drivers/passthrough/vtd/dmar.h b/xen/drivers/passthrough/vtd/dmar.h index 1a9c965e59..a44820a5e7 100644 --- a/xen/drivers/passthrough/vtd/dmar.h +++ b/xen/drivers/passthrough/vtd/dmar.h @@ -127,6 +127,34 @@ do { \ } \ } while (0) +#define IOMMU_FLUSH_WAIT(what, iommu, offset, op, cond, sts) \ +do { \ + static unsigned int __read_mostly threshold = 1; \ + s_time_t start = NOW(); \ + s_time_t timeout = start + DMAR_OPERATION_TIMEOUT * threshold; \ + \ + for ( ; ; ) \ + { \ + sts = op(iommu->reg, offset); \ + if ( cond ) \ + break; \ + if ( timeout && NOW() > timeout ) \ + { \ + threshold |= threshold << 1; \ + printk(XENLOG_WARNING VTDPREFIX \ + " IOMMU#%u: %s flush taking too long\n", \ + iommu->index, what); \ + timeout = 0; \ + } \ + cpu_relax(); \ + } \ + \ + if ( !timeout ) \ + printk(XENLOG_WARNING VTDPREFIX \ + " IOMMU#%u: %s flush took %lums\n", \ + iommu->index, what, (NOW() - start) / 10000000); \ +} while ( false ) + int vtd_hw_check(void); void disable_pmr(struct vtd_iommu *iommu); int is_igd_drhd(struct acpi_drhd_unit *drhd); diff --git a/xen/drivers/passthrough/vtd/iommu.c b/xen/drivers/passthrough/vtd/iommu.c index 32ff60bfe3..bef3d594c9 100644 --- a/xen/drivers/passthrough/vtd/iommu.c +++ b/xen/drivers/passthrough/vtd/iommu.c @@ -326,8 +326,8 @@ static void iommu_flush_write_buffer(struct vtd_iommu *iommu) dmar_writel(iommu->reg, DMAR_GCMD_REG, val | DMA_GCMD_WBF); /* Make sure hardware complete it */ - IOMMU_WAIT_OP(iommu, DMAR_GSTS_REG, dmar_readl, - !(val & DMA_GSTS_WBFS), val); + IOMMU_FLUSH_WAIT("write buffer", iommu, DMAR_GSTS_REG, dmar_readl, + !(val & DMA_GSTS_WBFS), val); spin_unlock_irqrestore(&iommu->register_lock, flags); } @@ -376,8 +376,8 @@ int vtd_flush_context_reg(struct vtd_iommu *iommu, uint16_t did, dmar_writeq(iommu->reg, DMAR_CCMD_REG, val); /* Make sure hardware complete it */ - IOMMU_WAIT_OP(iommu, DMAR_CCMD_REG, dmar_readq, - !(val & DMA_CCMD_ICC), val); + IOMMU_FLUSH_WAIT("context", iommu, DMAR_CCMD_REG, dmar_readq, + !(val & DMA_CCMD_ICC), val); spin_unlock_irqrestore(&iommu->register_lock, flags); /* flush context entry will implicitly flush write buffer */ @@ -454,8 +454,8 @@ int vtd_flush_iotlb_reg(struct vtd_iommu *iommu, uint16_t did, uint64_t addr, dmar_writeq(iommu->reg, tlb_offset + 8, val); /* Make sure hardware complete it */ - IOMMU_WAIT_OP(iommu, (tlb_offset + 8), dmar_readq, - !(val & DMA_TLB_IVT), val); + IOMMU_FLUSH_WAIT("iotlb", iommu, (tlb_offset + 8), dmar_readq, + !(val & DMA_TLB_IVT), val); spin_unlock_irqrestore(&iommu->register_lock, flags); /* check IOTLB invalidation granularity */ diff --git a/xen/drivers/passthrough/vtd/qinval.c b/xen/drivers/passthrough/vtd/qinval.c index 86a941ce27..1056ab1feb 100644 --- a/xen/drivers/passthrough/vtd/qinval.c +++ b/xen/drivers/passthrough/vtd/qinval.c @@ -29,8 +29,6 @@ #include "extern.h" #include "../ats.h" -#define VTD_QI_TIMEOUT 1 - static unsigned int __read_mostly qi_pg_order; static unsigned int __read_mostly qi_entry_nr; @@ -60,7 +58,11 @@ static unsigned int qinval_next_index(struct vtd_iommu *iommu) /* (tail+1 == head) indicates a full queue, wait for HW */ while ( ((tail + 1) & (qi_entry_nr - 1)) == ( dmar_readq(iommu->reg, DMAR_IQH_REG) >> QINVAL_INDEX_SHIFT ) ) + { + printk_once(XENLOG_ERR VTDPREFIX " IOMMU#%u: no QI slot available\n", + iommu->index); cpu_relax(); + } return tail; } @@ -180,23 +182,32 @@ static int __must_check queue_invalidate_wait(struct vtd_iommu *iommu, /* Now we don't support interrupt method */ if ( sw ) { - s_time_t timeout; - - /* In case all wait descriptor writes to same addr with same data */ - timeout = NOW() + MILLISECS(flush_dev_iotlb ? - iommu_dev_iotlb_timeout : VTD_QI_TIMEOUT); + static unsigned int __read_mostly threshold = 1; + s_time_t start = NOW(); + s_time_t timeout = start + (flush_dev_iotlb + ? iommu_dev_iotlb_timeout + : 100) * MILLISECS(threshold); while ( ACCESS_ONCE(*this_poll_slot) != QINVAL_STAT_DONE ) { - if ( NOW() > timeout ) + if ( timeout && NOW() > timeout ) { - print_qi_regs(iommu); + threshold |= threshold << 1; printk(XENLOG_WARNING VTDPREFIX - " Queue invalidate wait descriptor timed out\n"); - return -ETIMEDOUT; + " IOMMU#%u: QI%s wait descriptor taking too long\n", + iommu->index, flush_dev_iotlb ? " dev" : ""); + print_qi_regs(iommu); + timeout = 0; } cpu_relax(); } + + if ( !timeout ) + printk(XENLOG_WARNING VTDPREFIX + " IOMMU#%u: QI%s wait descriptor took %lums\n", + iommu->index, flush_dev_iotlb ? " dev" : "", + (NOW() - start) / 10000000); + return 0; } -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.14 From xen-changelog-bounces@lists.xenproject.org Wed Jun 09 06:11:54 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Wed, 09 Jun 2021 06:11:54 +0000 Received: from list by lists.xenproject.org with outflank-mailman.139033.257188 (Exim 4.92) (envelope-from ) id 1lqrRS-0005MX-Mg; Wed, 09 Jun 2021 06:11:54 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 139033.257188; Wed, 09 Jun 2021 06:11:54 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqrRS-0005MN-JJ; Wed, 09 Jun 2021 06:11:54 +0000 Received: by outflank-mailman (input) for mailman id 139033; Wed, 09 Jun 2021 06:11:53 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqrRR-0005M8-9p for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 06:11:53 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqrRR-0008GX-92 for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 06:11:53 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqrRR-0001QY-85 for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 06:11:53 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=6/NuXQ27qhzKUy6tOJaWsE0zz6a/3gygc7OnUKwYMmE=; b=QXqX2461do7Y0zAvVgfgRm32nf Jw/T4LniZTKdSgC6eBcExW807UQgxOX9j1ajqFXCuR/6vuK7q5RrWxN/G6cwONC1amfBsTitZhNVW oDG5ybJf3cLSYV+nKfCQ4Iy7FfYeqduCfp+RhB79ue9DKHwgk3CPnlzTtvQVr0KcJBas=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.14] AMD/IOMMU: wait for command slot to be available Message-Id: Date: Wed, 09 Jun 2021 06:11:53 +0000 commit 766b1f447314dd5e8f14c540440cbfc080f3b998 Author: Jan Beulich AuthorDate: Tue Jun 8 18:16:15 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 18:36:03 2021 +0100 AMD/IOMMU: wait for command slot to be available No caller cared about send_iommu_command() indicating unavailability of a slot. Hence if a sufficient number prior commands timed out, we did blindly assume that the requested command was submitted to the IOMMU when really it wasn't. This could mean both a hanging system (waiting for a command to complete that was never seen by the IOMMU) or blindly propagating success back to callers, making them believe they're fine to e.g. free previously unmapped pages. Fold the three involved functions into one, add spin waiting for an available slot along the lines of VT-d's qinval_next_index(), and as a consequence drop all error indicator return types/values. This is part of XSA-373 / CVE-2021-28692. Signed-off-by: Jan Beulich Reviewed-by: Paul Durrant (cherry picked from commit c4e44343579e2c3304d676404d81b2b9bd893c5b) --- xen/drivers/passthrough/amd/iommu_cmd.c | 41 ++++++++++++--------------------- 1 file changed, 15 insertions(+), 26 deletions(-) diff --git a/xen/drivers/passthrough/amd/iommu_cmd.c b/xen/drivers/passthrough/amd/iommu_cmd.c index 760babf20e..fa9603ee45 100644 --- a/xen/drivers/passthrough/amd/iommu_cmd.c +++ b/xen/drivers/passthrough/amd/iommu_cmd.c @@ -20,43 +20,32 @@ #include "iommu.h" #include "../ats.h" -static int queue_iommu_command(struct amd_iommu *iommu, u32 cmd[]) +static void send_iommu_command(struct amd_iommu *iommu, + const uint32_t cmd[4]) { - uint32_t tail, head; + uint32_t tail; tail = iommu->cmd_buffer.tail + sizeof(cmd_entry_t); if ( tail == iommu->cmd_buffer.size ) tail = 0; - head = readl(iommu->mmio_base + - IOMMU_CMD_BUFFER_HEAD_OFFSET) & IOMMU_RING_BUFFER_PTR_MASK; - if ( head != tail ) + while ( tail == (readl(iommu->mmio_base + + IOMMU_CMD_BUFFER_HEAD_OFFSET) & + IOMMU_RING_BUFFER_PTR_MASK) ) { - memcpy(iommu->cmd_buffer.buffer + iommu->cmd_buffer.tail, - cmd, sizeof(cmd_entry_t)); - - iommu->cmd_buffer.tail = tail; - return 1; + printk_once(XENLOG_ERR + "AMD IOMMU %04x:%02x:%02x.%u: no cmd slot available\n", + iommu->seg, PCI_BUS(iommu->bdf), + PCI_SLOT(iommu->bdf), PCI_FUNC(iommu->bdf)); + cpu_relax(); } - return 0; -} - -static void commit_iommu_command_buffer(struct amd_iommu *iommu) -{ - writel(iommu->cmd_buffer.tail, - iommu->mmio_base + IOMMU_CMD_BUFFER_TAIL_OFFSET); -} + memcpy(iommu->cmd_buffer.buffer + iommu->cmd_buffer.tail, + cmd, sizeof(cmd_entry_t)); -static int send_iommu_command(struct amd_iommu *iommu, u32 cmd[]) -{ - if ( queue_iommu_command(iommu, cmd) ) - { - commit_iommu_command_buffer(iommu); - return 1; - } + iommu->cmd_buffer.tail = tail; - return 0; + writel(tail, iommu->mmio_base + IOMMU_CMD_BUFFER_TAIL_OFFSET); } static void flush_command_buffer(struct amd_iommu *iommu) -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.14 From xen-changelog-bounces@lists.xenproject.org Wed Jun 09 06:12:04 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Wed, 09 Jun 2021 06:12:04 +0000 Received: from list by lists.xenproject.org with outflank-mailman.139034.257192 (Exim 4.92) (envelope-from ) id 1lqrRc-0005PJ-OC; Wed, 09 Jun 2021 06:12:04 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 139034.257192; Wed, 09 Jun 2021 06:12:04 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqrRc-0005PB-Kv; Wed, 09 Jun 2021 06:12:04 +0000 Received: by outflank-mailman (input) for mailman id 139034; Wed, 09 Jun 2021 06:12:03 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqrRb-0005Oo-Cp for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 06:12:03 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqrRb-0008Gu-Bz for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 06:12:03 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqrRb-0001Rh-B8 for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 06:12:03 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=a6fcSsa4/d03a0Zh74i5MdtHR6hLBN2QyBXBVOVB3yI=; b=J3Iy60LSM6nL8xB2omm4VobFbn mhc0MwdtA7LR7v1KiNprvoGjVsEjAgKeKLTkipJalmZuPVXPt19XUD4Q60aWQtV6wp22PrMt48M6O wUGlkcvaqqJDcuX7YmXLRZ4VWilZE9dPxCYdxc4Jx/EoA/1aZkfvwXuHs8/BNK5YHxqA=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.14] AMD/IOMMU: drop command completion timeout Message-Id: Date: Wed, 09 Jun 2021 06:12:03 +0000 commit 51278ce831a8fcc4ffb4c841e3442de3fd4ea76b Author: Jan Beulich AuthorDate: Tue Jun 8 18:19:29 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 18:36:03 2021 +0100 AMD/IOMMU: drop command completion timeout First and foremost - such timeouts were not signaled to callers, making them believe they're fine to e.g. free previously unmapped pages. Mirror VT-d's behavior: A fixed number of loop iterations is not a suitable way to detect timeouts in an environment (CPU and bus speeds) independent manner anyway. Furthermore, leaving an in-progress operation pending when it appears to take too long is problematic: If a command completed later, the signaling of its completion may instead be understood to signal a subsequently started command's completion. Log excessively long processing times (with a progressive threshold) to have some indication of problems in this area. Allow callers to specify a non-default timeout bias for this logging, using the same values as VT-d does, which in particular means a (by default) much larger value for device IO TLB invalidation. This is part of XSA-373 / CVE-2021-28692. Signed-off-by: Jan Beulich Reviewed-by: Paul Durrant (cherry picked from commit e4fee66043120c954fc309bbb37813604c1c0eb7) --- xen/drivers/passthrough/amd/iommu_cmd.c | 58 ++++++++++++++++++++------------- 1 file changed, 35 insertions(+), 23 deletions(-) diff --git a/xen/drivers/passthrough/amd/iommu_cmd.c b/xen/drivers/passthrough/amd/iommu_cmd.c index fa9603ee45..fc5377f5f2 100644 --- a/xen/drivers/passthrough/amd/iommu_cmd.c +++ b/xen/drivers/passthrough/amd/iommu_cmd.c @@ -48,10 +48,12 @@ static void send_iommu_command(struct amd_iommu *iommu, writel(tail, iommu->mmio_base + IOMMU_CMD_BUFFER_TAIL_OFFSET); } -static void flush_command_buffer(struct amd_iommu *iommu) +static void flush_command_buffer(struct amd_iommu *iommu, + unsigned int timeout_base) { - unsigned int cmd[4], status, loop_count; - bool comp_wait; + uint32_t cmd[4]; + s_time_t start, timeout; + static unsigned int __read_mostly threshold = 1; /* RW1C 'ComWaitInt' in status register */ writel(IOMMU_STATUS_COMP_WAIT_INT, @@ -67,22 +69,31 @@ static void flush_command_buffer(struct amd_iommu *iommu) IOMMU_COMP_WAIT_I_FLAG_SHIFT, &cmd[0]); send_iommu_command(iommu, cmd); - /* Make loop_count long enough for polling completion wait bit */ - loop_count = 1000; - do { - status = readl(iommu->mmio_base + IOMMU_STATUS_MMIO_OFFSET); - comp_wait = status & IOMMU_STATUS_COMP_WAIT_INT; - --loop_count; - } while ( !comp_wait && loop_count ); - - if ( comp_wait ) + start = NOW(); + timeout = start + (timeout_base ?: 100) * MILLISECS(threshold); + while ( !(readl(iommu->mmio_base + IOMMU_STATUS_MMIO_OFFSET) & + IOMMU_STATUS_COMP_WAIT_INT) ) { - /* RW1C 'ComWaitInt' in status register */ - writel(IOMMU_STATUS_COMP_WAIT_INT, - iommu->mmio_base + IOMMU_STATUS_MMIO_OFFSET); - return; + if ( timeout && NOW() > timeout ) + { + threshold |= threshold << 1; + printk(XENLOG_WARNING + "AMD IOMMU %04x:%02x:%02x.%u: %scompletion wait taking too long\n", + iommu->seg, PCI_BUS(iommu->bdf), + PCI_SLOT(iommu->bdf), PCI_FUNC(iommu->bdf), + timeout_base ? "iotlb " : ""); + timeout = 0; + } + cpu_relax(); } - AMD_IOMMU_DEBUG("Warning: ComWaitInt bit did not assert!\n"); + + if ( !timeout ) + printk(XENLOG_WARNING + "AMD IOMMU %04x:%02x:%02x.%u: %scompletion wait took %lums\n", + iommu->seg, PCI_BUS(iommu->bdf), + PCI_SLOT(iommu->bdf), PCI_FUNC(iommu->bdf), + timeout_base ? "iotlb " : "", + (NOW() - start) / 10000000); } /* Build low level iommu command messages */ @@ -294,7 +305,7 @@ void amd_iommu_flush_iotlb(u8 devfn, const struct pci_dev *pdev, /* send INVALIDATE_IOTLB_PAGES command */ spin_lock_irqsave(&iommu->lock, flags); invalidate_iotlb_pages(iommu, maxpend, 0, queueid, daddr, req_id, order); - flush_command_buffer(iommu); + flush_command_buffer(iommu, iommu_dev_iotlb_timeout); spin_unlock_irqrestore(&iommu->lock, flags); } @@ -331,7 +342,7 @@ static void _amd_iommu_flush_pages(struct domain *d, { spin_lock_irqsave(&iommu->lock, flags); invalidate_iommu_pages(iommu, daddr, dom_id, order); - flush_command_buffer(iommu); + flush_command_buffer(iommu, 0); spin_unlock_irqrestore(&iommu->lock, flags); } @@ -355,7 +366,7 @@ void amd_iommu_flush_device(struct amd_iommu *iommu, uint16_t bdf) ASSERT( spin_is_locked(&iommu->lock) ); invalidate_dev_table_entry(iommu, bdf); - flush_command_buffer(iommu); + flush_command_buffer(iommu, 0); } void amd_iommu_flush_intremap(struct amd_iommu *iommu, uint16_t bdf) @@ -363,7 +374,7 @@ void amd_iommu_flush_intremap(struct amd_iommu *iommu, uint16_t bdf) ASSERT( spin_is_locked(&iommu->lock) ); invalidate_interrupt_table(iommu, bdf); - flush_command_buffer(iommu); + flush_command_buffer(iommu, 0); } void amd_iommu_flush_all_caches(struct amd_iommu *iommu) @@ -371,7 +382,7 @@ void amd_iommu_flush_all_caches(struct amd_iommu *iommu) ASSERT( spin_is_locked(&iommu->lock) ); invalidate_iommu_all(iommu); - flush_command_buffer(iommu); + flush_command_buffer(iommu, 0); } void amd_iommu_send_guest_cmd(struct amd_iommu *iommu, u32 cmd[]) @@ -381,7 +392,8 @@ void amd_iommu_send_guest_cmd(struct amd_iommu *iommu, u32 cmd[]) spin_lock_irqsave(&iommu->lock, flags); send_iommu_command(iommu, cmd); - flush_command_buffer(iommu); + /* TBD: Timeout selection may require peeking into cmd[]. */ + flush_command_buffer(iommu, 0); spin_unlock_irqrestore(&iommu->lock, flags); } -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.14 From xen-changelog-bounces@lists.xenproject.org Wed Jun 09 06:12:14 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Wed, 09 Jun 2021 06:12:14 +0000 Received: from list by lists.xenproject.org with outflank-mailman.139035.257196 (Exim 4.92) (envelope-from ) id 1lqrRm-0005Sd-Q0; Wed, 09 Jun 2021 06:12:14 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 139035.257196; Wed, 09 Jun 2021 06:12:14 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqrRm-0005SR-Mh; Wed, 09 Jun 2021 06:12:14 +0000 Received: by outflank-mailman (input) for mailman id 139035; Wed, 09 Jun 2021 06:12:13 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqrRl-0005SH-GF for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 06:12:13 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqrRl-0008H4-FV for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 06:12:13 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqrRl-0001Se-Eh for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 06:12:13 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=upmQbTdsPP2h3d+MDd7Cdi3g7TNR5PqgTYbApzmcTUM=; b=Q6Nte+2JHlzUTZsoVuaWarrotn 1wT/opEbuDWb9/glX+DXK1o4S7y2xc0VKzkEZVfMaFXLuXm9qJGI50Kc27zJsVQTkqZor8RGozo5S 4XVGNCM2II9cJP4fBXgnv62wOQ/dr9Ji8vppfyamLw0Yb3dApiL5WC4A9kt8OnYpNoHs=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.14] x86/spec-ctrl: Protect against Speculative Code Store Bypass Message-Id: Date: Wed, 09 Jun 2021 06:12:13 +0000 commit fcf98eff7321a5ed8d9b4fccf44cde8ad7e2ec7e Author: Andrew Cooper AuthorDate: Thu Mar 11 14:39:11 2021 +0000 Commit: Andrew Cooper CommitDate: Tue Jun 8 18:36:03 2021 +0100 x86/spec-ctrl: Protect against Speculative Code Store Bypass Modern x86 processors have far-better-than-architecturally-guaranteed self modifying code detection. Typically, when a write hits an instruction in flight, a Machine Clear occurs to flush stale content in the frontend and backend. For self modifying code, before a write which hits an instruction in flight retires, the frontend can speculatively decode and execute the old instruction stream. Speculation of this form can suffer from type confusion in registers, and potentially leak data. Furthermore, updates are typically byte-wise, rather than atomic. Depending on timing, speculation can race ahead multiple times between individual writes, and execute the transiently-malformed instruction stream. Xen has stubs which are used in certain cases for emulation purposes. Inhibit speculation between updating the stub and executing it. This is XSA-375 / CVE-2021-0089. Signed-off-by: Andrew Cooper Reviewed-by: Jan Beulich (cherry picked from commit 45f59ed8865318bb0356954bad067f329677ce9e) --- xen/arch/x86/pv/emul-priv-op.c | 2 ++ xen/arch/x86/x86_emulate/x86_emulate.c | 1 + 2 files changed, 3 insertions(+) diff --git a/xen/arch/x86/pv/emul-priv-op.c b/xen/arch/x86/pv/emul-priv-op.c index 5c7b9117ae..5399fe382f 100644 --- a/xen/arch/x86/pv/emul-priv-op.c +++ b/xen/arch/x86/pv/emul-priv-op.c @@ -138,6 +138,8 @@ static io_emul_stub_t *io_emul_stub_setup(struct priv_op_ctxt *ctxt, u8 opcode, /* Runtime confirmation that we haven't clobbered an adjacent stub. */ BUG_ON(STUB_BUF_SIZE / 2 < (p - ctxt->io_emul_stub)); + block_speculation(); /* SCSB */ + /* Handy function-typed pointer to the stub. */ return (void *)stub_va; diff --git a/xen/arch/x86/x86_emulate/x86_emulate.c b/xen/arch/x86/x86_emulate/x86_emulate.c index 16d164904c..f186ae65fd 100644 --- a/xen/arch/x86/x86_emulate/x86_emulate.c +++ b/xen/arch/x86/x86_emulate/x86_emulate.c @@ -1256,6 +1256,7 @@ static inline int mkec(uint8_t e, int32_t ec, ...) # define invoke_stub(pre, post, constraints...) do { \ stub_exn.info = (union stub_exception_token) { .raw = ~0 }; \ stub_exn.line = __LINE__; /* Utility outweighs livepatching cost */ \ + block_speculation(); /* SCSB */ \ asm volatile ( pre "\n\tINDIRECT_CALL %[stub]\n\t" post "\n" \ ".Lret%=:\n\t" \ ".pushsection .fixup,\"ax\"\n" \ -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.14 From xen-changelog-bounces@lists.xenproject.org Wed Jun 09 06:12:24 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Wed, 09 Jun 2021 06:12:24 +0000 Received: from list by lists.xenproject.org with outflank-mailman.139036.257200 (Exim 4.92) (envelope-from ) id 1lqrRw-0005VI-RP; Wed, 09 Jun 2021 06:12:24 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 139036.257200; Wed, 09 Jun 2021 06:12:24 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqrRw-0005VA-OH; Wed, 09 Jun 2021 06:12:24 +0000 Received: by outflank-mailman (input) for mailman id 139036; Wed, 09 Jun 2021 06:12:23 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqrRv-0005Uv-JI for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 06:12:23 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqrRv-0008HF-Ia for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 06:12:23 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqrRv-0001Ta-Hc for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 06:12:23 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=ztVLECHx/DunTYF6oQjqyS6CEDvqx8co0fkLK514RP0=; b=HvIODvBrv8366U4U+Ub8nb06hO 29Y7Bl7tfu1twA4CwbvW9e1+xRxA9PzXeeP9vTIifdWE6PGbwFp8D8I2b2IwVaGTjNvnGv2rf1HAH DePQxhQLe3Fke1lEfNg+14di4J8HP4w3UWPt5cvzY0NWHB8c6hWwBmhx6G+oKZlaXa3s=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.14] x86/spec-ctrl: Mitigate TAA after S3 resume Message-Id: Date: Wed, 09 Jun 2021 06:12:23 +0000 commit 0ff7f9c5aa02cd2469a8fc03f1ed262f18933721 Author: Andrew Cooper AuthorDate: Thu May 20 01:21:39 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 18:36:03 2021 +0100 x86/spec-ctrl: Mitigate TAA after S3 resume The user chosen setting for MSR_TSX_CTRL needs restoring after S3. All APs get the correct setting via start_secondary(), but the BSP was missed out. This is XSA-377 / CVE-2021-28690. Fixes: 8c4330818f6 ("x86/spec-ctrl: Mitigate the TSX Asynchronous Abort sidechannel") Signed-off-by: Andrew Cooper Reviewed-by: Jan Beulich (cherry picked from commit 8cf276cb2e0b99b96333865873f56b0b31555ff1) --- xen/arch/x86/acpi/power.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/xen/arch/x86/acpi/power.c b/xen/arch/x86/acpi/power.c index 7f162a4df9..604cb7e222 100644 --- a/xen/arch/x86/acpi/power.c +++ b/xen/arch/x86/acpi/power.c @@ -286,6 +286,8 @@ static int enter_state(u32 state) microcode_update_one(); + tsx_init(); /* Needs microcode. May change HLE/RTM feature bits. */ + if ( !recheck_cpu_features(0) ) panic("Missing previously available feature(s)\n"); -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.14 From xen-changelog-bounces@lists.xenproject.org Wed Jun 09 10:55:12 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Wed, 09 Jun 2021 10:55:12 +0000 Received: from list by lists.xenproject.org with outflank-mailman.139203.257468 (Exim 4.92) (envelope-from ) id 1lqvrX-0000wX-7p; Wed, 09 Jun 2021 10:55:07 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 139203.257468; Wed, 09 Jun 2021 10:55:07 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqvrX-0000wP-4t; Wed, 09 Jun 2021 10:55:07 +0000 Received: by outflank-mailman (input) for mailman id 139203; Wed, 09 Jun 2021 10:55:05 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqvrV-0000wD-PA for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 10:55:05 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqvrV-0005S7-Na for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 10:55:05 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqvrV-00033h-ML for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 10:55:05 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=wdcYtBWhE/ic7rBQl3XMLjA4XhcN6l6AtFe5v7OmKUQ=; b=nFjNH8KYD42gDMdPtwzKvwugDV JFvPFkfjIcojLktVvKvw+Cak2gDTDEu5HmpxjgP0cWqduhYSEcAAFBLth13x1IqzDVGvqyqyRgX1v 0JIuPuNA75LDnse/S9e/Fa9LI8pRYQtqiWD9gDob3QVIrrvD3wloM5KCNf0cgMs6O/fY=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] MAINTAINERS: adjust x86/mm/shadow maintainers Message-Id: Date: Wed, 09 Jun 2021 10:55:05 +0000 commit 87d49efd3af7ea70e1b8250d619411e57da9e520 Author: Tim Deegan AuthorDate: Wed Jun 9 12:43:25 2021 +0200 Commit: Jan Beulich CommitDate: Wed Jun 9 12:43:25 2021 +0200 MAINTAINERS: adjust x86/mm/shadow maintainers Better reflect reality: Andrew and Jan are active maintainers and I review patches. Keep myself as a reviewer so I can help with historical context &c. Signed-off-by: Tim Deegan Acked-by: Jan Beulich Acked-by: Andrew Cooper --- MAINTAINERS | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/MAINTAINERS b/MAINTAINERS index d46b08a0d2..09a92534bf 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -591,7 +591,9 @@ F: xen/arch/x86/mm/mem_sharing.c F: tools/tests/mem-sharing/ X86 SHADOW PAGETABLES -M: Tim Deegan +M: Jan Beulich +M: Andrew Cooper +R: Tim Deegan S: Maintained F: xen/arch/x86/mm/shadow/ -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Wed Jun 09 10:55:17 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Wed, 09 Jun 2021 10:55:17 +0000 Received: from list by lists.xenproject.org with outflank-mailman.139204.257472 (Exim 4.92) (envelope-from ) id 1lqvrh-0000zH-9F; Wed, 09 Jun 2021 10:55:17 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 139204.257472; Wed, 09 Jun 2021 10:55:17 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqvrh-0000z9-6Q; Wed, 09 Jun 2021 10:55:17 +0000 Received: by outflank-mailman (input) for mailman id 139204; Wed, 09 Jun 2021 10:55:15 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqvrf-0000yi-Rc for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 10:55:15 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqvrf-0005SH-Qt for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 10:55:15 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqvrf-00034p-Pt for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 10:55:15 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=8ZqTu/rH2gAr2hT94TmaPOQ6edhWMWR9qXlK95uxuPw=; b=3CvuZwRo1ElUXptE97gv1pEWox igPBbiNmLNaW8dxlOF5PPGPIagea5SJkdGPGOWGREr6bQIcGLWY0kC+lSHQUmhMARyEmTnkdSDpL+ agNkvCbQF1UOU3avnTVNdy+XpJL7pQ31VFPgwDeo3kGDjuzD42zZkTVj/cRtx0Wdigr8=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] xen/char: default HAS_NS16550 to y only for X86 and ARM Message-Id: Date: Wed, 09 Jun 2021 10:55:15 +0000 commit 5151ce8ba1f155ecc6403f3a51e618a47a382262 Author: Connor Davis AuthorDate: Wed Jun 9 12:43:45 2021 +0200 Commit: Jan Beulich CommitDate: Wed Jun 9 12:43:45 2021 +0200 xen/char: default HAS_NS16550 to y only for X86 and ARM Defaulting to yes only for X86 and ARM reduces the requirements for a minimal build when porting new architectures. Signed-off-by: Connor Davis Acked-by: Jan Beulich Reviewed-by: Alistair Francis --- xen/drivers/char/Kconfig | 1 + 1 file changed, 1 insertion(+) diff --git a/xen/drivers/char/Kconfig b/xen/drivers/char/Kconfig index b572305657..2ff5b288e2 100644 --- a/xen/drivers/char/Kconfig +++ b/xen/drivers/char/Kconfig @@ -1,5 +1,6 @@ config HAS_NS16550 bool "NS16550 UART driver" if ARM + default n if RISCV default y help This selects the 16550-series UART support. For most systems, say Y. -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Wed Jun 09 10:55:27 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Wed, 09 Jun 2021 10:55:27 +0000 Received: from list by lists.xenproject.org with outflank-mailman.139205.257475 (Exim 4.92) (envelope-from ) id 1lqvrr-00012A-B8; Wed, 09 Jun 2021 10:55:27 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 139205.257475; Wed, 09 Jun 2021 10:55:27 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqvrr-000122-7u; Wed, 09 Jun 2021 10:55:27 +0000 Received: by outflank-mailman (input) for mailman id 139205; Wed, 09 Jun 2021 10:55:26 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqvrp-00011g-Uh for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 10:55:25 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqvrp-0005SY-Tt for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 10:55:25 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqvrp-00035R-Sz for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 10:55:25 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=4wEZLCQIVcTBe8jl/6pKnJSruQAb8FJtJ81liami9B8=; b=kl9U5f3VN54EalZ1B/9LXdB30T HbS15o1KJR+xzql6UHJeytxw0PzijImxrDS4NJisPiXQfl5zSfp1ffOcsG+0vOPfHq9VNZqDKWhyL nl6BjS87tXsWAxHOyFC63n5eX9K7x0L0YKTfY9iVh3rm5E0+2QKJvS6WVDkT39IrxeOs=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] xen: add files needed for minimal riscv build Message-Id: Date: Wed, 09 Jun 2021 10:55:25 +0000 commit f5035d480f7a7033f15765b67c19df86a8ef2c69 Author: Connor Davis AuthorDate: Wed Jun 9 12:45:38 2021 +0200 Commit: Jan Beulich CommitDate: Wed Jun 9 12:45:38 2021 +0200 xen: add files needed for minimal riscv build Add arch-specific makefiles and configs needed to build for riscv. Also add a minimal head.S that is a simple infinite loop. head.o can be built with $ make XEN_TARGET_ARCH=riscv64 SUBSYSTEMS=xen -C xen tiny64_defconfig $ make XEN_TARGET_ARCH=riscv64 SUBSYSTEMS=xen -C xen TARGET=riscv64/head.o No other TARGET is supported at the moment. Signed-off-by: Connor Davis Reviewed-by: Alistair Francis Acked-by: Bobby Eshleman --- MAINTAINERS | 9 +++++++ config/riscv64.mk | 5 ++++ xen/Makefile | 8 ++++-- xen/arch/riscv/Kconfig | 48 +++++++++++++++++++++++++++++++++ xen/arch/riscv/Kconfig.debug | 0 xen/arch/riscv/Makefile | 2 ++ xen/arch/riscv/Rules.mk | 0 xen/arch/riscv/arch.mk | 14 ++++++++++ xen/arch/riscv/configs/tiny64_defconfig | 13 +++++++++ xen/arch/riscv/riscv64/asm-offsets.c | 0 xen/arch/riscv/riscv64/head.S | 6 +++++ xen/include/asm-riscv/config.h | 47 ++++++++++++++++++++++++++++++++ 12 files changed, 150 insertions(+), 2 deletions(-) diff --git a/MAINTAINERS b/MAINTAINERS index 09a92534bf..7778c64fc7 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -456,6 +456,15 @@ F: tools/libs/light/libxl_nonetbuffer.c F: tools/hotplug/Linux/remus-netbuf-setup F: tools/hotplug/Linux/block-drbd-probe +RISCV +M: Bob Eshleman +M: Alistair Francis +R: Connor Davis +S: Supported +F: config/riscv64.mk +F: xen/arch/riscv/ +F: xen/include/asm-riscv/ + RTDS SCHEDULER M: Dario Faggioli M: Meng Xu diff --git a/config/riscv64.mk b/config/riscv64.mk new file mode 100644 index 0000000000..a5a21e5fa2 --- /dev/null +++ b/config/riscv64.mk @@ -0,0 +1,5 @@ +CONFIG_RISCV := y +CONFIG_RISCV_64 := y +CONFIG_RISCV_$(XEN_OS) := y + +CONFIG_XEN_INSTALL_SUFFIX := diff --git a/xen/Makefile b/xen/Makefile index 7ce7692354..89879fad4c 100644 --- a/xen/Makefile +++ b/xen/Makefile @@ -26,7 +26,9 @@ MAKEFLAGS += -rR EFI_MOUNTPOINT ?= $(BOOT_DIR)/efi ARCH=$(XEN_TARGET_ARCH) -SRCARCH=$(shell echo $(ARCH) | sed -e 's/x86.*/x86/' -e s'/arm\(32\|64\)/arm/g') +SRCARCH=$(shell echo $(ARCH) | \ + sed -e 's/x86.*/x86/' -e s'/arm\(32\|64\)/arm/g' \ + -e s'/riscv.*/riscv/g') # Don't break if the build process wasn't called from the top level # we need XEN_TARGET_ARCH to generate the proper config @@ -35,7 +37,8 @@ include $(XEN_ROOT)/Config.mk # Set ARCH/SUBARCH appropriately. export TARGET_SUBARCH := $(XEN_TARGET_ARCH) export TARGET_ARCH := $(shell echo $(XEN_TARGET_ARCH) | \ - sed -e 's/x86.*/x86/' -e s'/arm\(32\|64\)/arm/g') + sed -e 's/x86.*/x86/' -e s'/arm\(32\|64\)/arm/g' \ + -e s'/riscv.*/riscv/g') # Allow someone to change their config file export KCONFIG_CONFIG ?= .config @@ -335,6 +338,7 @@ _clean: delete-unfresh-files $(MAKE) $(clean) xsm $(MAKE) $(clean) crypto $(MAKE) $(clean) arch/arm + $(MAKE) $(clean) arch/riscv $(MAKE) $(clean) arch/x86 $(MAKE) $(clean) test $(MAKE) -f $(BASEDIR)/tools/kconfig/Makefile.kconfig ARCH=$(ARCH) SRCARCH=$(SRCARCH) clean diff --git a/xen/arch/riscv/Kconfig b/xen/arch/riscv/Kconfig new file mode 100644 index 0000000000..468e250c86 --- /dev/null +++ b/xen/arch/riscv/Kconfig @@ -0,0 +1,48 @@ +config RISCV + def_bool y + +config RISCV_64 + def_bool y + select 64BIT + +config ARCH_DEFCONFIG + string + default "arch/riscv/configs/tiny64_defconfig" + +menu "Architecture Features" + +source "arch/Kconfig" + +endmenu + +menu "ISA Selection" + +choice + prompt "Base ISA" + default RISCV_ISA_RV64IMA if RISCV_64 + help + This selects the base ISA extensions that Xen will target. + +config RISCV_ISA_RV64IMA + bool "RV64IMA" + help + Use the RV64I base ISA, plus the "M" and "A" extensions + for integer multiply/divide and atomic instructions, respectively. + +endchoice + +config RISCV_ISA_C + bool "Compressed extension" + default y + help + Add "C" to the ISA subsets that the toolchain is allowed to + emit when building Xen, which results in compressed instructions + in the Xen binary. + + If unsure, say Y. + +endmenu + +source "common/Kconfig" + +source "drivers/Kconfig" diff --git a/xen/arch/riscv/Kconfig.debug b/xen/arch/riscv/Kconfig.debug new file mode 100644 index 0000000000..e69de29bb2 diff --git a/xen/arch/riscv/Makefile b/xen/arch/riscv/Makefile new file mode 100644 index 0000000000..942e4ffbc1 --- /dev/null +++ b/xen/arch/riscv/Makefile @@ -0,0 +1,2 @@ +.PHONY: include +include: diff --git a/xen/arch/riscv/Rules.mk b/xen/arch/riscv/Rules.mk new file mode 100644 index 0000000000..e69de29bb2 diff --git a/xen/arch/riscv/arch.mk b/xen/arch/riscv/arch.mk new file mode 100644 index 0000000000..53dadb8975 --- /dev/null +++ b/xen/arch/riscv/arch.mk @@ -0,0 +1,14 @@ +######################################## +# RISCV-specific definitions + +CFLAGS-$(CONFIG_RISCV_64) += -mabi=lp64 + +riscv-march-$(CONFIG_RISCV_ISA_RV64IMA) := rv64ima +riscv-march-$(CONFIG_RISCV_ISA_C) := $(riscv-march-y)c + +# Note that -mcmodel=medany is used so that Xen can be mapped +# into the upper half _or_ the lower half of the address space. +# -mcmodel=medlow would force Xen into the lower half. + +CFLAGS += -march=$(riscv-march-y) -mstrict-align -mcmodel=medany +CFLAGS += -I$(BASEDIR)/include diff --git a/xen/arch/riscv/configs/tiny64_defconfig b/xen/arch/riscv/configs/tiny64_defconfig new file mode 100644 index 0000000000..3c9a2ff941 --- /dev/null +++ b/xen/arch/riscv/configs/tiny64_defconfig @@ -0,0 +1,13 @@ +# CONFIG_SCHED_CREDIT is not set +# CONFIG_SCHED_RTDS is not set +# CONFIG_SCHED_NULL is not set +# CONFIG_SCHED_ARINC653 is not set +# CONFIG_TRACEBUFFER is not set +# CONFIG_HYPFS is not set +# CONFIG_GRANT_TABLE is not set +# CONFIG_SPECULATIVE_HARDEN_ARRAY is not set + +CONFIG_RISCV_64=y +CONFIG_DEBUG=y +CONFIG_DEBUG_INFO=y +CONFIG_EXPERT=y diff --git a/xen/arch/riscv/riscv64/asm-offsets.c b/xen/arch/riscv/riscv64/asm-offsets.c new file mode 100644 index 0000000000..e69de29bb2 diff --git a/xen/arch/riscv/riscv64/head.S b/xen/arch/riscv/riscv64/head.S new file mode 100644 index 0000000000..0dbc27ba75 --- /dev/null +++ b/xen/arch/riscv/riscv64/head.S @@ -0,0 +1,6 @@ +#include + + .text + +ENTRY(start) + j start diff --git a/xen/include/asm-riscv/config.h b/xen/include/asm-riscv/config.h new file mode 100644 index 0000000000..e2ae21de61 --- /dev/null +++ b/xen/include/asm-riscv/config.h @@ -0,0 +1,47 @@ +#ifndef __RISCV_CONFIG_H__ +#define __RISCV_CONFIG_H__ + +#if defined(CONFIG_RISCV_64) +# define LONG_BYTEORDER 3 +# define ELFSIZE 64 +# define MAX_VIRT_CPUS 128u +#else +# error "Unsupported RISCV variant" +#endif + +#define BYTES_PER_LONG (1 << LONG_BYTEORDER) +#define BITS_PER_LONG (BYTES_PER_LONG << 3) +#define POINTER_ALIGN BYTES_PER_LONG + +#define BITS_PER_LLONG 64 + +/* xen_ulong_t is always 64 bits */ +#define BITS_PER_XEN_ULONG 64 + +#define CONFIG_RISCV_L1_CACHE_SHIFT 6 +#define CONFIG_PAGEALLOC_MAX_ORDER 18 +#define CONFIG_DOMU_MAX_ORDER 9 +#define CONFIG_HWDOM_MAX_ORDER 10 + +#define OPT_CONSOLE_STR "dtuart" +#define INVALID_VCPU_ID MAX_VIRT_CPUS + +/* Linkage for RISCV */ +#ifdef __ASSEMBLY__ +#define ALIGN .align 2 + +#define ENTRY(name) \ + .globl name; \ + ALIGN; \ + name: +#endif + +#endif /* __RISCV_CONFIG_H__ */ +/* + * Local variables: + * mode: C + * c-file-style: "BSD" + * c-basic-offset: 4 + * indent-tabs-mode: nil + * End: + */ -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Wed Jun 09 13:22:11 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Wed, 09 Jun 2021 13:22:11 +0000 Received: from list by lists.xenproject.org with outflank-mailman.139383.257759 (Exim 4.92) (envelope-from ) id 1lqy9m-0003vh-5E; Wed, 09 Jun 2021 13:22:06 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 139383.257759; Wed, 09 Jun 2021 13:22:06 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqy9m-0003vZ-2H; Wed, 09 Jun 2021 13:22:06 +0000 Received: by outflank-mailman (input) for mailman id 139383; Wed, 09 Jun 2021 13:22:05 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqy9l-0003vR-CY for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 13:22:05 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lqy9l-00086M-7O for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 13:22:05 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lqy9l-0006mq-5l for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 13:22:05 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=PHsf2qK+tSGoVJvekV/Z0HNBf4KnSvjBAS8ysoUz2uY=; b=vgkB6Mfz461SatJxkOwREh/jZO jStd/Y8t0ZXeETZMY4uEjUW3OrOeIsog1kmLa8F0OIsPW1MMYmtPZnhU5Nx2sgO2GwNHw2MpeFNN+ s+KkpfYeA5Q7aAlrQjmmsCUo9iZIaxxJiq1UyirptwApBcAW0+VihTnL7ecnvKAharOU=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] x86/tsx: Cope with TSX deprecation on SKL/KBL/CFL/WHL Message-Id: Date: Wed, 09 Jun 2021 13:22:05 +0000 commit 3e09045991cde360432bc7437103f8f8a6699359 Author: Andrew Cooper AuthorDate: Wed Sep 16 16:15:52 2020 +0100 Commit: Andrew Cooper CommitDate: Wed Jun 9 14:17:58 2021 +0100 x86/tsx: Cope with TSX deprecation on SKL/KBL/CFL/WHL The June 2021 microcode is formally de-featuring TSX on the older Skylake client CPUs. The workaround from the March 2019 microcode is being dropped, and replaced with additions to MSR_TSX_FORCE_ABORT to hide the HLE/RTM CPUID bits. With this microcode in place, TSX is disabled by default on these CPUs. Backwards compatibility is provided in the same way as for TAA - RTM force aborts, rather than suffering #UD, and the CPUID bits can be hidden to recover performance. Signed-off-by: Andrew Cooper Reviewed-by: Jan Beulich --- docs/misc/xen-command-line.pandoc | 13 +++++ tools/misc/xen-cpuid.c | 2 +- xen/arch/x86/tsx.c | 76 +++++++++++++++++++++++++++-- xen/include/asm-x86/cpufeature.h | 1 + xen/include/asm-x86/msr-index.h | 2 + xen/include/public/arch-x86/cpufeatureset.h | 1 + 6 files changed, 89 insertions(+), 6 deletions(-) diff --git a/docs/misc/xen-command-line.pandoc b/docs/misc/xen-command-line.pandoc index 1fae872626..3ece83a427 100644 --- a/docs/misc/xen-command-line.pandoc +++ b/docs/misc/xen-command-line.pandoc @@ -2309,6 +2309,12 @@ Several microcode updates are relevant: and formally retiring HLE from the architecture. The user can disable TSX to mitigate TAA, and elect to hide the HLE/RTM CPUID bits. + * June 2021, removing the workaround for March 2019 on client CPUs and + formally de-featured TSX on SKL/KBL/WHL/CFL (Note: SKX still retains the + March 2019 fix). Introduced the ability to hide the HLE/RTM CPUID bits. + PCR3 works fine, and TSX is disabled by default, but the user can re-enable + TSX at their own risk, accepting that the memory order erratum is unfixed. + On systems with the ability to configure TSX, this boolean offers system wide control of whether TSX is enabled or disabled. @@ -2326,6 +2332,13 @@ control of whether TSX is enabled or disabled. ordering errata default to `true` to enable working TSX. Alternatively, selecting `tsx=0` will disable TSX and restore PCR3 to a working state. + SKX and SKL/KBL/WHL/CFL on pre-June 2021 microcode default to `true`. + Alternatively, selecting `tsx=0` will disable TSX and restore PCR3 to a + working state. + + SKL/KBL/WHL/CFL on the June 2021 microcode or later default to `false`. + Alternatively, selecting `tsx=1` will re-enable TSX at the users own risk. + ### ucode > `= List of [ | scan=, nmi=, allow-same= ]` diff --git a/tools/misc/xen-cpuid.c b/tools/misc/xen-cpuid.c index d4bc83d8c9..735bcf8f0e 100644 --- a/tools/misc/xen-cpuid.c +++ b/tools/misc/xen-cpuid.c @@ -162,7 +162,7 @@ static const char *const str_7d0[32] = [ 4] = "fsrm", [ 8] = "avx512-vp2intersect", [ 9] = "srbds-ctrl", - [10] = "md-clear", + [10] = "md-clear", [11] = "rtm-always-abort", /* 12 */ [13] = "tsx-force-abort", [14] = "serialize", diff --git a/xen/arch/x86/tsx.c b/xen/arch/x86/tsx.c index 338191df7f..88adf08c49 100644 --- a/xen/arch/x86/tsx.c +++ b/xen/arch/x86/tsx.c @@ -59,6 +59,38 @@ void tsx_init(void) * ordering erratum, with at least the March 2019 microcode. */ + /* + * Probe for the June 2021 microcode which de-features TSX on + * client parts. (Note - this is a subset of parts impacted by + * the memory ordering errata.) + * + * RTM_ALWAYS_ABORT enumerates the new functionality, but is also + * read as zero if TSX_FORCE_ABORT.ENABLE_RTM has been set before + * we run. + * + * Undo this behaviour in Xen's view of the world. + */ + bool has_rtm_always_abort = cpu_has_rtm_always_abort; + + if ( !has_rtm_always_abort ) + { + uint64_t val; + + rdmsrl(MSR_TSX_FORCE_ABORT, val); + + if ( val & TSX_ENABLE_RTM ) + has_rtm_always_abort = true; + } + + /* + * Always force RTM_ALWAYS_ABORT, even if it currently visible. + * If the user explicitly opts to enable TSX, we'll set + * TSX_FORCE_ABORT.ENABLE_RTM and cause RTM_ALWAYS_ABORT to be + * hidden from the general CPUID scan later. + */ + if ( has_rtm_always_abort ) + setup_force_cpu_cap(X86_FEATURE_RTM_ALWAYS_ABORT); + /* * If no explicit tsx= option is provided, pick a default. * @@ -67,9 +99,16 @@ void tsx_init(void) * - parse_spec_ctrl() ran before any CPU details where know. * - We now know we're running on a CPU not affected by TAA (as * TSX_FORCE_ABORT is enumerated). + * - When RTM_ALWAYS_ABORT is enumerated, TSX malfunctions, so we + * only ever want it enabled by explicit user choice. + * + * Without RTM_ALWAYS_ABORT, leave TSX active. In particular, + * this includes SKX where TSX is still supported. + * + * With RTM_ALWAYS_ABORT, disable TSX. */ if ( opt_tsx < 0 ) - opt_tsx = 1; + opt_tsx = !cpu_has_rtm_always_abort; } /* @@ -90,7 +129,7 @@ void tsx_init(void) * Force the features to be visible in Xen's view if we see any of the * infrastructure capable of hiding them. */ - if ( cpu_has_tsx_ctrl ) + if ( cpu_has_tsx_ctrl || cpu_has_tsx_force_abort ) { setup_force_cpu_cap(X86_FEATURE_HLE); setup_force_cpu_cap(X86_FEATURE_RTM); @@ -131,9 +170,36 @@ void tsx_init(void) /* Check bottom bit only. Higher bits are various sentinels. */ rtm_disabled = !(opt_tsx & 1); - lo &= ~TSX_FORCE_ABORT_RTM; - if ( rtm_disabled ) - lo |= TSX_FORCE_ABORT_RTM; + lo &= ~(TSX_FORCE_ABORT_RTM | TSX_CPUID_CLEAR | TSX_ENABLE_RTM); + + if ( cpu_has_rtm_always_abort ) + { + /* + * June 2021 microcode, on a client part with TSX de-featured: + * - There are no mitigations for the TSX memory ordering errata. + * - Performance counter 3 works. (I.e. it isn't being used by + * microcode to work around the memory ordering errata.) + * - TSX_FORCE_ABORT.FORCE_ABORT_RTM is fixed read1/write-discard. + * - TSX_FORCE_ABORT.TSX_CPUID_CLEAR can be used to hide the + * HLE/RTM CPUID bits. + * - TSX_FORCE_ABORT.ENABLE_RTM may be used to opt in to + * re-enabling RTM, at the users own risk. + */ + lo |= rtm_disabled ? TSX_CPUID_CLEAR : TSX_ENABLE_RTM; + } + else + { + /* + * Either a server part where TSX isn't de-featured, or pre-June + * 2021 microcode: + * - By default, the TSX memory ordering errata is worked around + * in microcode at the cost of Performance Counter 3. + * - "Working TSX" vs "Working PCR3" can be selected by way of + * setting TSX_FORCE_ABORT.FORCE_ABORT_RTM. + */ + if ( rtm_disabled ) + lo |= TSX_FORCE_ABORT_RTM; + } wrmsr(MSR_TSX_FORCE_ABORT, lo, hi); } diff --git a/xen/include/asm-x86/cpufeature.h b/xen/include/asm-x86/cpufeature.h index 9f5ae3aa0d..a539a4bacd 100644 --- a/xen/include/asm-x86/cpufeature.h +++ b/xen/include/asm-x86/cpufeature.h @@ -131,6 +131,7 @@ #define cpu_has_avx512_4vnniw boot_cpu_has(X86_FEATURE_AVX512_4VNNIW) #define cpu_has_avx512_4fmaps boot_cpu_has(X86_FEATURE_AVX512_4FMAPS) #define cpu_has_avx512_vp2intersect boot_cpu_has(X86_FEATURE_AVX512_VP2INTERSECT) +#define cpu_has_rtm_always_abort boot_cpu_has(X86_FEATURE_RTM_ALWAYS_ABORT) #define cpu_has_tsx_force_abort boot_cpu_has(X86_FEATURE_TSX_FORCE_ABORT) #define cpu_has_serialize boot_cpu_has(X86_FEATURE_SERIALIZE) #define cpu_has_arch_caps boot_cpu_has(X86_FEATURE_ARCH_CAPS) diff --git a/xen/include/asm-x86/msr-index.h b/xen/include/asm-x86/msr-index.h index bd3a3a1e7f..9a772c12b8 100644 --- a/xen/include/asm-x86/msr-index.h +++ b/xen/include/asm-x86/msr-index.h @@ -61,6 +61,8 @@ #define MSR_TSX_FORCE_ABORT 0x0000010f #define TSX_FORCE_ABORT_RTM (_AC(1, ULL) << 0) +#define TSX_CPUID_CLEAR (_AC(1, ULL) << 1) +#define TSX_ENABLE_RTM (_AC(1, ULL) << 2) #define MSR_TSX_CTRL 0x00000122 #define TSX_CTRL_RTM_DISABLE (_AC(1, ULL) << 0) diff --git a/xen/include/public/arch-x86/cpufeatureset.h b/xen/include/public/arch-x86/cpufeatureset.h index b65af42436..380b51b1b3 100644 --- a/xen/include/public/arch-x86/cpufeatureset.h +++ b/xen/include/public/arch-x86/cpufeatureset.h @@ -265,6 +265,7 @@ XEN_CPUFEATURE(FSRM, 9*32+ 4) /*A Fast Short REP MOVS */ XEN_CPUFEATURE(AVX512_VP2INTERSECT, 9*32+8) /*a VP2INTERSECT{D,Q} insns */ XEN_CPUFEATURE(SRBDS_CTRL, 9*32+ 9) /* MSR_MCU_OPT_CTRL and RNGDS_MITG_DIS. */ XEN_CPUFEATURE(MD_CLEAR, 9*32+10) /*A VERW clears microarchitectural buffers */ +XEN_CPUFEATURE(RTM_ALWAYS_ABORT, 9*32+11) /*! June 2021 TSX defeaturing in microcode. */ XEN_CPUFEATURE(TSX_FORCE_ABORT, 9*32+13) /* MSR_TSX_FORCE_ABORT.RTM_ABORT */ XEN_CPUFEATURE(SERIALIZE, 9*32+14) /*a SERIALIZE insn */ XEN_CPUFEATURE(CET_IBT, 9*32+20) /* CET - Indirect Branch Tracking */ -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Wed Jun 09 16:33:10 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Wed, 09 Jun 2021 16:33:10 +0000 Received: from list by lists.xenproject.org with outflank-mailman.139476.257876 (Exim 4.92) (envelope-from ) id 1lr18a-0000ou-9h; Wed, 09 Jun 2021 16:33:04 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 139476.257876; Wed, 09 Jun 2021 16:33:04 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr18a-0000om-6l; Wed, 09 Jun 2021 16:33:04 +0000 Received: by outflank-mailman (input) for mailman id 139476; Wed, 09 Jun 2021 16:33:02 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr18Y-0000og-N2 for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 16:33:02 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr18Y-0003Z4-M0 for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 16:33:02 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lr18Y-0000cu-Kv for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 16:33:02 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=skb8A/6daK+Iy/o/ABQTIl0x8TQtnJhVkPua5dWWnWU=; b=j4mhCZOAiOC+yWMZ/DoeFUGC8+ xnLJvVcAAgcyYmwHrGWKkQgnlItE91Mcgssf3kcotH+yBVKk7/cf5112UDM3XQwZWwTCjrr9Nv2zo LjsMeaQGPKwhjz8cqIZLiPJ37W0n1s+E986TORI6eLTnYBPIH6Yjj4NzHIVIG6UopW7I=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.11] VT-d: size qinval queue dynamically Message-Id: Date: Wed, 09 Jun 2021 16:33:02 +0000 commit 37518c21cf97b2d333403f994a772cbbbe5bedfc Author: Jan Beulich AuthorDate: Tue Jun 8 19:15:06 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 19:16:38 2021 +0100 VT-d: size qinval queue dynamically With the present synchronous model, we need two slots for every operation (the operation itself and a wait descriptor). There can be one such pair of requests pending per CPU. To ensure that under all normal circumstances a slot is always available when one is requested, size the queue ring according to the number of present CPUs. This is part of XSA-373 / CVE-2021-28692. Signed-off-by: Jan Beulich Reviewed-by: Paul Durrant (cherry picked from commit cbfa62bb140c8f10a0ae96db3ce06e22b3b9d302) --- xen/drivers/passthrough/vtd/iommu.h | 10 +------- xen/drivers/passthrough/vtd/qinval.c | 44 ++++++++++++++++++++++++++++-------- 2 files changed, 35 insertions(+), 19 deletions(-) diff --git a/xen/drivers/passthrough/vtd/iommu.h b/xen/drivers/passthrough/vtd/iommu.h index e76b261c81..fc52a3249e 100644 --- a/xen/drivers/passthrough/vtd/iommu.h +++ b/xen/drivers/passthrough/vtd/iommu.h @@ -447,17 +447,9 @@ struct qinval_entry { }q; }; -/* Order of queue invalidation pages(max is 8) */ -#define QINVAL_PAGE_ORDER 2 - -#define QINVAL_ARCH_PAGE_ORDER (QINVAL_PAGE_ORDER + PAGE_SHIFT_4K - PAGE_SHIFT) -#define QINVAL_ARCH_PAGE_NR ( QINVAL_ARCH_PAGE_ORDER < 0 ? \ - 1 : \ - 1 << QINVAL_ARCH_PAGE_ORDER ) - /* Each entry is 16 bytes, so 2^8 entries per page */ #define QINVAL_ENTRY_ORDER ( PAGE_SHIFT - 4 ) -#define QINVAL_ENTRY_NR (1 << (QINVAL_PAGE_ORDER + 8)) +#define QINVAL_MAX_ENTRY_NR (1u << (7 + QINVAL_ENTRY_ORDER)) /* Status data flag */ #define QINVAL_STAT_INIT 0 diff --git a/xen/drivers/passthrough/vtd/qinval.c b/xen/drivers/passthrough/vtd/qinval.c index 09cbd36ebb..103afc15fe 100644 --- a/xen/drivers/passthrough/vtd/qinval.c +++ b/xen/drivers/passthrough/vtd/qinval.c @@ -31,6 +31,9 @@ #define VTD_QI_TIMEOUT 1 +static unsigned int __read_mostly qi_pg_order; +static unsigned int __read_mostly qi_entry_nr; + static int __must_check invalidate_sync(struct iommu *iommu); static void print_qi_regs(struct iommu *iommu) @@ -55,7 +58,7 @@ static unsigned int qinval_next_index(struct iommu *iommu) tail >>= QINVAL_INDEX_SHIFT; /* (tail+1 == head) indicates a full queue, wait for HW */ - while ( ( tail + 1 ) % QINVAL_ENTRY_NR == + while ( ((tail + 1) & (qi_entry_nr - 1)) == ( dmar_readq(iommu->reg, DMAR_IQH_REG) >> QINVAL_INDEX_SHIFT ) ) cpu_relax(); @@ -68,7 +71,7 @@ static void qinval_update_qtail(struct iommu *iommu, unsigned int index) /* Need hold register lock when update tail */ ASSERT( spin_is_locked(&iommu->register_lock) ); - val = (index + 1) % QINVAL_ENTRY_NR; + val = (index + 1) & (qi_entry_nr - 1); dmar_writeq(iommu->reg, DMAR_IQT_REG, (val << QINVAL_INDEX_SHIFT)); } @@ -417,7 +420,27 @@ int enable_qinval(struct iommu *iommu) if ( qi_ctrl->qinval_maddr == 0 ) { drhd = iommu_to_drhd(iommu); - qi_ctrl->qinval_maddr = alloc_pgtable_maddr(drhd, QINVAL_ARCH_PAGE_NR); + if ( !qi_entry_nr ) + { + /* + * With the present synchronous model, we need two slots for every + * operation (the operation itself and a wait descriptor). There + * can be one such pair of requests pending per CPU. One extra + * entry is needed as the ring is considered full when there's + * only one entry left. + */ + BUILD_BUG_ON(CONFIG_NR_CPUS * 2 >= QINVAL_MAX_ENTRY_NR); + qi_pg_order = get_order_from_bytes((num_present_cpus() * 2 + 1) << + (PAGE_SHIFT - + QINVAL_ENTRY_ORDER)); + qi_entry_nr = 1u << (qi_pg_order + QINVAL_ENTRY_ORDER); + + dprintk(XENLOG_INFO VTDPREFIX, + "QI: using %u-entry ring(s)\n", qi_entry_nr); + } + + qi_ctrl->qinval_maddr = + alloc_pgtable_maddr(drhd, qi_entry_nr >> QINVAL_ENTRY_ORDER); if ( qi_ctrl->qinval_maddr == 0 ) { dprintk(XENLOG_WARNING VTDPREFIX, @@ -431,15 +454,16 @@ int enable_qinval(struct iommu *iommu) spin_lock_irqsave(&iommu->register_lock, flags); - /* Setup Invalidation Queue Address(IQA) register with the - * address of the page we just allocated. QS field at - * bits[2:0] to indicate size of queue is one 4KB page. - * That's 256 entries. Queued Head (IQH) and Queue Tail (IQT) - * registers are automatically reset to 0 with write - * to IQA register. + /* + * Setup Invalidation Queue Address (IQA) register with the address of the + * pages we just allocated. The QS field at bits[2:0] indicates the size + * (page order) of the queue. + * + * Queued Head (IQH) and Queue Tail (IQT) registers are automatically + * reset to 0 with write to IQA register. */ dmar_writeq(iommu->reg, DMAR_IQA_REG, - qi_ctrl->qinval_maddr | QINVAL_PAGE_ORDER); + qi_ctrl->qinval_maddr | qi_pg_order); dmar_writeq(iommu->reg, DMAR_IQT_REG, 0); -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.11 From xen-changelog-bounces@lists.xenproject.org Wed Jun 09 16:33:14 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Wed, 09 Jun 2021 16:33:14 +0000 Received: from list by lists.xenproject.org with outflank-mailman.139477.257880 (Exim 4.92) (envelope-from ) id 1lr18k-0000ra-BY; Wed, 09 Jun 2021 16:33:14 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 139477.257880; Wed, 09 Jun 2021 16:33:14 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr18k-0000rR-8F; Wed, 09 Jun 2021 16:33:14 +0000 Received: by outflank-mailman (input) for mailman id 139477; Wed, 09 Jun 2021 16:33:12 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr18i-0000rA-Q9 for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 16:33:12 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr18i-0003Z8-PJ for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 16:33:12 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lr18i-0000dr-OG for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 16:33:12 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=IarTMSGJICbq8t7aEghVOuMRyp3JVbGrYC/VAMTW4IA=; b=twtTEAntaSnRYmceZatopgaApI ctochmg+/B5HMAR9Jw+6W8C7KxCH+XhKtsvj5YzE3kU+OZxL449bSscGxk/u93LqZoUP5N740vKRE QA7sdAWEN8Q2JqkNRatvbNhFZMDDaOlYbRuEoSJ6S2UqhTxB6Rslxy8px7NFVgXe+Jpo=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.11] AMD/IOMMU: size command buffer dynamically Message-Id: Date: Wed, 09 Jun 2021 16:33:12 +0000 commit 89d3cc62a50475a63a388487b8d4b574dfd426a5 Author: Jan Beulich AuthorDate: Tue Jun 8 19:15:18 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 19:16:38 2021 +0100 AMD/IOMMU: size command buffer dynamically With the present synchronous model, we need two slots for every operation (the operation itself and a wait command). There can be one such pair of commands pending per CPU. To ensure that under all normal circumstances a slot is always available when one is requested, size the command ring according to the number of present CPUs. This is part of XSA-373 / CVE-2021-28692. Signed-off-by: Jan Beulich Reviewed-by: Paul Durrant (cherry picked from commit df242851ddc93ac0b0a3a20ecab34acc29e3b36b) --- xen/drivers/passthrough/amd/iommu_cmd.c | 12 ++++-------- xen/drivers/passthrough/amd/iommu_init.c | 28 +++++++++++++++++++++++++--- xen/include/asm-x86/hvm/svm/amd-iommu-defs.h | 8 ++------ 3 files changed, 31 insertions(+), 17 deletions(-) diff --git a/xen/drivers/passthrough/amd/iommu_cmd.c b/xen/drivers/passthrough/amd/iommu_cmd.c index 08247fa354..2c464fcb58 100644 --- a/xen/drivers/passthrough/amd/iommu_cmd.c +++ b/xen/drivers/passthrough/amd/iommu_cmd.c @@ -24,8 +24,7 @@ static int queue_iommu_command(struct amd_iommu *iommu, u32 cmd[]) { - u32 tail, head, *cmd_buffer; - int i; + uint32_t tail, head; tail = iommu->cmd_buffer.tail; if ( ++tail == iommu->cmd_buffer.entries ) @@ -35,12 +34,9 @@ static int queue_iommu_command(struct amd_iommu *iommu, u32 cmd[]) IOMMU_CMD_BUFFER_HEAD_OFFSET)); if ( head != tail ) { - cmd_buffer = (u32 *)(iommu->cmd_buffer.buffer + - (iommu->cmd_buffer.tail * - IOMMU_CMD_BUFFER_ENTRY_SIZE)); - - for ( i = 0; i < IOMMU_CMD_BUFFER_U32_PER_ENTRY; i++ ) - cmd_buffer[i] = cmd[i]; + memcpy(iommu->cmd_buffer.buffer + + (iommu->cmd_buffer.tail * sizeof(cmd_entry_t)), + cmd, sizeof(cmd_entry_t)); iommu->cmd_buffer.tail = tail; return 1; diff --git a/xen/drivers/passthrough/amd/iommu_init.c b/xen/drivers/passthrough/amd/iommu_init.c index 7a089b4c60..cf792ef77f 100644 --- a/xen/drivers/passthrough/amd/iommu_init.c +++ b/xen/drivers/passthrough/amd/iommu_init.c @@ -136,7 +136,7 @@ static void register_iommu_cmd_buffer_in_mmio_space(struct amd_iommu *iommu) writel(entry, iommu->mmio_base + IOMMU_CMD_BUFFER_BASE_LOW_OFFSET); power_of2_entries = get_order_from_bytes(iommu->cmd_buffer.alloc_size) + - IOMMU_CMD_BUFFER_POWER_OF2_ENTRIES_PER_PAGE; + PAGE_SHIFT - IOMMU_CMD_BUFFER_ENTRY_ORDER; entry = 0; iommu_set_addr_hi_to_reg(&entry, addr_hi); @@ -1000,9 +1000,31 @@ static void * __init allocate_ring_buffer(struct ring_buffer *ring_buf, static void * __init allocate_cmd_buffer(struct amd_iommu *iommu) { /* allocate 'command buffer' in power of 2 increments of 4K */ + static unsigned int __read_mostly nr_ents; + + if ( !nr_ents ) + { + unsigned int order; + + /* + * With the present synchronous model, we need two slots for every + * operation (the operation itself and a wait command). There can be + * one such pair of requests pending per CPU. One extra entry is + * needed as the ring is considered full when there's only one entry + * left. + */ + BUILD_BUG_ON(CONFIG_NR_CPUS * 2 >= IOMMU_CMD_BUFFER_MAX_ENTRIES); + order = get_order_from_bytes((num_present_cpus() * 2 + 1) << + IOMMU_CMD_BUFFER_ENTRY_ORDER); + nr_ents = 1u << (order + PAGE_SHIFT - IOMMU_CMD_BUFFER_ENTRY_ORDER); + + AMD_IOMMU_DEBUG("using %u-entry cmd ring(s)\n", nr_ents); + } + + BUILD_BUG_ON(sizeof(cmd_entry_t) != (1u << IOMMU_CMD_BUFFER_ENTRY_ORDER)); + return allocate_ring_buffer(&iommu->cmd_buffer, sizeof(cmd_entry_t), - IOMMU_CMD_BUFFER_DEFAULT_ENTRIES, - "Command Buffer"); + nr_ents, "Command Buffer"); } static void * __init allocate_event_log(struct amd_iommu *iommu) diff --git a/xen/include/asm-x86/hvm/svm/amd-iommu-defs.h b/xen/include/asm-x86/hvm/svm/amd-iommu-defs.h index c479f0bb02..d2e285f03c 100644 --- a/xen/include/asm-x86/hvm/svm/amd-iommu-defs.h +++ b/xen/include/asm-x86/hvm/svm/amd-iommu-defs.h @@ -20,9 +20,6 @@ #ifndef _ASM_X86_64_AMD_IOMMU_DEFS_H #define _ASM_X86_64_AMD_IOMMU_DEFS_H -/* IOMMU Command Buffer entries: in power of 2 increments, minimum of 256 */ -#define IOMMU_CMD_BUFFER_DEFAULT_ENTRIES 512 - /* IOMMU Event Log entries: in power of 2 increments, minimum of 256 */ #define IOMMU_EVENT_LOG_DEFAULT_ENTRIES 512 @@ -185,9 +182,8 @@ #define IOMMU_CMD_BUFFER_LENGTH_MASK 0x0F000000 #define IOMMU_CMD_BUFFER_LENGTH_SHIFT 24 -#define IOMMU_CMD_BUFFER_ENTRY_SIZE 16 -#define IOMMU_CMD_BUFFER_POWER_OF2_ENTRIES_PER_PAGE 8 -#define IOMMU_CMD_BUFFER_U32_PER_ENTRY (IOMMU_CMD_BUFFER_ENTRY_SIZE / 4) +#define IOMMU_CMD_BUFFER_ENTRY_ORDER 4 +#define IOMMU_CMD_BUFFER_MAX_ENTRIES (1u << 15) #define IOMMU_CMD_OPCODE_MASK 0xF0000000 #define IOMMU_CMD_OPCODE_SHIFT 28 -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.11 From xen-changelog-bounces@lists.xenproject.org Wed Jun 09 16:33:24 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Wed, 09 Jun 2021 16:33:24 +0000 Received: from list by lists.xenproject.org with outflank-mailman.139478.257884 (Exim 4.92) (envelope-from ) id 1lr18u-0000uX-DA; Wed, 09 Jun 2021 16:33:24 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 139478.257884; Wed, 09 Jun 2021 16:33:24 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr18u-0000uP-9q; Wed, 09 Jun 2021 16:33:24 +0000 Received: by outflank-mailman (input) for mailman id 139478; Wed, 09 Jun 2021 16:33:22 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr18s-0000uB-Te for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 16:33:22 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr18s-0003ZR-Sk for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 16:33:22 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lr18s-0000gP-Rd for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 16:33:22 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=GdsXTusI/7yJ870dwUUwSjLZI+9+wikQ4Ykx9//BGeQ=; b=JxKsXzxLuVuBRxoE3RvYvOlp9Z b6pJyt/TFQycXC4NgQ11B4hkC90GyRcSxqSZImzwMKwNwakkvyW+U6cldOzkWp8l4k06ud4k8ULV6 HfbmN8RGkCicRF3WMM0iP8NzTLeyhg+s9eL8zZHZta62xc3eGQF7u+CMnzZcjypiInzM=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.11] VT-d: eliminate flush related timeouts Message-Id: Date: Wed, 09 Jun 2021 16:33:22 +0000 commit dc3e5441867de26b921c4e10241fb8ad1c9d3fad Author: Jan Beulich AuthorDate: Tue Jun 8 19:15:39 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 19:16:38 2021 +0100 VT-d: eliminate flush related timeouts Leaving an in-progress operation pending when it appears to take too long is problematic: If e.g. a QI command completed later, the write to the "poll slot" may instead be understood to signal a subsequently started command's completion. Also our accounting of the timeout period was actually wrong: We included the time it took for the command to actually make it to the front of the queue, which could be heavily affected by guests other than the one for which the flush is being performed. Do away with all timeout detection on all flush related code paths. Log excessively long processing times (with a progressive threshold) to have some indication of problems in this area. Additionally log (once) if qinval_next_index() didn't immediately find an available slot. Together with the earlier change sizing the queue(s) dynamically, we should now have a guarantee that with our fully synchronous model any demand for slots can actually be satisfied. This is part of XSA-373 / CVE-2021-28692. Signed-off-by: Jan Beulich Reviewed-by: Paul Durrant (cherry picked from commit 7c893abc4ee7e9af62297ba6fd5f27c89d0f33f5) --- xen/drivers/passthrough/vtd/dmar.h | 28 ++++++++++++++++++++++++++++ xen/drivers/passthrough/vtd/iommu.c | 12 ++++++------ xen/drivers/passthrough/vtd/qinval.c | 33 ++++++++++++++++++++++----------- 3 files changed, 56 insertions(+), 17 deletions(-) diff --git a/xen/drivers/passthrough/vtd/dmar.h b/xen/drivers/passthrough/vtd/dmar.h index 8f1b18ebb2..f67cba1386 100644 --- a/xen/drivers/passthrough/vtd/dmar.h +++ b/xen/drivers/passthrough/vtd/dmar.h @@ -127,6 +127,34 @@ do { \ } \ } while (0) +#define IOMMU_FLUSH_WAIT(what, iommu, offset, op, cond, sts) \ +do { \ + static unsigned int __read_mostly threshold = 1; \ + s_time_t start = NOW(); \ + s_time_t timeout = start + DMAR_OPERATION_TIMEOUT * threshold; \ + \ + for ( ; ; ) \ + { \ + sts = op(iommu->reg, offset); \ + if ( cond ) \ + break; \ + if ( timeout && NOW() > timeout ) \ + { \ + threshold |= threshold << 1; \ + printk(XENLOG_WARNING VTDPREFIX \ + " IOMMU#%u: %s flush taking too long\n", \ + iommu->index, what); \ + timeout = 0; \ + } \ + cpu_relax(); \ + } \ + \ + if ( !timeout ) \ + printk(XENLOG_WARNING VTDPREFIX \ + " IOMMU#%u: %s flush took %lums\n", \ + iommu->index, what, (NOW() - start) / 10000000); \ +} while ( false ) + int vtd_hw_check(void); void disable_pmr(struct iommu *iommu); int is_igd_drhd(struct acpi_drhd_unit *drhd); diff --git a/xen/drivers/passthrough/vtd/iommu.c b/xen/drivers/passthrough/vtd/iommu.c index f499c6f430..698b176a81 100644 --- a/xen/drivers/passthrough/vtd/iommu.c +++ b/xen/drivers/passthrough/vtd/iommu.c @@ -357,8 +357,8 @@ static void iommu_flush_write_buffer(struct iommu *iommu) dmar_writel(iommu->reg, DMAR_GCMD_REG, val | DMA_GCMD_WBF); /* Make sure hardware complete it */ - IOMMU_WAIT_OP(iommu, DMAR_GSTS_REG, dmar_readl, - !(val & DMA_GSTS_WBFS), val); + IOMMU_FLUSH_WAIT("write buffer", iommu, DMAR_GSTS_REG, dmar_readl, + !(val & DMA_GSTS_WBFS), val); spin_unlock_irqrestore(&iommu->register_lock, flags); } @@ -408,8 +408,8 @@ static int __must_check flush_context_reg(void *_iommu, u16 did, u16 source_id, dmar_writeq(iommu->reg, DMAR_CCMD_REG, val); /* Make sure hardware complete it */ - IOMMU_WAIT_OP(iommu, DMAR_CCMD_REG, dmar_readq, - !(val & DMA_CCMD_ICC), val); + IOMMU_FLUSH_WAIT("context", iommu, DMAR_CCMD_REG, dmar_readq, + !(val & DMA_CCMD_ICC), val); spin_unlock_irqrestore(&iommu->register_lock, flags); /* flush context entry will implicitly flush write buffer */ @@ -491,8 +491,8 @@ static int __must_check flush_iotlb_reg(void *_iommu, u16 did, u64 addr, dmar_writeq(iommu->reg, tlb_offset + 8, val); /* Make sure hardware complete it */ - IOMMU_WAIT_OP(iommu, (tlb_offset + 8), dmar_readq, - !(val & DMA_TLB_IVT), val); + IOMMU_FLUSH_WAIT("iotlb", iommu, (tlb_offset + 8), dmar_readq, + !(val & DMA_TLB_IVT), val); spin_unlock_irqrestore(&iommu->register_lock, flags); /* check IOTLB invalidation granularity */ diff --git a/xen/drivers/passthrough/vtd/qinval.c b/xen/drivers/passthrough/vtd/qinval.c index 103afc15fe..17c4235597 100644 --- a/xen/drivers/passthrough/vtd/qinval.c +++ b/xen/drivers/passthrough/vtd/qinval.c @@ -29,8 +29,6 @@ #include "extern.h" #include "../ats.h" -#define VTD_QI_TIMEOUT 1 - static unsigned int __read_mostly qi_pg_order; static unsigned int __read_mostly qi_entry_nr; @@ -60,7 +58,11 @@ static unsigned int qinval_next_index(struct iommu *iommu) /* (tail+1 == head) indicates a full queue, wait for HW */ while ( ((tail + 1) & (qi_entry_nr - 1)) == ( dmar_readq(iommu->reg, DMAR_IQH_REG) >> QINVAL_INDEX_SHIFT ) ) + { + printk_once(XENLOG_ERR VTDPREFIX " IOMMU#%u: no QI slot available\n", + iommu->index); cpu_relax(); + } return tail; } @@ -180,23 +182,32 @@ static int __must_check queue_invalidate_wait(struct iommu *iommu, /* Now we don't support interrupt method */ if ( sw ) { - s_time_t timeout; - - /* In case all wait descriptor writes to same addr with same data */ - timeout = NOW() + MILLISECS(flush_dev_iotlb ? - iommu_dev_iotlb_timeout : VTD_QI_TIMEOUT); + static unsigned int __read_mostly threshold = 1; + s_time_t start = NOW(); + s_time_t timeout = start + (flush_dev_iotlb + ? iommu_dev_iotlb_timeout + : 100) * MILLISECS(threshold); while ( ACCESS_ONCE(*this_poll_slot) != QINVAL_STAT_DONE ) { - if ( NOW() > timeout ) + if ( timeout && NOW() > timeout ) { - print_qi_regs(iommu); + threshold |= threshold << 1; printk(XENLOG_WARNING VTDPREFIX - " Queue invalidate wait descriptor timed out\n"); - return -ETIMEDOUT; + " IOMMU#%u: QI%s wait descriptor taking too long\n", + iommu->index, flush_dev_iotlb ? " dev" : ""); + print_qi_regs(iommu); + timeout = 0; } cpu_relax(); } + + if ( !timeout ) + printk(XENLOG_WARNING VTDPREFIX + " IOMMU#%u: QI%s wait descriptor took %lums\n", + iommu->index, flush_dev_iotlb ? " dev" : "", + (NOW() - start) / 10000000); + return 0; } -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.11 From xen-changelog-bounces@lists.xenproject.org Wed Jun 09 16:33:34 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Wed, 09 Jun 2021 16:33:34 +0000 Received: from list by lists.xenproject.org with outflank-mailman.139479.257889 (Exim 4.92) (envelope-from ) id 1lr194-0000xm-HW; Wed, 09 Jun 2021 16:33:34 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 139479.257889; Wed, 09 Jun 2021 16:33:34 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr194-0000xe-Dx; Wed, 09 Jun 2021 16:33:34 +0000 Received: by outflank-mailman (input) for mailman id 139479; Wed, 09 Jun 2021 16:33:33 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr193-0000xU-0J for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 16:33:33 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr192-0003Zb-Vf for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 16:33:32 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lr192-0000hQ-Up for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 16:33:32 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=s2baECs3VBdgPjJBnJxvPB2Tf6oLKMwxVZY0YcALTh4=; b=r3iGzZMn+t6WUv3U93uMR1GcnD Y8RldeUqAxdLHyF9vHYBnkTf0DKTNmqa8W0+wLT+MpFWe9kC+JmnX9EEFQ9iY8DXD33++K6Ma1OmN f8xygidmQHVgyN75EotfjlvCy///7QqRYAvQGHQwIE1Kz76LGWw84xIFJDi9zmSSiG7w=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.11] AMD/IOMMU: wait for command slot to be available Message-Id: Date: Wed, 09 Jun 2021 16:33:32 +0000 commit 9f8bf2a7ddab2a5c874072cfa430ffae9772de1d Author: Jan Beulich AuthorDate: Tue Jun 8 19:15:57 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 19:16:38 2021 +0100 AMD/IOMMU: wait for command slot to be available No caller cared about send_iommu_command() indicating unavailability of a slot. Hence if a sufficient number prior commands timed out, we did blindly assume that the requested command was submitted to the IOMMU when really it wasn't. This could mean both a hanging system (waiting for a command to complete that was never seen by the IOMMU) or blindly propagating success back to callers, making them believe they're fine to e.g. free previously unmapped pages. Fold the three involved functions into one, add spin waiting for an available slot along the lines of VT-d's qinval_next_index(), and as a consequence drop all error indicator return types/values. This is part of XSA-373 / CVE-2021-28692. Signed-off-by: Jan Beulich Reviewed-by: Paul Durrant (cherry picked from commit c4e44343579e2c3304d676404d81b2b9bd893c5b) --- xen/drivers/passthrough/amd/iommu_cmd.c | 42 ++++++++++++--------------------- 1 file changed, 15 insertions(+), 27 deletions(-) diff --git a/xen/drivers/passthrough/amd/iommu_cmd.c b/xen/drivers/passthrough/amd/iommu_cmd.c index 2c464fcb58..6625ee8518 100644 --- a/xen/drivers/passthrough/amd/iommu_cmd.c +++ b/xen/drivers/passthrough/amd/iommu_cmd.c @@ -22,48 +22,36 @@ #include #include "../ats.h" -static int queue_iommu_command(struct amd_iommu *iommu, u32 cmd[]) +static void send_iommu_command(struct amd_iommu *iommu, + const uint32_t cmd[4]) { - uint32_t tail, head; + uint32_t tail; tail = iommu->cmd_buffer.tail; if ( ++tail == iommu->cmd_buffer.entries ) tail = 0; - head = iommu_get_rb_pointer(readl(iommu->mmio_base + - IOMMU_CMD_BUFFER_HEAD_OFFSET)); - if ( head != tail ) + while ( tail == iommu_get_rb_pointer(readl(iommu->mmio_base + + IOMMU_CMD_BUFFER_HEAD_OFFSET)) ) { - memcpy(iommu->cmd_buffer.buffer + - (iommu->cmd_buffer.tail * sizeof(cmd_entry_t)), - cmd, sizeof(cmd_entry_t)); - - iommu->cmd_buffer.tail = tail; - return 1; + printk_once(XENLOG_ERR + "AMD IOMMU %04x:%02x:%02x.%u: no cmd slot available\n", + iommu->seg, PCI_BUS(iommu->bdf), + PCI_SLOT(iommu->bdf), PCI_FUNC(iommu->bdf)); + cpu_relax(); } - return 0; -} + memcpy(iommu->cmd_buffer.buffer + + (iommu->cmd_buffer.tail * sizeof(cmd_entry_t)), + cmd, sizeof(cmd_entry_t)); -static void commit_iommu_command_buffer(struct amd_iommu *iommu) -{ - u32 tail = 0; + iommu->cmd_buffer.tail = tail; + tail = 0; iommu_set_rb_pointer(&tail, iommu->cmd_buffer.tail); writel(tail, iommu->mmio_base+IOMMU_CMD_BUFFER_TAIL_OFFSET); } -int send_iommu_command(struct amd_iommu *iommu, u32 cmd[]) -{ - if ( queue_iommu_command(iommu, cmd) ) - { - commit_iommu_command_buffer(iommu); - return 1; - } - - return 0; -} - static void flush_command_buffer(struct amd_iommu *iommu) { u32 cmd[4], status; -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.11 From xen-changelog-bounces@lists.xenproject.org Wed Jun 09 16:33:44 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Wed, 09 Jun 2021 16:33:44 +0000 Received: from list by lists.xenproject.org with outflank-mailman.139480.257892 (Exim 4.92) (envelope-from ) id 1lr19E-00010z-Ij; Wed, 09 Jun 2021 16:33:44 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 139480.257892; Wed, 09 Jun 2021 16:33:44 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr19E-00010n-Fc; Wed, 09 Jun 2021 16:33:44 +0000 Received: by outflank-mailman (input) for mailman id 139480; Wed, 09 Jun 2021 16:33:43 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr19D-00010g-36 for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 16:33:43 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr19D-0003Zt-2O for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 16:33:43 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lr19D-0000iN-1X for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 16:33:43 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=rKhbweVixbHl5Jl9aoI1Gjpei9w7eW8aLDyFaRipjH0=; b=FmkL/oDrlBPE1ONanRhnXS3vQi f8AMDcRUuoh2CrHbCzHZeLZqzYXChwCQTp8HJAUZixYXzasojMGhV1sebfg86B/RaV6BLdX3iSe8e BIKQ9xaLEghDbiUlZvWp60WkIUPWvRIiuI3cfCOrqvNVLGrw+Tf0WBAo2N5vwTHGRD5M=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.11] AMD/IOMMU: drop command completion timeout Message-Id: Date: Wed, 09 Jun 2021 16:33:43 +0000 commit 944da2f5b8530433aeaa46493db48f3ddb0aa302 Author: Jan Beulich AuthorDate: Tue Jun 8 19:16:07 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 19:16:38 2021 +0100 AMD/IOMMU: drop command completion timeout First and foremost - such timeouts were not signaled to callers, making them believe they're fine to e.g. free previously unmapped pages. Mirror VT-d's behavior: A fixed number of loop iterations is not a suitable way to detect timeouts in an environment (CPU and bus speeds) independent manner anyway. Furthermore, leaving an in-progress operation pending when it appears to take too long is problematic: If a command completed later, the signaling of its completion may instead be understood to signal a subsequently started command's completion. Log excessively long processing times (with a progressive threshold) to have some indication of problems in this area. Allow callers to specify a non-default timeout bias for this logging, using the same values as VT-d does, which in particular means a (by default) much larger value for device IO TLB invalidation. This is part of XSA-373 / CVE-2021-28692. Signed-off-by: Jan Beulich Reviewed-by: Paul Durrant (cherry picked from commit e4fee66043120c954fc309bbb37813604c1c0eb7) --- xen/drivers/passthrough/amd/iommu_cmd.c | 60 +++++++++++++++++++-------------- 1 file changed, 35 insertions(+), 25 deletions(-) diff --git a/xen/drivers/passthrough/amd/iommu_cmd.c b/xen/drivers/passthrough/amd/iommu_cmd.c index 6625ee8518..ebdd0f5529 100644 --- a/xen/drivers/passthrough/amd/iommu_cmd.c +++ b/xen/drivers/passthrough/amd/iommu_cmd.c @@ -52,10 +52,12 @@ static void send_iommu_command(struct amd_iommu *iommu, writel(tail, iommu->mmio_base+IOMMU_CMD_BUFFER_TAIL_OFFSET); } -static void flush_command_buffer(struct amd_iommu *iommu) +static void flush_command_buffer(struct amd_iommu *iommu, + unsigned int timeout_base) { - u32 cmd[4], status; - int loop_count, comp_wait; + uint32_t cmd[4]; + s_time_t start, timeout; + static unsigned int __read_mostly threshold = 1; /* RW1C 'ComWaitInt' in status register */ writel(IOMMU_STATUS_COMP_WAIT_INT_MASK, @@ -71,24 +73,31 @@ static void flush_command_buffer(struct amd_iommu *iommu) IOMMU_COMP_WAIT_I_FLAG_SHIFT, &cmd[0]); send_iommu_command(iommu, cmd); - /* Make loop_count long enough for polling completion wait bit */ - loop_count = 1000; - do { - status = readl(iommu->mmio_base + IOMMU_STATUS_MMIO_OFFSET); - comp_wait = get_field_from_reg_u32(status, - IOMMU_STATUS_COMP_WAIT_INT_MASK, - IOMMU_STATUS_COMP_WAIT_INT_SHIFT); - --loop_count; - } while ( !comp_wait && loop_count ); - - if ( comp_wait ) + start = NOW(); + timeout = start + (timeout_base ?: 100) * MILLISECS(threshold); + while ( !(readl(iommu->mmio_base + IOMMU_STATUS_MMIO_OFFSET) & + IOMMU_STATUS_COMP_WAIT_INT_MASK) ) { - /* RW1C 'ComWaitInt' in status register */ - writel(IOMMU_STATUS_COMP_WAIT_INT_MASK, - iommu->mmio_base + IOMMU_STATUS_MMIO_OFFSET); - return; + if ( timeout && NOW() > timeout ) + { + threshold |= threshold << 1; + printk(XENLOG_WARNING + "AMD IOMMU %04x:%02x:%02x.%u: %scompletion wait taking too long\n", + iommu->seg, PCI_BUS(iommu->bdf), + PCI_SLOT(iommu->bdf), PCI_FUNC(iommu->bdf), + timeout_base ? "iotlb " : ""); + timeout = 0; + } + cpu_relax(); } - AMD_IOMMU_DEBUG("Warning: ComWaitInt bit did not assert!\n"); + + if ( !timeout ) + printk(XENLOG_WARNING + "AMD IOMMU %04x:%02x:%02x.%u: %scompletion wait took %lums\n", + iommu->seg, PCI_BUS(iommu->bdf), + PCI_SLOT(iommu->bdf), PCI_FUNC(iommu->bdf), + timeout_base ? "iotlb " : "", + (NOW() - start) / 10000000); } /* Build low level iommu command messages */ @@ -300,7 +309,7 @@ void amd_iommu_flush_iotlb(u8 devfn, const struct pci_dev *pdev, /* send INVALIDATE_IOTLB_PAGES command */ spin_lock_irqsave(&iommu->lock, flags); invalidate_iotlb_pages(iommu, maxpend, 0, queueid, gaddr, req_id, order); - flush_command_buffer(iommu); + flush_command_buffer(iommu, iommu_dev_iotlb_timeout); spin_unlock_irqrestore(&iommu->lock, flags); } @@ -337,7 +346,7 @@ static void _amd_iommu_flush_pages(struct domain *d, { spin_lock_irqsave(&iommu->lock, flags); invalidate_iommu_pages(iommu, gaddr, dom_id, order); - flush_command_buffer(iommu); + flush_command_buffer(iommu, 0); spin_unlock_irqrestore(&iommu->lock, flags); } @@ -361,7 +370,7 @@ void amd_iommu_flush_device(struct amd_iommu *iommu, uint16_t bdf) ASSERT( spin_is_locked(&iommu->lock) ); invalidate_dev_table_entry(iommu, bdf); - flush_command_buffer(iommu); + flush_command_buffer(iommu, 0); } void amd_iommu_flush_intremap(struct amd_iommu *iommu, uint16_t bdf) @@ -369,7 +378,7 @@ void amd_iommu_flush_intremap(struct amd_iommu *iommu, uint16_t bdf) ASSERT( spin_is_locked(&iommu->lock) ); invalidate_interrupt_table(iommu, bdf); - flush_command_buffer(iommu); + flush_command_buffer(iommu, 0); } void amd_iommu_flush_all_caches(struct amd_iommu *iommu) @@ -377,7 +386,7 @@ void amd_iommu_flush_all_caches(struct amd_iommu *iommu) ASSERT( spin_is_locked(&iommu->lock) ); invalidate_iommu_all(iommu); - flush_command_buffer(iommu); + flush_command_buffer(iommu, 0); } void amd_iommu_send_guest_cmd(struct amd_iommu *iommu, u32 cmd[]) @@ -387,7 +396,8 @@ void amd_iommu_send_guest_cmd(struct amd_iommu *iommu, u32 cmd[]) spin_lock_irqsave(&iommu->lock, flags); send_iommu_command(iommu, cmd); - flush_command_buffer(iommu); + /* TBD: Timeout selection may require peeking into cmd[]. */ + flush_command_buffer(iommu, 0); spin_unlock_irqrestore(&iommu->lock, flags); } -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.11 From xen-changelog-bounces@lists.xenproject.org Wed Jun 09 16:33:54 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Wed, 09 Jun 2021 16:33:54 +0000 Received: from list by lists.xenproject.org with outflank-mailman.139481.257896 (Exim 4.92) (envelope-from ) id 1lr19O-00013o-KX; Wed, 09 Jun 2021 16:33:54 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 139481.257896; Wed, 09 Jun 2021 16:33:54 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr19O-00013g-HD; Wed, 09 Jun 2021 16:33:54 +0000 Received: by outflank-mailman (input) for mailman id 139481; Wed, 09 Jun 2021 16:33:53 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr19N-00013R-6i for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 16:33:53 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr19N-0003a3-5v for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 16:33:53 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lr19N-0000jY-50 for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 16:33:53 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=b66/VxC9ho5XL/2Vb+PGEqHThaOd524MpLolMMekUYI=; b=4EfCCykoK52yvPr8BArL3U5//Z +TkTwfoYnY60ord7IB6Um5rxvT6lFDd3X/PQSQaSptvvE3b5ctqvyjHO7jbKeQSqGbsEyGKP+eE9e qyT3sVGTfcSNp7Jz0tEDvuyORVppab2NWGJc+JXCLVG6T/dRLgQItTAdb9/uFjLZXLZQ=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.11] x86/spec-ctrl: Protect against Speculative Code Store Bypass Message-Id: Date: Wed, 09 Jun 2021 16:33:53 +0000 commit 7937627baf95e3f142639df7cb750658ca1871f3 Author: Andrew Cooper AuthorDate: Thu Mar 11 14:39:11 2021 +0000 Commit: Andrew Cooper CommitDate: Tue Jun 8 19:16:38 2021 +0100 x86/spec-ctrl: Protect against Speculative Code Store Bypass Modern x86 processors have far-better-than-architecturally-guaranteed self modifying code detection. Typically, when a write hits an instruction in flight, a Machine Clear occurs to flush stale content in the frontend and backend. For self modifying code, before a write which hits an instruction in flight retires, the frontend can speculatively decode and execute the old instruction stream. Speculation of this form can suffer from type confusion in registers, and potentially leak data. Furthermore, updates are typically byte-wise, rather than atomic. Depending on timing, speculation can race ahead multiple times between individual writes, and execute the transiently-malformed instruction stream. Xen has stubs which are used in certain cases for emulation purposes. Inhibit speculation between updating the stub and executing it. This is XSA-375 / CVE-2021-0089. Signed-off-by: Andrew Cooper Reviewed-by: Jan Beulich (cherry picked from commit 45f59ed8865318bb0356954bad067f329677ce9e) --- xen/arch/x86/pv/emul-priv-op.c | 2 ++ xen/arch/x86/x86_emulate/x86_emulate.c | 1 + 2 files changed, 3 insertions(+) diff --git a/xen/arch/x86/pv/emul-priv-op.c b/xen/arch/x86/pv/emul-priv-op.c index 755f00db33..fdba01ce96 100644 --- a/xen/arch/x86/pv/emul-priv-op.c +++ b/xen/arch/x86/pv/emul-priv-op.c @@ -101,6 +101,8 @@ static io_emul_stub_t *io_emul_stub_setup(struct priv_op_ctxt *ctxt, u8 opcode, BUILD_BUG_ON(STUB_BUF_SIZE / 2 < MAX(9, /* Default emul stub */ 5 + IOEMUL_QUIRK_STUB_BYTES)); + asm volatile ( "lfence" ::: "memory" ); /* SCSB */ + /* Handy function-typed pointer to the stub. */ return (void *)stub_va; } diff --git a/xen/arch/x86/x86_emulate/x86_emulate.c b/xen/arch/x86/x86_emulate/x86_emulate.c index 1f0c208fc5..19594ab45d 100644 --- a/xen/arch/x86/x86_emulate/x86_emulate.c +++ b/xen/arch/x86/x86_emulate/x86_emulate.c @@ -1008,6 +1008,7 @@ static inline int mkec(uint8_t e, int32_t ec, ...) # define invoke_stub(pre, post, constraints...) do { \ stub_exn.info = (union stub_exception_token) { .raw = ~0 }; \ stub_exn.line = __LINE__; /* Utility outweighs livepatching cost */ \ + asm volatile ( "lfence" ::: "memory" ); /* SCSB */ \ asm volatile ( pre "\n\tINDIRECT_CALL %[stub]\n\t" post "\n" \ ".Lret%=:\n\t" \ ".pushsection .fixup,\"ax\"\n" \ -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.11 From xen-changelog-bounces@lists.xenproject.org Wed Jun 09 16:34:04 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Wed, 09 Jun 2021 16:34:04 +0000 Received: from list by lists.xenproject.org with outflank-mailman.139482.257900 (Exim 4.92) (envelope-from ) id 1lr19Y-000175-Le; Wed, 09 Jun 2021 16:34:04 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 139482.257900; Wed, 09 Jun 2021 16:34:04 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr19Y-00016x-Ij; Wed, 09 Jun 2021 16:34:04 +0000 Received: by outflank-mailman (input) for mailman id 139482; Wed, 09 Jun 2021 16:34:03 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr19X-00016M-9f for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 16:34:03 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr19X-0003ag-8s for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 16:34:03 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lr19X-0000kf-80 for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 16:34:03 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=0j756BVK76/rRSFjp4ZkvtG9EPs963AZrmX0zyBKRno=; b=Y5NoxAUGwCxeNu1OurQKSjdOcy dpjeDH0g0HSrLwj+IlCllIO2qz34Ni3B/JhKq9A6OOJvq89f++eRUlnGZqyySZ7vA8AokLnRESjQy ruoyLgOFhRRi/cxb2Qhxd/mND+2qAeMlSbRxIU4CQunHDbl6lW/+BgIiVKvp7TqFcDjo=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.11] x86/spec-ctrl: Mitigate TAA after S3 resume Message-Id: Date: Wed, 09 Jun 2021 16:34:03 +0000 commit ef32c7afa2731b758226d6e10a1e489b1a15fc41 Author: Andrew Cooper AuthorDate: Thu May 20 01:21:39 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 19:16:38 2021 +0100 x86/spec-ctrl: Mitigate TAA after S3 resume The user chosen setting for MSR_TSX_CTRL needs restoring after S3. All APs get the correct setting via start_secondary(), but the BSP was missed out. This is XSA-377 / CVE-2021-28690. Fixes: 8c4330818f6 ("x86/spec-ctrl: Mitigate the TSX Asynchronous Abort sidechannel") Signed-off-by: Andrew Cooper Reviewed-by: Jan Beulich (cherry picked from commit 8cf276cb2e0b99b96333865873f56b0b31555ff1) --- xen/arch/x86/acpi/power.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/xen/arch/x86/acpi/power.c b/xen/arch/x86/acpi/power.c index 30e1bd5cd3..451cba622c 100644 --- a/xen/arch/x86/acpi/power.c +++ b/xen/arch/x86/acpi/power.c @@ -259,6 +259,8 @@ static int enter_state(u32 state) microcode_resume_cpu(0); + tsx_init(); /* Needs microcode. May change HLE/RTM feature bits. */ + if ( !recheck_cpu_features(0) ) panic("Missing previously available feature(s)."); -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.11 From xen-changelog-bounces@lists.xenproject.org Wed Jun 09 19:11:10 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Wed, 09 Jun 2021 19:11:10 +0000 Received: from list by lists.xenproject.org with outflank-mailman.139532.257983 (Exim 4.92) (envelope-from ) id 1lr3bU-00018B-BU; Wed, 09 Jun 2021 19:11:04 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 139532.257983; Wed, 09 Jun 2021 19:11:04 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr3bU-000183-8g; Wed, 09 Jun 2021 19:11:04 +0000 Received: by outflank-mailman (input) for mailman id 139532; Wed, 09 Jun 2021 19:11:02 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr3bS-00017v-Pa for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 19:11:02 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr3bS-0006WJ-Oa for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 19:11:02 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lr3bS-0000TK-Nd for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 19:11:02 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=05Q6GZiokFPvJjtuQu1aWaZ0F3xRsLHAR2BdrKA53sM=; b=jPj0FhuxI8TMiTjWCR3tebLIHM wrYVXduhWyTMJZyUBAe5+ui7fHq4wb53fjkJ3vP7XY6v5Z+3Md78enHhKLWecU8VSsSP8QWPoqFqV UHOyo5aWpeAk2yUSkwA1zwfVv59o76rRcMpUGL9KJTmp2CqrB3UsLduGgqdG0SVGIpBI=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.13] xen/arm: Create dom0less domUs earlier Message-Id: Date: Wed, 09 Jun 2021 19:11:02 +0000 commit 95197d46f5e291b0bf12cf617e602bbc67fd633f Author: Julien Grall AuthorDate: Mon May 17 17:47:13 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 18:53:03 2021 +0100 xen/arm: Create dom0less domUs earlier In a follow-up patch we will need to unallocate the boot modules before heap_init_late() is called. The modules will contain the domUs kernel and initramfs. Therefore Xen will need to create extra domUs (used by dom0less) before heap_init_late(). This has two consequences on dom0less: 1) Domains will not be unpaused as soon as they are created but once all have been created. However, Xen doesn't guarantee an order to unpause, so this is not something one could rely on. 2) The memory allocated for a domU will not be scrubbed anymore when an admin select bootscrub=on. This is not something we advertised, but if this is a concern we can introduce either force scrub for all domUs or a per-domain flag in the DT. The behavior for bootscrub=off and bootscrub=idle (default) has not changed. This is part of XSA-372 / CVE-2021-28693. Signed-off-by: Julien Grall Reviewed-by: Jan Beulich Reviewed-by: Stefano Stabellini Tested-by: Stefano Stabellini (cherry picked from commit 371347c5b64da699d9f5a0edda5dc496fd2b7a5c) --- xen/arch/arm/domain_build.c | 2 -- xen/arch/arm/setup.c | 9 +++++---- 2 files changed, 5 insertions(+), 6 deletions(-) diff --git a/xen/arch/arm/domain_build.c b/xen/arch/arm/domain_build.c index 8c5ff11cdc..2b37ea019d 100644 --- a/xen/arch/arm/domain_build.c +++ b/xen/arch/arm/domain_build.c @@ -2509,8 +2509,6 @@ void __init create_domUs(void) if ( construct_domU(d, node) != 0 ) panic("Could not set up domain %s\n", dt_node_name(node)); - - domain_unpause_by_systemcontroller(d); } } diff --git a/xen/arch/arm/setup.c b/xen/arch/arm/setup.c index 3c899cd4a0..9a5aaeed3b 100644 --- a/xen/arch/arm/setup.c +++ b/xen/arch/arm/setup.c @@ -784,7 +784,7 @@ void __init start_xen(unsigned long boot_phys_offset, int cpus, i; const char *cmdline; struct bootmodule *xen_bootmodule; - struct domain *dom0; + struct domain *dom0, *d; struct xen_domctl_createdomain dom0_cfg = { .flags = XEN_DOMCTL_CDF_hvm | XEN_DOMCTL_CDF_hap, .max_evtchn_port = -1, @@ -972,6 +972,8 @@ void __init start_xen(unsigned long boot_phys_offset, if ( construct_dom0(dom0) != 0) panic("Could not set up DOM0 guest OS\n"); + create_domUs(); + heap_init_late(); init_trace_bufs(); @@ -985,9 +987,8 @@ void __init start_xen(unsigned long boot_phys_offset, system_state = SYS_STATE_active; - create_domUs(); - - domain_unpause_by_systemcontroller(dom0); + for_each_domain( d ) + domain_unpause_by_systemcontroller(d); /* Switch on to the dynamically allocated stack for the idle vcpu * since the static one we're running on is about to be freed. */ -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.13 From xen-changelog-bounces@lists.xenproject.org Wed Jun 09 19:11:14 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Wed, 09 Jun 2021 19:11:14 +0000 Received: from list by lists.xenproject.org with outflank-mailman.139533.257986 (Exim 4.92) (envelope-from ) id 1lr3be-0001AV-DM; Wed, 09 Jun 2021 19:11:14 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 139533.257986; Wed, 09 Jun 2021 19:11:14 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr3be-0001AN-A9; Wed, 09 Jun 2021 19:11:14 +0000 Received: by outflank-mailman (input) for mailman id 139533; Wed, 09 Jun 2021 19:11:12 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr3bc-00019w-TL for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 19:11:12 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr3bc-0006WU-Ri for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 19:11:12 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lr3bc-0000Tu-Qi for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 19:11:12 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=MgnXGnP+8oNfeWOaTXA5Fbl8l7dhk/8tz+vVxWQZSfI=; b=y5Bcp95EYXzJcQyI69MDDW4KoX fyipURyIRVw1owzwSZkjW8+/giK9CBMl4c4eKtnnkQ6dxNE/cQNp+OlbuhugbZNy31+HYlvCdITeg o7pGIDt+cgMQ9G3fnmBFgOEawdd8EZtZ6PWZmd9JWb0jQlihJZ4KNxHAIbpTY/8sKz4Y=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.13] xen/arm: Boot modules should always be scrubbed if bootscrub={on, idle} Message-Id: Date: Wed, 09 Jun 2021 19:11:12 +0000 commit def4352113d21acf4d49579c4ed3b86c6f2c922e Author: Julien Grall AuthorDate: Sat Apr 17 17:38:28 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 18:53:04 2021 +0100 xen/arm: Boot modules should always be scrubbed if bootscrub={on, idle} The function to initialize the pages (see init_heap_pages()) will request scrub when the admin request idle bootscrub (default) and state == SYS_STATE_active. When bootscrub=on, Xen will scrub any free pages in heap_init_late(). Currently, the boot modules (e.g. kernels, initramfs) will be discarded/ freed after heap_init_late() is called and system_state switched to SYS_STATE_active. This means the pages associated with the boot modules will not get scrubbed before getting re-purposed. If the memory is assigned to an untrusted domU, it may be able to retrieve secrets from the modules. This is part of XSA-372 / CVE-2021-28693. Fixes: 1774e9b1df27 ("xen/arm: introduce create_domUs") Signed-off-by: Julien Grall Reviewed-by: Jan Beulich Reviewed-by: Stefano Stabellini Tested-by: Stefano Stabellini (cherry picked from commit fd5dc41ceaed9cfcfa011cdfd50f264c89277a90) --- xen/arch/arm/setup.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-) diff --git a/xen/arch/arm/setup.c b/xen/arch/arm/setup.c index 9a5aaeed3b..de21fbce0e 100644 --- a/xen/arch/arm/setup.c +++ b/xen/arch/arm/setup.c @@ -74,7 +74,6 @@ static __used void init_done(void) /* Must be done past setting system_state. */ unregister_init_virtual_region(); - discard_initial_modules(); free_init_memory(); startup_cpu_idle_loop(); } @@ -974,6 +973,12 @@ void __init start_xen(unsigned long boot_phys_offset, create_domUs(); + /* + * This needs to be called **before** heap_init_late() so modules + * will be scrubbed (unless suppressed). + */ + discard_initial_modules(); + heap_init_late(); init_trace_bufs(); -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.13 From xen-changelog-bounces@lists.xenproject.org Wed Jun 09 19:11:24 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Wed, 09 Jun 2021 19:11:24 +0000 Received: from list by lists.xenproject.org with outflank-mailman.139534.257991 (Exim 4.92) (envelope-from ) id 1lr3bo-0001F9-GF; Wed, 09 Jun 2021 19:11:24 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 139534.257991; Wed, 09 Jun 2021 19:11:24 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr3bo-0001F1-D9; Wed, 09 Jun 2021 19:11:24 +0000 Received: by outflank-mailman (input) for mailman id 139534; Wed, 09 Jun 2021 19:11:23 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr3bm-0001Eg-VU for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 19:11:22 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr3bm-0006Wl-Um for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 19:11:22 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lr3bm-0000UW-Tp for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 19:11:22 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=9PhpKKnaPgl4CZdiQ9L9kjxm9LAU0xlWn1Rjf22sllQ=; b=FB2wq5Bxk/pYKccE7Pc7kAjcY1 F/+jCjthAOepD8yMRBUdMmChVggzj119cwL/niHc/R9ANj7D0DoDGCM4cg5WbekZ2r+8vDIdLvhDX l5CD6mk+9+OBRGFf1VewdVFIEy6qmHExsSEO7RaW/pUZt5EkVM7xoHX4jSgfnu+tPAcg=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.13] VT-d: size qinval queue dynamically Message-Id: Date: Wed, 09 Jun 2021 19:11:22 +0000 commit 9e3c8b1448d50a3154882a686719f8acc0dad6e7 Author: Jan Beulich AuthorDate: Tue Jun 8 18:55:50 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 18:55:58 2021 +0100 VT-d: size qinval queue dynamically With the present synchronous model, we need two slots for every operation (the operation itself and a wait descriptor). There can be one such pair of requests pending per CPU. To ensure that under all normal circumstances a slot is always available when one is requested, size the queue ring according to the number of present CPUs. This is part of XSA-373 / CVE-2021-28692. Signed-off-by: Jan Beulich Reviewed-by: Paul Durrant (cherry picked from commit cbfa62bb140c8f10a0ae96db3ce06e22b3b9d302) --- xen/drivers/passthrough/vtd/iommu.h | 10 +------- xen/drivers/passthrough/vtd/qinval.c | 46 +++++++++++++++++++++++++++--------- 2 files changed, 36 insertions(+), 20 deletions(-) diff --git a/xen/drivers/passthrough/vtd/iommu.h b/xen/drivers/passthrough/vtd/iommu.h index 57737b1a4a..e002dc6f01 100644 --- a/xen/drivers/passthrough/vtd/iommu.h +++ b/xen/drivers/passthrough/vtd/iommu.h @@ -450,17 +450,9 @@ struct qinval_entry { }q; }; -/* Order of queue invalidation pages(max is 8) */ -#define QINVAL_PAGE_ORDER 2 - -#define QINVAL_ARCH_PAGE_ORDER (QINVAL_PAGE_ORDER + PAGE_SHIFT_4K - PAGE_SHIFT) -#define QINVAL_ARCH_PAGE_NR ( QINVAL_ARCH_PAGE_ORDER < 0 ? \ - 1 : \ - 1 << QINVAL_ARCH_PAGE_ORDER ) - /* Each entry is 16 bytes, so 2^8 entries per page */ #define QINVAL_ENTRY_ORDER ( PAGE_SHIFT - 4 ) -#define QINVAL_ENTRY_NR (1 << (QINVAL_PAGE_ORDER + 8)) +#define QINVAL_MAX_ENTRY_NR (1u << (7 + QINVAL_ENTRY_ORDER)) /* Status data flag */ #define QINVAL_STAT_INIT 0 diff --git a/xen/drivers/passthrough/vtd/qinval.c b/xen/drivers/passthrough/vtd/qinval.c index dfcbb66c8a..86a941ce27 100644 --- a/xen/drivers/passthrough/vtd/qinval.c +++ b/xen/drivers/passthrough/vtd/qinval.c @@ -31,6 +31,9 @@ #define VTD_QI_TIMEOUT 1 +static unsigned int __read_mostly qi_pg_order; +static unsigned int __read_mostly qi_entry_nr; + static int __must_check invalidate_sync(struct vtd_iommu *iommu); static void print_qi_regs(struct vtd_iommu *iommu) @@ -55,7 +58,7 @@ static unsigned int qinval_next_index(struct vtd_iommu *iommu) tail >>= QINVAL_INDEX_SHIFT; /* (tail+1 == head) indicates a full queue, wait for HW */ - while ( ( tail + 1 ) % QINVAL_ENTRY_NR == + while ( ((tail + 1) & (qi_entry_nr - 1)) == ( dmar_readq(iommu->reg, DMAR_IQH_REG) >> QINVAL_INDEX_SHIFT ) ) cpu_relax(); @@ -68,7 +71,7 @@ static void qinval_update_qtail(struct vtd_iommu *iommu, unsigned int index) /* Need hold register lock when update tail */ ASSERT( spin_is_locked(&iommu->register_lock) ); - val = (index + 1) % QINVAL_ENTRY_NR; + val = (index + 1) & (qi_entry_nr - 1); dmar_writeq(iommu->reg, DMAR_IQT_REG, (val << QINVAL_INDEX_SHIFT)); } @@ -403,8 +406,28 @@ int enable_qinval(struct vtd_iommu *iommu) if ( iommu->qinval_maddr == 0 ) { - iommu->qinval_maddr = alloc_pgtable_maddr(QINVAL_ARCH_PAGE_NR, - iommu->node); + if ( !qi_entry_nr ) + { + /* + * With the present synchronous model, we need two slots for every + * operation (the operation itself and a wait descriptor). There + * can be one such pair of requests pending per CPU. One extra + * entry is needed as the ring is considered full when there's + * only one entry left. + */ + BUILD_BUG_ON(CONFIG_NR_CPUS * 2 >= QINVAL_MAX_ENTRY_NR); + qi_pg_order = get_order_from_bytes((num_present_cpus() * 2 + 1) << + (PAGE_SHIFT - + QINVAL_ENTRY_ORDER)); + qi_entry_nr = 1u << (qi_pg_order + QINVAL_ENTRY_ORDER); + + dprintk(XENLOG_INFO VTDPREFIX, + "QI: using %u-entry ring(s)\n", qi_entry_nr); + } + + iommu->qinval_maddr = + alloc_pgtable_maddr(qi_entry_nr >> QINVAL_ENTRY_ORDER, + iommu->node); if ( iommu->qinval_maddr == 0 ) { dprintk(XENLOG_WARNING VTDPREFIX, @@ -418,15 +441,16 @@ int enable_qinval(struct vtd_iommu *iommu) spin_lock_irqsave(&iommu->register_lock, flags); - /* Setup Invalidation Queue Address(IQA) register with the - * address of the page we just allocated. QS field at - * bits[2:0] to indicate size of queue is one 4KB page. - * That's 256 entries. Queued Head (IQH) and Queue Tail (IQT) - * registers are automatically reset to 0 with write - * to IQA register. + /* + * Setup Invalidation Queue Address (IQA) register with the address of the + * pages we just allocated. The QS field at bits[2:0] indicates the size + * (page order) of the queue. + * + * Queued Head (IQH) and Queue Tail (IQT) registers are automatically + * reset to 0 with write to IQA register. */ dmar_writeq(iommu->reg, DMAR_IQA_REG, - iommu->qinval_maddr | QINVAL_PAGE_ORDER); + iommu->qinval_maddr | qi_pg_order); dmar_writeq(iommu->reg, DMAR_IQT_REG, 0); -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.13 From xen-changelog-bounces@lists.xenproject.org Wed Jun 09 19:11:34 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Wed, 09 Jun 2021 19:11:34 +0000 Received: from list by lists.xenproject.org with outflank-mailman.139536.257995 (Exim 4.92) (envelope-from ) id 1lr3by-0001Ir-HW; Wed, 09 Jun 2021 19:11:34 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 139536.257995; Wed, 09 Jun 2021 19:11:34 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr3by-0001Ij-Ee; Wed, 09 Jun 2021 19:11:34 +0000 Received: by outflank-mailman (input) for mailman id 139536; Wed, 09 Jun 2021 19:11:33 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr3bx-0001IH-2j for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 19:11:33 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr3bx-0006X9-1j for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 19:11:33 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lr3bx-0000V3-0h for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 19:11:33 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=KuCSTyafJ/5fDeyR167y8Hn8a8Ktdv8g9IhZSFji4Ak=; b=n5TkuSZrMs9/L5AUbw4XfRar0j Cm9/fIMJflF8fuIm6V7T31zTI8WlTUIFDC4e3wwAU/B1+5kFFXgiODy5BYvKpJnGDssCE4ijLZu5M Dfy1L38aU+jOa/Z2iHoycNk/4JvwtXxNsAtPOiYE4t9ldThvBJCMwtvyhs1RWQ4zi5i8=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.13] AMD/IOMMU: size command buffer dynamically Message-Id: Date: Wed, 09 Jun 2021 19:11:33 +0000 commit 83c0f6bbb25769b9ac711ea6b218d5ee858a360e Author: Jan Beulich AuthorDate: Tue Jun 8 18:56:24 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 18:56:26 2021 +0100 AMD/IOMMU: size command buffer dynamically With the present synchronous model, we need two slots for every operation (the operation itself and a wait command). There can be one such pair of commands pending per CPU. To ensure that under all normal circumstances a slot is always available when one is requested, size the command ring according to the number of present CPUs. This is part of XSA-373 / CVE-2021-28692. Signed-off-by: Jan Beulich Reviewed-by: Paul Durrant (cherry picked from commit df242851ddc93ac0b0a3a20ecab34acc29e3b36b) --- xen/drivers/passthrough/amd/iommu_cmd.c | 4 ++-- xen/drivers/passthrough/amd/iommu_init.c | 28 +++++++++++++++++++++++++--- xen/include/asm-x86/hvm/svm/amd-iommu-defs.h | 7 ++----- 3 files changed, 29 insertions(+), 10 deletions(-) diff --git a/xen/drivers/passthrough/amd/iommu_cmd.c b/xen/drivers/passthrough/amd/iommu_cmd.c index af3a1fb865..c9c7d7b290 100644 --- a/xen/drivers/passthrough/amd/iommu_cmd.c +++ b/xen/drivers/passthrough/amd/iommu_cmd.c @@ -35,8 +35,8 @@ static int queue_iommu_command(struct amd_iommu *iommu, u32 cmd[]) if ( head != tail ) { memcpy(iommu->cmd_buffer.buffer + - (iommu->cmd_buffer.tail * IOMMU_CMD_BUFFER_ENTRY_SIZE), - cmd, IOMMU_CMD_BUFFER_ENTRY_SIZE); + (iommu->cmd_buffer.tail * sizeof(cmd_entry_t)), + cmd, sizeof(cmd_entry_t)); iommu->cmd_buffer.tail = tail; return 1; diff --git a/xen/drivers/passthrough/amd/iommu_init.c b/xen/drivers/passthrough/amd/iommu_init.c index 566e6defa1..2209eba066 100644 --- a/xen/drivers/passthrough/amd/iommu_init.c +++ b/xen/drivers/passthrough/amd/iommu_init.c @@ -125,7 +125,7 @@ static void register_iommu_cmd_buffer_in_mmio_space(struct amd_iommu *iommu) writel(entry, iommu->mmio_base + IOMMU_CMD_BUFFER_BASE_LOW_OFFSET); power_of2_entries = get_order_from_bytes(iommu->cmd_buffer.alloc_size) + - IOMMU_CMD_BUFFER_POWER_OF2_ENTRIES_PER_PAGE; + PAGE_SHIFT - IOMMU_CMD_BUFFER_ENTRY_ORDER; entry = 0; iommu_set_addr_hi_to_reg(&entry, addr_hi); @@ -1050,9 +1050,31 @@ static void *__init allocate_ring_buffer(struct ring_buffer *ring_buf, static void * __init allocate_cmd_buffer(struct amd_iommu *iommu) { /* allocate 'command buffer' in power of 2 increments of 4K */ + static unsigned int __read_mostly nr_ents; + + if ( !nr_ents ) + { + unsigned int order; + + /* + * With the present synchronous model, we need two slots for every + * operation (the operation itself and a wait command). There can be + * one such pair of requests pending per CPU. One extra entry is + * needed as the ring is considered full when there's only one entry + * left. + */ + BUILD_BUG_ON(CONFIG_NR_CPUS * 2 >= IOMMU_CMD_BUFFER_MAX_ENTRIES); + order = get_order_from_bytes((num_present_cpus() * 2 + 1) << + IOMMU_CMD_BUFFER_ENTRY_ORDER); + nr_ents = 1u << (order + PAGE_SHIFT - IOMMU_CMD_BUFFER_ENTRY_ORDER); + + AMD_IOMMU_DEBUG("using %u-entry cmd ring(s)\n", nr_ents); + } + + BUILD_BUG_ON(sizeof(cmd_entry_t) != (1u << IOMMU_CMD_BUFFER_ENTRY_ORDER)); + return allocate_ring_buffer(&iommu->cmd_buffer, sizeof(cmd_entry_t), - IOMMU_CMD_BUFFER_DEFAULT_ENTRIES, - "Command Buffer", false); + nr_ents, "Command Buffer", false); } static void * __init allocate_event_log(struct amd_iommu *iommu) diff --git a/xen/include/asm-x86/hvm/svm/amd-iommu-defs.h b/xen/include/asm-x86/hvm/svm/amd-iommu-defs.h index 6049507a30..a54d6e9fc6 100644 --- a/xen/include/asm-x86/hvm/svm/amd-iommu-defs.h +++ b/xen/include/asm-x86/hvm/svm/amd-iommu-defs.h @@ -20,9 +20,6 @@ #ifndef _ASM_X86_64_AMD_IOMMU_DEFS_H #define _ASM_X86_64_AMD_IOMMU_DEFS_H -/* IOMMU Command Buffer entries: in power of 2 increments, minimum of 256 */ -#define IOMMU_CMD_BUFFER_DEFAULT_ENTRIES 512 - /* IOMMU Event Log entries: in power of 2 increments, minimum of 256 */ #define IOMMU_EVENT_LOG_DEFAULT_ENTRIES 512 @@ -168,8 +165,8 @@ struct amd_iommu_dte { #define IOMMU_CMD_BUFFER_LENGTH_MASK 0x0F000000 #define IOMMU_CMD_BUFFER_LENGTH_SHIFT 24 -#define IOMMU_CMD_BUFFER_ENTRY_SIZE 16 -#define IOMMU_CMD_BUFFER_POWER_OF2_ENTRIES_PER_PAGE 8 +#define IOMMU_CMD_BUFFER_ENTRY_ORDER 4 +#define IOMMU_CMD_BUFFER_MAX_ENTRIES (1u << 15) #define IOMMU_CMD_OPCODE_MASK 0xF0000000 #define IOMMU_CMD_OPCODE_SHIFT 28 -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.13 From xen-changelog-bounces@lists.xenproject.org Wed Jun 09 19:11:43 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Wed, 09 Jun 2021 19:11:43 +0000 Received: from list by lists.xenproject.org with outflank-mailman.139537.257999 (Exim 4.92) (envelope-from ) id 1lr3c7-0001MF-J9; Wed, 09 Jun 2021 19:11:43 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 139537.257999; Wed, 09 Jun 2021 19:11:43 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr3c7-0001M7-GB; Wed, 09 Jun 2021 19:11:43 +0000 Received: by outflank-mailman (input) for mailman id 139537; Wed, 09 Jun 2021 19:11:43 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr3c7-0001Lx-5l for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 19:11:43 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr3c7-0006XR-51 for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 19:11:43 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lr3c7-0000Vi-3w for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 19:11:43 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=O575mBUlOAwvBwKsmPwvuX2zF9Ac0SdQv78yU/KlrB0=; b=TN9PIhutiy/D6yQjX7R9j9nFxi z+jziGhgexGfMR9XES3Z5hyNTNJqeejGL/CnpoErs8CTTEdD3mu2LYt/lZIKo9Pc0ODnocIQNIpQd ITaPFaQ92YdNsqwXzh9zIEvS4AA9FKm9S+Q6he2rL6hNFLYfDFS2j4QPvFeHkYYc9taQ=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.13] VT-d: eliminate flush related timeouts Message-Id: Date: Wed, 09 Jun 2021 19:11:43 +0000 commit d3d8a298f7fd27371414fc771786c2a0f1564152 Author: Jan Beulich AuthorDate: Tue Jun 8 18:56:39 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 18:56:42 2021 +0100 VT-d: eliminate flush related timeouts Leaving an in-progress operation pending when it appears to take too long is problematic: If e.g. a QI command completed later, the write to the "poll slot" may instead be understood to signal a subsequently started command's completion. Also our accounting of the timeout period was actually wrong: We included the time it took for the command to actually make it to the front of the queue, which could be heavily affected by guests other than the one for which the flush is being performed. Do away with all timeout detection on all flush related code paths. Log excessively long processing times (with a progressive threshold) to have some indication of problems in this area. Additionally log (once) if qinval_next_index() didn't immediately find an available slot. Together with the earlier change sizing the queue(s) dynamically, we should now have a guarantee that with our fully synchronous model any demand for slots can actually be satisfied. This is part of XSA-373 / CVE-2021-28692. Signed-off-by: Jan Beulich Reviewed-by: Paul Durrant (cherry picked from commit 7c893abc4ee7e9af62297ba6fd5f27c89d0f33f5) --- xen/drivers/passthrough/vtd/dmar.h | 28 ++++++++++++++++++++++++++++ xen/drivers/passthrough/vtd/iommu.c | 12 ++++++------ xen/drivers/passthrough/vtd/qinval.c | 33 ++++++++++++++++++++++----------- 3 files changed, 56 insertions(+), 17 deletions(-) diff --git a/xen/drivers/passthrough/vtd/dmar.h b/xen/drivers/passthrough/vtd/dmar.h index 1a9c965e59..a44820a5e7 100644 --- a/xen/drivers/passthrough/vtd/dmar.h +++ b/xen/drivers/passthrough/vtd/dmar.h @@ -127,6 +127,34 @@ do { \ } \ } while (0) +#define IOMMU_FLUSH_WAIT(what, iommu, offset, op, cond, sts) \ +do { \ + static unsigned int __read_mostly threshold = 1; \ + s_time_t start = NOW(); \ + s_time_t timeout = start + DMAR_OPERATION_TIMEOUT * threshold; \ + \ + for ( ; ; ) \ + { \ + sts = op(iommu->reg, offset); \ + if ( cond ) \ + break; \ + if ( timeout && NOW() > timeout ) \ + { \ + threshold |= threshold << 1; \ + printk(XENLOG_WARNING VTDPREFIX \ + " IOMMU#%u: %s flush taking too long\n", \ + iommu->index, what); \ + timeout = 0; \ + } \ + cpu_relax(); \ + } \ + \ + if ( !timeout ) \ + printk(XENLOG_WARNING VTDPREFIX \ + " IOMMU#%u: %s flush took %lums\n", \ + iommu->index, what, (NOW() - start) / 10000000); \ +} while ( false ) + int vtd_hw_check(void); void disable_pmr(struct vtd_iommu *iommu); int is_igd_drhd(struct acpi_drhd_unit *drhd); diff --git a/xen/drivers/passthrough/vtd/iommu.c b/xen/drivers/passthrough/vtd/iommu.c index 34ee5d22f1..dfd867d40b 100644 --- a/xen/drivers/passthrough/vtd/iommu.c +++ b/xen/drivers/passthrough/vtd/iommu.c @@ -320,8 +320,8 @@ static void iommu_flush_write_buffer(struct vtd_iommu *iommu) dmar_writel(iommu->reg, DMAR_GCMD_REG, val | DMA_GCMD_WBF); /* Make sure hardware complete it */ - IOMMU_WAIT_OP(iommu, DMAR_GSTS_REG, dmar_readl, - !(val & DMA_GSTS_WBFS), val); + IOMMU_FLUSH_WAIT("write buffer", iommu, DMAR_GSTS_REG, dmar_readl, + !(val & DMA_GSTS_WBFS), val); spin_unlock_irqrestore(&iommu->register_lock, flags); } @@ -370,8 +370,8 @@ int vtd_flush_context_reg(struct vtd_iommu *iommu, uint16_t did, dmar_writeq(iommu->reg, DMAR_CCMD_REG, val); /* Make sure hardware complete it */ - IOMMU_WAIT_OP(iommu, DMAR_CCMD_REG, dmar_readq, - !(val & DMA_CCMD_ICC), val); + IOMMU_FLUSH_WAIT("context", iommu, DMAR_CCMD_REG, dmar_readq, + !(val & DMA_CCMD_ICC), val); spin_unlock_irqrestore(&iommu->register_lock, flags); /* flush context entry will implicitly flush write buffer */ @@ -448,8 +448,8 @@ int vtd_flush_iotlb_reg(struct vtd_iommu *iommu, uint16_t did, uint64_t addr, dmar_writeq(iommu->reg, tlb_offset + 8, val); /* Make sure hardware complete it */ - IOMMU_WAIT_OP(iommu, (tlb_offset + 8), dmar_readq, - !(val & DMA_TLB_IVT), val); + IOMMU_FLUSH_WAIT("iotlb", iommu, (tlb_offset + 8), dmar_readq, + !(val & DMA_TLB_IVT), val); spin_unlock_irqrestore(&iommu->register_lock, flags); /* check IOTLB invalidation granularity */ diff --git a/xen/drivers/passthrough/vtd/qinval.c b/xen/drivers/passthrough/vtd/qinval.c index 86a941ce27..1056ab1feb 100644 --- a/xen/drivers/passthrough/vtd/qinval.c +++ b/xen/drivers/passthrough/vtd/qinval.c @@ -29,8 +29,6 @@ #include "extern.h" #include "../ats.h" -#define VTD_QI_TIMEOUT 1 - static unsigned int __read_mostly qi_pg_order; static unsigned int __read_mostly qi_entry_nr; @@ -60,7 +58,11 @@ static unsigned int qinval_next_index(struct vtd_iommu *iommu) /* (tail+1 == head) indicates a full queue, wait for HW */ while ( ((tail + 1) & (qi_entry_nr - 1)) == ( dmar_readq(iommu->reg, DMAR_IQH_REG) >> QINVAL_INDEX_SHIFT ) ) + { + printk_once(XENLOG_ERR VTDPREFIX " IOMMU#%u: no QI slot available\n", + iommu->index); cpu_relax(); + } return tail; } @@ -180,23 +182,32 @@ static int __must_check queue_invalidate_wait(struct vtd_iommu *iommu, /* Now we don't support interrupt method */ if ( sw ) { - s_time_t timeout; - - /* In case all wait descriptor writes to same addr with same data */ - timeout = NOW() + MILLISECS(flush_dev_iotlb ? - iommu_dev_iotlb_timeout : VTD_QI_TIMEOUT); + static unsigned int __read_mostly threshold = 1; + s_time_t start = NOW(); + s_time_t timeout = start + (flush_dev_iotlb + ? iommu_dev_iotlb_timeout + : 100) * MILLISECS(threshold); while ( ACCESS_ONCE(*this_poll_slot) != QINVAL_STAT_DONE ) { - if ( NOW() > timeout ) + if ( timeout && NOW() > timeout ) { - print_qi_regs(iommu); + threshold |= threshold << 1; printk(XENLOG_WARNING VTDPREFIX - " Queue invalidate wait descriptor timed out\n"); - return -ETIMEDOUT; + " IOMMU#%u: QI%s wait descriptor taking too long\n", + iommu->index, flush_dev_iotlb ? " dev" : ""); + print_qi_regs(iommu); + timeout = 0; } cpu_relax(); } + + if ( !timeout ) + printk(XENLOG_WARNING VTDPREFIX + " IOMMU#%u: QI%s wait descriptor took %lums\n", + iommu->index, flush_dev_iotlb ? " dev" : "", + (NOW() - start) / 10000000); + return 0; } -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.13 From xen-changelog-bounces@lists.xenproject.org Wed Jun 09 19:11:53 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Wed, 09 Jun 2021 19:11:53 +0000 Received: from list by lists.xenproject.org with outflank-mailman.139539.258003 (Exim 4.92) (envelope-from ) id 1lr3cH-0001PI-LB; Wed, 09 Jun 2021 19:11:53 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 139539.258003; Wed, 09 Jun 2021 19:11:53 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr3cH-0001PA-Hk; Wed, 09 Jun 2021 19:11:53 +0000 Received: by outflank-mailman (input) for mailman id 139539; Wed, 09 Jun 2021 19:11:53 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr3cH-0001P4-8Z for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 19:11:53 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr3cH-0006Xi-7u for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 19:11:53 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lr3cH-0000WW-7D for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 19:11:53 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=IOYitGnNxLv/QEvHj+sXm0/wcxld+2xe5B0ZOdUqJDA=; b=RSeRAwve+nTwupi9YGUzWxWmlz LvU2U6URpsNAmUlEymHjI8kyKQ7tuc3sjZ3o676l5y1i0d7c6K9jt3RzalKYV4PWn7AZ41WlXz8QO EQ9a4wn24JeFsRStXfZX89+b4/HC4QK8neYnQbVby32Y6pO1sloWYf95B+13vWdSuDvk=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.13] AMD/IOMMU: wait for command slot to be available Message-Id: Date: Wed, 09 Jun 2021 19:11:53 +0000 commit 0a3eb9bbc2188f36692a570c85816c6d3d0da697 Author: Jan Beulich AuthorDate: Tue Jun 8 18:58:05 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 18:58:12 2021 +0100 AMD/IOMMU: wait for command slot to be available No caller cared about send_iommu_command() indicating unavailability of a slot. Hence if a sufficient number prior commands timed out, we did blindly assume that the requested command was submitted to the IOMMU when really it wasn't. This could mean both a hanging system (waiting for a command to complete that was never seen by the IOMMU) or blindly propagating success back to callers, making them believe they're fine to e.g. free previously unmapped pages. Fold the three involved functions into one, add spin waiting for an available slot along the lines of VT-d's qinval_next_index(), and as a consequence drop all error indicator return types/values. This is part of XSA-373 / CVE-2021-28692. Signed-off-by: Jan Beulich Reviewed-by: Paul Durrant (cherry picked from commit c4e44343579e2c3304d676404d81b2b9bd893c5b) --- xen/drivers/passthrough/amd/iommu_cmd.c | 42 ++++++++++++--------------------- 1 file changed, 15 insertions(+), 27 deletions(-) diff --git a/xen/drivers/passthrough/amd/iommu_cmd.c b/xen/drivers/passthrough/amd/iommu_cmd.c index c9c7d7b290..3e48fc3733 100644 --- a/xen/drivers/passthrough/amd/iommu_cmd.c +++ b/xen/drivers/passthrough/amd/iommu_cmd.c @@ -22,48 +22,36 @@ #include #include "../ats.h" -static int queue_iommu_command(struct amd_iommu *iommu, u32 cmd[]) +static void send_iommu_command(struct amd_iommu *iommu, + const uint32_t cmd[4]) { - uint32_t tail, head; + uint32_t tail; tail = iommu->cmd_buffer.tail; if ( ++tail == iommu->cmd_buffer.entries ) tail = 0; - head = iommu_get_rb_pointer(readl(iommu->mmio_base + - IOMMU_CMD_BUFFER_HEAD_OFFSET)); - if ( head != tail ) + while ( tail == iommu_get_rb_pointer(readl(iommu->mmio_base + + IOMMU_CMD_BUFFER_HEAD_OFFSET)) ) { - memcpy(iommu->cmd_buffer.buffer + - (iommu->cmd_buffer.tail * sizeof(cmd_entry_t)), - cmd, sizeof(cmd_entry_t)); - - iommu->cmd_buffer.tail = tail; - return 1; + printk_once(XENLOG_ERR + "AMD IOMMU %04x:%02x:%02x.%u: no cmd slot available\n", + iommu->seg, PCI_BUS(iommu->bdf), + PCI_SLOT(iommu->bdf), PCI_FUNC(iommu->bdf)); + cpu_relax(); } - return 0; -} + memcpy(iommu->cmd_buffer.buffer + + (iommu->cmd_buffer.tail * sizeof(cmd_entry_t)), + cmd, sizeof(cmd_entry_t)); -static void commit_iommu_command_buffer(struct amd_iommu *iommu) -{ - u32 tail = 0; + iommu->cmd_buffer.tail = tail; + tail = 0; iommu_set_rb_pointer(&tail, iommu->cmd_buffer.tail); writel(tail, iommu->mmio_base+IOMMU_CMD_BUFFER_TAIL_OFFSET); } -int send_iommu_command(struct amd_iommu *iommu, u32 cmd[]) -{ - if ( queue_iommu_command(iommu, cmd) ) - { - commit_iommu_command_buffer(iommu); - return 1; - } - - return 0; -} - static void flush_command_buffer(struct amd_iommu *iommu) { u32 cmd[4], status; -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.13 From xen-changelog-bounces@lists.xenproject.org Wed Jun 09 19:12:04 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Wed, 09 Jun 2021 19:12:04 +0000 Received: from list by lists.xenproject.org with outflank-mailman.139540.258007 (Exim 4.92) (envelope-from ) id 1lr3cS-0001TH-OD; Wed, 09 Jun 2021 19:12:04 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 139540.258007; Wed, 09 Jun 2021 19:12:04 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr3cS-0001T9-Ki; Wed, 09 Jun 2021 19:12:04 +0000 Received: by outflank-mailman (input) for mailman id 139540; Wed, 09 Jun 2021 19:12:03 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr3cR-0001St-Bf for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 19:12:03 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr3cR-0006YN-Au for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 19:12:03 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lr3cR-0000Xf-A1 for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 19:12:03 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=dWpMmWq5K46YSHAMWN2cAMkrVwWteZcnBO8siOt3Bb0=; b=BBIiPHNf6dkE6XXScREu7UiDCj U4L5CerQknrjReyYozvOSXf73057/48WDHjtxge3iTYnqMtE3Mtp4+e3bdbUjvLFW39IeL4AvThEe zgd2muKODlaHYlV7Cej0hIvx8YT43POi/Az+LmEI3Zt8jNjxyAqcnC6C4lHkxr/ALGv0=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.13] AMD/IOMMU: drop command completion timeout Message-Id: Date: Wed, 09 Jun 2021 19:12:03 +0000 commit f7993297b3469f4396af8ed14ba376982cabd64b Author: Jan Beulich AuthorDate: Tue Jun 8 18:58:24 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 18:58:26 2021 +0100 AMD/IOMMU: drop command completion timeout First and foremost - such timeouts were not signaled to callers, making them believe they're fine to e.g. free previously unmapped pages. Mirror VT-d's behavior: A fixed number of loop iterations is not a suitable way to detect timeouts in an environment (CPU and bus speeds) independent manner anyway. Furthermore, leaving an in-progress operation pending when it appears to take too long is problematic: If a command completed later, the signaling of its completion may instead be understood to signal a subsequently started command's completion. Log excessively long processing times (with a progressive threshold) to have some indication of problems in this area. Allow callers to specify a non-default timeout bias for this logging, using the same values as VT-d does, which in particular means a (by default) much larger value for device IO TLB invalidation. This is part of XSA-373 / CVE-2021-28692. Signed-off-by: Jan Beulich Reviewed-by: Paul Durrant (cherry picked from commit e4fee66043120c954fc309bbb37813604c1c0eb7) --- xen/drivers/passthrough/amd/iommu_cmd.c | 60 +++++++++++++++++++-------------- 1 file changed, 35 insertions(+), 25 deletions(-) diff --git a/xen/drivers/passthrough/amd/iommu_cmd.c b/xen/drivers/passthrough/amd/iommu_cmd.c index 3e48fc3733..801b27af58 100644 --- a/xen/drivers/passthrough/amd/iommu_cmd.c +++ b/xen/drivers/passthrough/amd/iommu_cmd.c @@ -52,10 +52,12 @@ static void send_iommu_command(struct amd_iommu *iommu, writel(tail, iommu->mmio_base+IOMMU_CMD_BUFFER_TAIL_OFFSET); } -static void flush_command_buffer(struct amd_iommu *iommu) +static void flush_command_buffer(struct amd_iommu *iommu, + unsigned int timeout_base) { - u32 cmd[4], status; - int loop_count, comp_wait; + uint32_t cmd[4]; + s_time_t start, timeout; + static unsigned int __read_mostly threshold = 1; /* RW1C 'ComWaitInt' in status register */ writel(IOMMU_STATUS_COMP_WAIT_INT_MASK, @@ -71,24 +73,31 @@ static void flush_command_buffer(struct amd_iommu *iommu) IOMMU_COMP_WAIT_I_FLAG_SHIFT, &cmd[0]); send_iommu_command(iommu, cmd); - /* Make loop_count long enough for polling completion wait bit */ - loop_count = 1000; - do { - status = readl(iommu->mmio_base + IOMMU_STATUS_MMIO_OFFSET); - comp_wait = get_field_from_reg_u32(status, - IOMMU_STATUS_COMP_WAIT_INT_MASK, - IOMMU_STATUS_COMP_WAIT_INT_SHIFT); - --loop_count; - } while ( !comp_wait && loop_count ); - - if ( comp_wait ) + start = NOW(); + timeout = start + (timeout_base ?: 100) * MILLISECS(threshold); + while ( !(readl(iommu->mmio_base + IOMMU_STATUS_MMIO_OFFSET) & + IOMMU_STATUS_COMP_WAIT_INT_MASK) ) { - /* RW1C 'ComWaitInt' in status register */ - writel(IOMMU_STATUS_COMP_WAIT_INT_MASK, - iommu->mmio_base + IOMMU_STATUS_MMIO_OFFSET); - return; + if ( timeout && NOW() > timeout ) + { + threshold |= threshold << 1; + printk(XENLOG_WARNING + "AMD IOMMU %04x:%02x:%02x.%u: %scompletion wait taking too long\n", + iommu->seg, PCI_BUS(iommu->bdf), + PCI_SLOT(iommu->bdf), PCI_FUNC(iommu->bdf), + timeout_base ? "iotlb " : ""); + timeout = 0; + } + cpu_relax(); } - AMD_IOMMU_DEBUG("Warning: ComWaitInt bit did not assert!\n"); + + if ( !timeout ) + printk(XENLOG_WARNING + "AMD IOMMU %04x:%02x:%02x.%u: %scompletion wait took %lums\n", + iommu->seg, PCI_BUS(iommu->bdf), + PCI_SLOT(iommu->bdf), PCI_FUNC(iommu->bdf), + timeout_base ? "iotlb " : "", + (NOW() - start) / 10000000); } /* Build low level iommu command messages */ @@ -300,7 +309,7 @@ void amd_iommu_flush_iotlb(u8 devfn, const struct pci_dev *pdev, /* send INVALIDATE_IOTLB_PAGES command */ spin_lock_irqsave(&iommu->lock, flags); invalidate_iotlb_pages(iommu, maxpend, 0, queueid, daddr, req_id, order); - flush_command_buffer(iommu); + flush_command_buffer(iommu, iommu_dev_iotlb_timeout); spin_unlock_irqrestore(&iommu->lock, flags); } @@ -337,7 +346,7 @@ static void _amd_iommu_flush_pages(struct domain *d, { spin_lock_irqsave(&iommu->lock, flags); invalidate_iommu_pages(iommu, daddr, dom_id, order); - flush_command_buffer(iommu); + flush_command_buffer(iommu, 0); spin_unlock_irqrestore(&iommu->lock, flags); } @@ -361,7 +370,7 @@ void amd_iommu_flush_device(struct amd_iommu *iommu, uint16_t bdf) ASSERT( spin_is_locked(&iommu->lock) ); invalidate_dev_table_entry(iommu, bdf); - flush_command_buffer(iommu); + flush_command_buffer(iommu, 0); } void amd_iommu_flush_intremap(struct amd_iommu *iommu, uint16_t bdf) @@ -369,7 +378,7 @@ void amd_iommu_flush_intremap(struct amd_iommu *iommu, uint16_t bdf) ASSERT( spin_is_locked(&iommu->lock) ); invalidate_interrupt_table(iommu, bdf); - flush_command_buffer(iommu); + flush_command_buffer(iommu, 0); } void amd_iommu_flush_all_caches(struct amd_iommu *iommu) @@ -377,7 +386,7 @@ void amd_iommu_flush_all_caches(struct amd_iommu *iommu) ASSERT( spin_is_locked(&iommu->lock) ); invalidate_iommu_all(iommu); - flush_command_buffer(iommu); + flush_command_buffer(iommu, 0); } void amd_iommu_send_guest_cmd(struct amd_iommu *iommu, u32 cmd[]) @@ -387,7 +396,8 @@ void amd_iommu_send_guest_cmd(struct amd_iommu *iommu, u32 cmd[]) spin_lock_irqsave(&iommu->lock, flags); send_iommu_command(iommu, cmd); - flush_command_buffer(iommu); + /* TBD: Timeout selection may require peeking into cmd[]. */ + flush_command_buffer(iommu, 0); spin_unlock_irqrestore(&iommu->lock, flags); } -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.13 From xen-changelog-bounces@lists.xenproject.org Wed Jun 09 19:12:14 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Wed, 09 Jun 2021 19:12:14 +0000 Received: from list by lists.xenproject.org with outflank-mailman.139541.258011 (Exim 4.92) (envelope-from ) id 1lr3cc-0001WH-Q0; Wed, 09 Jun 2021 19:12:14 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 139541.258011; Wed, 09 Jun 2021 19:12:14 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr3cc-0001W6-Md; Wed, 09 Jun 2021 19:12:14 +0000 Received: by outflank-mailman (input) for mailman id 139541; Wed, 09 Jun 2021 19:12:13 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr3cb-0001Vt-F7 for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 19:12:13 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr3cb-0006Ya-ES for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 19:12:13 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lr3cb-0000YV-Db for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 19:12:13 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=94FqHee56kb434Lg0fbXSrIk2n/p4/j8qnL+Al3MWKo=; b=28WRBATR+y6UqkWtNx2qYc9XUN RhWHq9Ro1YTstBuIBX6ugDo3dTt3LufyHOLCuDJTfiY9DE3LL2O/Yc7MhLLZgcd9b63A8gP+fFNNP og6r16nD/4fUa6Rw5d3FUj4H+vYZaf1OuQBGHego9tg2p/Ofv2OBS/BUQetqYTj5urGY=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.13] x86/spec-ctrl: Protect against Speculative Code Store Bypass Message-Id: Date: Wed, 09 Jun 2021 19:12:13 +0000 commit 97af34f9c36cb77132d91dfbc0cf8d502185764d Author: Andrew Cooper AuthorDate: Thu Mar 11 14:39:11 2021 +0000 Commit: Andrew Cooper CommitDate: Tue Jun 8 18:58:34 2021 +0100 x86/spec-ctrl: Protect against Speculative Code Store Bypass Modern x86 processors have far-better-than-architecturally-guaranteed self modifying code detection. Typically, when a write hits an instruction in flight, a Machine Clear occurs to flush stale content in the frontend and backend. For self modifying code, before a write which hits an instruction in flight retires, the frontend can speculatively decode and execute the old instruction stream. Speculation of this form can suffer from type confusion in registers, and potentially leak data. Furthermore, updates are typically byte-wise, rather than atomic. Depending on timing, speculation can race ahead multiple times between individual writes, and execute the transiently-malformed instruction stream. Xen has stubs which are used in certain cases for emulation purposes. Inhibit speculation between updating the stub and executing it. This is XSA-375 / CVE-2021-0089. Signed-off-by: Andrew Cooper Reviewed-by: Jan Beulich (cherry picked from commit 45f59ed8865318bb0356954bad067f329677ce9e) --- xen/arch/x86/pv/emul-priv-op.c | 2 ++ xen/arch/x86/x86_emulate/x86_emulate.c | 1 + 2 files changed, 3 insertions(+) diff --git a/xen/arch/x86/pv/emul-priv-op.c b/xen/arch/x86/pv/emul-priv-op.c index 6dc4f92a84..59c15ca0e7 100644 --- a/xen/arch/x86/pv/emul-priv-op.c +++ b/xen/arch/x86/pv/emul-priv-op.c @@ -97,6 +97,8 @@ static io_emul_stub_t *io_emul_stub_setup(struct priv_op_ctxt *ctxt, u8 opcode, BUILD_BUG_ON(STUB_BUF_SIZE / 2 < MAX(9, /* Default emul stub */ 5 + IOEMUL_QUIRK_STUB_BYTES)); + block_speculation(); /* SCSB */ + /* Handy function-typed pointer to the stub. */ return (void *)stub_va; } diff --git a/xen/arch/x86/x86_emulate/x86_emulate.c b/xen/arch/x86/x86_emulate/x86_emulate.c index 2dee940adb..195924076c 100644 --- a/xen/arch/x86/x86_emulate/x86_emulate.c +++ b/xen/arch/x86/x86_emulate/x86_emulate.c @@ -1172,6 +1172,7 @@ static inline int mkec(uint8_t e, int32_t ec, ...) # define invoke_stub(pre, post, constraints...) do { \ stub_exn.info = (union stub_exception_token) { .raw = ~0 }; \ stub_exn.line = __LINE__; /* Utility outweighs livepatching cost */ \ + block_speculation(); /* SCSB */ \ asm volatile ( pre "\n\tINDIRECT_CALL %[stub]\n\t" post "\n" \ ".Lret%=:\n\t" \ ".pushsection .fixup,\"ax\"\n" \ -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.13 From xen-changelog-bounces@lists.xenproject.org Wed Jun 09 19:12:24 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Wed, 09 Jun 2021 19:12:24 +0000 Received: from list by lists.xenproject.org with outflank-mailman.139542.258015 (Exim 4.92) (envelope-from ) id 1lr3cm-0001ZZ-RC; Wed, 09 Jun 2021 19:12:24 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 139542.258015; Wed, 09 Jun 2021 19:12:24 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr3cm-0001ZR-OC; Wed, 09 Jun 2021 19:12:24 +0000 Received: by outflank-mailman (input) for mailman id 139542; Wed, 09 Jun 2021 19:12:23 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr3cl-0001Z9-IJ for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 19:12:23 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr3cl-0006Yl-HX for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 19:12:23 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lr3cl-0000a6-Gb for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 19:12:23 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=lrq81ccXGWnjxp7lRbTqSra5/AfcEeHigM35Vhbe+pA=; b=MUST/qA1gC6shmxMHP/Xj+34+Y ycgpDJrC/ld1Id32SK6PdvjPd/Xm00qoVDbUGnxxJJXLlX5igRgMa+/pLnO8yeoQ+yCH2kj3KJ8WD KqxZpVvLbwcWe+F3l24h8egW5WbR6mSKA4Ej6Uy0AqELI170D2etAIn+FIJk2jiSrOFs=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.13] x86/spec-ctrl: Mitigate TAA after S3 resume Message-Id: Date: Wed, 09 Jun 2021 19:12:23 +0000 commit 9bd6416528f9080cead0f8c22441f5568dbd0bf3 Author: Andrew Cooper AuthorDate: Thu May 20 01:21:39 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 18:58:34 2021 +0100 x86/spec-ctrl: Mitigate TAA after S3 resume The user chosen setting for MSR_TSX_CTRL needs restoring after S3. All APs get the correct setting via start_secondary(), but the BSP was missed out. This is XSA-377 / CVE-2021-28690. Fixes: 8c4330818f6 ("x86/spec-ctrl: Mitigate the TSX Asynchronous Abort sidechannel") Signed-off-by: Andrew Cooper Reviewed-by: Jan Beulich (cherry picked from commit 8cf276cb2e0b99b96333865873f56b0b31555ff1) --- xen/arch/x86/acpi/power.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/xen/arch/x86/acpi/power.c b/xen/arch/x86/acpi/power.c index ec679c4d3a..19b04aed22 100644 --- a/xen/arch/x86/acpi/power.c +++ b/xen/arch/x86/acpi/power.c @@ -288,6 +288,8 @@ static int enter_state(u32 state) microcode_update_one(); + tsx_init(); /* Needs microcode. May change HLE/RTM feature bits. */ + if ( !recheck_cpu_features(0) ) panic("Missing previously available feature(s)\n"); -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.13 From xen-changelog-bounces@lists.xenproject.org Wed Jun 09 22:11:13 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Wed, 09 Jun 2021 22:11:13 +0000 Received: from list by lists.xenproject.org with outflank-mailman.139560.258053 (Exim 4.92) (envelope-from ) id 1lr6Pp-0000en-Qq; Wed, 09 Jun 2021 22:11:13 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 139560.258053; Wed, 09 Jun 2021 22:11:13 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr6Pp-0000ef-Nt; Wed, 09 Jun 2021 22:11:13 +0000 Received: by outflank-mailman (input) for mailman id 139560; Wed, 09 Jun 2021 22:11:12 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr6Po-0000eX-JK for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 22:11:12 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr6Po-0001AQ-IY for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 22:11:12 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lr6Po-0004qm-HX for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 22:11:12 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=0ueTCcFKaQrzUFJAcAov/oUZR6b2PzeflA+AyZCpHzI=; b=pO3FFbFUkyD65jL3ANfInpA4U6 k7KnpcPC7etFls+vFKKR0tTca9YT4GxlVeBI96gtx55sCY1eKnngizfI+BHeO/NTJP7NB45Ii8ybd WYy1QOxC5iyrCygZ4yEOftAWHnd1jQH79mJyIJT6uNFJFSg4GIATlL0bdTiIzNj851sI=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.12] xen/arm: Boot modules should always be scrubbed if bootscrub={on, idle} Message-Id: Date: Wed, 09 Jun 2021 22:11:12 +0000 commit aa8866c1e0a92342166c2786ed4ae2455eb7ce61 Author: Julien Grall AuthorDate: Sat Apr 17 17:38:28 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 19:14:03 2021 +0100 xen/arm: Boot modules should always be scrubbed if bootscrub={on, idle} The function to initialize the pages (see init_heap_pages()) will request scrub when the admin request idle bootscrub (default) and state == SYS_STATE_active. When bootscrub=on, Xen will scrub any free pages in heap_init_late(). Currently, the boot modules (e.g. kernels, initramfs) will be discarded/ freed after heap_init_late() is called and system_state switched to SYS_STATE_active. This means the pages associated with the boot modules will not get scrubbed before getting re-purposed. If the memory is assigned to an untrusted domU, it may be able to retrieve secrets from the modules. This is part of XSA-372 / CVE-2021-28693. Fixes: 1774e9b1df27 ("xen/arm: introduce create_domUs") Signed-off-by: Julien Grall Reviewed-by: Jan Beulich Reviewed-by: Stefano Stabellini Tested-by: Stefano Stabellini (cherry picked from commit fd5dc41ceaed9cfcfa011cdfd50f264c89277a90) --- xen/arch/arm/setup.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-) diff --git a/xen/arch/arm/setup.c b/xen/arch/arm/setup.c index 0e54e9c73e..ba95c06d89 100644 --- a/xen/arch/arm/setup.c +++ b/xen/arch/arm/setup.c @@ -73,7 +73,6 @@ static __used void init_done(void) /* Must be done past setting system_state. */ unregister_init_virtual_region(); - discard_initial_modules(); free_init_memory(); startup_cpu_idle_loop(); } @@ -904,6 +903,12 @@ void __init start_xen(unsigned long boot_phys_offset, create_domUs(); + /* + * This needs to be called **before** heap_init_late() so modules + * will be scrubbed (unless suppressed). + */ + discard_initial_modules(); + heap_init_late(); init_trace_bufs(); -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.12 From xen-changelog-bounces@lists.xenproject.org Wed Jun 09 22:11:13 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Wed, 09 Jun 2021 22:11:13 +0000 Received: from list by lists.xenproject.org with outflank-mailman.139559.258048 (Exim 4.92) (envelope-from ) id 1lr6Pf-0000dO-PS; Wed, 09 Jun 2021 22:11:03 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 139559.258048; Wed, 09 Jun 2021 22:11:03 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr6Pf-0000dG-MQ; Wed, 09 Jun 2021 22:11:03 +0000 Received: by outflank-mailman (input) for mailman id 139559; Wed, 09 Jun 2021 22:11:02 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr6Pe-0000dA-Hd for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 22:11:02 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr6Pe-0001AH-FH for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 22:11:02 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lr6Pe-0004q7-EG for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 22:11:02 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=vrCQa9QmfmPWRu6Zo4ZTOCXPTFDmT985EkQzFG0cqrs=; b=sMXHeu3Rto+gLkVUhjS7mSxJHw CjAknH0Q0Z0wxQdB+CGrcLp7nGcrtTS6+sl2X+QBouZ1k3Z9cgAiPYmfUzCjPw8y/APsJYXJBTNRu /aHICdjIcxCBdmH74hxQU/1RZjIaGaNR8NHIwrBv9wYc1ery8CRzDoxj/ZpTPYiux1kk=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.12] xen/arm: Create dom0less domUs earlier Message-Id: Date: Wed, 09 Jun 2021 22:11:02 +0000 commit 2c395704e6fedcb1d2dee71c59ce37259aaad3b9 Author: Julien Grall AuthorDate: Mon May 17 17:47:13 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 19:14:03 2021 +0100 xen/arm: Create dom0less domUs earlier In a follow-up patch we will need to unallocate the boot modules before heap_init_late() is called. The modules will contain the domUs kernel and initramfs. Therefore Xen will need to create extra domUs (used by dom0less) before heap_init_late(). This has two consequences on dom0less: 1) Domains will not be unpaused as soon as they are created but once all have been created. However, Xen doesn't guarantee an order to unpause, so this is not something one could rely on. 2) The memory allocated for a domU will not be scrubbed anymore when an admin select bootscrub=on. This is not something we advertised, but if this is a concern we can introduce either force scrub for all domUs or a per-domain flag in the DT. The behavior for bootscrub=off and bootscrub=idle (default) has not changed. This is part of XSA-372 / CVE-2021-28693. Signed-off-by: Julien Grall Reviewed-by: Jan Beulich Reviewed-by: Stefano Stabellini Tested-by: Stefano Stabellini (cherry picked from commit 371347c5b64da699d9f5a0edda5dc496fd2b7a5c) --- xen/arch/arm/domain_build.c | 2 -- xen/arch/arm/setup.c | 9 +++++---- 2 files changed, 5 insertions(+), 6 deletions(-) diff --git a/xen/arch/arm/domain_build.c b/xen/arch/arm/domain_build.c index d9836779d1..6c5a6db144 100644 --- a/xen/arch/arm/domain_build.c +++ b/xen/arch/arm/domain_build.c @@ -2092,8 +2092,6 @@ void __init create_domUs(void) if ( construct_domU(d, node) != 0 ) panic("Could not set up domain %s\n", dt_node_name(node)); - - domain_unpause_by_systemcontroller(d); } } diff --git a/xen/arch/arm/setup.c b/xen/arch/arm/setup.c index d986f84f8d..0e54e9c73e 100644 --- a/xen/arch/arm/setup.c +++ b/xen/arch/arm/setup.c @@ -736,7 +736,7 @@ void __init start_xen(unsigned long boot_phys_offset, int cpus, i; const char *cmdline; struct bootmodule *xen_bootmodule; - struct domain *dom0; + struct domain *dom0, *d; struct xen_domctl_createdomain dom0_cfg = { .flags = XEN_DOMCTL_CDF_hvm_guest | XEN_DOMCTL_CDF_hap, .max_evtchn_port = -1, @@ -902,6 +902,8 @@ void __init start_xen(unsigned long boot_phys_offset, if ( construct_dom0(dom0) != 0) panic("Could not set up DOM0 guest OS\n"); + create_domUs(); + heap_init_late(); init_trace_bufs(); @@ -915,9 +917,8 @@ void __init start_xen(unsigned long boot_phys_offset, system_state = SYS_STATE_active; - create_domUs(); - - domain_unpause_by_systemcontroller(dom0); + for_each_domain( d ) + domain_unpause_by_systemcontroller(d); /* Switch on to the dynamically allocated stack for the idle vcpu * since the static one we're running on is about to be freed. */ -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.12 From xen-changelog-bounces@lists.xenproject.org Wed Jun 09 22:11:23 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Wed, 09 Jun 2021 22:11:23 +0000 Received: from list by lists.xenproject.org with outflank-mailman.139562.258057 (Exim 4.92) (envelope-from ) id 1lr6Pz-0000ho-T5; Wed, 09 Jun 2021 22:11:23 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 139562.258057; Wed, 09 Jun 2021 22:11:23 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr6Pz-0000hh-PS; Wed, 09 Jun 2021 22:11:23 +0000 Received: by outflank-mailman (input) for mailman id 139562; Wed, 09 Jun 2021 22:11:22 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr6Py-0000hW-MV for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 22:11:22 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr6Py-0001Ai-Lh for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 22:11:22 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lr6Py-0004rQ-Kn for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 22:11:22 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=XmUMbteh2VjTaSz9AxCCafe9A30rK7VMjR45h4kN48o=; b=dAwRmcOgxA97dfV5DKI5miEdA/ jsaTwlbGrpyMoPK8d8gmKVw5mh/vNWtZMQdnLaa9G0V4jwpVgTdSM8zHA73s5KXJomcfBQdJQQ123 pxhQ/uDZV57xj0MBWiAeWZTcHYyhx21cuR/mS6/U8ZloDnM2mRBqAHXU0DnKVoXgqPXY=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.12] VT-d: size qinval queue dynamically Message-Id: Date: Wed, 09 Jun 2021 22:11:22 +0000 commit 95d23c7563636e6304931a559b6610387ef2bf4b Author: Jan Beulich AuthorDate: Tue Jun 8 19:10:30 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 19:14:03 2021 +0100 VT-d: size qinval queue dynamically With the present synchronous model, we need two slots for every operation (the operation itself and a wait descriptor). There can be one such pair of requests pending per CPU. To ensure that under all normal circumstances a slot is always available when one is requested, size the queue ring according to the number of present CPUs. This is part of XSA-373 / CVE-2021-28692. Signed-off-by: Jan Beulich Reviewed-by: Paul Durrant (cherry picked from commit cbfa62bb140c8f10a0ae96db3ce06e22b3b9d302) --- xen/drivers/passthrough/vtd/iommu.h | 10 +------- xen/drivers/passthrough/vtd/qinval.c | 44 ++++++++++++++++++++++++++++-------- 2 files changed, 35 insertions(+), 19 deletions(-) diff --git a/xen/drivers/passthrough/vtd/iommu.h b/xen/drivers/passthrough/vtd/iommu.h index c9290a3996..3fa6a4ff13 100644 --- a/xen/drivers/passthrough/vtd/iommu.h +++ b/xen/drivers/passthrough/vtd/iommu.h @@ -450,17 +450,9 @@ struct qinval_entry { }q; }; -/* Order of queue invalidation pages(max is 8) */ -#define QINVAL_PAGE_ORDER 2 - -#define QINVAL_ARCH_PAGE_ORDER (QINVAL_PAGE_ORDER + PAGE_SHIFT_4K - PAGE_SHIFT) -#define QINVAL_ARCH_PAGE_NR ( QINVAL_ARCH_PAGE_ORDER < 0 ? \ - 1 : \ - 1 << QINVAL_ARCH_PAGE_ORDER ) - /* Each entry is 16 bytes, so 2^8 entries per page */ #define QINVAL_ENTRY_ORDER ( PAGE_SHIFT - 4 ) -#define QINVAL_ENTRY_NR (1 << (QINVAL_PAGE_ORDER + 8)) +#define QINVAL_MAX_ENTRY_NR (1u << (7 + QINVAL_ENTRY_ORDER)) /* Status data flag */ #define QINVAL_STAT_INIT 0 diff --git a/xen/drivers/passthrough/vtd/qinval.c b/xen/drivers/passthrough/vtd/qinval.c index 09cbd36ebb..103afc15fe 100644 --- a/xen/drivers/passthrough/vtd/qinval.c +++ b/xen/drivers/passthrough/vtd/qinval.c @@ -31,6 +31,9 @@ #define VTD_QI_TIMEOUT 1 +static unsigned int __read_mostly qi_pg_order; +static unsigned int __read_mostly qi_entry_nr; + static int __must_check invalidate_sync(struct iommu *iommu); static void print_qi_regs(struct iommu *iommu) @@ -55,7 +58,7 @@ static unsigned int qinval_next_index(struct iommu *iommu) tail >>= QINVAL_INDEX_SHIFT; /* (tail+1 == head) indicates a full queue, wait for HW */ - while ( ( tail + 1 ) % QINVAL_ENTRY_NR == + while ( ((tail + 1) & (qi_entry_nr - 1)) == ( dmar_readq(iommu->reg, DMAR_IQH_REG) >> QINVAL_INDEX_SHIFT ) ) cpu_relax(); @@ -68,7 +71,7 @@ static void qinval_update_qtail(struct iommu *iommu, unsigned int index) /* Need hold register lock when update tail */ ASSERT( spin_is_locked(&iommu->register_lock) ); - val = (index + 1) % QINVAL_ENTRY_NR; + val = (index + 1) & (qi_entry_nr - 1); dmar_writeq(iommu->reg, DMAR_IQT_REG, (val << QINVAL_INDEX_SHIFT)); } @@ -417,7 +420,27 @@ int enable_qinval(struct iommu *iommu) if ( qi_ctrl->qinval_maddr == 0 ) { drhd = iommu_to_drhd(iommu); - qi_ctrl->qinval_maddr = alloc_pgtable_maddr(drhd, QINVAL_ARCH_PAGE_NR); + if ( !qi_entry_nr ) + { + /* + * With the present synchronous model, we need two slots for every + * operation (the operation itself and a wait descriptor). There + * can be one such pair of requests pending per CPU. One extra + * entry is needed as the ring is considered full when there's + * only one entry left. + */ + BUILD_BUG_ON(CONFIG_NR_CPUS * 2 >= QINVAL_MAX_ENTRY_NR); + qi_pg_order = get_order_from_bytes((num_present_cpus() * 2 + 1) << + (PAGE_SHIFT - + QINVAL_ENTRY_ORDER)); + qi_entry_nr = 1u << (qi_pg_order + QINVAL_ENTRY_ORDER); + + dprintk(XENLOG_INFO VTDPREFIX, + "QI: using %u-entry ring(s)\n", qi_entry_nr); + } + + qi_ctrl->qinval_maddr = + alloc_pgtable_maddr(drhd, qi_entry_nr >> QINVAL_ENTRY_ORDER); if ( qi_ctrl->qinval_maddr == 0 ) { dprintk(XENLOG_WARNING VTDPREFIX, @@ -431,15 +454,16 @@ int enable_qinval(struct iommu *iommu) spin_lock_irqsave(&iommu->register_lock, flags); - /* Setup Invalidation Queue Address(IQA) register with the - * address of the page we just allocated. QS field at - * bits[2:0] to indicate size of queue is one 4KB page. - * That's 256 entries. Queued Head (IQH) and Queue Tail (IQT) - * registers are automatically reset to 0 with write - * to IQA register. + /* + * Setup Invalidation Queue Address (IQA) register with the address of the + * pages we just allocated. The QS field at bits[2:0] indicates the size + * (page order) of the queue. + * + * Queued Head (IQH) and Queue Tail (IQT) registers are automatically + * reset to 0 with write to IQA register. */ dmar_writeq(iommu->reg, DMAR_IQA_REG, - qi_ctrl->qinval_maddr | QINVAL_PAGE_ORDER); + qi_ctrl->qinval_maddr | qi_pg_order); dmar_writeq(iommu->reg, DMAR_IQT_REG, 0); -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.12 From xen-changelog-bounces@lists.xenproject.org Wed Jun 09 22:11:34 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Wed, 09 Jun 2021 22:11:34 +0000 Received: from list by lists.xenproject.org with outflank-mailman.139563.258061 (Exim 4.92) (envelope-from ) id 1lr6Q9-0000mI-WD; Wed, 09 Jun 2021 22:11:34 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 139563.258061; Wed, 09 Jun 2021 22:11:33 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr6Q9-0000mA-TF; Wed, 09 Jun 2021 22:11:33 +0000 Received: by outflank-mailman (input) for mailman id 139563; Wed, 09 Jun 2021 22:11:32 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr6Q8-0000ln-Ph for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 22:11:32 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr6Q8-0001Az-Ou for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 22:11:32 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lr6Q8-0004s6-O3 for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 22:11:32 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=mvEtQCRCo4KDB0a6878LkcQ40+8O6Je2bv3cWa+fJDQ=; b=Y2Zv2066NIdmYkrQSxDFNm0yX/ B8PI5E0s6wHIm6QPA10V/FAMu4x7JLPgUXjaVme12dvX3hGznWCnUx0BXkZmxqhL7hra1X7agpB9N 6wDvQ6PFduprnbOrRO9FJfPs9Iht3Mm3Ls5F+ef+qaYxesI3jAZZ+6nwRXMXYdX46J+M=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.12] AMD/IOMMU: size command buffer dynamically Message-Id: Date: Wed, 09 Jun 2021 22:11:32 +0000 commit ced413bdfddbfbaf72caf79333962fe199bea6d9 Author: Jan Beulich AuthorDate: Tue Jun 8 19:10:44 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 19:14:03 2021 +0100 AMD/IOMMU: size command buffer dynamically With the present synchronous model, we need two slots for every operation (the operation itself and a wait command). There can be one such pair of commands pending per CPU. To ensure that under all normal circumstances a slot is always available when one is requested, size the command ring according to the number of present CPUs. This is part of XSA-373 / CVE-2021-28692. Signed-off-by: Jan Beulich Reviewed-by: Paul Durrant (cherry picked from commit df242851ddc93ac0b0a3a20ecab34acc29e3b36b) --- xen/drivers/passthrough/amd/iommu_cmd.c | 4 ++-- xen/drivers/passthrough/amd/iommu_init.c | 28 +++++++++++++++++++++++++--- xen/include/asm-x86/hvm/svm/amd-iommu-defs.h | 7 ++----- 3 files changed, 29 insertions(+), 10 deletions(-) diff --git a/xen/drivers/passthrough/amd/iommu_cmd.c b/xen/drivers/passthrough/amd/iommu_cmd.c index af3a1fb865..c9c7d7b290 100644 --- a/xen/drivers/passthrough/amd/iommu_cmd.c +++ b/xen/drivers/passthrough/amd/iommu_cmd.c @@ -35,8 +35,8 @@ static int queue_iommu_command(struct amd_iommu *iommu, u32 cmd[]) if ( head != tail ) { memcpy(iommu->cmd_buffer.buffer + - (iommu->cmd_buffer.tail * IOMMU_CMD_BUFFER_ENTRY_SIZE), - cmd, IOMMU_CMD_BUFFER_ENTRY_SIZE); + (iommu->cmd_buffer.tail * sizeof(cmd_entry_t)), + cmd, sizeof(cmd_entry_t)); iommu->cmd_buffer.tail = tail; return 1; diff --git a/xen/drivers/passthrough/amd/iommu_init.c b/xen/drivers/passthrough/amd/iommu_init.c index 534d6bb889..98298f0bf0 100644 --- a/xen/drivers/passthrough/amd/iommu_init.c +++ b/xen/drivers/passthrough/amd/iommu_init.c @@ -136,7 +136,7 @@ static void register_iommu_cmd_buffer_in_mmio_space(struct amd_iommu *iommu) writel(entry, iommu->mmio_base + IOMMU_CMD_BUFFER_BASE_LOW_OFFSET); power_of2_entries = get_order_from_bytes(iommu->cmd_buffer.alloc_size) + - IOMMU_CMD_BUFFER_POWER_OF2_ENTRIES_PER_PAGE; + PAGE_SHIFT - IOMMU_CMD_BUFFER_ENTRY_ORDER; entry = 0; iommu_set_addr_hi_to_reg(&entry, addr_hi); @@ -1000,9 +1000,31 @@ static void * __init allocate_ring_buffer(struct ring_buffer *ring_buf, static void * __init allocate_cmd_buffer(struct amd_iommu *iommu) { /* allocate 'command buffer' in power of 2 increments of 4K */ + static unsigned int __read_mostly nr_ents; + + if ( !nr_ents ) + { + unsigned int order; + + /* + * With the present synchronous model, we need two slots for every + * operation (the operation itself and a wait command). There can be + * one such pair of requests pending per CPU. One extra entry is + * needed as the ring is considered full when there's only one entry + * left. + */ + BUILD_BUG_ON(CONFIG_NR_CPUS * 2 >= IOMMU_CMD_BUFFER_MAX_ENTRIES); + order = get_order_from_bytes((num_present_cpus() * 2 + 1) << + IOMMU_CMD_BUFFER_ENTRY_ORDER); + nr_ents = 1u << (order + PAGE_SHIFT - IOMMU_CMD_BUFFER_ENTRY_ORDER); + + AMD_IOMMU_DEBUG("using %u-entry cmd ring(s)\n", nr_ents); + } + + BUILD_BUG_ON(sizeof(cmd_entry_t) != (1u << IOMMU_CMD_BUFFER_ENTRY_ORDER)); + return allocate_ring_buffer(&iommu->cmd_buffer, sizeof(cmd_entry_t), - IOMMU_CMD_BUFFER_DEFAULT_ENTRIES, - "Command Buffer"); + nr_ents, "Command Buffer"); } static void * __init allocate_event_log(struct amd_iommu *iommu) diff --git a/xen/include/asm-x86/hvm/svm/amd-iommu-defs.h b/xen/include/asm-x86/hvm/svm/amd-iommu-defs.h index a217245249..d9ec20cc90 100644 --- a/xen/include/asm-x86/hvm/svm/amd-iommu-defs.h +++ b/xen/include/asm-x86/hvm/svm/amd-iommu-defs.h @@ -20,9 +20,6 @@ #ifndef _ASM_X86_64_AMD_IOMMU_DEFS_H #define _ASM_X86_64_AMD_IOMMU_DEFS_H -/* IOMMU Command Buffer entries: in power of 2 increments, minimum of 256 */ -#define IOMMU_CMD_BUFFER_DEFAULT_ENTRIES 512 - /* IOMMU Event Log entries: in power of 2 increments, minimum of 256 */ #define IOMMU_EVENT_LOG_DEFAULT_ENTRIES 512 @@ -184,8 +181,8 @@ #define IOMMU_CMD_BUFFER_LENGTH_MASK 0x0F000000 #define IOMMU_CMD_BUFFER_LENGTH_SHIFT 24 -#define IOMMU_CMD_BUFFER_ENTRY_SIZE 16 -#define IOMMU_CMD_BUFFER_POWER_OF2_ENTRIES_PER_PAGE 8 +#define IOMMU_CMD_BUFFER_ENTRY_ORDER 4 +#define IOMMU_CMD_BUFFER_MAX_ENTRIES (1u << 15) #define IOMMU_CMD_OPCODE_MASK 0xF0000000 #define IOMMU_CMD_OPCODE_SHIFT 28 -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.12 From xen-changelog-bounces@lists.xenproject.org Wed Jun 09 22:11:44 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Wed, 09 Jun 2021 22:11:44 +0000 Received: from list by lists.xenproject.org with outflank-mailman.139564.258065 (Exim 4.92) (envelope-from ) id 1lr6QK-0000pF-1a; Wed, 09 Jun 2021 22:11:44 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 139564.258065; Wed, 09 Jun 2021 22:11:44 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr6QJ-0000p7-Un; Wed, 09 Jun 2021 22:11:43 +0000 Received: by outflank-mailman (input) for mailman id 139564; Wed, 09 Jun 2021 22:11:42 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr6QI-0000oz-TC for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 22:11:42 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr6QI-0001BA-S6 for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 22:11:42 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lr6QI-0004sj-RA for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 22:11:42 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=KbliazqsiETxnGrKyKVdS4x7um7WOYIJ6VMvArdGXT8=; b=D83rENAyUcTxIW/8YAC47aj5fJ 811Jxhx5IpnFz2zaQlN9CO/xO12CyXUpGWfQ3BAVWD/i5rF1Nhd9wFHMDnAAdvbadgWfDKV9WLDmD HT/euVcCubLr2AmyHduLXYKPHfvDKyUpQEctfhO2M4jEh2rawcZ0llw2nm/bbnagUN3M=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.12] VT-d: eliminate flush related timeouts Message-Id: Date: Wed, 09 Jun 2021 22:11:42 +0000 commit e318c1330fbfb3388bdb8e5501a0ab89d6aa5a5a Author: Jan Beulich AuthorDate: Tue Jun 8 19:10:58 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 19:14:03 2021 +0100 VT-d: eliminate flush related timeouts Leaving an in-progress operation pending when it appears to take too long is problematic: If e.g. a QI command completed later, the write to the "poll slot" may instead be understood to signal a subsequently started command's completion. Also our accounting of the timeout period was actually wrong: We included the time it took for the command to actually make it to the front of the queue, which could be heavily affected by guests other than the one for which the flush is being performed. Do away with all timeout detection on all flush related code paths. Log excessively long processing times (with a progressive threshold) to have some indication of problems in this area. Additionally log (once) if qinval_next_index() didn't immediately find an available slot. Together with the earlier change sizing the queue(s) dynamically, we should now have a guarantee that with our fully synchronous model any demand for slots can actually be satisfied. This is part of XSA-373 / CVE-2021-28692. Signed-off-by: Jan Beulich Reviewed-by: Paul Durrant (cherry picked from commit 7c893abc4ee7e9af62297ba6fd5f27c89d0f33f5) --- xen/drivers/passthrough/vtd/dmar.h | 28 ++++++++++++++++++++++++++++ xen/drivers/passthrough/vtd/iommu.c | 12 ++++++------ xen/drivers/passthrough/vtd/qinval.c | 33 ++++++++++++++++++++++----------- 3 files changed, 56 insertions(+), 17 deletions(-) diff --git a/xen/drivers/passthrough/vtd/dmar.h b/xen/drivers/passthrough/vtd/dmar.h index 95bb132471..8438595556 100644 --- a/xen/drivers/passthrough/vtd/dmar.h +++ b/xen/drivers/passthrough/vtd/dmar.h @@ -127,6 +127,34 @@ do { \ } \ } while (0) +#define IOMMU_FLUSH_WAIT(what, iommu, offset, op, cond, sts) \ +do { \ + static unsigned int __read_mostly threshold = 1; \ + s_time_t start = NOW(); \ + s_time_t timeout = start + DMAR_OPERATION_TIMEOUT * threshold; \ + \ + for ( ; ; ) \ + { \ + sts = op(iommu->reg, offset); \ + if ( cond ) \ + break; \ + if ( timeout && NOW() > timeout ) \ + { \ + threshold |= threshold << 1; \ + printk(XENLOG_WARNING VTDPREFIX \ + " IOMMU#%u: %s flush taking too long\n", \ + iommu->index, what); \ + timeout = 0; \ + } \ + cpu_relax(); \ + } \ + \ + if ( !timeout ) \ + printk(XENLOG_WARNING VTDPREFIX \ + " IOMMU#%u: %s flush took %lums\n", \ + iommu->index, what, (NOW() - start) / 10000000); \ +} while ( false ) + int vtd_hw_check(void); void disable_pmr(struct iommu *iommu); int is_igd_drhd(struct acpi_drhd_unit *drhd); diff --git a/xen/drivers/passthrough/vtd/iommu.c b/xen/drivers/passthrough/vtd/iommu.c index f36fded6fe..89e0cab531 100644 --- a/xen/drivers/passthrough/vtd/iommu.c +++ b/xen/drivers/passthrough/vtd/iommu.c @@ -357,8 +357,8 @@ static void iommu_flush_write_buffer(struct iommu *iommu) dmar_writel(iommu->reg, DMAR_GCMD_REG, val | DMA_GCMD_WBF); /* Make sure hardware complete it */ - IOMMU_WAIT_OP(iommu, DMAR_GSTS_REG, dmar_readl, - !(val & DMA_GSTS_WBFS), val); + IOMMU_FLUSH_WAIT("write buffer", iommu, DMAR_GSTS_REG, dmar_readl, + !(val & DMA_GSTS_WBFS), val); spin_unlock_irqrestore(&iommu->register_lock, flags); } @@ -408,8 +408,8 @@ static int __must_check flush_context_reg(void *_iommu, u16 did, u16 source_id, dmar_writeq(iommu->reg, DMAR_CCMD_REG, val); /* Make sure hardware complete it */ - IOMMU_WAIT_OP(iommu, DMAR_CCMD_REG, dmar_readq, - !(val & DMA_CCMD_ICC), val); + IOMMU_FLUSH_WAIT("context", iommu, DMAR_CCMD_REG, dmar_readq, + !(val & DMA_CCMD_ICC), val); spin_unlock_irqrestore(&iommu->register_lock, flags); /* flush context entry will implicitly flush write buffer */ @@ -491,8 +491,8 @@ static int __must_check flush_iotlb_reg(void *_iommu, u16 did, u64 addr, dmar_writeq(iommu->reg, tlb_offset + 8, val); /* Make sure hardware complete it */ - IOMMU_WAIT_OP(iommu, (tlb_offset + 8), dmar_readq, - !(val & DMA_TLB_IVT), val); + IOMMU_FLUSH_WAIT("iotlb", iommu, (tlb_offset + 8), dmar_readq, + !(val & DMA_TLB_IVT), val); spin_unlock_irqrestore(&iommu->register_lock, flags); /* check IOTLB invalidation granularity */ diff --git a/xen/drivers/passthrough/vtd/qinval.c b/xen/drivers/passthrough/vtd/qinval.c index 103afc15fe..17c4235597 100644 --- a/xen/drivers/passthrough/vtd/qinval.c +++ b/xen/drivers/passthrough/vtd/qinval.c @@ -29,8 +29,6 @@ #include "extern.h" #include "../ats.h" -#define VTD_QI_TIMEOUT 1 - static unsigned int __read_mostly qi_pg_order; static unsigned int __read_mostly qi_entry_nr; @@ -60,7 +58,11 @@ static unsigned int qinval_next_index(struct iommu *iommu) /* (tail+1 == head) indicates a full queue, wait for HW */ while ( ((tail + 1) & (qi_entry_nr - 1)) == ( dmar_readq(iommu->reg, DMAR_IQH_REG) >> QINVAL_INDEX_SHIFT ) ) + { + printk_once(XENLOG_ERR VTDPREFIX " IOMMU#%u: no QI slot available\n", + iommu->index); cpu_relax(); + } return tail; } @@ -180,23 +182,32 @@ static int __must_check queue_invalidate_wait(struct iommu *iommu, /* Now we don't support interrupt method */ if ( sw ) { - s_time_t timeout; - - /* In case all wait descriptor writes to same addr with same data */ - timeout = NOW() + MILLISECS(flush_dev_iotlb ? - iommu_dev_iotlb_timeout : VTD_QI_TIMEOUT); + static unsigned int __read_mostly threshold = 1; + s_time_t start = NOW(); + s_time_t timeout = start + (flush_dev_iotlb + ? iommu_dev_iotlb_timeout + : 100) * MILLISECS(threshold); while ( ACCESS_ONCE(*this_poll_slot) != QINVAL_STAT_DONE ) { - if ( NOW() > timeout ) + if ( timeout && NOW() > timeout ) { - print_qi_regs(iommu); + threshold |= threshold << 1; printk(XENLOG_WARNING VTDPREFIX - " Queue invalidate wait descriptor timed out\n"); - return -ETIMEDOUT; + " IOMMU#%u: QI%s wait descriptor taking too long\n", + iommu->index, flush_dev_iotlb ? " dev" : ""); + print_qi_regs(iommu); + timeout = 0; } cpu_relax(); } + + if ( !timeout ) + printk(XENLOG_WARNING VTDPREFIX + " IOMMU#%u: QI%s wait descriptor took %lums\n", + iommu->index, flush_dev_iotlb ? " dev" : "", + (NOW() - start) / 10000000); + return 0; } -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.12 From xen-changelog-bounces@lists.xenproject.org Wed Jun 09 22:11:54 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Wed, 09 Jun 2021 22:11:54 +0000 Received: from list by lists.xenproject.org with outflank-mailman.139565.258069 (Exim 4.92) (envelope-from ) id 1lr6QU-0000s2-3N; Wed, 09 Jun 2021 22:11:54 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 139565.258069; Wed, 09 Jun 2021 22:11:54 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr6QU-0000rr-05; Wed, 09 Jun 2021 22:11:54 +0000 Received: by outflank-mailman (input) for mailman id 139565; Wed, 09 Jun 2021 22:11:53 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr6QS-0000rb-Vq for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 22:11:52 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr6QS-0001BL-V3 for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 22:11:52 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lr6QS-0004tQ-UE for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 22:11:52 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=D+wIwG6mZpBlLukittvtBc8DPMmELQ7DcdB2CvCDFc8=; b=TOEekYBJ7i+ktHFiTTQF37tdCv KKwKlz5IZH+yTR4LuExJG/FnBZ4v2xpkqlswlH9XUMeGb/aasdnCcMdBxv/ahJloS5wDW6ZbyJde0 ChWkSg59/dAUefmhKYpHhDFfM0nNz+QO7G/Ebl1djUNcH3ii1B6Kg4whEqxghTfp0TDI=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.12] AMD/IOMMU: wait for command slot to be available Message-Id: Date: Wed, 09 Jun 2021 22:11:52 +0000 commit e5f3be9d6ef53d4dfef0f7a550ce6cd93b9629e2 Author: Jan Beulich AuthorDate: Tue Jun 8 19:11:18 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 19:14:03 2021 +0100 AMD/IOMMU: wait for command slot to be available No caller cared about send_iommu_command() indicating unavailability of a slot. Hence if a sufficient number prior commands timed out, we did blindly assume that the requested command was submitted to the IOMMU when really it wasn't. This could mean both a hanging system (waiting for a command to complete that was never seen by the IOMMU) or blindly propagating success back to callers, making them believe they're fine to e.g. free previously unmapped pages. Fold the three involved functions into one, add spin waiting for an available slot along the lines of VT-d's qinval_next_index(), and as a consequence drop all error indicator return types/values. This is part of XSA-373 / CVE-2021-28692. Signed-off-by: Jan Beulich Reviewed-by: Paul Durrant (cherry picked from commit c4e44343579e2c3304d676404d81b2b9bd893c5b) --- xen/drivers/passthrough/amd/iommu_cmd.c | 42 ++++++++++++--------------------- 1 file changed, 15 insertions(+), 27 deletions(-) diff --git a/xen/drivers/passthrough/amd/iommu_cmd.c b/xen/drivers/passthrough/amd/iommu_cmd.c index c9c7d7b290..3e48fc3733 100644 --- a/xen/drivers/passthrough/amd/iommu_cmd.c +++ b/xen/drivers/passthrough/amd/iommu_cmd.c @@ -22,48 +22,36 @@ #include #include "../ats.h" -static int queue_iommu_command(struct amd_iommu *iommu, u32 cmd[]) +static void send_iommu_command(struct amd_iommu *iommu, + const uint32_t cmd[4]) { - uint32_t tail, head; + uint32_t tail; tail = iommu->cmd_buffer.tail; if ( ++tail == iommu->cmd_buffer.entries ) tail = 0; - head = iommu_get_rb_pointer(readl(iommu->mmio_base + - IOMMU_CMD_BUFFER_HEAD_OFFSET)); - if ( head != tail ) + while ( tail == iommu_get_rb_pointer(readl(iommu->mmio_base + + IOMMU_CMD_BUFFER_HEAD_OFFSET)) ) { - memcpy(iommu->cmd_buffer.buffer + - (iommu->cmd_buffer.tail * sizeof(cmd_entry_t)), - cmd, sizeof(cmd_entry_t)); - - iommu->cmd_buffer.tail = tail; - return 1; + printk_once(XENLOG_ERR + "AMD IOMMU %04x:%02x:%02x.%u: no cmd slot available\n", + iommu->seg, PCI_BUS(iommu->bdf), + PCI_SLOT(iommu->bdf), PCI_FUNC(iommu->bdf)); + cpu_relax(); } - return 0; -} + memcpy(iommu->cmd_buffer.buffer + + (iommu->cmd_buffer.tail * sizeof(cmd_entry_t)), + cmd, sizeof(cmd_entry_t)); -static void commit_iommu_command_buffer(struct amd_iommu *iommu) -{ - u32 tail = 0; + iommu->cmd_buffer.tail = tail; + tail = 0; iommu_set_rb_pointer(&tail, iommu->cmd_buffer.tail); writel(tail, iommu->mmio_base+IOMMU_CMD_BUFFER_TAIL_OFFSET); } -int send_iommu_command(struct amd_iommu *iommu, u32 cmd[]) -{ - if ( queue_iommu_command(iommu, cmd) ) - { - commit_iommu_command_buffer(iommu); - return 1; - } - - return 0; -} - static void flush_command_buffer(struct amd_iommu *iommu) { u32 cmd[4], status; -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.12 From xen-changelog-bounces@lists.xenproject.org Wed Jun 09 22:12:05 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Wed, 09 Jun 2021 22:12:05 +0000 Received: from list by lists.xenproject.org with outflank-mailman.139566.258073 (Exim 4.92) (envelope-from ) id 1lr6Qf-0000vM-5Q; Wed, 09 Jun 2021 22:12:05 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 139566.258073; Wed, 09 Jun 2021 22:12:05 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr6Qf-0000vE-1i; Wed, 09 Jun 2021 22:12:05 +0000 Received: by outflank-mailman (input) for mailman id 139566; Wed, 09 Jun 2021 22:12:03 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr6Qd-0000uu-2p for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 22:12:03 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr6Qd-0001Bz-1x for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 22:12:03 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lr6Qd-0004uP-0w for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 22:12:03 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=7eqO08FTHim9om3fF1+ykIq+UvTbpRBIR54q0/MgW+o=; b=02SqDJODQ1c6zas1LNLj3PBDw7 7ycPWDJFIU6jfuSf9SLBNxJJPtxBxfaCuXOAP7M/n1Y84PFHj4vk7C4ksOi4R+sPs45iifwje3U3K RBLgmpRy5Ru8xIqm8P8hQohWHmUoEt7BkUrncRIgtvA8iSKnJ6DYeVXNbsDSG1PLNRDg=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.12] AMD/IOMMU: drop command completion timeout Message-Id: Date: Wed, 09 Jun 2021 22:12:03 +0000 commit d024fe144d072ec1e3c7e03a7992a147cfdbd35e Author: Jan Beulich AuthorDate: Tue Jun 8 19:11:49 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 19:14:03 2021 +0100 AMD/IOMMU: drop command completion timeout First and foremost - such timeouts were not signaled to callers, making them believe they're fine to e.g. free previously unmapped pages. Mirror VT-d's behavior: A fixed number of loop iterations is not a suitable way to detect timeouts in an environment (CPU and bus speeds) independent manner anyway. Furthermore, leaving an in-progress operation pending when it appears to take too long is problematic: If a command completed later, the signaling of its completion may instead be understood to signal a subsequently started command's completion. Log excessively long processing times (with a progressive threshold) to have some indication of problems in this area. Allow callers to specify a non-default timeout bias for this logging, using the same values as VT-d does, which in particular means a (by default) much larger value for device IO TLB invalidation. This is part of XSA-373 / CVE-2021-28692. Signed-off-by: Jan Beulich Reviewed-by: Paul Durrant (cherry picked from commit e4fee66043120c954fc309bbb37813604c1c0eb7) --- xen/drivers/passthrough/amd/iommu_cmd.c | 60 +++++++++++++++++++-------------- 1 file changed, 35 insertions(+), 25 deletions(-) diff --git a/xen/drivers/passthrough/amd/iommu_cmd.c b/xen/drivers/passthrough/amd/iommu_cmd.c index 3e48fc3733..801b27af58 100644 --- a/xen/drivers/passthrough/amd/iommu_cmd.c +++ b/xen/drivers/passthrough/amd/iommu_cmd.c @@ -52,10 +52,12 @@ static void send_iommu_command(struct amd_iommu *iommu, writel(tail, iommu->mmio_base+IOMMU_CMD_BUFFER_TAIL_OFFSET); } -static void flush_command_buffer(struct amd_iommu *iommu) +static void flush_command_buffer(struct amd_iommu *iommu, + unsigned int timeout_base) { - u32 cmd[4], status; - int loop_count, comp_wait; + uint32_t cmd[4]; + s_time_t start, timeout; + static unsigned int __read_mostly threshold = 1; /* RW1C 'ComWaitInt' in status register */ writel(IOMMU_STATUS_COMP_WAIT_INT_MASK, @@ -71,24 +73,31 @@ static void flush_command_buffer(struct amd_iommu *iommu) IOMMU_COMP_WAIT_I_FLAG_SHIFT, &cmd[0]); send_iommu_command(iommu, cmd); - /* Make loop_count long enough for polling completion wait bit */ - loop_count = 1000; - do { - status = readl(iommu->mmio_base + IOMMU_STATUS_MMIO_OFFSET); - comp_wait = get_field_from_reg_u32(status, - IOMMU_STATUS_COMP_WAIT_INT_MASK, - IOMMU_STATUS_COMP_WAIT_INT_SHIFT); - --loop_count; - } while ( !comp_wait && loop_count ); - - if ( comp_wait ) + start = NOW(); + timeout = start + (timeout_base ?: 100) * MILLISECS(threshold); + while ( !(readl(iommu->mmio_base + IOMMU_STATUS_MMIO_OFFSET) & + IOMMU_STATUS_COMP_WAIT_INT_MASK) ) { - /* RW1C 'ComWaitInt' in status register */ - writel(IOMMU_STATUS_COMP_WAIT_INT_MASK, - iommu->mmio_base + IOMMU_STATUS_MMIO_OFFSET); - return; + if ( timeout && NOW() > timeout ) + { + threshold |= threshold << 1; + printk(XENLOG_WARNING + "AMD IOMMU %04x:%02x:%02x.%u: %scompletion wait taking too long\n", + iommu->seg, PCI_BUS(iommu->bdf), + PCI_SLOT(iommu->bdf), PCI_FUNC(iommu->bdf), + timeout_base ? "iotlb " : ""); + timeout = 0; + } + cpu_relax(); } - AMD_IOMMU_DEBUG("Warning: ComWaitInt bit did not assert!\n"); + + if ( !timeout ) + printk(XENLOG_WARNING + "AMD IOMMU %04x:%02x:%02x.%u: %scompletion wait took %lums\n", + iommu->seg, PCI_BUS(iommu->bdf), + PCI_SLOT(iommu->bdf), PCI_FUNC(iommu->bdf), + timeout_base ? "iotlb " : "", + (NOW() - start) / 10000000); } /* Build low level iommu command messages */ @@ -300,7 +309,7 @@ void amd_iommu_flush_iotlb(u8 devfn, const struct pci_dev *pdev, /* send INVALIDATE_IOTLB_PAGES command */ spin_lock_irqsave(&iommu->lock, flags); invalidate_iotlb_pages(iommu, maxpend, 0, queueid, daddr, req_id, order); - flush_command_buffer(iommu); + flush_command_buffer(iommu, iommu_dev_iotlb_timeout); spin_unlock_irqrestore(&iommu->lock, flags); } @@ -337,7 +346,7 @@ static void _amd_iommu_flush_pages(struct domain *d, { spin_lock_irqsave(&iommu->lock, flags); invalidate_iommu_pages(iommu, daddr, dom_id, order); - flush_command_buffer(iommu); + flush_command_buffer(iommu, 0); spin_unlock_irqrestore(&iommu->lock, flags); } @@ -361,7 +370,7 @@ void amd_iommu_flush_device(struct amd_iommu *iommu, uint16_t bdf) ASSERT( spin_is_locked(&iommu->lock) ); invalidate_dev_table_entry(iommu, bdf); - flush_command_buffer(iommu); + flush_command_buffer(iommu, 0); } void amd_iommu_flush_intremap(struct amd_iommu *iommu, uint16_t bdf) @@ -369,7 +378,7 @@ void amd_iommu_flush_intremap(struct amd_iommu *iommu, uint16_t bdf) ASSERT( spin_is_locked(&iommu->lock) ); invalidate_interrupt_table(iommu, bdf); - flush_command_buffer(iommu); + flush_command_buffer(iommu, 0); } void amd_iommu_flush_all_caches(struct amd_iommu *iommu) @@ -377,7 +386,7 @@ void amd_iommu_flush_all_caches(struct amd_iommu *iommu) ASSERT( spin_is_locked(&iommu->lock) ); invalidate_iommu_all(iommu); - flush_command_buffer(iommu); + flush_command_buffer(iommu, 0); } void amd_iommu_send_guest_cmd(struct amd_iommu *iommu, u32 cmd[]) @@ -387,7 +396,8 @@ void amd_iommu_send_guest_cmd(struct amd_iommu *iommu, u32 cmd[]) spin_lock_irqsave(&iommu->lock, flags); send_iommu_command(iommu, cmd); - flush_command_buffer(iommu); + /* TBD: Timeout selection may require peeking into cmd[]. */ + flush_command_buffer(iommu, 0); spin_unlock_irqrestore(&iommu->lock, flags); } -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.12 From xen-changelog-bounces@lists.xenproject.org Wed Jun 09 22:12:15 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Wed, 09 Jun 2021 22:12:15 +0000 Received: from list by lists.xenproject.org with outflank-mailman.139567.258076 (Exim 4.92) (envelope-from ) id 1lr6Qp-0000yq-8i; Wed, 09 Jun 2021 22:12:15 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 139567.258076; Wed, 09 Jun 2021 22:12:15 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr6Qp-0000yi-59; Wed, 09 Jun 2021 22:12:15 +0000 Received: by outflank-mailman (input) for mailman id 139567; Wed, 09 Jun 2021 22:12:13 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr6Qn-0000yH-6a for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 22:12:13 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr6Qn-0001C9-5o for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 22:12:13 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lr6Qn-0004v8-4u for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 22:12:13 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=SIkLrarIxAkxLpjr4Qb7NAnFfz3LaHdmgPa3qym49zE=; b=3Y/lNvlemu0cp8SiFJtC2wMR5j fRIYGP5wnpfbLTX2ToGwUtzeri29HQ+3Jt/lCtmUWDzsTbU52IskfbHkQ9DdPxCXwwdHP1/+aMxIJ hgOysJJBM3zKTWCkZJxMB7xG2rMYOUhbuovfIhdAAwUAjth4y1TOO5cd3fT08GWBYoKU=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.12] x86/spec-ctrl: Protect against Speculative Code Store Bypass Message-Id: Date: Wed, 09 Jun 2021 22:12:13 +0000 commit 99f2c4672368481307a52c24ff1a630a7451ba5a Author: Andrew Cooper AuthorDate: Thu Mar 11 14:39:11 2021 +0000 Commit: Andrew Cooper CommitDate: Tue Jun 8 19:14:03 2021 +0100 x86/spec-ctrl: Protect against Speculative Code Store Bypass Modern x86 processors have far-better-than-architecturally-guaranteed self modifying code detection. Typically, when a write hits an instruction in flight, a Machine Clear occurs to flush stale content in the frontend and backend. For self modifying code, before a write which hits an instruction in flight retires, the frontend can speculatively decode and execute the old instruction stream. Speculation of this form can suffer from type confusion in registers, and potentially leak data. Furthermore, updates are typically byte-wise, rather than atomic. Depending on timing, speculation can race ahead multiple times between individual writes, and execute the transiently-malformed instruction stream. Xen has stubs which are used in certain cases for emulation purposes. Inhibit speculation between updating the stub and executing it. This is XSA-375 / CVE-2021-0089. Signed-off-by: Andrew Cooper Reviewed-by: Jan Beulich (cherry picked from commit 45f59ed8865318bb0356954bad067f329677ce9e) --- xen/arch/x86/pv/emul-priv-op.c | 2 ++ xen/arch/x86/x86_emulate/x86_emulate.c | 1 + 2 files changed, 3 insertions(+) diff --git a/xen/arch/x86/pv/emul-priv-op.c b/xen/arch/x86/pv/emul-priv-op.c index 933036ea34..e9147e24c7 100644 --- a/xen/arch/x86/pv/emul-priv-op.c +++ b/xen/arch/x86/pv/emul-priv-op.c @@ -97,6 +97,8 @@ static io_emul_stub_t *io_emul_stub_setup(struct priv_op_ctxt *ctxt, u8 opcode, BUILD_BUG_ON(STUB_BUF_SIZE / 2 < MAX(9, /* Default emul stub */ 5 + IOEMUL_QUIRK_STUB_BYTES)); + asm volatile ( "lfence" ::: "memory" ); /* SCSB */ + /* Handy function-typed pointer to the stub. */ return (void *)stub_va; } diff --git a/xen/arch/x86/x86_emulate/x86_emulate.c b/xen/arch/x86/x86_emulate/x86_emulate.c index 09de6ee15d..c7c55fa572 100644 --- a/xen/arch/x86/x86_emulate/x86_emulate.c +++ b/xen/arch/x86/x86_emulate/x86_emulate.c @@ -1093,6 +1093,7 @@ static inline int mkec(uint8_t e, int32_t ec, ...) # define invoke_stub(pre, post, constraints...) do { \ stub_exn.info = (union stub_exception_token) { .raw = ~0 }; \ stub_exn.line = __LINE__; /* Utility outweighs livepatching cost */ \ + asm volatile ( "lfence" ::: "memory" ); /* SCSB */ \ asm volatile ( pre "\n\tINDIRECT_CALL %[stub]\n\t" post "\n" \ ".Lret%=:\n\t" \ ".pushsection .fixup,\"ax\"\n" \ -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.12 From xen-changelog-bounces@lists.xenproject.org Wed Jun 09 22:12:25 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Wed, 09 Jun 2021 22:12:25 +0000 Received: from list by lists.xenproject.org with outflank-mailman.139568.258081 (Exim 4.92) (envelope-from ) id 1lr6Qz-000124-A7; Wed, 09 Jun 2021 22:12:25 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 139568.258081; Wed, 09 Jun 2021 22:12:25 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr6Qz-00011w-75; Wed, 09 Jun 2021 22:12:25 +0000 Received: by outflank-mailman (input) for mailman id 139568; Wed, 09 Jun 2021 22:12:23 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr6Qx-00011F-9k for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 22:12:23 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr6Qx-0001CL-8f for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 22:12:23 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lr6Qx-0004vn-7z for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 22:12:23 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=PxWdNx7fK162u8Z4EM8GtmEN2QV2Lbm1aQf1csPbwo8=; b=cCFXSa0WjQDna0ggr7NaOkF+9l /r9CwL1RjJrW6rJ5Ylf0tsHN8u6I3blDuA7UuA3GN2ZfnbIrTbZz2xbuasoKiQ+GpHb/7FJZeVw4B IdbEVQ3ozda4AG3G6GIdwiGVGZ385bwXZgt99pz1Ai9Vs/u7n4t8haIRibBYev0OaCxM=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.12] x86/spec-ctrl: Mitigate TAA after S3 resume Message-Id: Date: Wed, 09 Jun 2021 22:12:23 +0000 commit ea20eee97e9e0861127a8070cc7b9ae3557b09fb Author: Andrew Cooper AuthorDate: Thu May 20 01:21:39 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 19:14:03 2021 +0100 x86/spec-ctrl: Mitigate TAA after S3 resume The user chosen setting for MSR_TSX_CTRL needs restoring after S3. All APs get the correct setting via start_secondary(), but the BSP was missed out. This is XSA-377 / CVE-2021-28690. Fixes: 8c4330818f6 ("x86/spec-ctrl: Mitigate the TSX Asynchronous Abort sidechannel") Signed-off-by: Andrew Cooper Reviewed-by: Jan Beulich (cherry picked from commit 8cf276cb2e0b99b96333865873f56b0b31555ff1) --- xen/arch/x86/acpi/power.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/xen/arch/x86/acpi/power.c b/xen/arch/x86/acpi/power.c index a07aa3b9ed..660363a3df 100644 --- a/xen/arch/x86/acpi/power.c +++ b/xen/arch/x86/acpi/power.c @@ -259,6 +259,8 @@ static int enter_state(u32 state) microcode_resume_cpu(0); + tsx_init(); /* Needs microcode. May change HLE/RTM feature bits. */ + if ( !recheck_cpu_features(0) ) panic("Missing previously available feature(s)\n"); -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.12 From xen-changelog-bounces@lists.xenproject.org Wed Jun 09 23:55:10 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Wed, 09 Jun 2021 23:55:10 +0000 Received: from list by lists.xenproject.org with outflank-mailman.139594.258133 (Exim 4.92) (envelope-from ) id 1lr82M-00030i-A6; Wed, 09 Jun 2021 23:55:06 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 139594.258133; Wed, 09 Jun 2021 23:55:06 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr82M-00030a-7C; Wed, 09 Jun 2021 23:55:06 +0000 Received: by outflank-mailman (input) for mailman id 139594; Wed, 09 Jun 2021 23:55:04 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr82K-00030U-Pv for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 23:55:04 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lr82K-0002tC-Ms for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 23:55:04 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lr82K-0003Ym-Ll for xen-changelog@lists.xenproject.org; Wed, 09 Jun 2021 23:55:04 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=/YpABR6ckWrzwv5+x7TToJwHSIoRDdRWNY5KVBpfCVA=; b=o2pAcx1HIqRKW/cDCgiqsL3ENd 8I7l8JZ8DGZDulTaQvnFc1U/OHQkQ3gWigpk6JX0JgLknD0MiXTC5A6FiT55iLv42tW8qW3QqRevj FN4SkSnXc7+kLYfC4vSMTQ9bl/39e3J64u4EOrJF6A0rSDwZlR9F2QAqwhSdZMaw6J74=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] xen/arm32: SPSR_hyp/SPSR Message-Id: Date: Wed, 09 Jun 2021 23:55:04 +0000 commit dfcffb128be46a3e413eaa941744536fe53c94b6 Author: Stefano Stabellini AuthorDate: Wed Jun 9 10:37:59 2021 -0700 Commit: Stefano Stabellini CommitDate: Wed Jun 9 16:52:29 2021 -0700 xen/arm32: SPSR_hyp/SPSR SPSR_hyp is not meant to be accessed from Hyp mode (EL2); accesses trigger UNPREDICTABLE behaviour. Xen should read/write SPSR instead. See: ARM DDI 0487D.b page G8-5993. This fixes booting Xen/arm32 on QEMU. Signed-off-by: Stefano Stabellini Reviewed-by: Julien Grall Reviewed-by: Edgar E. Iglesias Tested-by: Edgar E. Iglesias --- xen/arch/arm/arm32/entry.S | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/xen/arch/arm/arm32/entry.S b/xen/arch/arm/arm32/entry.S index f2f1bc7a31..4e421109db 100644 --- a/xen/arch/arm/arm32/entry.S +++ b/xen/arch/arm/arm32/entry.S @@ -170,7 +170,7 @@ ENDPROC(prepare_context_from_guest) mrc CP32(r11, HSR) /* Save exception syndrome */ str r11, [sp, #UREGS_hsr] - mrs r11, SPSR_hyp + mrs r11, SPSR str r11, [sp, #UREGS_cpsr] /* @@ -395,7 +395,7 @@ return_to_hypervisor: ldr r11, [sp, #UREGS_pc] msr ELR_hyp, r11 ldr r11, [sp, #UREGS_cpsr] - msr SPSR_hyp, r11 + msr SPSR, r11 #ifdef CONFIG_ARM32_HARDEN_BRANCH_PREDICTOR /* * Hardening branch predictor may require to setup a different -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Thu Jun 10 15:00:12 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 10 Jun 2021 15:00:12 +0000 Received: from list by lists.xenproject.org with outflank-mailman.139955.258674 (Exim 4.92) (envelope-from ) id 1lrMAA-0003HQ-FS; Thu, 10 Jun 2021 15:00:06 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 139955.258674; Thu, 10 Jun 2021 15:00:06 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lrMAA-0003HH-Am; Thu, 10 Jun 2021 15:00:06 +0000 Received: by outflank-mailman (input) for mailman id 139955; Thu, 10 Jun 2021 15:00:05 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lrMA9-0003BF-9g for xen-changelog@lists.xenproject.org; Thu, 10 Jun 2021 15:00:05 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lrMA9-0000T7-7e for xen-changelog@lists.xenproject.org; Thu, 10 Jun 2021 15:00:05 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lrMA9-0004xw-6V for xen-changelog@lists.xenproject.org; Thu, 10 Jun 2021 15:00:05 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=9inR2/wWtBGfpyoC+G/Lnp9jdzC5x0M6BP9c4Nivnxs=; b=ELqhcVRirpPFwgakkWYJ+sKzjz ShmqXNFora65LveLCZoQn/fwGevDxKa2CF34rULirtu0UqagkzznNCi1m2yEyfUYYxdQQXMuRJ8sJ h2w6oPT0cloluU7Ri1JEqleyxb1X5nMtjY5cBRre84/IAxhbkf73c69QeTBWKVwVuGV0=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] x86: please Clang in arch_set_info_guest() Message-Id: Date: Thu, 10 Jun 2021 15:00:05 +0000 commit 2bb17a45b1814b0b6aa4646eff58e16f876281fd Author: Jan Beulich AuthorDate: Thu Jun 10 16:56:24 2021 +0200 Commit: Jan Beulich CommitDate: Thu Jun 10 16:56:24 2021 +0200 x86: please Clang in arch_set_info_guest() Clang 10 reports domain.c:1328:10: error: variable 'cr3_mfn' is used uninitialized whenever 'if' condition is false [-Werror,-Wsometimes-uninitialized] if ( !compat ) ^~~~~~~ domain.c:1334:34: note: uninitialized use occurs here cr3_page = get_page_from_mfn(cr3_mfn, d); ^~~~~~~ domain.c:1328:5: note: remove the 'if' if its condition is always true if ( !compat ) ^~~~~~~~~~~~~~ domain.c:1042:18: note: initialize the variable 'cr3_mfn' to silence this warning mfn_t cr3_mfn; ^ = 0 domain.c:1189:14: error: variable 'fail' is used uninitialized whenever 'if' condition is false [-Werror,-Wsometimes-uninitialized] if ( !compat ) ^~~~~~~ domain.c:1211:9: note: uninitialized use occurs here fail |= v->arch.pv.gdt_ents != c(gdt_ents); ^~~~ domain.c:1189:9: note: remove the 'if' if its condition is always true if ( !compat ) ^~~~~~~~~~~~~~ domain.c:1187:18: note: initialize the variable 'fail' to silence this warning bool fail; ^ = false despite this being a build with -O2 in effect, and despite "compat" being constant "false" when CONFIG_COMPAT (and hence CONFIG_PV32) is not defined, as it gets set at the top of the function from the result of is_pv_32bit_domain(). Re-arrange the two "offending" if()s such that when COMPAT=n the respective variables will be seen as unconditionally initialized. The original aim was to have the !compat cases first, though. Reported-by: Andrew Cooper Signed-off-by: Jan Beulich Acked-by: Andrew Cooper --- xen/arch/x86/domain.c | 28 ++++++++++++++-------------- 1 file changed, 14 insertions(+), 14 deletions(-) diff --git a/xen/arch/x86/domain.c b/xen/arch/x86/domain.c index 50a27197b5..ef1812dc14 100644 --- a/xen/arch/x86/domain.c +++ b/xen/arch/x86/domain.c @@ -1186,7 +1186,17 @@ int arch_set_info_guest( unsigned long pfn = pagetable_get_pfn(v->arch.guest_table); bool fail; - if ( !compat ) +#ifdef CONFIG_COMPAT + if ( compat ) + { + l4_pgentry_t *l4tab = map_domain_page(_mfn(pfn)); + + pfn = l4e_get_pfn(*l4tab); + unmap_domain_page(l4tab); + fail = compat_pfn_to_cr3(pfn) != c.cmp->ctrlreg[3]; + } + else +#endif { fail = xen_pfn_to_cr3(pfn) != c.nat->ctrlreg[3]; if ( pagetable_is_null(v->arch.guest_table_user) ) @@ -1197,16 +1207,6 @@ int arch_set_info_guest( fail |= xen_pfn_to_cr3(pfn) != c.nat->ctrlreg[1]; } } -#ifdef CONFIG_COMPAT - else - { - l4_pgentry_t *l4tab = map_domain_page(_mfn(pfn)); - - pfn = l4e_get_pfn(*l4tab); - unmap_domain_page(l4tab); - fail = compat_pfn_to_cr3(pfn) != c.cmp->ctrlreg[3]; - } -#endif fail |= v->arch.pv.gdt_ents != c(gdt_ents); for ( i = 0; !fail && i < nr_gdt_frames; ++i ) @@ -1325,12 +1325,12 @@ int arch_set_info_guest( set_bit(_VPF_in_reset, &v->pause_flags); - if ( !compat ) - cr3_mfn = _mfn(xen_cr3_to_pfn(c.nat->ctrlreg[3])); #ifdef CONFIG_COMPAT - else + if ( compat ) cr3_mfn = _mfn(compat_cr3_to_pfn(c.cmp->ctrlreg[3])); + else #endif + cr3_mfn = _mfn(xen_cr3_to_pfn(c.nat->ctrlreg[3])); cr3_page = get_page_from_mfn(cr3_mfn, d); if ( !cr3_page ) -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Thu Jun 10 16:55:11 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 10 Jun 2021 16:55:11 +0000 Received: from list by lists.xenproject.org with outflank-mailman.139994.258739 (Exim 4.92) (envelope-from ) id 1lrNxQ-0007oN-4P; Thu, 10 Jun 2021 16:55:04 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 139994.258739; Thu, 10 Jun 2021 16:55:04 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lrNxQ-0007oF-1U; Thu, 10 Jun 2021 16:55:04 +0000 Received: by outflank-mailman (input) for mailman id 139994; Thu, 10 Jun 2021 16:55:02 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lrNxO-0007o9-Ja for xen-changelog@lists.xenproject.org; Thu, 10 Jun 2021 16:55:02 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lrNxO-0002xf-Gc for xen-changelog@lists.xenproject.org; Thu, 10 Jun 2021 16:55:02 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lrNxO-0008Qw-FA for xen-changelog@lists.xenproject.org; Thu, 10 Jun 2021 16:55:02 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=cJpBbDCcsjADNBhrVCiq3A6Hqbb3wjLkdzyZFczHO58=; b=TXqCDxCfJMw2JZvMXOljrCUaVb SGPvDBuw5ngbnLdw+Vbr8/qgqNg5nyZ0bIUBEwi7Sp7DMV4D+0Xyvz2SUyQGcwE2e2T9j5r3MjIRF V1G833R5FK4CeA97WjljASPT3GXOxBFsxNb66Gc1X/6LrPcFTbQ/urL/iZ00XLt6/DKA=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.15] AMD/IOMMU: wait for command slot to be available Message-Id: Date: Thu, 10 Jun 2021 16:55:02 +0000 commit 874dac9b27c09a1fe976dccd62d4509c0389ffea Author: Jan Beulich AuthorDate: Tue Jun 8 18:33:49 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 18:34:05 2021 +0100 AMD/IOMMU: wait for command slot to be available No caller cared about send_iommu_command() indicating unavailability of a slot. Hence if a sufficient number prior commands timed out, we did blindly assume that the requested command was submitted to the IOMMU when really it wasn't. This could mean both a hanging system (waiting for a command to complete that was never seen by the IOMMU) or blindly propagating success back to callers, making them believe they're fine to e.g. free previously unmapped pages. Fold the three involved functions into one, add spin waiting for an available slot along the lines of VT-d's qinval_next_index(), and as a consequence drop all error indicator return types/values. This is part of XSA-373 / CVE-2021-28692. Signed-off-by: Jan Beulich Reviewed-by: Paul Durrant (cherry picked from commit c4e44343579e2c3304d676404d81b2b9bd893c5b) --- xen/drivers/passthrough/amd/iommu_cmd.c | 39 +++++++++++---------------------- 1 file changed, 13 insertions(+), 26 deletions(-) diff --git a/xen/drivers/passthrough/amd/iommu_cmd.c b/xen/drivers/passthrough/amd/iommu_cmd.c index 54d7528f96..135883aaf8 100644 --- a/xen/drivers/passthrough/amd/iommu_cmd.c +++ b/xen/drivers/passthrough/amd/iommu_cmd.c @@ -20,43 +20,30 @@ #include "iommu.h" #include "../ats.h" -static int queue_iommu_command(struct amd_iommu *iommu, u32 cmd[]) +static void send_iommu_command(struct amd_iommu *iommu, + const uint32_t cmd[4]) { - uint32_t tail, head; + uint32_t tail; tail = iommu->cmd_buffer.tail + sizeof(cmd_entry_t); if ( tail == iommu->cmd_buffer.size ) tail = 0; - head = readl(iommu->mmio_base + - IOMMU_CMD_BUFFER_HEAD_OFFSET) & IOMMU_RING_BUFFER_PTR_MASK; - if ( head != tail ) + while ( tail == (readl(iommu->mmio_base + + IOMMU_CMD_BUFFER_HEAD_OFFSET) & + IOMMU_RING_BUFFER_PTR_MASK) ) { - memcpy(iommu->cmd_buffer.buffer + iommu->cmd_buffer.tail, - cmd, sizeof(cmd_entry_t)); - - iommu->cmd_buffer.tail = tail; - return 1; + printk_once(XENLOG_ERR "AMD IOMMU %pp: no cmd slot available\n", + &PCI_SBDF2(iommu->seg, iommu->bdf)); + cpu_relax(); } - return 0; -} - -static void commit_iommu_command_buffer(struct amd_iommu *iommu) -{ - writel(iommu->cmd_buffer.tail, - iommu->mmio_base + IOMMU_CMD_BUFFER_TAIL_OFFSET); -} + memcpy(iommu->cmd_buffer.buffer + iommu->cmd_buffer.tail, + cmd, sizeof(cmd_entry_t)); -static int send_iommu_command(struct amd_iommu *iommu, u32 cmd[]) -{ - if ( queue_iommu_command(iommu, cmd) ) - { - commit_iommu_command_buffer(iommu); - return 1; - } + iommu->cmd_buffer.tail = tail; - return 0; + writel(tail, iommu->mmio_base + IOMMU_CMD_BUFFER_TAIL_OFFSET); } static void flush_command_buffer(struct amd_iommu *iommu) -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.15 From xen-changelog-bounces@lists.xenproject.org Thu Jun 10 16:55:14 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 10 Jun 2021 16:55:14 +0000 Received: from list by lists.xenproject.org with outflank-mailman.139995.258743 (Exim 4.92) (envelope-from ) id 1lrNxa-0007pv-5p; Thu, 10 Jun 2021 16:55:14 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 139995.258743; Thu, 10 Jun 2021 16:55:14 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lrNxa-0007po-30; Thu, 10 Jun 2021 16:55:14 +0000 Received: by outflank-mailman (input) for mailman id 139995; Thu, 10 Jun 2021 16:55:12 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lrNxY-0007pe-Kg for xen-changelog@lists.xenproject.org; Thu, 10 Jun 2021 16:55:12 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lrNxY-0002yG-Jq for xen-changelog@lists.xenproject.org; Thu, 10 Jun 2021 16:55:12 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lrNxY-0008SF-In for xen-changelog@lists.xenproject.org; Thu, 10 Jun 2021 16:55:12 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=fPhJbnC9xYrMXpwrjSTR2JWffFA/64aKGIFgh9/Ad8k=; b=j55iuRL7nIr8XLmnLz7Dkg35e0 DbrQg55JFrS7gFk1O6XgQ0+uJ7/MZEZ+bio+Q4LDhGBO4NGEyPXrPW85SpGOses9IlbF+o/y9QIvn Md2CU5b+u04Fihs14dswLVUySYCA5AtD7gs9Wsc51LF28laANULXXPT0p0Nb/RUQbkpE=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.15] AMD/IOMMU: drop command completion timeout Message-Id: Date: Thu, 10 Jun 2021 16:55:12 +0000 commit a339ceaa8f17e827ee5eb25f05ad6f52ba8d6b1c Author: Jan Beulich AuthorDate: Tue Jun 8 18:34:36 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 18:34:38 2021 +0100 AMD/IOMMU: drop command completion timeout First and foremost - such timeouts were not signaled to callers, making them believe they're fine to e.g. free previously unmapped pages. Mirror VT-d's behavior: A fixed number of loop iterations is not a suitable way to detect timeouts in an environment (CPU and bus speeds) independent manner anyway. Furthermore, leaving an in-progress operation pending when it appears to take too long is problematic: If a command completed later, the signaling of its completion may instead be understood to signal a subsequently started command's completion. Log excessively long processing times (with a progressive threshold) to have some indication of problems in this area. Allow callers to specify a non-default timeout bias for this logging, using the same values as VT-d does, which in particular means a (by default) much larger value for device IO TLB invalidation. This is part of XSA-373 / CVE-2021-28692. Signed-off-by: Jan Beulich Reviewed-by: Paul Durrant (cherry picked from commit e4fee66043120c954fc309bbb37813604c1c0eb7) --- xen/drivers/passthrough/amd/iommu_cmd.c | 56 +++++++++++++++++++-------------- 1 file changed, 33 insertions(+), 23 deletions(-) diff --git a/xen/drivers/passthrough/amd/iommu_cmd.c b/xen/drivers/passthrough/amd/iommu_cmd.c index 135883aaf8..92a688dc9d 100644 --- a/xen/drivers/passthrough/amd/iommu_cmd.c +++ b/xen/drivers/passthrough/amd/iommu_cmd.c @@ -46,10 +46,12 @@ static void send_iommu_command(struct amd_iommu *iommu, writel(tail, iommu->mmio_base + IOMMU_CMD_BUFFER_TAIL_OFFSET); } -static void flush_command_buffer(struct amd_iommu *iommu) +static void flush_command_buffer(struct amd_iommu *iommu, + unsigned int timeout_base) { - unsigned int cmd[4], status, loop_count; - bool comp_wait; + uint32_t cmd[4]; + s_time_t start, timeout; + static unsigned int __read_mostly threshold = 1; /* RW1C 'ComWaitInt' in status register */ writel(IOMMU_STATUS_COMP_WAIT_INT, @@ -65,22 +67,29 @@ static void flush_command_buffer(struct amd_iommu *iommu) IOMMU_COMP_WAIT_I_FLAG_SHIFT, &cmd[0]); send_iommu_command(iommu, cmd); - /* Make loop_count long enough for polling completion wait bit */ - loop_count = 1000; - do { - status = readl(iommu->mmio_base + IOMMU_STATUS_MMIO_OFFSET); - comp_wait = status & IOMMU_STATUS_COMP_WAIT_INT; - --loop_count; - } while ( !comp_wait && loop_count ); - - if ( comp_wait ) + start = NOW(); + timeout = start + (timeout_base ?: 100) * MILLISECS(threshold); + while ( !(readl(iommu->mmio_base + IOMMU_STATUS_MMIO_OFFSET) & + IOMMU_STATUS_COMP_WAIT_INT) ) { - /* RW1C 'ComWaitInt' in status register */ - writel(IOMMU_STATUS_COMP_WAIT_INT, - iommu->mmio_base + IOMMU_STATUS_MMIO_OFFSET); - return; + if ( timeout && NOW() > timeout ) + { + threshold |= threshold << 1; + printk(XENLOG_WARNING + "AMD IOMMU %pp: %scompletion wait taking too long\n", + &PCI_SBDF2(iommu->seg, iommu->bdf), + timeout_base ? "iotlb " : ""); + timeout = 0; + } + cpu_relax(); } - AMD_IOMMU_DEBUG("Warning: ComWaitInt bit did not assert!\n"); + + if ( !timeout ) + printk(XENLOG_WARNING + "AMD IOMMU %pp: %scompletion wait took %lums\n", + &PCI_SBDF2(iommu->seg, iommu->bdf), + timeout_base ? "iotlb " : "", + (NOW() - start) / 10000000); } /* Build low level iommu command messages */ @@ -291,7 +300,7 @@ void amd_iommu_flush_iotlb(u8 devfn, const struct pci_dev *pdev, /* send INVALIDATE_IOTLB_PAGES command */ spin_lock_irqsave(&iommu->lock, flags); invalidate_iotlb_pages(iommu, maxpend, 0, queueid, daddr, req_id, order); - flush_command_buffer(iommu); + flush_command_buffer(iommu, iommu_dev_iotlb_timeout); spin_unlock_irqrestore(&iommu->lock, flags); } @@ -328,7 +337,7 @@ static void _amd_iommu_flush_pages(struct domain *d, { spin_lock_irqsave(&iommu->lock, flags); invalidate_iommu_pages(iommu, daddr, dom_id, order); - flush_command_buffer(iommu); + flush_command_buffer(iommu, 0); spin_unlock_irqrestore(&iommu->lock, flags); } @@ -352,7 +361,7 @@ void amd_iommu_flush_device(struct amd_iommu *iommu, uint16_t bdf) ASSERT( spin_is_locked(&iommu->lock) ); invalidate_dev_table_entry(iommu, bdf); - flush_command_buffer(iommu); + flush_command_buffer(iommu, 0); } void amd_iommu_flush_intremap(struct amd_iommu *iommu, uint16_t bdf) @@ -360,7 +369,7 @@ void amd_iommu_flush_intremap(struct amd_iommu *iommu, uint16_t bdf) ASSERT( spin_is_locked(&iommu->lock) ); invalidate_interrupt_table(iommu, bdf); - flush_command_buffer(iommu); + flush_command_buffer(iommu, 0); } void amd_iommu_flush_all_caches(struct amd_iommu *iommu) @@ -368,7 +377,7 @@ void amd_iommu_flush_all_caches(struct amd_iommu *iommu) ASSERT( spin_is_locked(&iommu->lock) ); invalidate_iommu_all(iommu); - flush_command_buffer(iommu); + flush_command_buffer(iommu, 0); } void amd_iommu_send_guest_cmd(struct amd_iommu *iommu, u32 cmd[]) @@ -378,7 +387,8 @@ void amd_iommu_send_guest_cmd(struct amd_iommu *iommu, u32 cmd[]) spin_lock_irqsave(&iommu->lock, flags); send_iommu_command(iommu, cmd); - flush_command_buffer(iommu); + /* TBD: Timeout selection may require peeking into cmd[]. */ + flush_command_buffer(iommu, 0); spin_unlock_irqrestore(&iommu->lock, flags); } -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.15 From xen-changelog-bounces@lists.xenproject.org Fri Jun 11 10:00:10 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Fri, 11 Jun 2021 10:00:10 +0000 Received: from list by lists.xenproject.org with outflank-mailman.140334.259322 (Exim 4.92) (envelope-from ) id 1lrdxO-00008H-2Z; Fri, 11 Jun 2021 10:00:06 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 140334.259322; Fri, 11 Jun 2021 10:00:06 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lrdxN-00007x-VO; Fri, 11 Jun 2021 10:00:05 +0000 Received: by outflank-mailman (input) for mailman id 140334; Fri, 11 Jun 2021 10:00:05 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lrdxN-0008PW-2E for xen-changelog@lists.xenproject.org; Fri, 11 Jun 2021 10:00:05 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lrdxM-0002dn-UL for xen-changelog@lists.xenproject.org; Fri, 11 Jun 2021 10:00:04 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lrdxM-0007FL-SZ for xen-changelog@lists.xenproject.org; Fri, 11 Jun 2021 10:00:04 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=7Q/WBnBuFFN/GsI6fHIecQAcvYCQcuqVIOYM8hPImC0=; b=ff6ehLzj1oKT3Hs+m4ZrI8IkJO va2lKze5GE8CGf1p4epu27mMmsUsAMG3Jy1OVZh3Lbb/yXEkzwFg+MDpQofHcrgG4iSQN+tXcRhUh he67h8MRoMyNfcPMTY0J0brGLKNwCRIbDDbVtPcNCKqptM0FAMFaU2+5WmLAFQADfAEQ=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] tools/libs/store: cleanup libxenstore interface Message-Id: Date: Fri, 11 Jun 2021 10:00:04 +0000 commit d2cad41defe4e0e9987549fbc8ebdf9ae138f90f Author: Juergen Gross AuthorDate: Wed May 12 16:48:32 2021 +0200 Commit: Ian Jackson CommitDate: Fri Jun 11 10:56:47 2021 +0100 tools/libs/store: cleanup libxenstore interface There are some internals in the libxenstore interface which should be removed. Move those functions into xs_lib.c and the related definitions into xs_lib.h. Remove the functions from the mapfile. Add xs_lib.o to xenstore_client as some of the internal functions are needed there. Bump the libxenstore version to 4.0 as the change is incompatible. Note that the removed functions should not result in any problem as they ought to be used by xenstored or xenstore_client only. Avoid an enum as part of a structure as the size of an enum is compiler implementation dependent. Signed-off-by: Juergen Gross Acked-by: Ian Jackson --- tools/include/xenstore_lib.h | 54 +++--------------- tools/libs/store/Makefile | 4 +- tools/libs/store/libxenstore.map | 10 +--- tools/libs/store/xs.c | 112 +----------------------------------- tools/xenstore/Makefile | 4 +- tools/xenstore/utils.h | 11 ++++ tools/xenstore/xenstore_client.c | 2 + tools/xenstore/xenstored_control.c | 1 + tools/xenstore/xenstored_core.c | 19 +++---- tools/xenstore/xenstored_core.h | 6 +- tools/xenstore/xenstored_watch.c | 2 +- tools/xenstore/xs_lib.c | 114 ++++++++++++++++++++++++++++++++++++- tools/xenstore/xs_lib.h | 50 ++++++++++++++++ tools/xenstore/xs_tdb_dump.c | 2 +- 14 files changed, 204 insertions(+), 187 deletions(-) diff --git a/tools/include/xenstore_lib.h b/tools/include/xenstore_lib.h index 4c9b6d1685..2266009ec1 100644 --- a/tools/include/xenstore_lib.h +++ b/tools/include/xenstore_lib.h @@ -26,42 +26,26 @@ #include #include -/* Bitmask of permissions. */ -enum xs_perm_type { - XS_PERM_NONE = 0, - XS_PERM_READ = 1, - XS_PERM_WRITE = 2, - /* Internal use. */ - XS_PERM_ENOENT_OK = 4, - XS_PERM_OWNER = 8, - XS_PERM_IGNORE = 16, -}; - struct xs_permissions { unsigned int id; - enum xs_perm_type perms; -}; - -/* Header of the node record in tdb. */ -struct xs_tdb_record_hdr { - uint64_t generation; - uint32_t num_perms; - uint32_t datalen; - uint32_t childlen; - struct xs_permissions perms[0]; + unsigned int perms; /* Bitmask of permissions. */ +#define XS_PERM_NONE 0x00 +#define XS_PERM_READ 0x01 +#define XS_PERM_WRITE 0x02 + /* Internal use. */ +#define XS_PERM_ENOENT_OK 0x04 +#define XS_PERM_OWNER 0x08 +#define XS_PERM_IGNORE 0x10 }; /* Each 10 bits takes ~ 3 digits, plus one, plus one for nul terminator. */ #define MAX_STRLEN(x) ((sizeof(x) * CHAR_BIT + CHAR_BIT-1) / 10 * 3 + 2) /* Path for various daemon things: env vars can override. */ -const char *xs_daemon_rootdir(void); const char *xs_daemon_rundir(void); const char *xs_daemon_socket(void); const char *xs_daemon_socket_ro(void); -const char *xs_domain_dev(void); -const char *xs_daemon_tdb(void); /* Simple write function: loops for you. */ bool xs_write_all(int fd, const void *data, unsigned int len); @@ -70,26 +54,4 @@ bool xs_write_all(int fd, const void *data, unsigned int len); bool xs_strings_to_perms(struct xs_permissions *perms, unsigned int num, const char *strings); -/* Convert permissions to a string (up to len MAX_STRLEN(unsigned int)+1). */ -bool xs_perm_to_string(const struct xs_permissions *perm, - char *buffer, size_t buf_len); - -/* Given a string and a length, count how many strings (nul terms). */ -unsigned int xs_count_strings(const char *strings, unsigned int len); - -/* Sanitising (quoting) possibly-binary strings. */ -struct expanding_buffer { - char *buf; - int avail; -}; - -/* Ensure that given expanding buffer has at least min_avail characters. */ -char *expanding_buffer_ensure(struct expanding_buffer *, int min_avail); - -/* sanitise_value() may return NULL if malloc fails. */ -char *sanitise_value(struct expanding_buffer *, const char *val, unsigned len); - -/* *out_len_r on entry is ignored; out must be at least strlen(in)+1 bytes. */ -void unsanitise_value(char *out, unsigned *out_len_r, const char *in); - #endif /* XENSTORE_LIB_H */ diff --git a/tools/libs/store/Makefile b/tools/libs/store/Makefile index bee57b5629..43b018aa8c 100644 --- a/tools/libs/store/Makefile +++ b/tools/libs/store/Makefile @@ -1,8 +1,8 @@ XEN_ROOT=$(CURDIR)/../../.. include $(XEN_ROOT)/tools/Rules.mk -MAJOR = 3.0 -MINOR = 3 +MAJOR = 4 +MINOR = 0 ifeq ($(CONFIG_Linux),y) APPEND_LDFLAGS += -ldl diff --git a/tools/libs/store/libxenstore.map b/tools/libs/store/libxenstore.map index 9854305a2c..7e6c7bdd30 100644 --- a/tools/libs/store/libxenstore.map +++ b/tools/libs/store/libxenstore.map @@ -1,4 +1,4 @@ -VERS_3.0.3 { +VERS_4.0 { global: xs_open; xs_close; @@ -32,18 +32,10 @@ VERS_3.0.3 { xs_control_command; xs_debug_command; xs_suspend_evtchn_port; - xs_daemon_rootdir; xs_daemon_rundir; xs_daemon_socket; xs_daemon_socket_ro; - xs_domain_dev; - xs_daemon_tdb; xs_write_all; xs_strings_to_perms; - xs_perm_to_string; - xs_count_strings; - expanding_buffer_ensure; - sanitise_value; - unsanitise_value; local: *; /* Do not expose anything by default */ }; diff --git a/tools/libs/store/xs.c b/tools/libs/store/xs.c index c91377c27f..7a9a8b1656 100644 --- a/tools/libs/store/xs.c +++ b/tools/libs/store/xs.c @@ -34,6 +34,7 @@ #include #include #include "xenstore.h" +#include "xs_lib.h" #include "list.h" #include "utils.h" @@ -1358,117 +1359,6 @@ static void *read_thread(void *arg) } #endif -char *expanding_buffer_ensure(struct expanding_buffer *ebuf, int min_avail) -{ - int want; - char *got; - - if (ebuf->avail >= min_avail) - return ebuf->buf; - - if (min_avail >= INT_MAX/3) - return 0; - - want = ebuf->avail + min_avail + 10; - got = realloc(ebuf->buf, want); - if (!got) - return 0; - - ebuf->buf = got; - ebuf->avail = want; - return ebuf->buf; -} - -char *sanitise_value(struct expanding_buffer *ebuf, - const char *val, unsigned len) -{ - int used, remain, c; - unsigned char *ip; - -#define ADD(c) (ebuf->buf[used++] = (c)) -#define ADDF(f,c) (used += sprintf(ebuf->buf+used, (f), (c))) - - assert(len < INT_MAX/5); - - ip = (unsigned char *)val; - used = 0; - remain = len; - - if (!expanding_buffer_ensure(ebuf, remain + 1)) - return NULL; - - while (remain-- > 0) { - c= *ip++; - - if (c >= ' ' && c <= '~' && c != '\\') { - ADD(c); - continue; - } - - if (!expanding_buffer_ensure(ebuf, used + remain + 5)) - /* for "\\nnn\0" */ - return 0; - - ADD('\\'); - switch (c) { - case '\t': ADD('t'); break; - case '\n': ADD('n'); break; - case '\r': ADD('r'); break; - case '\\': ADD('\\'); break; - default: - if (c < 010) ADDF("%03o", c); - else ADDF("x%02x", c); - } - } - - ADD(0); - assert(used <= ebuf->avail); - return ebuf->buf; - -#undef ADD -#undef ADDF -} - -void unsanitise_value(char *out, unsigned *out_len_r, const char *in) -{ - const char *ip; - char *op; - unsigned c; - int n; - - for (ip = in, op = out; (c = *ip++); *op++ = c) { - if (c == '\\') { - c = *ip++; - -#define GETF(f) do { \ - n = 0; \ - sscanf(ip, f "%n", &c, &n); \ - ip += n; \ - } while (0) - - switch (c) { - case 't': c= '\t'; break; - case 'n': c= '\n'; break; - case 'r': c= '\r'; break; - case '\\': c= '\\'; break; - case 'x': GETF("%2x"); break; - case '0': case '4': - case '1': case '5': - case '2': case '6': - case '3': case '7': --ip; GETF("%3o"); break; - case 0: --ip; break; - default:; - } -#undef GETF - } - } - - *op = 0; - - if (out_len_r) - *out_len_r = op - out; -} - /* * Local variables: * mode: C diff --git a/tools/xenstore/Makefile b/tools/xenstore/Makefile index 8ced12c27e..292b478fa1 100644 --- a/tools/xenstore/Makefile +++ b/tools/xenstore/Makefile @@ -78,8 +78,8 @@ xenstored.a: $(XENSTORED_OBJS) $(CLIENTS): xenstore ln -f xenstore $@ -xenstore: xenstore_client.o - $(CC) $< $(LDFLAGS) $(LDLIBS_libxenstore) $(LDLIBS_libxentoolcore) $(SOCKET_LIBS) -o $@ $(APPEND_LDFLAGS) +xenstore: xenstore_client.o xs_lib.o + $(CC) $^ $(LDFLAGS) $(LDLIBS_libxenstore) $(LDLIBS_libxentoolcore) $(SOCKET_LIBS) -o $@ $(APPEND_LDFLAGS) xenstore-control: xenstore_control.o $(CC) $< $(LDFLAGS) $(LDLIBS_libxenstore) $(LDLIBS_libxenctrl) $(LDLIBS_libxenguest) $(LDLIBS_libxentoolcore) $(SOCKET_LIBS) -o $@ $(APPEND_LDFLAGS) diff --git a/tools/xenstore/utils.h b/tools/xenstore/utils.h index 87713a8e5d..9d012b97c1 100644 --- a/tools/xenstore/utils.h +++ b/tools/xenstore/utils.h @@ -7,6 +7,17 @@ #include +#include "xenstore_lib.h" + +/* Header of the node record in tdb. */ +struct xs_tdb_record_hdr { + uint64_t generation; + uint32_t num_perms; + uint32_t datalen; + uint32_t childlen; + struct xs_permissions perms[0]; +}; + /* Is A == B ? */ #define streq(a,b) (strcmp((a),(b)) == 0) diff --git a/tools/xenstore/xenstore_client.c b/tools/xenstore/xenstore_client.c index ddbafc5175..0628ba275e 100644 --- a/tools/xenstore/xenstore_client.c +++ b/tools/xenstore/xenstore_client.c @@ -22,6 +22,8 @@ #include +#include "xs_lib.h" + #define PATH_SEP '/' #define MAX_PATH_LEN 256 diff --git a/tools/xenstore/xenstored_control.c b/tools/xenstore/xenstored_control.c index 2c63ba7ead..0d57f9f940 100644 --- a/tools/xenstore/xenstored_control.c +++ b/tools/xenstore/xenstored_control.c @@ -34,6 +34,7 @@ #include "utils.h" #include "talloc.h" +#include "xs_lib.h" #include "xenstored_core.h" #include "xenstored_control.h" #include "xenstored_domain.h" diff --git a/tools/xenstore/xenstored_core.c b/tools/xenstore/xenstored_core.c index df1e54c1f8..883a1a582a 100644 --- a/tools/xenstore/xenstored_core.c +++ b/tools/xenstore/xenstored_core.c @@ -46,7 +46,7 @@ #include "utils.h" #include "list.h" #include "talloc.h" -#include "xenstore_lib.h" +#include "xs_lib.h" #include "xenstored_core.h" #include "xenstored_watch.h" #include "xenstored_transaction.h" @@ -542,11 +542,11 @@ static int write_node(struct connection *conn, struct node *node, return write_node_raw(conn, &key, node, no_quota_check); } -enum xs_perm_type perm_for_conn(struct connection *conn, - const struct node_perms *perms) +unsigned int perm_for_conn(struct connection *conn, + const struct node_perms *perms) { unsigned int i; - enum xs_perm_type mask = XS_PERM_READ|XS_PERM_WRITE|XS_PERM_OWNER; + unsigned int mask = XS_PERM_READ|XS_PERM_WRITE|XS_PERM_OWNER; /* Owners and tools get it all... */ if (!domain_is_unprivileged(conn) || perms->p[0].id == conn->id @@ -584,7 +584,7 @@ char *get_parent(const void *ctx, const char *node) * Temporary memory allocations are done with ctx. */ static int ask_parents(struct connection *conn, const void *ctx, - const char *name, enum xs_perm_type *perm) + const char *name, unsigned int *perm) { struct node *node; @@ -618,10 +618,9 @@ static int ask_parents(struct connection *conn, const void *ctx, * Temporary memory allocations are done with ctx. */ static int errno_from_parents(struct connection *conn, const void *ctx, - const char *node, int errnum, - enum xs_perm_type perm) + const char *node, int errnum, unsigned int perm) { - enum xs_perm_type parent_perm = XS_PERM_NONE; + unsigned int parent_perm = XS_PERM_NONE; /* We always tell them about memory failures. */ if (errnum == ENOMEM) @@ -641,7 +640,7 @@ static int errno_from_parents(struct connection *conn, const void *ctx, static struct node *get_node(struct connection *conn, const void *ctx, const char *name, - enum xs_perm_type perm) + unsigned int perm) { struct node *node; @@ -873,7 +872,7 @@ static struct node *get_node_canonicalized(struct connection *conn, const void *ctx, const char *name, char **canonical_name, - enum xs_perm_type perm) + unsigned int perm) { char *tmp_name; diff --git a/tools/xenstore/xenstored_core.h b/tools/xenstore/xenstored_core.h index da43ec0e02..bb36111ecc 100644 --- a/tools/xenstore/xenstored_core.h +++ b/tools/xenstore/xenstored_core.h @@ -189,8 +189,8 @@ void send_ack(struct connection *conn, enum xsd_sockmsg_type type); char *canonicalize(struct connection *conn, const void *ctx, const char *node); /* Get access permissions. */ -enum xs_perm_type perm_for_conn(struct connection *conn, - const struct node_perms *perms); +unsigned int perm_for_conn(struct connection *conn, + const struct node_perms *perms); /* Write a node to the tdb data base. */ int write_node_raw(struct connection *conn, TDB_DATA *key, struct node *node, @@ -204,8 +204,6 @@ struct connection *new_connection(const struct interface_funcs *funcs); struct connection *get_connection_by_id(unsigned int conn_id); void check_store(void); void corrupt(struct connection *conn, const char *fmt, ...); -enum xs_perm_type perm_for_conn(struct connection *conn, - const struct node_perms *perms); /* Is this a valid node name? */ bool is_valid_nodename(const char *node); diff --git a/tools/xenstore/xenstored_watch.c b/tools/xenstore/xenstored_watch.c index db89e0141f..aca0a71bad 100644 --- a/tools/xenstore/xenstored_watch.c +++ b/tools/xenstore/xenstored_watch.c @@ -124,7 +124,7 @@ static bool watch_permitted(struct connection *conn, const void *ctx, const char *name, struct node *node, struct node_perms *perms) { - enum xs_perm_type perm; + unsigned int perm; struct node *parent; char *parent_name; diff --git a/tools/xenstore/xs_lib.c b/tools/xenstore/xs_lib.c index 80c03acbea..10fa4c3ad0 100644 --- a/tools/xenstore/xs_lib.c +++ b/tools/xenstore/xs_lib.c @@ -16,12 +16,13 @@ License along with this library; If not, see . */ +#include #include #include #include #include #include -#include "xenstore_lib.h" +#include "xs_lib.h" /* Common routines for the Xen store daemon and client library. */ @@ -179,3 +180,114 @@ unsigned int xs_count_strings(const char *strings, unsigned int len) return num; } + +char *expanding_buffer_ensure(struct expanding_buffer *ebuf, int min_avail) +{ + int want; + char *got; + + if (ebuf->avail >= min_avail) + return ebuf->buf; + + if (min_avail >= INT_MAX/3) + return 0; + + want = ebuf->avail + min_avail + 10; + got = realloc(ebuf->buf, want); + if (!got) + return 0; + + ebuf->buf = got; + ebuf->avail = want; + return ebuf->buf; +} + +char *sanitise_value(struct expanding_buffer *ebuf, + const char *val, unsigned len) +{ + int used, remain, c; + unsigned char *ip; + +#define ADD(c) (ebuf->buf[used++] = (c)) +#define ADDF(f,c) (used += sprintf(ebuf->buf+used, (f), (c))) + + assert(len < INT_MAX/5); + + ip = (unsigned char *)val; + used = 0; + remain = len; + + if (!expanding_buffer_ensure(ebuf, remain + 1)) + return NULL; + + while (remain-- > 0) { + c= *ip++; + + if (c >= ' ' && c <= '~' && c != '\\') { + ADD(c); + continue; + } + + if (!expanding_buffer_ensure(ebuf, used + remain + 5)) + /* for "\\nnn\0" */ + return 0; + + ADD('\\'); + switch (c) { + case '\t': ADD('t'); break; + case '\n': ADD('n'); break; + case '\r': ADD('r'); break; + case '\\': ADD('\\'); break; + default: + if (c < 010) ADDF("%03o", c); + else ADDF("x%02x", c); + } + } + + ADD(0); + assert(used <= ebuf->avail); + return ebuf->buf; + +#undef ADD +#undef ADDF +} + +void unsanitise_value(char *out, unsigned *out_len_r, const char *in) +{ + const char *ip; + char *op; + unsigned c; + int n; + + for (ip = in, op = out; (c = *ip++); *op++ = c) { + if (c == '\\') { + c = *ip++; + +#define GETF(f) do { \ + n = 0; \ + sscanf(ip, f "%n", &c, &n); \ + ip += n; \ + } while (0) + + switch (c) { + case 't': c= '\t'; break; + case 'n': c= '\n'; break; + case 'r': c= '\r'; break; + case '\\': c= '\\'; break; + case 'x': GETF("%2x"); break; + case '0': case '4': + case '1': case '5': + case '2': case '6': + case '3': case '7': --ip; GETF("%3o"); break; + case 0: --ip; break; + default:; + } +#undef GETF + } + } + + *op = 0; + + if (out_len_r) + *out_len_r = op - out; +} diff --git a/tools/xenstore/xs_lib.h b/tools/xenstore/xs_lib.h new file mode 100644 index 0000000000..efa05997d6 --- /dev/null +++ b/tools/xenstore/xs_lib.h @@ -0,0 +1,50 @@ +/* + Common routines between Xen store user library and daemon. + Copyright (C) 2005 Rusty Russell IBM Corporation + + This library is free software; you can redistribute it and/or + modify it under the terms of the GNU Lesser General Public + License as published by the Free Software Foundation; either + version 2.1 of the License, or (at your option) any later version. + + This library is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public + License along with this library; If not, see . +*/ + +#ifndef XS_LIB_H +#define XS_LIB_H + +#include "xenstore_lib.h" + +const char *xs_daemon_rootdir(void); +const char *xs_domain_dev(void); +const char *xs_daemon_tdb(void); + +/* Convert permissions to a string (up to len MAX_STRLEN(unsigned int)+1). */ +bool xs_perm_to_string(const struct xs_permissions *perm, + char *buffer, size_t buf_len); + +/* Given a string and a length, count how many strings (nul terms). */ +unsigned int xs_count_strings(const char *strings, unsigned int len); + +/* Sanitising (quoting) possibly-binary strings. */ +struct expanding_buffer { + char *buf; + int avail; +}; + +/* Ensure that given expanding buffer has at least min_avail characters. */ +char *expanding_buffer_ensure(struct expanding_buffer *, int min_avail); + +/* sanitise_value() may return NULL if malloc fails. */ +char *sanitise_value(struct expanding_buffer *, const char *val, unsigned len); + +/* *out_len_r on entry is ignored; out must be at least strlen(in)+1 bytes. */ +void unsanitise_value(char *out, unsigned *out_len_r, const char *in); + +#endif /* XS_LIB_H */ diff --git a/tools/xenstore/xs_tdb_dump.c b/tools/xenstore/xs_tdb_dump.c index f74676cf1c..5d2db392b4 100644 --- a/tools/xenstore/xs_tdb_dump.c +++ b/tools/xenstore/xs_tdb_dump.c @@ -17,7 +17,7 @@ static uint32_t total_size(struct xs_tdb_record_hdr *hdr) + hdr->datalen + hdr->childlen; } -static char perm_to_char(enum xs_perm_type perm) +static char perm_to_char(unsigned int perm) { return perm == XS_PERM_READ ? 'r' : perm == XS_PERM_WRITE ? 'w' : -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Fri Jun 11 22:11:13 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Fri, 11 Jun 2021 22:11:13 +0000 Received: from list by lists.xenproject.org with outflank-mailman.140669.259866 (Exim 4.92) (envelope-from ) id 1lrpMo-0003vS-Lq; Fri, 11 Jun 2021 22:11:06 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 140669.259866; Fri, 11 Jun 2021 22:11:06 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lrpMo-0003vK-I5; Fri, 11 Jun 2021 22:11:06 +0000 Received: by outflank-mailman (input) for mailman id 140669; Fri, 11 Jun 2021 22:11:04 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lrpMm-0003vE-OQ for xen-changelog@lists.xenproject.org; Fri, 11 Jun 2021 22:11:04 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lrpMm-0007Ta-NX for xen-changelog@lists.xenproject.org; Fri, 11 Jun 2021 22:11:04 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lrpMm-0005GF-MU for xen-changelog@lists.xenproject.org; Fri, 11 Jun 2021 22:11:04 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=lobV2Bh1FKYjUK6JTRA2WF03cpJ8GZiUCR96UrloR70=; b=7LfoXIaTCDzSlNqgnoqLDbb4jh tqtKizU+O5YfXx7yCf85RibSKqUxJ6nFO0gNJyGCg1WFe3MKeMYhTdNfJdJogzKSVAfLXk7BK2Q+R CeFzDdh6tukRy2uVvd/bAdaY/vi3ciXl8huiCurRFzXxL62QI8nsJNRFhBQwJwONroi8=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] Arm32: MSR to SPSR needs qualification Message-Id: Date: Fri, 11 Jun 2021 22:11:04 +0000 commit 93031fbe9f4c341a2e7950a088025ea550291433 Author: Jan Beulich AuthorDate: Fri Jun 11 15:04:24 2021 +0200 Commit: Stefano Stabellini CommitDate: Fri Jun 11 15:05:24 2021 -0700 Arm32: MSR to SPSR needs qualification The Arm ARM's description of MSR (ARM DDI 0406C.d section B9.3.12) doesn't even allow for plain "SPSR" here, and while gas accepts this, it takes it to mean SPSR_cf. Yet surely all of SPSR wants updating on this path, not just the lowest and highest 8 bits. Fixes: dfcffb128be4 ("xen/arm32: SPSR_hyp/SPSR") Signed-off-by: Jan Beulich Reviewed-by: Stefano Stabellini --- xen/arch/arm/arm32/entry.S | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/xen/arch/arm/arm32/entry.S b/xen/arch/arm/arm32/entry.S index 4e421109db..38826142ad 100644 --- a/xen/arch/arm/arm32/entry.S +++ b/xen/arch/arm/arm32/entry.S @@ -395,7 +395,7 @@ return_to_hypervisor: ldr r11, [sp, #UREGS_pc] msr ELR_hyp, r11 ldr r11, [sp, #UREGS_cpsr] - msr SPSR, r11 + msr SPSR_cxsf, r11 #ifdef CONFIG_ARM32_HARDEN_BRANCH_PREDICTOR /* * Hardening branch predictor may require to setup a different -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Mon Jun 14 10:11:12 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Mon, 14 Jun 2021 10:11:12 +0000 Received: from list by lists.xenproject.org with outflank-mailman.141307.261052 (Exim 4.92) (envelope-from ) id 1lsjYf-0001Gy-VG; Mon, 14 Jun 2021 10:11:05 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 141307.261052; Mon, 14 Jun 2021 10:11:05 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lsjYf-0001Gp-Rm; Mon, 14 Jun 2021 10:11:05 +0000 Received: by outflank-mailman (input) for mailman id 141307; Mon, 14 Jun 2021 10:11:04 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lsjYe-0001Gj-EZ for xen-changelog@lists.xenproject.org; Mon, 14 Jun 2021 10:11:04 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lsjYe-0001gT-50 for xen-changelog@lists.xenproject.org; Mon, 14 Jun 2021 10:11:04 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lsjYe-0005Za-3x for xen-changelog@lists.xenproject.org; Mon, 14 Jun 2021 10:11:04 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=uO5pvR+fFO4/VKf/VW7Kj+6LbZgdD+Plc5BkDopoCKE=; b=lwtQezRiMJl4qrd5l49B4n6Z7z muAkYeyHQx1oilB4NEANh71z79ijM1IrmafuTiX7B9EsXFtO5k7UrcqW0ZlHZ0Qka13nZcGqH1YhT GU1uxvw6ny6j4XX81/NenIXwPh7NXQAvU7q2zJCmHaPbM/6weiGMcwknHYY/DqvzLunQ=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] xen/grant-table: Simplify the update to the per-vCPU maptrack freelist Message-Id: Date: Mon, 14 Jun 2021 10:11:04 +0000 commit 4f1858763b7b1aeb79fa7c818eca98c96943aa69 Author: Julien Grall AuthorDate: Mon Jun 14 11:08:30 2021 +0100 Commit: Julien Grall CommitDate: Mon Jun 14 11:09:20 2021 +0100 xen/grant-table: Simplify the update to the per-vCPU maptrack freelist Since XSA-228 (commit 02cbeeb62075 "gnttab: split maptrack lock to make it fulfill its purpose again"), v->maptrack_head, v->maptrack_tail and the content of the freelist are accessed with the lock v->maptrack_freelist_lock held. Therefore it is not necessary to update the fields using cmpxchg() and also read them atomically. Note that there are two cases where v->maptrack_tail is accessed without the lock. They both happen in get_maptrack_handle() when initializing the free list of the current vCPU. Therefore there is no possible race. The code is now reworked to remove any use of cmpxch() and read_atomic() when accessing the fields v->maptrack_{head, tail} as wel as the freelist. Take the opportunity to add a comment on top of the lock definition and explain what it protects. Signed-off-by: Julien Grall Reviewed-by: Jan Beulich --- xen/common/grant_table.c | 66 ++++++++++++++++++++---------------------------- xen/include/xen/sched.h | 8 +++++- 2 files changed, 34 insertions(+), 40 deletions(-) diff --git a/xen/common/grant_table.c b/xen/common/grant_table.c index ab30e2e8cf..fab77ab9cc 100644 --- a/xen/common/grant_table.c +++ b/xen/common/grant_table.c @@ -543,33 +543,27 @@ double_gt_unlock(struct grant_table *lgt, struct grant_table *rgt) static inline grant_handle_t _get_maptrack_handle(struct grant_table *t, struct vcpu *v) { - unsigned int head, next, prev_head; + unsigned int head, next; spin_lock(&v->maptrack_freelist_lock); - do { - /* No maptrack pages allocated for this VCPU yet? */ - head = read_atomic(&v->maptrack_head); - if ( unlikely(head == MAPTRACK_TAIL) ) - { - spin_unlock(&v->maptrack_freelist_lock); - return INVALID_MAPTRACK_HANDLE; - } - - /* - * Always keep one entry in the free list to make it easier to - * add free entries to the tail. - */ - next = read_atomic(&maptrack_entry(t, head).ref); - if ( unlikely(next == MAPTRACK_TAIL) ) - { - spin_unlock(&v->maptrack_freelist_lock); - return INVALID_MAPTRACK_HANDLE; - } + /* No maptrack pages allocated for this VCPU yet? */ + head = v->maptrack_head; + if ( unlikely(head == MAPTRACK_TAIL) ) + { + spin_unlock(&v->maptrack_freelist_lock); + return INVALID_MAPTRACK_HANDLE; + } - prev_head = head; - head = cmpxchg(&v->maptrack_head, prev_head, next); - } while ( head != prev_head ); + /* + * Always keep one entry in the free list to make it easier to + * add free entries to the tail. + */ + next = maptrack_entry(t, head).ref; + if ( unlikely(next == MAPTRACK_TAIL) ) + head = INVALID_MAPTRACK_HANDLE; + else + v->maptrack_head = next; spin_unlock(&v->maptrack_freelist_lock); @@ -623,7 +617,7 @@ put_maptrack_handle( { struct domain *currd = current->domain; struct vcpu *v; - unsigned int prev_tail, cur_tail; + unsigned int tail; /* 1. Set entry to be a tail. */ maptrack_entry(t, handle).ref = MAPTRACK_TAIL; @@ -633,14 +627,11 @@ put_maptrack_handle( spin_lock(&v->maptrack_freelist_lock); - cur_tail = read_atomic(&v->maptrack_tail); - do { - prev_tail = cur_tail; - cur_tail = cmpxchg(&v->maptrack_tail, prev_tail, handle); - } while ( cur_tail != prev_tail ); + tail = v->maptrack_tail; + v->maptrack_tail = handle; /* 3. Update the old tail entry to point to the new entry. */ - write_atomic(&maptrack_entry(t, prev_tail).ref, handle); + maptrack_entry(t, tail).ref = handle; spin_unlock(&v->maptrack_freelist_lock); } @@ -650,7 +641,7 @@ get_maptrack_handle( struct grant_table *lgt) { struct vcpu *curr = current; - unsigned int i, head; + unsigned int i; grant_handle_t handle; struct grant_mapping *new_mt = NULL; @@ -686,7 +677,7 @@ get_maptrack_handle( maptrack_entry(lgt, handle).ref = MAPTRACK_TAIL; curr->maptrack_tail = handle; if ( curr->maptrack_head == MAPTRACK_TAIL ) - write_atomic(&curr->maptrack_head, handle); + curr->maptrack_head = handle; spin_unlock(&curr->maptrack_freelist_lock); } return steal_maptrack_handle(lgt, curr); @@ -707,7 +698,7 @@ get_maptrack_handle( new_mt[i].vcpu = curr->vcpu_id; } - /* Set tail directly if this is the first page for this VCPU. */ + /* Set tail directly if this is the first page for the local vCPU. */ if ( curr->maptrack_tail == MAPTRACK_TAIL ) curr->maptrack_tail = handle + MAPTRACK_PER_PAGE - 1; @@ -716,13 +707,10 @@ get_maptrack_handle( lgt->maptrack_limit += MAPTRACK_PER_PAGE; spin_unlock(&lgt->maptrack_lock); - spin_lock(&curr->maptrack_freelist_lock); - - do { - new_mt[i - 1].ref = read_atomic(&curr->maptrack_head); - head = cmpxchg(&curr->maptrack_head, new_mt[i - 1].ref, handle + 1); - } while ( head != new_mt[i - 1].ref ); + spin_lock(&curr->maptrack_freelist_lock); + new_mt[i - 1].ref = curr->maptrack_head; + curr->maptrack_head = handle + 1; spin_unlock(&curr->maptrack_freelist_lock); return handle; diff --git a/xen/include/xen/sched.h b/xen/include/xen/sched.h index 3982167144..28146ee404 100644 --- a/xen/include/xen/sched.h +++ b/xen/include/xen/sched.h @@ -255,7 +255,13 @@ struct vcpu /* VCPU paused by system controller. */ int controller_pause_count; - /* Grant table map tracking. */ + /* + * Grant table map tracking. The lock maptrack_freelist_lock + * protects: + * - entries in the freelist + * - maptrack_head + * - maptrack_tail + */ spinlock_t maptrack_freelist_lock; unsigned int maptrack_head; unsigned int maptrack_tail; -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Mon Jun 14 12:44:12 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Mon, 14 Jun 2021 12:44:12 +0000 Received: from list by lists.xenproject.org with outflank-mailman.141405.261194 (Exim 4.92) (envelope-from ) id 1lslwk-000576-Ek; Mon, 14 Jun 2021 12:44:06 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 141405.261194; Mon, 14 Jun 2021 12:44:06 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lslwk-00056y-Bn; Mon, 14 Jun 2021 12:44:06 +0000 Received: by outflank-mailman (input) for mailman id 141405; Mon, 14 Jun 2021 12:44:05 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lslwj-00056p-Al for xen-changelog@lists.xenproject.org; Mon, 14 Jun 2021 12:44:05 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lslwj-0004El-9e for xen-changelog@lists.xenproject.org; Mon, 14 Jun 2021 12:44:05 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lslwj-0000mP-8O for xen-changelog@lists.xenproject.org; Mon, 14 Jun 2021 12:44:05 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=yDvE54a2ftjJTkfEaqUh3r71Ws9i9kUCGHn8dwBAtL8=; b=BRw4lcW6S0bju0pwC3lItGUypq rDYPstPxpPYFA2vn3Qmxn0FKQo0ozQhwmd3zb/4vR9VeptcLGDyRy9DjC+kuypCByswjgcKi+1EVX tpz9k8vRPedHW2i/PN+4WEkVT6UOcQpody29IXHjt/nubZ9cnU8dvpXJXW3Rmcegz9QQ=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] revert "tools/libs/guest: fix max_pfn setting in map_p2m()" Message-Id: Date: Mon, 14 Jun 2021 12:44:05 +0000 commit 60d82f4bb0a22eb8765d33f87207afab25bc2396 Author: Juergen Gross AuthorDate: Mon Jun 14 14:39:52 2021 +0200 Commit: Jan Beulich CommitDate: Mon Jun 14 14:39:52 2021 +0200 revert "tools/libs/guest: fix max_pfn setting in map_p2m()" The reasoning for commit 7bd8989ab77b6a ("tools/libs/guest: fix max_pfn setting in map_p2m()") was wrong. The max_pfn field in shared_info is misnamed, it has the semantics of num_pfns, which is hidden at least partially in Linux, as the kernel is (wrongly) treating it like the highest used pfn in some places. So revert above commit. Fixes: 7bd8989ab77b6a ("tools/libs/guest: fix max_pfn setting in map_p2m()") Signed-off-by: Juergen Gross --- tools/libs/guest/xg_sr_save_x86_pv.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/tools/libs/guest/xg_sr_save_x86_pv.c b/tools/libs/guest/xg_sr_save_x86_pv.c index dae7f2817f..4964f1f7b8 100644 --- a/tools/libs/guest/xg_sr_save_x86_pv.c +++ b/tools/libs/guest/xg_sr_save_x86_pv.c @@ -468,7 +468,7 @@ static int map_p2m(struct xc_sr_context *ctx) ctx->x86.pv.p2m_generation = ~0ULL; ctx->x86.pv.max_pfn = GET_FIELD(ctx->x86.pv.shinfo, arch.max_pfn, - ctx->x86.pv.width); + ctx->x86.pv.width) - 1; p2m_cr3 = GET_FIELD(ctx->x86.pv.shinfo, arch.p2m_cr3, ctx->x86.pv.width); return p2m_cr3 ? map_p2m_list(ctx, p2m_cr3) : map_p2m_tree(ctx); -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Mon Jun 14 12:44:16 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Mon, 14 Jun 2021 12:44:16 +0000 Received: from list by lists.xenproject.org with outflank-mailman.141406.261198 (Exim 4.92) (envelope-from ) id 1lslwu-00058y-GG; Mon, 14 Jun 2021 12:44:16 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 141406.261198; Mon, 14 Jun 2021 12:44:16 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lslwu-00058q-DL; Mon, 14 Jun 2021 12:44:16 +0000 Received: by outflank-mailman (input) for mailman id 141406; Mon, 14 Jun 2021 12:44:15 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lslwt-00058e-E5 for xen-changelog@lists.xenproject.org; Mon, 14 Jun 2021 12:44:15 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lslwt-0004Ep-DB for xen-changelog@lists.xenproject.org; Mon, 14 Jun 2021 12:44:15 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lslwt-0000nD-CD for xen-changelog@lists.xenproject.org; Mon, 14 Jun 2021 12:44:15 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=5sCv4JeJwB5RIerV0QjqWaAHg8r9tLMfFuzpSeG5WnA=; b=SG6lCo17+ofr2b6UHVR0zX0WY5 71LuQHVORerQ11+hCZuFlRxkYqq6d4t8x1OxRYDw7Id9XXYTyIC6OCDu95Uz5a8LoBjiU3j1zDPcl HDPdue6WWKu5FUFx35aZv9GYyE8hay20ocX6HS5rN4Hdf3OGb7SlBt08QfdyUIS1IkF4=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] MAINTAINERS: add myself as tools/libs reviewer Message-Id: Date: Mon, 14 Jun 2021 12:44:15 +0000 commit f7079d7ef69f6bf38d6ec3bda294ed5eabcf98ba Author: Juergen Gross AuthorDate: Mon Jun 14 14:40:19 2021 +0200 Commit: Jan Beulich CommitDate: Mon Jun 14 14:40:19 2021 +0200 MAINTAINERS: add myself as tools/libs reviewer I have touched most of the Xen libraries in the past, and there is a clear lack of reviewer band width in the tools area. Add myself as a tools/libs reviewer for that reason. Signed-off-by: Juergen Gross Acked-by: Ian Jackson --- MAINTAINERS | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/MAINTAINERS b/MAINTAINERS index 7778c64fc7..39750bb75d 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -374,6 +374,13 @@ F: xen/include/{kexec,kimage}.h F: xen/arch/x86/machine_kexec.c F: xen/arch/x86/x86_64/kexec_reloc.S +LIBS: +M: Ian Jackson +M: Wei Liu +R: Juergen Gross +S: Supported +F: tools/libs/ + LIBXENLIGHT M: Ian Jackson M: Wei Liu -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Mon Jun 14 13:22:11 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Mon, 14 Jun 2021 13:22:11 +0000 Received: from list by lists.xenproject.org with outflank-mailman.141475.261310 (Exim 4.92) (envelope-from ) id 1lsmXU-000646-Tz; Mon, 14 Jun 2021 13:22:04 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 141475.261310; Mon, 14 Jun 2021 13:22:04 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lsmXU-00063y-RB; Mon, 14 Jun 2021 13:22:04 +0000 Received: by outflank-mailman (input) for mailman id 141475; Mon, 14 Jun 2021 13:22:04 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lsmXU-00063s-F6 for xen-changelog@lists.xenproject.org; Mon, 14 Jun 2021 13:22:04 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lsmXU-0004xR-Bu for xen-changelog@lists.xenproject.org; Mon, 14 Jun 2021 13:22:04 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lsmXU-0004B5-Ax for xen-changelog@lists.xenproject.org; Mon, 14 Jun 2021 13:22:04 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=fVcFwa3R9g2G1sByfiAMFhcL40VZlpmF+MuCdXXtqIk=; b=v9hGLZeVSldwUYP5NLakxu1Qye VzdydumzpsEMdbCJOVyLT3PkJVgsZRnX6TJDVD+ujm71x8oA6ZE9YCSwHbSknbbu8ki4ky+C7GEl+ lzcxmQPkTRfrsHY1szu2dRDg31tijZqVWILjm98VTWgxduZJzI8B8rDaDgNyKlHFV9Wc=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] xen/arm32: avoid .rodata to be marked as executable Message-Id: Date: Mon, 14 Jun 2021 13:22:04 +0000 commit 163f47c14737cfa9dfb3240deea356b08caf7614 Author: Jan Beulich AuthorDate: Fri Jun 11 11:19:15 2021 +0200 Commit: Julien Grall CommitDate: Mon Jun 14 14:14:11 2021 +0100 xen/arm32: avoid .rodata to be marked as executable The section .proc.info lives in .rodata as it doesn't contain any executable code. However, the section is still marked as executable as the consequence .rodata will also be marked executable. Xen doesn't use the ELF permissions to decide the page-table mapping permission. However, this will confuse disassemblers. '#execinstr' is now removed on all the pushsection dealing with .proc.info Signed-off-by: Jan Beulich [julieng: Rework the commit message] Acked-by: Julien Grall --- xen/arch/arm/arm32/proc-v7.S | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/xen/arch/arm/arm32/proc-v7.S b/xen/arch/arm/arm32/proc-v7.S index 1efde2d72d..c90a31d80f 100644 --- a/xen/arch/arm/arm32/proc-v7.S +++ b/xen/arch/arm/arm32/proc-v7.S @@ -29,7 +29,7 @@ brahma15mp_init: mcr CP32(r0, ACTLR) mov pc, lr - .section ".proc.info", #alloc, #execinstr + .section ".proc.info", #alloc .type __v7_ca15mp_proc_info, #object __v7_ca15mp_proc_info: .long 0x410FC0F0 /* Cortex-A15 */ @@ -38,7 +38,7 @@ __v7_ca15mp_proc_info: .long caxx_processor .size __v7_ca15mp_proc_info, . - __v7_ca15mp_proc_info - .section ".proc.info", #alloc, #execinstr + .section ".proc.info", #alloc .type __v7_ca7mp_proc_info, #object __v7_ca7mp_proc_info: .long 0x410FC070 /* Cortex-A7 */ @@ -47,7 +47,7 @@ __v7_ca7mp_proc_info: .long caxx_processor .size __v7_ca7mp_proc_info, . - __v7_ca7mp_proc_info - .section ".proc.info", #alloc, #execinstr + .section ".proc.info", #alloc .type __v7_brahma15mp_proc_info, #object __v7_brahma15mp_proc_info: .long 0x420F00F0 /* Broadcom Brahma-B15 */ -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Mon Jun 14 16:55:10 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Mon, 14 Jun 2021 16:55:10 +0000 Received: from list by lists.xenproject.org with outflank-mailman.141733.261706 (Exim 4.92) (envelope-from ) id 1lspre-0002u3-7u; Mon, 14 Jun 2021 16:55:06 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 141733.261706; Mon, 14 Jun 2021 16:55:06 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lspre-0002tv-4f; Mon, 14 Jun 2021 16:55:06 +0000 Received: by outflank-mailman (input) for mailman id 141733; Mon, 14 Jun 2021 16:55:05 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lsprd-0002th-Gl for xen-changelog@lists.xenproject.org; Mon, 14 Jun 2021 16:55:05 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lsprd-0000ns-Fz for xen-changelog@lists.xenproject.org; Mon, 14 Jun 2021 16:55:05 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lsprd-0008Fl-FF for xen-changelog@lists.xenproject.org; Mon, 14 Jun 2021 16:55:05 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=n2jSNyAxjXFlfVcTFri3iR6x+b+4GmQNtpfO9ilPgsI=; b=KMui6hG/iEGu/MOfFt7A5vHJVN 6kPp29TuASyqWsNmW7FaBlzb3b/iNVG+yhWN9F5gjV/yhuagAsfcmipZwMXyUVd0PfciDL8T8vigy SnhNL3OMPyQGX+X7Kd4vl9kTut/EdrVuWto58lWh9XQDu8Vvu++c6iv0dA12jqxpKaHo=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] Arm: avoid .init.data to be marked as executable Message-Id: Date: Mon, 14 Jun 2021 16:55:05 +0000 commit 8c9ed863738ff9e8b91975d6aa4464e7e8324eb7 Author: Jan Beulich AuthorDate: Mon Jun 14 15:52:36 2021 +0200 Commit: Julien Grall CommitDate: Mon Jun 14 17:53:52 2021 +0100 Arm: avoid .init.data to be marked as executable This confuses disassemblers, at the very least. Move .altinstr_replacement to .init.text. The previously redundant ALIGN() now gets converted to page alignment, such that the hypervisor mapping won't have this as executable (it'll instead get mapped r/w, which I'm told is intended to be adjusted at some point). Note that for the actual patching logic's purposes this part of .init.text _has_ to live after _einittext (or before _sinittext), or else branch_insn_requires_update() would produce wrong results. Also, to have .altinstr_replacement have consistent attributes in the object files, add "x" to the one instance where it was missing. Signed-off-by: Jan Beulich Acked-by: Julien Grall --- xen/arch/arm/xen.lds.S | 4 ++-- xen/include/asm-arm/alternative.h | 2 +- 2 files changed, 3 insertions(+), 3 deletions(-) diff --git a/xen/arch/arm/xen.lds.S b/xen/arch/arm/xen.lds.S index fc1d39a886..b773f91f1c 100644 --- a/xen/arch/arm/xen.lds.S +++ b/xen/arch/arm/xen.lds.S @@ -148,6 +148,8 @@ SECTIONS _sinittext = .; *(.init.text) _einittext = .; + . = ALIGN(PAGE_SIZE); /* Avoid mapping alt insns executable */ + *(.altinstr_replacement) } :text . = ALIGN(PAGE_SIZE); .init.data : { @@ -169,8 +171,6 @@ SECTIONS __alt_instructions = .; *(.altinstructions) __alt_instructions_end = .; - . = ALIGN(4); - *(.altinstr_replacement) #ifdef CONFIG_DEBUG_LOCK_PROFILE . = ALIGN(POINTER_ALIGN); diff --git a/xen/include/asm-arm/alternative.h b/xen/include/asm-arm/alternative.h index 92e888aff8..1eb4b60fbb 100644 --- a/xen/include/asm-arm/alternative.h +++ b/xen/include/asm-arm/alternative.h @@ -67,7 +67,7 @@ int apply_alternatives(const struct alt_instr *start, const struct alt_instr *en ALTINSTR_ENTRY(feature,cb) \ ".popsection\n" \ " .if " __stringify(cb) " == 0\n" \ - ".pushsection .altinstr_replacement, \"a\"\n" \ + ".pushsection .altinstr_replacement, \"ax\"\n" \ "663:\n\t" \ newinstr "\n" \ "664:\n\t" \ -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Tue Jun 15 13:11:13 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 15 Jun 2021 13:11:13 +0000 Received: from list by lists.xenproject.org with outflank-mailman.142008.262137 (Exim 4.92) (envelope-from ) id 1lt8qQ-0005JI-U4; Tue, 15 Jun 2021 13:11:06 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 142008.262137; Tue, 15 Jun 2021 13:11:06 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lt8qQ-0005J9-QQ; Tue, 15 Jun 2021 13:11:06 +0000 Received: by outflank-mailman (input) for mailman id 142008; Tue, 15 Jun 2021 13:11:05 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lt8qP-0005J3-CM for xen-changelog@lists.xenproject.org; Tue, 15 Jun 2021 13:11:05 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lt8qP-0003Ec-AM for xen-changelog@lists.xenproject.org; Tue, 15 Jun 2021 13:11:05 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lt8qP-0004h6-9F for xen-changelog@lists.xenproject.org; Tue, 15 Jun 2021 13:11:05 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=Z4xsuyiBvKdXJjnaU1BucDllNpGUszMaRZpDI4LXR/c=; b=MOOc0gQ16W5V5j+Ju8WdjkOyMv TMKWMebe5HcdnzmYRxukoIjWRs1+UIFI/OAPIQAkfPMX3s6qLWqtfUGuevoA3Uh6ZyItwrY1mFVUb zUvuHmuqnkZbimjbBK0q5fDBcudgRbYfX+TFTLGq2FnZY7zVFPAiPj8jKjdzc5TNfqmk=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] libxl: Replace deprecated QMP command by "query-cpus-fast" Message-Id: Date: Tue, 15 Jun 2021 13:11:05 +0000 commit d5f54009dba11d04bfe2a28eee47b994de66b84a Author: Anthony PERARD AuthorDate: Tue May 11 10:28:03 2021 +0100 Commit: Julien Grall CommitDate: Tue Jun 15 13:56:26 2021 +0100 libxl: Replace deprecated QMP command by "query-cpus-fast" We use the deprecated QMP command "query-cpus" which is removed in the QEMU 6.0 release. There's a replacement which is "query-cpus-fast", and have been available since QEMU 2.12 (April 2018). This patch try the new command first and when the command isn't available, it fall back to the deprecated one so libxl still works with older QEMU versions. Signed-off-by: Anthony PERARD Reviewed-by: Jason Andryuk --- tools/libs/light/libxl_domain.c | 103 ++++++++++++++++++++++++++++++++++++++-- 1 file changed, 98 insertions(+), 5 deletions(-) diff --git a/tools/libs/light/libxl_domain.c b/tools/libs/light/libxl_domain.c index 5d4ec90711..8c003aa7cb 100644 --- a/tools/libs/light/libxl_domain.c +++ b/tools/libs/light/libxl_domain.c @@ -1740,6 +1740,35 @@ out: return rc; } +static int qmp_parse_query_cpus_fast(libxl__gc *gc, + libxl_domid domid, + const libxl__json_object *response, + libxl_bitmap *const map) +{ + int i; + const libxl__json_object *cpu; + + libxl_bitmap_set_none(map); + /* Parse response to QMP command "query-cpus-fast": + * [ { 'cpu-index': 'int',...} ] + */ + for (i = 0; (cpu = libxl__json_array_get(response, i)); i++) { + unsigned int cpu_index; + const libxl__json_object *o; + + o = libxl__json_map_get("cpu-index", cpu, JSON_INTEGER); + if (!o) { + LOGD(ERROR, domid, "Failed to retrieve CPU index."); + return ERROR_QEMU_API; + } + + cpu_index = libxl__json_object_get_integer(o); + libxl_bitmap_set(map, cpu_index); + } + + return 0; +} + static int qmp_parse_query_cpus(libxl__gc *gc, libxl_domid domid, const libxl__json_object *response, @@ -1778,8 +1807,13 @@ typedef struct set_vcpuonline_state { int index; /* for loop on final_map */ } set_vcpuonline_state; +static void set_vcpuonline_qmp_cpus_fast_queried(libxl__egc *, + libxl__ev_qmp *, const libxl__json_object *, int rc); static void set_vcpuonline_qmp_cpus_queried(libxl__egc *, libxl__ev_qmp *, const libxl__json_object *, int rc); +static void set_vcpuonline_qmp_query_cpus_parse(libxl__egc *, + libxl__ev_qmp *qmp, const libxl__json_object *, + bool query_cpus_fast, int rc); static void set_vcpuonline_qmp_add_cpu(libxl__egc *, libxl__ev_qmp *, const libxl__json_object *response, int rc); static void set_vcpuonline_timeout(libxl__egc *egc, @@ -1840,8 +1874,8 @@ int libxl_set_vcpuonline(libxl_ctx *ctx, uint32_t domid, set_vcpuonline_timeout, LIBXL_QMP_CMD_TIMEOUT * 1000); if (rc) goto out; - qmp->callback = set_vcpuonline_qmp_cpus_queried; - rc = libxl__ev_qmp_send(egc, qmp, "query-cpus", NULL); + qmp->callback = set_vcpuonline_qmp_cpus_fast_queried; + rc = libxl__ev_qmp_send(egc, qmp, "query-cpus-fast", NULL); if (rc) goto out; return AO_INPROGRESS; default: @@ -1860,11 +1894,39 @@ out: return AO_INPROGRESS; } +static void set_vcpuonline_qmp_cpus_fast_queried(libxl__egc *egc, + libxl__ev_qmp *qmp, const libxl__json_object *response, int rc) +{ + EGC_GC; + set_vcpuonline_state *svos = CONTAINER_OF(qmp, *svos, qmp); + + if (rc == ERROR_QMP_COMMAND_NOT_FOUND) { + /* Try again, we probably talking to a QEMU older than 2.12 */ + qmp->callback = set_vcpuonline_qmp_cpus_queried; + rc = libxl__ev_qmp_send(egc, qmp, "query-cpus", NULL); + if (rc) goto out; + return; + } + +out: + set_vcpuonline_qmp_query_cpus_parse(egc, qmp, response, true, rc); +} + static void set_vcpuonline_qmp_cpus_queried(libxl__egc *egc, libxl__ev_qmp *qmp, const libxl__json_object *response, int rc) { EGC_GC; set_vcpuonline_state *svos = CONTAINER_OF(qmp, *svos, qmp); + + set_vcpuonline_qmp_query_cpus_parse(egc, qmp, response, false, rc); +} + +static void set_vcpuonline_qmp_query_cpus_parse(libxl__egc *egc, + libxl__ev_qmp *qmp, const libxl__json_object *response, + bool query_cpus_fast, int rc) +{ + EGC_GC; + set_vcpuonline_state *svos = CONTAINER_OF(qmp, *svos, qmp); int i; libxl_bitmap current_map; @@ -1876,7 +1938,11 @@ static void set_vcpuonline_qmp_cpus_queried(libxl__egc *egc, if (rc) goto out; libxl_bitmap_alloc(CTX, ¤t_map, svos->info.vcpu_max_id + 1); - rc = qmp_parse_query_cpus(gc, qmp->domid, response, ¤t_map); + if (query_cpus_fast) { + rc = qmp_parse_query_cpus_fast(gc, qmp->domid, response, ¤t_map); + } else { + rc = qmp_parse_query_cpus(gc, qmp->domid, response, ¤t_map); + } if (rc) goto out; libxl_bitmap_copy_alloc(CTX, final_map, svos->cpumap); @@ -2121,6 +2187,9 @@ typedef struct { static void retrieve_domain_configuration_lock_acquired( libxl__egc *egc, libxl__ev_slowlock *, int rc); +static void retrieve_domain_configuration_cpu_fast_queried( + libxl__egc *egc, libxl__ev_qmp *qmp, + const libxl__json_object *response, int rc); static void retrieve_domain_configuration_cpu_queried( libxl__egc *egc, libxl__ev_qmp *qmp, const libxl__json_object *response, int rc); @@ -2198,8 +2267,8 @@ static void retrieve_domain_configuration_lock_acquired( if (rc) goto out; libxl_bitmap_alloc(CTX, &rdcs->qemuu_cpus, d_config->b_info.max_vcpus); - rdcs->qmp.callback = retrieve_domain_configuration_cpu_queried; - rc = libxl__ev_qmp_send(egc, &rdcs->qmp, "query-cpus", NULL); + rdcs->qmp.callback = retrieve_domain_configuration_cpu_fast_queried; + rc = libxl__ev_qmp_send(egc, &rdcs->qmp, "query-cpus-fast", NULL); if (rc) goto out; has_callback = true; } @@ -2210,6 +2279,30 @@ out: retrieve_domain_configuration_end(egc, rdcs, rc); } +static void retrieve_domain_configuration_cpu_fast_queried( + libxl__egc *egc, libxl__ev_qmp *qmp, + const libxl__json_object *response, int rc) +{ + EGC_GC; + retrieve_domain_configuration_state *rdcs = + CONTAINER_OF(qmp, *rdcs, qmp); + + if (rc == ERROR_QMP_COMMAND_NOT_FOUND) { + /* Try again, we probably talking to a QEMU older than 2.12 */ + rdcs->qmp.callback = retrieve_domain_configuration_cpu_queried; + rc = libxl__ev_qmp_send(egc, &rdcs->qmp, "query-cpus", NULL); + if (rc) goto out; + return; + } + + if (rc) goto out; + + rc = qmp_parse_query_cpus_fast(gc, qmp->domid, response, &rdcs->qemuu_cpus); + +out: + retrieve_domain_configuration_end(egc, rdcs, rc); +} + static void retrieve_domain_configuration_cpu_queried( libxl__egc *egc, libxl__ev_qmp *qmp, const libxl__json_object *response, int rc) -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Tue Jun 15 13:11:16 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 15 Jun 2021 13:11:16 +0000 Received: from list by lists.xenproject.org with outflank-mailman.142009.262141 (Exim 4.92) (envelope-from ) id 1lt8qa-0005L6-Uu; Tue, 15 Jun 2021 13:11:16 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 142009.262141; Tue, 15 Jun 2021 13:11:16 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lt8qa-0005Ky-S2; Tue, 15 Jun 2021 13:11:16 +0000 Received: by outflank-mailman (input) for mailman id 142009; Tue, 15 Jun 2021 13:11:15 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lt8qZ-0005Ko-Et for xen-changelog@lists.xenproject.org; Tue, 15 Jun 2021 13:11:15 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lt8qZ-0003Eg-E0 for xen-changelog@lists.xenproject.org; Tue, 15 Jun 2021 13:11:15 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lt8qZ-0004hm-Cn for xen-changelog@lists.xenproject.org; Tue, 15 Jun 2021 13:11:15 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=7xGhIaSnoGU9iJYDqhuvbFpNNLAZ+WQlf+eORQqaPMI=; b=xwN5t2eB4KjvfWb6AwkaiCS9pr YeBmj0ZVxbou5OtkwiRE6zApl5sy8oyKPP+lZAGR4TpH5+0lyYqIU3o5RxApK1FCuhugfpS3EMGc/ LpT+NIGGDgcpJLtWVt47IlqZIn2YONMMzytRPdq0vcbJeYLuc2z6UjZW1uSwEd844bB4=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] libxl: Replace QEMU's command line short-form boolean option Message-Id: Date: Tue, 15 Jun 2021 13:11:15 +0000 commit f3f778c81769075ac0eb93b98d4b2803e7936453 Author: Anthony PERARD AuthorDate: Tue May 11 10:28:04 2021 +0100 Commit: Julien Grall CommitDate: Tue Jun 15 13:56:26 2021 +0100 libxl: Replace QEMU's command line short-form boolean option Short-form boolean options are deprecated in QEMU 6.0. Upstream commit that deprecate those: ccd3b3b8112b ("qemu-option: warn for short-form boolean options"). Signed-off-by: Anthony PERARD Reviewed-by: Jason Andryuk --- tools/libs/light/libxl_dm.c | 14 +++++++------- 1 file changed, 7 insertions(+), 7 deletions(-) diff --git a/tools/libs/light/libxl_dm.c b/tools/libs/light/libxl_dm.c index 3599a82ef0..0a0c1ef7c6 100644 --- a/tools/libs/light/libxl_dm.c +++ b/tools/libs/light/libxl_dm.c @@ -977,14 +977,14 @@ static char *dm_spice_options(libxl__gc *gc, if (spice->host) opt = GCSPRINTF("%s,addr=%s", opt, spice->host); if (libxl_defbool_val(spice->disable_ticketing)) - opt = GCSPRINTF("%s,disable-ticketing", opt); + opt = GCSPRINTF("%s,disable-ticketing=on", opt); else opt = GCSPRINTF("%s,password=%s", opt, spice->passwd); opt = GCSPRINTF("%s,agent-mouse=%s", opt, libxl_defbool_val(spice->agent_mouse) ? "on" : "off"); if (!libxl_defbool_val(spice->clipboard_sharing)) - opt = GCSPRINTF("%s,disable-copy-paste", opt); + opt = GCSPRINTF("%s,disable-copy-paste=on", opt); if (spice->image_compression) opt = GCSPRINTF("%s,image-compression=%s", opt, @@ -1224,7 +1224,7 @@ static int libxl__build_device_model_args_new(libxl__gc *gc, flexarray_append(dm_args, "-chardev"); if (state->dm_monitor_fd >= 0) { flexarray_append(dm_args, - GCSPRINTF("socket,id=libxl-cmd,fd=%d,server,nowait", + GCSPRINTF("socket,id=libxl-cmd,fd=%d,server=on,wait=off", state->dm_monitor_fd)); /* @@ -1237,7 +1237,7 @@ static int libxl__build_device_model_args_new(libxl__gc *gc, } else { flexarray_append(dm_args, GCSPRINTF("socket,id=libxl-cmd," - "path=%s,server,nowait", + "path=%s,server=on,wait=off", libxl__qemu_qmp_path(gc, guest_domid))); } @@ -1247,7 +1247,7 @@ static int libxl__build_device_model_args_new(libxl__gc *gc, flexarray_append(dm_args, "-chardev"); flexarray_append(dm_args, GCSPRINTF("socket,id=libxenstat-cmd," - "path=%s/qmp-libxenstat-%d,server,nowait", + "path=%s/qmp-libxenstat-%d,server=on,wait=off", libxl__run_dir_path(), guest_domid)); flexarray_append(dm_args, "-mon"); @@ -1264,7 +1264,7 @@ static int libxl__build_device_model_args_new(libxl__gc *gc, case LIBXL_CHANNEL_CONNECTION_SOCKET: path = guest_config->channels[i].u.socket.path; chardev = GCSPRINTF("socket,id=libxl-channel%d,path=%s," - "server,nowait", devid, path); + "server=on,wait=off", devid, path); break; default: /* We've forgotten to add the clause */ @@ -1577,7 +1577,7 @@ static int libxl__build_device_model_args_new(libxl__gc *gc, nics[i].colo_##sock_port) { \ flexarray_append(dm_args, "-chardev"); \ flexarray_append(dm_args, \ - GCSPRINTF("socket,id=%s,host=%s,port=%s,server,nowait", \ + GCSPRINTF("socket,id=%s,host=%s,port=%s,server=on,wait=off", \ nics[i].colo_##sock_id, \ nics[i].colo_##sock_ip, \ nics[i].colo_##sock_port)); \ -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Tue Jun 15 13:11:27 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 15 Jun 2021 13:11:27 +0000 Received: from list by lists.xenproject.org with outflank-mailman.142010.262145 (Exim 4.92) (envelope-from ) id 1lt8ql-0005O4-0P; Tue, 15 Jun 2021 13:11:27 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 142010.262145; Tue, 15 Jun 2021 13:11:26 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lt8qk-0005Nv-Ta; Tue, 15 Jun 2021 13:11:26 +0000 Received: by outflank-mailman (input) for mailman id 142010; Tue, 15 Jun 2021 13:11:25 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lt8qj-0005Ne-Ji for xen-changelog@lists.xenproject.org; Tue, 15 Jun 2021 13:11:25 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lt8qj-0003Ex-It for xen-changelog@lists.xenproject.org; Tue, 15 Jun 2021 13:11:25 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lt8qj-0004ij-GH for xen-changelog@lists.xenproject.org; Tue, 15 Jun 2021 13:11:25 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=606UofqnzrUkagnqHPQ6V5o5JhsgwRT9tEj+ENFL0G8=; b=lXzLK4yYkyVdlnwbr5GHktSxR7 6Q0vbRMeeeqgEs1esairNkuEetHZ+oB3uX8r/8ViY81gXIbQlP1entUpcweQMsb1dzS6MahB71PwU mmny6yf+d+fDoJ9vPBfRrp3eHX2pe5i+pJ5ztjeHoP1+qe8iMguPUeZmN9sfKyrpu+kw=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] libxl: Replace deprecated "cpu-add" QMP command by "device_add" Message-Id: Date: Tue, 15 Jun 2021 13:11:25 +0000 commit 4e217db45e83fc3173382306c8b03da86099a25d Author: Anthony PERARD AuthorDate: Tue May 11 10:28:05 2021 +0100 Commit: Julien Grall CommitDate: Tue Jun 15 13:56:26 2021 +0100 libxl: Replace deprecated "cpu-add" QMP command by "device_add" The command "cpu-add" for CPU hotplug is deprecated and has been removed from QEMU 6.0 (April 2021). We need to add cpus with the command "device_add" now. In order to find out which parameters to pass to "device_add" we first make a call to "query-hotpluggable-cpus" which list the cpus drivers and properties. The algorithm to figure out which CPU to add, and by extension if any CPU needs to be hotplugged, is in the function that adds the cpus. Because of that, the command "query-hotpluggable-cpus" is always called, even when not needed. In case we are using a version of QEMU older than 2.7 (Sept 2016) which don't have "query-hotpluggable-cpus", we fallback to using "cpu-add". Signed-off-by: Anthony PERARD Reviewed-by: Jason Andryuk --- tools/libs/light/libxl_domain.c | 89 ++++++++++++++++++++++++++++++++++++++++- 1 file changed, 87 insertions(+), 2 deletions(-) diff --git a/tools/libs/light/libxl_domain.c b/tools/libs/light/libxl_domain.c index 8c003aa7cb..c00c36c928 100644 --- a/tools/libs/light/libxl_domain.c +++ b/tools/libs/light/libxl_domain.c @@ -1805,6 +1805,7 @@ typedef struct set_vcpuonline_state { libxl_dominfo info; libxl_bitmap final_map; int index; /* for loop on final_map */ + const char *cpu_driver; } set_vcpuonline_state; static void set_vcpuonline_qmp_cpus_fast_queried(libxl__egc *, @@ -1814,6 +1815,10 @@ static void set_vcpuonline_qmp_cpus_queried(libxl__egc *, static void set_vcpuonline_qmp_query_cpus_parse(libxl__egc *, libxl__ev_qmp *qmp, const libxl__json_object *, bool query_cpus_fast, int rc); +static void set_vcpuonline_qmp_query_hotpluggable_cpus(libxl__egc *egc, + libxl__ev_qmp *qmp, const libxl__json_object *response, int rc); +static void set_vcpuonline_qmp_device_add_cpu(libxl__egc *, + libxl__ev_qmp *, const libxl__json_object *response, int rc); static void set_vcpuonline_qmp_add_cpu(libxl__egc *, libxl__ev_qmp *, const libxl__json_object *response, int rc); static void set_vcpuonline_timeout(libxl__egc *egc, @@ -1951,13 +1956,54 @@ static void set_vcpuonline_qmp_query_cpus_parse(libxl__egc *egc, libxl_bitmap_reset(final_map, i); } + qmp->callback = set_vcpuonline_qmp_query_hotpluggable_cpus; + rc = libxl__ev_qmp_send(egc, qmp, "query-hotpluggable-cpus", NULL); + out: libxl_bitmap_dispose(¤t_map); + if (rc) + set_vcpuonline_done(egc, svos, rc); /* must be last */ +} + +static void set_vcpuonline_qmp_query_hotpluggable_cpus(libxl__egc *egc, + libxl__ev_qmp *qmp, const libxl__json_object *response, int rc) +{ + set_vcpuonline_state *svos = CONTAINER_OF(qmp, *svos, qmp); + const libxl__json_object *cpu; + const libxl__json_object *cpu_driver; + + if (rc == ERROR_QMP_COMMAND_NOT_FOUND) { + /* We are probably connected to a version of QEMU older than 2.7, + * let's fallback to using "cpu-add" command. */ + svos->index = -1; + set_vcpuonline_qmp_add_cpu(egc, qmp, NULL, 0); /* must be last */ + return; + } + + if (rc) goto out; + + /* Parse response to QMP command "query-hotpluggable-cpus" + * [ { 'type': 'str', ... ] + * + * We are looking for the driver name for CPU to be hotplug. We'll + * assume that cpus property are core-id=0, thread-id=0 and + * socket-id=$cpu_index, as we start qemu with "-smp %d,maxcpus=%d", so + * we don't parse the properties listed for each hotpluggable cpus. + */ + + cpu = libxl__json_array_get(response, 0); + cpu_driver = libxl__json_map_get("type", cpu, JSON_STRING); + svos->cpu_driver = libxl__json_object_get_string(cpu_driver); + + if (!svos->cpu_driver) + rc = ERROR_QEMU_API; + +out: svos->index = -1; - set_vcpuonline_qmp_add_cpu(egc, qmp, NULL, rc); /* must be last */ + set_vcpuonline_qmp_device_add_cpu(egc, qmp, NULL, rc); /* must be last */ } -static void set_vcpuonline_qmp_add_cpu(libxl__egc *egc, +static void set_vcpuonline_qmp_device_add_cpu(libxl__egc *egc, libxl__ev_qmp *qmp, const libxl__json_object *response, int rc) { STATE_AO_GC(qmp->ao); @@ -1969,6 +2015,45 @@ static void set_vcpuonline_qmp_add_cpu(libxl__egc *egc, if (rc) goto out; + while (libxl_bitmap_cpu_valid(map, ++svos->index)) { + if (libxl_bitmap_test(map, svos->index)) { + qmp->callback = set_vcpuonline_qmp_device_add_cpu; + libxl__qmp_param_add_string(gc, &args, "id", GCSPRINTF("cpu-%d", svos->index)); + libxl__qmp_param_add_string(gc, &args, "driver", svos->cpu_driver); + /* We'll assume that we start QEMU with -smp %d,maxcpus=%d, so + * that "core-id" and "thread-id" are always 0 so that + * "socket-id" correspond the cpu index. + * Those properties are otherwise listed by + * "query-hotpluggable-cpus". */ + libxl__qmp_param_add_integer(gc, &args, "socket-id", svos->index); + libxl__qmp_param_add_integer(gc, &args, "core-id", 0); + libxl__qmp_param_add_integer(gc, &args, "thread-id", 0); + rc = libxl__ev_qmp_send(egc, qmp, "device_add", args); + if (rc) goto out; + return; + } + } + +out: + set_vcpuonline_done(egc, svos, rc); +} + +/* Fallback function for QEMU older than 2.7, when + * 'query-hotpluggable-cpus' wasn't available and vcpu object couldn't be + * added with 'device_add'. */ +static void set_vcpuonline_qmp_add_cpu(libxl__egc *egc, libxl__ev_qmp *qmp, + const libxl__json_object *response, + int rc) +{ + STATE_AO_GC(qmp->ao); + set_vcpuonline_state *svos = CONTAINER_OF(qmp, *svos, qmp); + libxl__json_object *args = NULL; + + /* Convenience aliases */ + libxl_bitmap *map = &svos->final_map; + + if (rc) goto out; + while (libxl_bitmap_cpu_valid(map, ++svos->index)) { if (libxl_bitmap_test(map, svos->index)) { qmp->callback = set_vcpuonline_qmp_add_cpu; -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Tue Jun 15 13:11:37 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 15 Jun 2021 13:11:37 +0000 Received: from list by lists.xenproject.org with outflank-mailman.142011.262149 (Exim 4.92) (envelope-from ) id 1lt8qv-0005SM-1s; Tue, 15 Jun 2021 13:11:37 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 142011.262149; Tue, 15 Jun 2021 13:11:37 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lt8qu-0005SE-VD; Tue, 15 Jun 2021 13:11:36 +0000 Received: by outflank-mailman (input) for mailman id 142011; Tue, 15 Jun 2021 13:11:35 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lt8qt-0005Rz-Mw for xen-changelog@lists.xenproject.org; Tue, 15 Jun 2021 13:11:35 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lt8qt-0003FC-M5 for xen-changelog@lists.xenproject.org; Tue, 15 Jun 2021 13:11:35 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lt8qt-0004jV-LD for xen-changelog@lists.xenproject.org; Tue, 15 Jun 2021 13:11:35 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=t9oZwua62FeJqTnmXfojxNSEj29vnSGBb85iRucVua0=; b=pdsU9muh59NRN5qZc6zSgq+cix fAdDkEPu2cgXdUcUJ1PwhAs61C4GiLMKTxSjEm8a10nrc4PqGgE4bQA9VBtUYgI7cmDrKEkUAgaQg GejtKoBuzcxYWKJbia1Uazkww/twK3gyadGR9mx88mFhk3Uam6yQR42TjoYH+sNjysEs=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] libxl: Use -device for cd-rom drives Message-Id: Date: Tue, 15 Jun 2021 13:11:35 +0000 commit 85760c03d664400368a3f76ae0225307c25049a7 Author: Anthony PERARD AuthorDate: Tue May 11 10:28:06 2021 +0100 Commit: Julien Grall CommitDate: Tue Jun 15 13:56:26 2021 +0100 libxl: Use -device for cd-rom drives This allows to set an `id` on the device instead of only the drive. We are going to need the `id` with the "eject" and "blockdev-change-media" QMP command as using `device` parameter on those is deprecated. (`device` is the `id` of the `-drive` on the command line). We set the same `id` on both -device and -drive as QEMU doesn't complain and we can then either do "eject id=$id" or "eject device=$id". Using "-drive + -device" instead of only "-drive" has been available since at least QEMU 0.15, and seems to be the preferred way as it separates the host part (-drive which describe the disk image location and format) from the guest part (-device which describe the emulated device). More information in qemu.git/docs/qdev-device-use.txt . Changing the command line during migration for the cdrom seems fine. Also the documentation about migration in QEMU explains that the device state ID is "been formed from a bus name and device address", so second IDE bus and first device address on bus is still thus and doesn't matter if written "-drive if=ide,index=2" or "-drive ide-cd,bus=ide.1,unit=0". See qemu.git/docs/devel/migration.rst . Signed-off-by: Anthony PERARD Reviewed-by: Jason Andryuk --- tools/libs/light/libxl_dm.c | 16 +++++++++++++--- 1 file changed, 13 insertions(+), 3 deletions(-) diff --git a/tools/libs/light/libxl_dm.c b/tools/libs/light/libxl_dm.c index 0a0c1ef7c6..5b01cf2841 100644 --- a/tools/libs/light/libxl_dm.c +++ b/tools/libs/light/libxl_dm.c @@ -1913,6 +1913,7 @@ static int libxl__build_device_model_args_new(libxl__gc *gc, } if (disks[i].is_cdrom) { + const char *drive_id; if (disk > 4) { LOGD(WARN, guest_domid, "Emulated CDROM can be only one of the first 4 disks.\n" "Disk %s will be available via PV drivers but not as an " @@ -1920,13 +1921,22 @@ static int libxl__build_device_model_args_new(libxl__gc *gc, disks[i].vdev); continue; } - drive = libxl__sprintf(gc, - "if=ide,index=%d,readonly=on,media=cdrom,id=ide-%i", - disk, dev_number); + + drive_id = GCSPRINTF("ide-%i", dev_number); + drive = GCSPRINTF("if=none,readonly=on,id=%s", drive_id); if (target_path) drive = libxl__sprintf(gc, "%s,file=%s,format=%s", drive, target_path, format); + + flexarray_vappend(dm_args, + "-drive", drive, + "-device", + GCSPRINTF("ide-cd,id=%s,drive=%s,bus=ide.%u,unit=%u", + drive_id, drive_id, + disk / 2, disk % 2), + NULL); + continue; } else { /* * Explicit sd disks are passed through as is. -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Tue Jun 15 13:11:47 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 15 Jun 2021 13:11:47 +0000 Received: from list by lists.xenproject.org with outflank-mailman.142012.262154 (Exim 4.92) (envelope-from ) id 1lt8r5-0005VL-40; Tue, 15 Jun 2021 13:11:47 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 142012.262154; Tue, 15 Jun 2021 13:11:47 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lt8r5-0005VD-0S; Tue, 15 Jun 2021 13:11:47 +0000 Received: by outflank-mailman (input) for mailman id 142012; Tue, 15 Jun 2021 13:11:45 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lt8r3-0005V1-Q9 for xen-changelog@lists.xenproject.org; Tue, 15 Jun 2021 13:11:45 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lt8r3-0003FK-PK for xen-changelog@lists.xenproject.org; Tue, 15 Jun 2021 13:11:45 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lt8r3-0004kA-OH for xen-changelog@lists.xenproject.org; Tue, 15 Jun 2021 13:11:45 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=mqWII0UeM9p0mUBnEuO7QvyiBGrIU6O8icZ9drmN3sU=; b=GgR5leRKnS198cjnSbDc1YGjpl ncbWCecI1AlUnVpmSIIsjS+FMzDVuoa3sbBiSP+hFJC0oGACte1iKqGuHbRh/6AxJau5BtCYlDtmG TM+41u9q4K1NpyXy21biAB/I6kUgFaBuF7yDhbOOADL2bHwx1Mhdt1KPGIv0PCZcOrKw=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] libxl: Assert qmp_ev's state in qmp_ev_qemu_compare_version Message-Id: Date: Tue, 15 Jun 2021 13:11:45 +0000 commit 0ff26a3225d69ffec76fe5aca8296852fa951204 Author: Anthony PERARD AuthorDate: Tue May 11 10:28:07 2021 +0100 Commit: Julien Grall CommitDate: Tue Jun 15 13:56:26 2021 +0100 libxl: Assert qmp_ev's state in qmp_ev_qemu_compare_version We are supposed to read the version information only when qmp_ev is in state "Connected" (that correspond to state==qmp_state_connected), assert it so that the function isn't used too early. Signed-off-by: Anthony PERARD Reviewed-by: Jason Andryuk --- tools/libs/light/libxl_qmp.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/tools/libs/light/libxl_qmp.c b/tools/libs/light/libxl_qmp.c index 9b638e6f54..d0967c9f02 100644 --- a/tools/libs/light/libxl_qmp.c +++ b/tools/libs/light/libxl_qmp.c @@ -292,6 +292,8 @@ static int qmp_handle_response(libxl__gc *gc, libxl__qmp_handler *qmp, static int qmp_ev_qemu_compare_version(libxl__ev_qmp *ev, int major, int minor, int micro) { + assert(ev->state == qmp_state_connected); + #define CHECK_VERSION(level) do { \ if (ev->qemu_version.level > (level)) return +1; \ if (ev->qemu_version.level < (level)) return -1; \ -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Tue Jun 15 13:11:57 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 15 Jun 2021 13:11:57 +0000 Received: from list by lists.xenproject.org with outflank-mailman.142013.262156 (Exim 4.92) (envelope-from ) id 1lt8rF-0005ZB-61; Tue, 15 Jun 2021 13:11:57 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 142013.262156; Tue, 15 Jun 2021 13:11:57 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lt8rF-0005Z3-38; Tue, 15 Jun 2021 13:11:57 +0000 Received: by outflank-mailman (input) for mailman id 142013; Tue, 15 Jun 2021 13:11:55 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lt8rD-0005Yi-UH for xen-changelog@lists.xenproject.org; Tue, 15 Jun 2021 13:11:55 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lt8rD-0003Fp-TW for xen-changelog@lists.xenproject.org; Tue, 15 Jun 2021 13:11:55 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lt8rD-0004l3-SN for xen-changelog@lists.xenproject.org; Tue, 15 Jun 2021 13:11:55 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=79mSEKtIh/lpJC+kT/WfTUzYus6L2rrj8Rv2LSy5ZMA=; b=KE55yiRhFR99E/GobX+1qeSc5z qVqXSwwqXaszntYIfBE26YlJKi1EYMhXXtS2lVuV4YG/0MT6iLINLmaTAQDOnKEU8TLkDZHjZPnTT tv7jPocpkP6xt6XzWnwTh77cNVgGslRZB4jf9guP2cEdCnW2Mw7oF+Q83+GojwiujlhQ=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] libxl: Export libxl__qmp_ev_qemu_compare_version Message-Id: Date: Tue, 15 Jun 2021 13:11:55 +0000 commit 7c313e8365eb663311a0cf39f77b4f5880244765 Author: Anthony PERARD AuthorDate: Tue May 11 10:28:08 2021 +0100 Commit: Julien Grall CommitDate: Tue Jun 15 13:56:26 2021 +0100 libxl: Export libxl__qmp_ev_qemu_compare_version We are going to want to check QEMU's version in other places where we can use libxl__ev_qmp_send. Signed-off-by: Anthony PERARD Reviewed-by: Jason Andryuk --- tools/libs/light/libxl_internal.h | 8 ++++++++ tools/libs/light/libxl_qmp.c | 4 ++-- 2 files changed, 10 insertions(+), 2 deletions(-) diff --git a/tools/libs/light/libxl_internal.h b/tools/libs/light/libxl_internal.h index 44a2f3c8fe..0b4671318c 100644 --- a/tools/libs/light/libxl_internal.h +++ b/tools/libs/light/libxl_internal.h @@ -492,6 +492,14 @@ _hidden int libxl__ev_qmp_send(libxl__egc *egc, libxl__ev_qmp *ev, const char *cmd, libxl__json_object *args); _hidden void libxl__ev_qmp_dispose(libxl__gc *gc, libxl__ev_qmp *ev); +/* return values: + * < 0 if qemu's version < asked version + * = 0 if qemu's version == asked version + * > 0 if qemu's version > asked version + */ +_hidden int libxl__qmp_ev_qemu_compare_version(libxl__ev_qmp *ev, int major, + int minor, int micro); + typedef enum { /* initial state */ qmp_state_disconnected = 1, diff --git a/tools/libs/light/libxl_qmp.c b/tools/libs/light/libxl_qmp.c index d0967c9f02..fb146a54cb 100644 --- a/tools/libs/light/libxl_qmp.c +++ b/tools/libs/light/libxl_qmp.c @@ -289,7 +289,7 @@ static int qmp_handle_response(libxl__gc *gc, libxl__qmp_handler *qmp, * = 0 if qemu's version == asked version * > 0 if qemu's version > asked version */ -static int qmp_ev_qemu_compare_version(libxl__ev_qmp *ev, int major, +int libxl__qmp_ev_qemu_compare_version(libxl__ev_qmp *ev, int major, int minor, int micro) { assert(ev->state == qmp_state_connected); @@ -1073,7 +1073,7 @@ static void dm_state_save_to_fdset(libxl__egc *egc, libxl__ev_qmp *ev, int fdset /* The `live` parameter was added to QEMU 2.11. It signals QEMU that * the save operation is for a live migration rather than for taking a * snapshot. */ - if (qmp_ev_qemu_compare_version(ev, 2, 11, 0) >= 0) + if (libxl__qmp_ev_qemu_compare_version(ev, 2, 11, 0) >= 0) libxl__qmp_param_add_bool(gc, &args, "live", dsps->live); QMP_PARAMETERS_SPRINTF(&args, "filename", "/dev/fdset/%d", fdset); rc = libxl__ev_qmp_send(egc, ev, "xen-save-devices-state", args); -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Tue Jun 15 13:12:07 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 15 Jun 2021 13:12:07 +0000 Received: from list by lists.xenproject.org with outflank-mailman.142014.262162 (Exim 4.92) (envelope-from ) id 1lt8rP-0005bm-88; Tue, 15 Jun 2021 13:12:07 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 142014.262162; Tue, 15 Jun 2021 13:12:07 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lt8rP-0005be-4d; Tue, 15 Jun 2021 13:12:07 +0000 Received: by outflank-mailman (input) for mailman id 142014; Tue, 15 Jun 2021 13:12:06 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lt8rO-0005bW-1n for xen-changelog@lists.xenproject.org; Tue, 15 Jun 2021 13:12:06 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lt8rO-0003GJ-12 for xen-changelog@lists.xenproject.org; Tue, 15 Jun 2021 13:12:06 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lt8rN-0004me-WF for xen-changelog@lists.xenproject.org; Tue, 15 Jun 2021 13:12:06 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=gSZyRNZglAy+65+1AC07ulqRXK4uoT1kaoSOJYY9kmM=; b=YPgELIxSlyBwfT0/YJk9/RHGLK 4EOMBRNStjUysc+JLzj5TWWGxaIaEiqflfZwANEv+PnjKO5p+P4Hw18MXO9KSO0LWI6Zy9U2YjVUd 2Z42lkrr1JA6g+rRPqF1STai0DTZxzYKID3jsCipnlfDyh+L44PJXbhWIZ527H/8NgOM=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] libxl: Use `id` with the "eject" QMP command Message-Id: Date: Tue, 15 Jun 2021 13:12:05 +0000 commit 0c0b3a7e4a2d65fd252b89b46bdcdb048bb24b6c Author: Anthony PERARD AuthorDate: Tue May 11 10:28:09 2021 +0100 Commit: Julien Grall CommitDate: Tue Jun 15 13:56:26 2021 +0100 libxl: Use `id` with the "eject" QMP command `device` parameter is deprecated since QEMU 2.8. This requires changes to the command line introduced by: "libxl: Use -device for cd-rom drives" Signed-off-by: Anthony PERARD Reviewed-by: Jason Andryuk --- tools/libs/light/libxl_disk.c | 43 ++++++++++++++++++++++++++++++++++++------- 1 file changed, 36 insertions(+), 7 deletions(-) diff --git a/tools/libs/light/libxl_disk.c b/tools/libs/light/libxl_disk.c index 411ffeaca6..faabdea7a4 100644 --- a/tools/libs/light/libxl_disk.c +++ b/tools/libs/light/libxl_disk.c @@ -656,6 +656,8 @@ typedef struct { static void cdrom_insert_lock_acquired(libxl__egc *, libxl__ev_slowlock *, int rc); +static void cdrom_insert_qmp_connected(libxl__egc *, libxl__ev_qmp *, + const libxl__json_object *, int rc); static void cdrom_insert_ejected(libxl__egc *egc, libxl__ev_qmp *, const libxl__json_object *, int rc); static void cdrom_insert_addfd_cb(libxl__egc *egc, libxl__ev_qmp *, @@ -770,13 +772,12 @@ static void cdrom_insert_lock_acquired(libxl__egc *egc, */ if (cis->dm_ver == LIBXL_DEVICE_MODEL_VERSION_QEMU_XEN) { - libxl__json_object *args = NULL; - int devid = libxl__device_disk_dev_number(cis->disk->vdev, - NULL, NULL); - - QMP_PARAMETERS_SPRINTF(&args, "device", "ide-%i", devid); - cis->qmp.callback = cdrom_insert_ejected; - rc = libxl__ev_qmp_send(egc, &cis->qmp, "eject", args); + /* Before running the "eject" command, we need to know QEMU's + * version to find out which command to issue. + * cis->qmp isn't in Connected state yet, so run a dummy command + * to have QEMU's version available. */ + cis->qmp.callback = cdrom_insert_qmp_connected; + rc = libxl__ev_qmp_send(egc, &cis->qmp, "query-version", NULL); if (rc) goto out; } else { cdrom_insert_ejected(egc, &cis->qmp, NULL, 0); /* must be last */ @@ -787,6 +788,34 @@ out: cdrom_insert_done(egc, cis, rc); /* must be last */ } +static void cdrom_insert_qmp_connected(libxl__egc *egc, libxl__ev_qmp *qmp, + const libxl__json_object *response, + int rc) +{ + libxl__cdrom_insert_state *cis = CONTAINER_OF(qmp, *cis, qmp); + STATE_AO_GC(cis->ao); + libxl__json_object *args = NULL; + int devid = libxl__device_disk_dev_number(cis->disk->vdev, + NULL, NULL); + + if (rc) goto out; + + /* Using `device` parameter is deprecated since QEMU 2.8, we should + * use `id` now. They both have different meaning but we set the + * same `id` on -drive and -device on the command line. + */ + if (libxl__qmp_ev_qemu_compare_version(qmp, 2, 8, 0) >= 0) + QMP_PARAMETERS_SPRINTF(&args, "id", "ide-%i", devid); + else + QMP_PARAMETERS_SPRINTF(&args, "device", "ide-%i", devid); + qmp->callback = cdrom_insert_ejected; + rc = libxl__ev_qmp_send(egc, qmp, "eject", args); + if (rc) goto out; + return; +out: + cdrom_insert_done(egc, cis, rc); /* must be last */ +} + static void cdrom_insert_ejected(libxl__egc *egc, libxl__ev_qmp *qmp, const libxl__json_object *response, -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Tue Jun 15 13:12:17 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 15 Jun 2021 13:12:17 +0000 Received: from list by lists.xenproject.org with outflank-mailman.142015.262165 (Exim 4.92) (envelope-from ) id 1lt8rZ-0005f1-9J; Tue, 15 Jun 2021 13:12:17 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 142015.262165; Tue, 15 Jun 2021 13:12:17 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lt8rZ-0005et-6A; Tue, 15 Jun 2021 13:12:17 +0000 Received: by outflank-mailman (input) for mailman id 142015; Tue, 15 Jun 2021 13:12:16 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lt8rY-0005eh-4s for xen-changelog@lists.xenproject.org; Tue, 15 Jun 2021 13:12:16 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lt8rY-0003GU-47 for xen-changelog@lists.xenproject.org; Tue, 15 Jun 2021 13:12:16 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lt8rY-0004oD-3M for xen-changelog@lists.xenproject.org; Tue, 15 Jun 2021 13:12:16 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=pkUjWYkhsRqFEiAxsMrExe316vpZ7hUrzA2c2inn7zo=; b=5+lqZx2jiB3EOXHnE1MfrAcd+P 7HwikKFazeyAb0sOImeTzWD1vqHMWBOgseNfhs6FpvNTdvanUw0BHfrswmzo8p9ZI4/oAleMC2/rn gyfDwXH/Ce0pPB/vFJA+gGRwML1ytLqwlMApjDh+OSvp4usBmBKxuvfl9lwb/Iwc0+iI=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] libxl: Replace QMP command "change" by "blockdev-change-media" Message-Id: Date: Tue, 15 Jun 2021 13:12:16 +0000 commit fe6630ddc4e8a8fbf8dd28a1bc58e3881393f9c1 Author: Anthony PERARD AuthorDate: Tue May 11 10:28:10 2021 +0100 Commit: Julien Grall CommitDate: Tue Jun 15 13:56:26 2021 +0100 libxl: Replace QMP command "change" by "blockdev-change-media" "change" command as been removed in QEMU 6.0. We can use "blockdev-change-medium" instead. Using `id` with "blockdev-change-medium" requires a change to the QEMU command line, introduced by: "libxl: Use -device for cd-rom drives" Signed-off-by: Anthony PERARD Reviewed-by: Jason Andryuk --- tools/libs/light/libxl_disk.c | 24 +++++++++++++++++++----- 1 file changed, 19 insertions(+), 5 deletions(-) diff --git a/tools/libs/light/libxl_disk.c b/tools/libs/light/libxl_disk.c index faabdea7a4..93936d0dd0 100644 --- a/tools/libs/light/libxl_disk.c +++ b/tools/libs/light/libxl_disk.c @@ -962,12 +962,26 @@ static void cdrom_insert_addfd_cb(libxl__egc *egc, fdset = libxl__json_object_get_integer(o); devid = libxl__device_disk_dev_number(disk->vdev, NULL, NULL); - QMP_PARAMETERS_SPRINTF(&args, "device", "ide-%i", devid); - QMP_PARAMETERS_SPRINTF(&args, "target", "/dev/fdset/%d", fdset); - libxl__qmp_param_add_string(gc, &args, "arg", - libxl__qemu_disk_format_string(disk->format)); qmp->callback = cdrom_insert_inserted; - rc = libxl__ev_qmp_send(egc, qmp, "change", args); + + /* "change" is deprecated since QEMU 2.5 and the `device` parameter for + * for "blockdev-change-medium" is deprecated in QEMU 2.8. + * But `id` is only available in 2.8 we'll start using the new command + * with `id` with QEMU 2.8. + */ + if (libxl__qmp_ev_qemu_compare_version(qmp, 2, 8, 0) >= 0) { + QMP_PARAMETERS_SPRINTF(&args, "id", "ide-%i", devid); + QMP_PARAMETERS_SPRINTF(&args, "filename", "/dev/fdset/%d", fdset); + libxl__qmp_param_add_string(gc, &args, "format", + libxl__qemu_disk_format_string(disk->format)); + rc = libxl__ev_qmp_send(egc, qmp, "blockdev-change-medium", args); + } else { + QMP_PARAMETERS_SPRINTF(&args, "device", "ide-%i", devid); + QMP_PARAMETERS_SPRINTF(&args, "target", "/dev/fdset/%d", fdset); + libxl__qmp_param_add_string(gc, &args, "arg", + libxl__qemu_disk_format_string(disk->format)); + rc = libxl__ev_qmp_send(egc, qmp, "change", args); + } out: if (rc) cdrom_insert_done(egc, cis, rc); /* must be last */ -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Tue Jun 15 13:22:08 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 15 Jun 2021 13:22:08 +0000 Received: from list by lists.xenproject.org with outflank-mailman.142023.262180 (Exim 4.92) (envelope-from ) id 1lt914-00079M-8Z; Tue, 15 Jun 2021 13:22:06 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 142023.262180; Tue, 15 Jun 2021 13:22:06 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lt914-00079E-5Q; Tue, 15 Jun 2021 13:22:06 +0000 Received: by outflank-mailman (input) for mailman id 142023; Tue, 15 Jun 2021 13:22:04 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lt912-000795-Ls for xen-changelog@lists.xenproject.org; Tue, 15 Jun 2021 13:22:04 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lt912-0003R5-Kl for xen-changelog@lists.xenproject.org; Tue, 15 Jun 2021 13:22:04 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lt912-0005UD-Jh for xen-changelog@lists.xenproject.org; Tue, 15 Jun 2021 13:22:04 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=dGKk7LzzlCSKVydO7tCj3XdTbWYS4v6UHzE4W6CMgew=; b=wKNx0d/LPRFynZzM+pdiBniqLt 5F8LbnbaDMdVRUZLDib+//ql/ztxqOm0GDEhwn9mSjktJvujGx9WFXQ4RaaLw+uviQgwwzJE0EKIz CMPETZIgF817PUt7YGOB6tyiDguF1OJ8hdpWnyMTImZ0ev2TBbX0MCx4oRwdIW5+njqE=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] x86/vpt: fully init timers before putting onto list Message-Id: Date: Tue, 15 Jun 2021 13:22:04 +0000 commit 6d622f3a96bbd76ce8422c6e3805e6609417ec76 Author: Jan Beulich AuthorDate: Tue Jun 15 15:14:20 2021 +0200 Commit: Jan Beulich CommitDate: Tue Jun 15 15:14:20 2021 +0200 x86/vpt: fully init timers before putting onto list With pt_vcpu_lock() no longer acquiring the pt_migrate lock, parties iterating the list and acting on the timers of the list entries will no longer be kept from entering their loops by create_periodic_time()'s holding of that lock. Therefore at least init_timer() needs calling ahead of list insertion, but keep this and set_timer() together. Fixes: 8113b02f0bf8 ("x86/vpt: do not take pt_migrate rwlock in some cases") Reported-by: Igor Druzhinin Signed-off-by: Jan Beulich Reviewed-by: Roger Pau Monné Reviewed-by: Boris Ostrovsky --- xen/arch/x86/hvm/vpt.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/xen/arch/x86/hvm/vpt.c b/xen/arch/x86/hvm/vpt.c index 4cc0a0848b..6fdc3e19fe 100644 --- a/xen/arch/x86/hvm/vpt.c +++ b/xen/arch/x86/hvm/vpt.c @@ -554,14 +554,14 @@ void create_periodic_time( pt->cb = cb; pt->priv = data; + init_timer(&pt->timer, pt_timer_fn, pt, v->processor); + set_timer(&pt->timer, pt->scheduled); + pt_vcpu_lock(v); pt->on_list = 1; list_add(&pt->list, &v->arch.hvm.tm_list); pt_vcpu_unlock(v); - init_timer(&pt->timer, pt_timer_fn, pt, v->processor); - set_timer(&pt->timer, pt->scheduled); - write_unlock(&v->domain->arch.hvm.pl_time->pt_migrate); } -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Tue Jun 15 13:22:16 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 15 Jun 2021 13:22:16 +0000 Received: from list by lists.xenproject.org with outflank-mailman.142024.262184 (Exim 4.92) (envelope-from ) id 1lt91E-0007Ba-9y; Tue, 15 Jun 2021 13:22:16 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 142024.262184; Tue, 15 Jun 2021 13:22:16 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lt91E-0007BS-6v; Tue, 15 Jun 2021 13:22:16 +0000 Received: by outflank-mailman (input) for mailman id 142024; Tue, 15 Jun 2021 13:22:14 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lt91C-0007BG-Oa for xen-changelog@lists.xenproject.org; Tue, 15 Jun 2021 13:22:14 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lt91C-0003RD-Ns for xen-changelog@lists.xenproject.org; Tue, 15 Jun 2021 13:22:14 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lt91C-0005Wg-Mv for xen-changelog@lists.xenproject.org; Tue, 15 Jun 2021 13:22:14 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=2MXTOxHcztOMQzSke0yrGitcrZfHI7oZkH/9Ed1VhTs=; b=AAoWCM734wQO6swexa7LKwwEif e0wGRB5oPm5NFKCwdOef8Tr4Vx0OM1bx501vOwcYkiEu0a66RoemfhcC2lW+IskjptIh4Y21hSJOs pfduD1u5jhgtWjidfZ7p4xaWvrkt+wXoL2/tNMcCpn3j+nMxO4I3Y2pPdaHc2QEkx46o=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] x86: move .altinstr_replacement past _einittext Message-Id: Date: Tue, 15 Jun 2021 13:22:14 +0000 commit 93c5f98296fc78de79d621418a1e62fd413e73d1 Author: Jan Beulich AuthorDate: Tue Jun 15 15:15:26 2021 +0200 Commit: Jan Beulich CommitDate: Tue Jun 15 15:15:26 2021 +0200 x86: move .altinstr_replacement past _einittext This section's contents do not represent part of actual hypervisor text, so shouldn't be included in what is_kernel_inittext() or (while still booting) is_active_kernel_text() report "true" for. Keep them in .init.text though, as there's no real reason to have a separate section for this in the final binary. Signed-off-by: Jan Beulich Acked-by: Andrew Cooper --- xen/arch/x86/xen.lds.S | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/xen/arch/x86/xen.lds.S b/xen/arch/x86/xen.lds.S index 34f647a44e..9c6c1c8005 100644 --- a/xen/arch/x86/xen.lds.S +++ b/xen/arch/x86/xen.lds.S @@ -185,13 +185,13 @@ SECTIONS #endif _sinittext = .; *(.init.text) + _einittext = .; /* * Here are the replacement instructions. The linker sticks them * as binary blobs. The .altinstructions has enough data to get * the address and the length of them to patch the kernel safely. */ *(.altinstr_replacement) - _einittext = .; #ifdef EFI /* EFI wants to merge all of .init.* ELF doesn't. */ . = ALIGN(SMP_CACHE_BYTES); -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Tue Jun 15 17:11:10 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 15 Jun 2021 17:11:10 +0000 Received: from list by lists.xenproject.org with outflank-mailman.142351.262677 (Exim 4.92) (envelope-from ) id 1ltCag-0003pP-3l; Tue, 15 Jun 2021 17:11:06 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 142351.262677; Tue, 15 Jun 2021 17:11:06 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1ltCag-0003pH-0i; Tue, 15 Jun 2021 17:11:06 +0000 Received: by outflank-mailman (input) for mailman id 142351; Tue, 15 Jun 2021 17:11:04 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1ltCae-0003pB-Nd for xen-changelog@lists.xenproject.org; Tue, 15 Jun 2021 17:11:04 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1ltCae-00087t-LT for xen-changelog@lists.xenproject.org; Tue, 15 Jun 2021 17:11:04 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1ltCae-0002ec-JS for xen-changelog@lists.xenproject.org; Tue, 15 Jun 2021 17:11:04 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=ZB/Zikr3sOnDq7Ji2n/dWS+OQjMUbuIfKK44Hus+j28=; b=OQfJa6+xtOMJ5F+YqVAglCh32v kTjXLFttLMLgM/qjCC8zS2L25KKQaSEYJGcBurJyO1BBFfS1qd192i26m5QZHkhjEI1jaa37Zmnd7 i8cd0Wh2VxbKDKcUkmrOdoEJj+T2oRk9o/4hEL4hA32mqo0VOc1m4s4zaoOdEiHyC2tg=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] libs/foreignmemory: Fix osdep_xenforeignmemory_map prototype Message-Id: Date: Tue, 15 Jun 2021 17:11:04 +0000 commit 5d3e4ebb5c71477d74a0c503438545a0126d3863 Author: Anthony PERARD AuthorDate: Tue Jun 1 16:41:47 2021 +0100 Commit: Julien Grall CommitDate: Tue Jun 15 18:07:58 2021 +0100 libs/foreignmemory: Fix osdep_xenforeignmemory_map prototype Commit cf8c4d3d13b8 made some preparation to have one day variable-length-array argument, but didn't declare the array in the function prototype the same way as in the function definition. And now GCC 11 complains about it. Fixes: cf8c4d3d13b8 ("tools/libs/foreignmemory: pull array length argument to map forward") Signed-off-by: Anthony PERARD Reviewed-by: Jan Beulich --- tools/libs/foreignmemory/private.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/tools/libs/foreignmemory/private.h b/tools/libs/foreignmemory/private.h index 1ee3626dd2..5bb0cefb09 100644 --- a/tools/libs/foreignmemory/private.h +++ b/tools/libs/foreignmemory/private.h @@ -32,7 +32,7 @@ int osdep_xenforeignmemory_close(xenforeignmemory_handle *fmem); void *osdep_xenforeignmemory_map(xenforeignmemory_handle *fmem, uint32_t dom, void *addr, int prot, int flags, size_t num, - const xen_pfn_t arr[num], int err[num]); + const xen_pfn_t arr[/*num*/], int err[/*num*/]); int osdep_xenforeignmemory_unmap(xenforeignmemory_handle *fmem, void *addr, size_t num); -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Tue Jun 15 19:55:11 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 15 Jun 2021 19:55:11 +0000 Received: from list by lists.xenproject.org with outflank-mailman.142385.262731 (Exim 4.92) (envelope-from ) id 1ltF9N-000302-Rw; Tue, 15 Jun 2021 19:55:05 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 142385.262731; Tue, 15 Jun 2021 19:55:05 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1ltF9N-0002zv-OA; Tue, 15 Jun 2021 19:55:05 +0000 Received: by outflank-mailman (input) for mailman id 142385; Tue, 15 Jun 2021 19:55:05 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1ltF9N-0002zp-4D for xen-changelog@lists.xenproject.org; Tue, 15 Jun 2021 19:55:05 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1ltF9N-0002U2-2N for xen-changelog@lists.xenproject.org; Tue, 15 Jun 2021 19:55:05 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1ltF9N-00011I-1Q for xen-changelog@lists.xenproject.org; Tue, 15 Jun 2021 19:55:05 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=HW9LP+QT7aHIgiVQacEG+KyGZsPhwyklUNPj++dSekk=; b=i6NOXYzO8IKNH4hU2YYC8RZc9q gk1htLCj6t9noK+BcM2iEQv9W/d3uYhsuZ3+xyKycAN8j9ue226suujYdkc/JQ4ujqerHW1EdT2Xn PSt8/EUIrQ2oB9psRJpTupVCh3wfc466H+8lwzD+r3CfNl/WmqB7uIi+BR37j7bENQMM=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] x86/platform: Improve MSR permission handling for XENPF_resource_op Message-Id: Date: Tue, 15 Jun 2021 19:55:05 +0000 commit 2cf3b4b92ab1a34e8cb1e859196e1492c89299de Author: Andrew Cooper AuthorDate: Thu Jun 10 11:01:06 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 15 20:50:32 2021 +0100 x86/platform: Improve MSR permission handling for XENPF_resource_op The logic to disallow writes to the TSC is out-of-place, and should be in check_resource_access() rather than in resource_access(). Split the existing allow_access_msr() into two - msr_{read,write}_allowed() - and move all permissions checks here. Furthermore, guard access to MSR_IA32_CMT_{EVTSEL,CTR} to prohibit their use on hardware which is lacking the QoS Monitoring feature. Introduce cpu_has_pqe to help with the logic. Signed-off-by: Andrew Cooper Reviewed-by: Jan Beulich --- xen/arch/x86/platform_hypercall.c | 41 ++++++++++++++++++++++++++++----------- xen/arch/x86/psr.c | 2 +- xen/include/asm-x86/cpufeature.h | 1 + 3 files changed, 32 insertions(+), 12 deletions(-) diff --git a/xen/arch/x86/platform_hypercall.c b/xen/arch/x86/platform_hypercall.c index 23fadbc782..41d8e59563 100644 --- a/xen/arch/x86/platform_hypercall.c +++ b/xen/arch/x86/platform_hypercall.c @@ -64,17 +64,33 @@ long cpu_frequency_change_helper(void *data) return cpu_frequency_change((uint64_t)data); } -static bool allow_access_msr(unsigned int msr) +static bool msr_read_allowed(unsigned int msr) { switch ( msr ) { - /* MSR for CMT, refer to chapter 17.14 of Intel SDM. */ case MSR_IA32_CMT_EVTSEL: case MSR_IA32_CMT_CTR: + return cpu_has_pqe; + case MSR_IA32_TSC: return true; } + if ( ppin_msr && msr == ppin_msr ) + return true; + + return false; +} + +static bool msr_write_allowed(unsigned int msr) +{ + switch ( msr ) + { + case MSR_IA32_CMT_EVTSEL: + case MSR_IA32_CMT_CTR: + return cpu_has_pqe; + } + return false; } @@ -96,15 +112,19 @@ void check_resource_access(struct resource_access *ra) switch ( entry->u.cmd ) { case XEN_RESOURCE_OP_MSR_READ: - if ( ppin_msr && entry->idx == ppin_msr ) - break; - /* fall through */ + if ( entry->idx >> 32 ) + ret = -EINVAL; + else if ( !msr_read_allowed(entry->idx) ) + ret = -EPERM; + break; + case XEN_RESOURCE_OP_MSR_WRITE: if ( entry->idx >> 32 ) ret = -EINVAL; - else if ( !allow_access_msr(entry->idx) ) - ret = -EACCES; + else if ( !msr_write_allowed(entry->idx) ) + ret = -EPERM; break; + default: ret = -EOPNOTSUPP; break; @@ -163,12 +183,11 @@ void resource_access(void *info) } } break; + case XEN_RESOURCE_OP_MSR_WRITE: - if ( unlikely(entry->idx == MSR_IA32_TSC) ) - ret = -EPERM; - else - ret = wrmsr_safe(entry->idx, entry->val); + ret = wrmsr_safe(entry->idx, entry->val); break; + default: BUG(); break; diff --git a/xen/arch/x86/psr.c b/xen/arch/x86/psr.c index d7f8864651..d805b85dc6 100644 --- a/xen/arch/x86/psr.c +++ b/xen/arch/x86/psr.c @@ -1558,7 +1558,7 @@ static void psr_cpu_init(void) struct cpuid_leaf regs; uint32_t feat_mask; - if ( !psr_alloc_feat_enabled() || !boot_cpu_has(X86_FEATURE_PQE) ) + if ( !psr_alloc_feat_enabled() || !cpu_has_pqe ) goto assoc_init; if ( boot_cpu_data.cpuid_level < PSR_CPUID_LEVEL_CAT ) diff --git a/xen/include/asm-x86/cpufeature.h b/xen/include/asm-x86/cpufeature.h index a539a4bacd..5f6b83f71c 100644 --- a/xen/include/asm-x86/cpufeature.h +++ b/xen/include/asm-x86/cpufeature.h @@ -94,6 +94,7 @@ #define cpu_has_bmi2 boot_cpu_has(X86_FEATURE_BMI2) #define cpu_has_invpcid boot_cpu_has(X86_FEATURE_INVPCID) #define cpu_has_rtm boot_cpu_has(X86_FEATURE_RTM) +#define cpu_has_pqe boot_cpu_has(X86_FEATURE_PQE) #define cpu_has_fpu_sel (!boot_cpu_has(X86_FEATURE_NO_FPU_SEL)) #define cpu_has_mpx boot_cpu_has(X86_FEATURE_MPX) #define cpu_has_avx512f boot_cpu_has(X86_FEATURE_AVX512F) -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Tue Jun 15 19:55:15 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 15 Jun 2021 19:55:15 +0000 Received: from list by lists.xenproject.org with outflank-mailman.142386.262735 (Exim 4.92) (envelope-from ) id 1ltF9X-00032P-V8; Tue, 15 Jun 2021 19:55:15 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 142386.262735; Tue, 15 Jun 2021 19:55:15 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1ltF9X-00032H-SC; Tue, 15 Jun 2021 19:55:15 +0000 Received: by outflank-mailman (input) for mailman id 142386; Tue, 15 Jun 2021 19:55:15 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1ltF9X-00032B-6A for xen-changelog@lists.xenproject.org; Tue, 15 Jun 2021 19:55:15 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1ltF9X-0002U7-5L for xen-changelog@lists.xenproject.org; Tue, 15 Jun 2021 19:55:15 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1ltF9X-000129-4R for xen-changelog@lists.xenproject.org; Tue, 15 Jun 2021 19:55:15 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=v95woW8ljFYwVOvNySBtlrEfmtrpmrTQn0OrwDBQGmk=; b=yJUn49CePviiTPVPjbRUPpLa1r PHpCKs4cQplQk3et4OdOWwkDBzGMPpzbi5PT+D78OIFbMMG/c1mx6WhEUwXOZI97hmWpeFhzGmiAZ 1zLFKtVLp325CCqtTLtgxLvmxRYetfl2fm9tRxEFHYARRKxIsBpNHEDvdz8dkxv0PDjU=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] x86/platform: Permit reading the TSX control MSRs via XENPF_resource_op Message-Id: Date: Tue, 15 Jun 2021 19:55:15 +0000 commit 3ccaa174faeb8fbd68f79c670e32c6689cfc7ffc Author: Andrew Cooper AuthorDate: Thu Jun 10 11:02:47 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 15 20:50:32 2021 +0100 x86/platform: Permit reading the TSX control MSRs via XENPF_resource_op We are going to want this to write some tests with. Signed-off-by: Andrew Cooper Reviewed-by: Jan Beulich --- xen/arch/x86/platform_hypercall.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/xen/arch/x86/platform_hypercall.c b/xen/arch/x86/platform_hypercall.c index 41d8e59563..284c2dfb9e 100644 --- a/xen/arch/x86/platform_hypercall.c +++ b/xen/arch/x86/platform_hypercall.c @@ -74,6 +74,12 @@ static bool msr_read_allowed(unsigned int msr) case MSR_IA32_TSC: return true; + + case MSR_TSX_FORCE_ABORT: + return cpu_has_tsx_force_abort; + + case MSR_TSX_CTRL: + return cpu_has_tsx_ctrl; } if ( ppin_msr && msr == ppin_msr ) -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Tue Jun 15 19:55:27 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 15 Jun 2021 19:55:27 +0000 Received: from list by lists.xenproject.org with outflank-mailman.142387.262739 (Exim 4.92) (envelope-from ) id 1ltF9j-00035F-0Z; Tue, 15 Jun 2021 19:55:27 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 142387.262739; Tue, 15 Jun 2021 19:55:26 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1ltF9i-000356-Tn; Tue, 15 Jun 2021 19:55:26 +0000 Received: by outflank-mailman (input) for mailman id 142387; Tue, 15 Jun 2021 19:55:25 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1ltF9h-00034t-9G for xen-changelog@lists.xenproject.org; Tue, 15 Jun 2021 19:55:25 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1ltF9h-0002UR-8O for xen-changelog@lists.xenproject.org; Tue, 15 Jun 2021 19:55:25 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1ltF9h-00012u-7O for xen-changelog@lists.xenproject.org; Tue, 15 Jun 2021 19:55:25 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=+QaBOiMfIA8cFBxnSTV24fRs7Sz/bNn/2ahEFwZlZSQ=; b=CMzVqoYTSig3L0bb1B7Pycltev LVUQr8LuKt1anS43LmXD49JeCSiXqxZg1tBfFtbIBjcOSGPkEpnTfFXOGFOOUwoC2V12Mg6mbRNtt 4/TToN2uI5QPhrjzKtcKEDapsg4hRLxhlMyzFJj391+Hp0E4rekucWvJDVJehu+mslKw=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] x86/msr: Expose MSR_ARCH_CAPS in the raw and host policies Message-Id: Date: Tue, 15 Jun 2021 19:55:25 +0000 commit b672695e748869b2e2e2cb03f671f12003d2b079 Author: Andrew Cooper AuthorDate: Fri Jun 11 11:37:53 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 15 20:50:32 2021 +0100 x86/msr: Expose MSR_ARCH_CAPS in the raw and host policies MSR_ARCH_CAPS is still not supported for guests yet (other than the hardware domain), until the toolstack learns how to construct an MSR policy. However, we want access to the host ARCH_CAPS_TSX_CTRL value in particular for testing purposes. Signed-off-by: Andrew Cooper Reviewed-by: Jan Beulich --- xen/arch/x86/msr.c | 15 +++++++++++++++ 1 file changed, 15 insertions(+) diff --git a/xen/arch/x86/msr.c b/xen/arch/x86/msr.c index 374f92b2c5..72bbe46949 100644 --- a/xen/arch/x86/msr.c +++ b/xen/arch/x86/msr.c @@ -47,8 +47,13 @@ struct msr_policy __read_mostly hvm_def_msr_policy; static void __init calculate_raw_policy(void) { + struct msr_policy *mp = &raw_msr_policy; + /* 0x000000ce MSR_INTEL_PLATFORM_INFO */ /* Was already added by probe_cpuid_faulting() */ + + if ( cpu_has_arch_caps ) + rdmsrl(MSR_ARCH_CAPABILITIES, mp->arch_caps.raw); } static void __init calculate_host_policy(void) @@ -60,6 +65,12 @@ static void __init calculate_host_policy(void) /* 0x000000ce MSR_INTEL_PLATFORM_INFO */ /* probe_cpuid_faulting() sanity checks presence of MISC_FEATURES_ENABLES */ mp->platform_info.cpuid_faulting = cpu_has_cpuid_faulting; + + /* Temporary, until we have known_features[] for feature bits in MSRs. */ + mp->arch_caps.raw &= + (ARCH_CAPS_RDCL_NO | ARCH_CAPS_IBRS_ALL | ARCH_CAPS_RSBA | + ARCH_CAPS_SKIP_L1DFL | ARCH_CAPS_SSB_NO | ARCH_CAPS_MDS_NO | + ARCH_CAPS_IF_PSCHANGE_MC_NO | ARCH_CAPS_TSX_CTRL | ARCH_CAPS_TAA_NO); } static void __init calculate_pv_max_policy(void) @@ -67,6 +78,8 @@ static void __init calculate_pv_max_policy(void) struct msr_policy *mp = &pv_max_msr_policy; *mp = host_msr_policy; + + mp->arch_caps.raw = 0; /* Not supported yet. */ } static void __init calculate_pv_def_policy(void) @@ -84,6 +97,8 @@ static void __init calculate_hvm_max_policy(void) /* It's always possible to emulate CPUID faulting for HVM guests */ mp->platform_info.cpuid_faulting = true; + + mp->arch_caps.raw = 0; /* Not supported yet. */ } static void __init calculate_hvm_def_policy(void) -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Tue Jun 15 19:55:37 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 15 Jun 2021 19:55:37 +0000 Received: from list by lists.xenproject.org with outflank-mailman.142388.262743 (Exim 4.92) (envelope-from ) id 1ltF9t-00037y-1z; Tue, 15 Jun 2021 19:55:37 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 142388.262743; Tue, 15 Jun 2021 19:55:37 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1ltF9s-00037q-VK; Tue, 15 Jun 2021 19:55:36 +0000 Received: by outflank-mailman (input) for mailman id 142388; Tue, 15 Jun 2021 19:55:35 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1ltF9r-00037W-CC for xen-changelog@lists.xenproject.org; Tue, 15 Jun 2021 19:55:35 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1ltF9r-0002Uh-BO for xen-changelog@lists.xenproject.org; Tue, 15 Jun 2021 19:55:35 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1ltF9r-00013V-AR for xen-changelog@lists.xenproject.org; Tue, 15 Jun 2021 19:55:35 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=oHix19XDIb6AMNW3uoR3Q8n+/gnyGvOE3OgLiO/Dvfc=; b=ABy42Z6dFSfk5SuuV26irjzoSe IYUz4VYIiqZmJTVznfu5T6vrwyocdrv7gm6DglWGIY3eXSWu6SsZPIMuO9JShhp2Ei7E5jz/JVwlE CLiDrwnypiigCp3m1Znze5sbrgmC1MXueCPJ7OSQkZ3goNGgPbYS1Q+kqThwpT7ofdtM=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] libs/guest: Move struct xc_cpu_policy into xg_private.h Message-Id: Date: Tue, 15 Jun 2021 19:55:35 +0000 commit f7ad9ee7cafa679946beaa40d5eb3eab500e86b5 Author: Andrew Cooper AuthorDate: Thu Jun 10 12:57:42 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 15 20:50:32 2021 +0100 libs/guest: Move struct xc_cpu_policy into xg_private.h ... so tests can peek at the internals, without the structure being generally available to users of the library. Signed-off-by: Andrew Cooper Acked-by: Jan Beulich --- tools/libs/guest/xg_cpuid_x86.c | 11 +---------- tools/libs/guest/xg_private.h | 11 +++++++++++ 2 files changed, 12 insertions(+), 10 deletions(-) diff --git a/tools/libs/guest/xg_cpuid_x86.c b/tools/libs/guest/xg_cpuid_x86.c index ec5a47fde4..e01d657e03 100644 --- a/tools/libs/guest/xg_cpuid_x86.c +++ b/tools/libs/guest/xg_cpuid_x86.c @@ -22,7 +22,7 @@ #include #include #include -#include "xc_private.h" +#include "xg_private.h" #include "xc_bitops.h" #include #include @@ -34,18 +34,9 @@ enum { #include -#include - #define bitmaskof(idx) (1u << ((idx) & 31)) #define featureword_of(idx) ((idx) >> 5) -struct xc_cpu_policy { - struct cpuid_policy cpuid; - struct msr_policy msr; - xen_cpuid_leaf_t leaves[CPUID_MAX_SERIALISED_LEAVES]; - xen_msr_entry_t entries[MSR_MAX_SERIALISED_ENTRIES]; -}; - int xc_get_cpu_levelling_caps(xc_interface *xch, uint32_t *caps) { DECLARE_SYSCTL; diff --git a/tools/libs/guest/xg_private.h b/tools/libs/guest/xg_private.h index 03d765da21..28441ee13f 100644 --- a/tools/libs/guest/xg_private.h +++ b/tools/libs/guest/xg_private.h @@ -168,4 +168,15 @@ int pin_table(xc_interface *xch, unsigned int type, unsigned long mfn, #define M2P_SIZE(_m) ROUNDUP(((_m) * sizeof(xen_pfn_t)), M2P_SHIFT) #define M2P_CHUNKS(_m) (M2P_SIZE((_m)) >> M2P_SHIFT) +#if defined(__x86_64__) || defined(__i386__) +#include + +struct xc_cpu_policy { + struct cpuid_policy cpuid; + struct msr_policy msr; + xen_cpuid_leaf_t leaves[CPUID_MAX_SERIALISED_LEAVES]; + xen_msr_entry_t entries[MSR_MAX_SERIALISED_ENTRIES]; +}; +#endif /* x86 */ + #endif /* XG_PRIVATE_H */ -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Tue Jun 15 19:55:47 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 15 Jun 2021 19:55:47 +0000 Received: from list by lists.xenproject.org with outflank-mailman.142389.262747 (Exim 4.92) (envelope-from ) id 1ltFA3-0003BU-4F; Tue, 15 Jun 2021 19:55:47 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 142389.262747; Tue, 15 Jun 2021 19:55:47 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1ltFA3-0003BM-0a; Tue, 15 Jun 2021 19:55:47 +0000 Received: by outflank-mailman (input) for mailman id 142389; Tue, 15 Jun 2021 19:55:45 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1ltFA1-0003B4-FF for xen-changelog@lists.xenproject.org; Tue, 15 Jun 2021 19:55:45 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1ltFA1-0002Us-EN for xen-changelog@lists.xenproject.org; Tue, 15 Jun 2021 19:55:45 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1ltFA1-00014C-DY for xen-changelog@lists.xenproject.org; Tue, 15 Jun 2021 19:55:45 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=PwsohcaHHBLfmUyyhpQ/Jgs2JGlL1phgLNMLYY++D5w=; b=1CO6YwbJBN8bakZcKEemgQzaTu kKJdTCmZ045hS6OzP7RatEtKUdoxqs+Cgw9d/YpLVGwZRMqJjTStd22SCX92QLiErJp3FmS3nP3zd oXxuZFQN1fbEgOyU/eInxO+HeTNfgLurcGLim/0SoP5vfhfk1Y99xNHhf5VU1mW0dHXk=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] tests: Introduce a TSX test Message-Id: Date: Tue, 15 Jun 2021 19:55:45 +0000 commit 4bcf6433eed3d9cbc00865ec62380a33ca832dac Author: Andrew Cooper AuthorDate: Thu Jun 10 12:34:45 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 15 20:50:32 2021 +0100 tests: Introduce a TSX test See the comment at the top of test-tsx.c for details. This covers various complexities encountered while trying to address the recent TSX deprecation on client parts. A sample run on KabyLake with latest microcode and default tsx= looks like this: root@host# ./test-tsx TSX tests Got 8 CPUs Testing MSR_TSX_FORCE_ABORT consistency CPU0 val 0x3 Testing MSR_TSX_CTRL consistency Testing RTM behaviour Got Abort Testing PV default/max policies Max: RTM 1, HLE 1, TSX_FORCE_ABORT 0, RTM_ALWAYS_ABORT 0, TSX_CTRL 0 Def: RTM 0, HLE 0, TSX_FORCE_ABORT 0, RTM_ALWAYS_ABORT 0, TSX_CTRL 0 Testing HVM default/max policies Max: RTM 1, HLE 1, TSX_FORCE_ABORT 0, RTM_ALWAYS_ABORT 0, TSX_CTRL 0 Def: RTM 0, HLE 0, TSX_FORCE_ABORT 0, RTM_ALWAYS_ABORT 0, TSX_CTRL 0 Testing PV guest Created d7 Cur: RTM 0, HLE 0, TSX_FORCE_ABORT 0, RTM_ALWAYS_ABORT 0, TSX_CTRL 0 Cur: RTM 1, HLE 1, TSX_FORCE_ABORT 0, RTM_ALWAYS_ABORT 0, TSX_CTRL 0 Testing HVM guest Created d8 Cur: RTM 0, HLE 0, TSX_FORCE_ABORT 0, RTM_ALWAYS_ABORT 0, TSX_CTRL 0 Cur: RTM 1, HLE 1, TSX_FORCE_ABORT 0, RTM_ALWAYS_ABORT 0, TSX_CTRL 0 Signed-off-by: Andrew Cooper Acked-by: Jan Beulich --- tools/tests/Makefile | 1 + tools/tests/tsx/.gitignore | 1 + tools/tests/tsx/Makefile | 45 ++++ tools/tests/tsx/test-tsx.c | 538 +++++++++++++++++++++++++++++++++++++++++++++ 4 files changed, 585 insertions(+) diff --git a/tools/tests/Makefile b/tools/tests/Makefile index 8746aabe6b..25531a984a 100644 --- a/tools/tests/Makefile +++ b/tools/tests/Makefile @@ -5,6 +5,7 @@ SUBDIRS-y := SUBDIRS-y += resource SUBDIRS-$(CONFIG_X86) += cpu-policy SUBDIRS-$(CONFIG_X86) += mce-test +SUBDIRS-$(CONFIG_X86) += tsx ifneq ($(clang),y) SUBDIRS-$(CONFIG_X86) += x86_emulator endif diff --git a/tools/tests/tsx/.gitignore b/tools/tests/tsx/.gitignore new file mode 100644 index 0000000000..97ec4db7ff --- /dev/null +++ b/tools/tests/tsx/.gitignore @@ -0,0 +1 @@ +test-tsx diff --git a/tools/tests/tsx/Makefile b/tools/tests/tsx/Makefile new file mode 100644 index 0000000000..d7d2a5d95e --- /dev/null +++ b/tools/tests/tsx/Makefile @@ -0,0 +1,45 @@ +XEN_ROOT = $(CURDIR)/../../.. +include $(XEN_ROOT)/tools/Rules.mk + +TARGET := test-tsx + +.PHONY: all +all: $(TARGET) + +.PHONY: clean +clean: + $(RM) -- *.o $(TARGET) $(DEPS_RM) + +.PHONY: distclean +distclean: clean + $(RM) -- *~ + +.PHONY: install +install: all + $(INSTALL_DIR) $(DESTDIR)$(LIBEXEC_BIN) + $(INSTALL_PROG) $(TARGET) $(DESTDIR)$(LIBEXEC_BIN) + +.PHONY: uninstall +uninstall: + $(RM) -- $(DESTDIR)$(LIBEXEC_BIN)/$(TARGET) + +.PHONY: uninstall +uninstall: + +CFLAGS += -Werror +CFLAGS += -I$(XEN_ROOT)/tools/libs/ctrl -I$(XEN_ROOT)/tools/libs/guest +CFLAGS += $(CFLAGS_xeninclude) +CFLAGS += $(CFLAGS_libxenctrl) +CFLAGS += $(CFLAGS_libxenguest) +CFLAGS += $(APPEND_CFLAGS) + +LDFLAGS += $(LDLIBS_libxenctrl) +LDFLAGS += $(LDLIBS_libxenguest) +LDFLAGS += $(APPEND_LDFLAGS) + +%.o: Makefile + +$(TARGET): test-tsx.o + $(CC) -o $@ $< $(LDFLAGS) + +-include $(DEPS_INCLUDE) diff --git a/tools/tests/tsx/test-tsx.c b/tools/tests/tsx/test-tsx.c new file mode 100644 index 0000000000..fab99c135e --- /dev/null +++ b/tools/tests/tsx/test-tsx.c @@ -0,0 +1,538 @@ +/* + * TSX settings and consistency tests + * + * This tests various behaviours and invariants with regards to TSX. It + * ideally wants running for several microcode versions, and all applicable + * tsx= commandline settings, on a single CPU, including after an S3 + * suspend/resume event. + * + * It tests specifically: + * - The consistency of MSR_TSX_CTRL/MSR_TSX_FORCE_ABORT values across the + * system, and their accessibility WRT data in the host CPU policy. + * - The actual behaviour of RTM on the system. + * - Cross-check the default/max policies based on the actual RTM behaviour. + * - Create some guests, check their defaults, and check that the defaults + * can be changed. + */ + +#define _GNU_SOURCE + +#include +#include +#include +#include +#include +#include +#include +#include + +#include +#include +#include + +#include "xg_private.h" + +enum { +#define XEN_CPUFEATURE(name, value) X86_FEATURE_##name = value, +#include +}; +#define bitmaskof(idx) (1u << ((idx) & 31)) + +#define MSR_ARCH_CAPABILITIES 0x0000010a +#define ARCH_CAPS_TSX_CTRL (1 << 7) +#define MSR_TSX_FORCE_ABORT 0x0000010f +#define MSR_TSX_CTRL 0x00000122 + +static unsigned int nr_failures; +#define fail(fmt, ...) \ +({ \ + nr_failures++; \ + (void)printf(fmt, ##__VA_ARGS__); \ +}) + +static xc_interface *xch; + +/* + * Policies, arranged as an array for easy collection of all of them. We + * don't care about the raw policy (index 0) so reuse that for the guest + * policy. + */ +static struct xc_cpu_policy policies[6]; +#define guest_policy policies[0] +#define host policies[XEN_SYSCTL_cpu_policy_host] +#define pv_max policies[XEN_SYSCTL_cpu_policy_pv_max] +#define hvm_max policies[XEN_SYSCTL_cpu_policy_hvm_max] +#define pv_default policies[XEN_SYSCTL_cpu_policy_pv_default] +#define hvm_default policies[XEN_SYSCTL_cpu_policy_hvm_default] + +static bool xen_has_pv = true, xen_has_hvm = true; + +static xc_physinfo_t physinfo; + +static enum rtm_behaviour { + RTM_UD, + RTM_OK, + RTM_ABORT, +} rtm_behaviour; + +/* + * Test a specific TSX MSR for consistency across the system, taking into + * account whether it ought to be accessible or not. + * + * We can't query offline CPUs, so skip those if encountered. We don't care + * particularly for the exact MSR value, but we do care that it is the same + * everywhere. + */ +static void test_tsx_msr_consistency(unsigned int msr, bool accessible) +{ + uint64_t cpu0_val = ~0; + + for ( unsigned int cpu = 0; cpu <= physinfo.max_cpu_id; ++cpu ) + { + xc_resource_entry_t ent = { + .u.cmd = XEN_RESOURCE_OP_MSR_READ, + .idx = msr, + }; + xc_resource_op_t op = { + .cpu = cpu, + .entries = &ent, + .nr_entries = 1, + }; + int rc = xc_resource_op(xch, 1, &op); + + if ( rc < 0 ) + { + /* Don't emit a message for offline CPUs */ + if ( errno != ENODEV ) + fail(" xc_resource_op() for CPU%u failed: rc %d, errno %d - %s\n", + cpu, rc, errno, strerror(errno)); + continue; + } + + if ( accessible ) + { + if ( rc != 1 ) + { + fail(" Expected 1 result, got %d\n", rc); + continue; + } + if ( ent.u.ret != 0 ) + { + fail(" Expected ok, got %d\n", ent.u.ret); + continue; + } + } + else + { + if ( rc != 0 ) + fail(" Expected 0 results, got %u\n", rc); + else if ( ent.u.ret != -EPERM ) + fail(" Expected -EPERM, got %d\n", ent.u.ret); + continue; + } + + if ( cpu == 0 ) + { + cpu0_val = ent.val; + printf(" CPU0 val %#"PRIx64"\n", cpu0_val); + } + else if ( ent.val != cpu0_val ) + fail(" CPU%u val %#"PRIx64" differs from CPU0 %#"PRIx64"\n", + cpu, ent.val, cpu0_val); + } +} + +/* + * Check all TSX MSRs, and in particular that their accessibility matches what + * is expressed in the host CPU policy. + */ +static void test_tsx_msrs(void) +{ + printf("Testing MSR_TSX_FORCE_ABORT consistency\n"); + test_tsx_msr_consistency( + MSR_TSX_FORCE_ABORT, host.cpuid.feat.tsx_force_abort); + + printf("Testing MSR_TSX_CTRL consistency\n"); + test_tsx_msr_consistency( + MSR_TSX_CTRL, host.msr.arch_caps.tsx_ctrl); +} + +/* + * Probe for how RTM behaves, deliberately not inspecting CPUID. + * Distinguishes between "no support at all" (i.e. XBEGIN suffers #UD), + * working ok, and appearing to always abort. + */ +static enum rtm_behaviour probe_rtm_behaviour(void) +{ + for ( unsigned int i = 0; i < 1000; ++i ) + { + /* + * Opencoding the RTM infrastructure from immintrin.h, because we + * still support older versions of GCC. Also so we can include #UD + * detection logic. + */ +#define XBEGIN_STARTED -1 +#define XBEGIN_UD -2 + unsigned int status = XBEGIN_STARTED; + + asm volatile ( ".Lxbegin: .byte 0xc7,0xf8,0,0,0,0" /* XBEGIN 1f; 1: */ + : "+a" (status) :: "memory" ); + if ( status == XBEGIN_STARTED ) + { + asm volatile ( ".byte 0x0f,0x01,0xd5" ::: "memory" ); /* XEND */ + return RTM_OK; + } + else if ( status == XBEGIN_UD ) + return RTM_UD; + } + + return RTM_ABORT; +} + +static struct sigaction old_sigill; + +static void sigill_handler(int signo, siginfo_t *info, void *extra) +{ + extern const char xbegin_label[] asm(".Lxbegin"); + + if ( info->si_addr == xbegin_label && + memcmp(info->si_addr, "\xc7\xf8\x00\x00\x00\x00", 6) == 0 ) + { + ucontext_t *context = extra; + + /* + * Found the XBEGIN instruction. Step over it, and update `status` to + * signal #UD. + */ +#if defined(__linux__) +# ifdef __x86_64__ + context->uc_mcontext.gregs[REG_RIP] += 6; + context->uc_mcontext.gregs[REG_RAX] = XBEGIN_UD; +# else + context->uc_mcontext.gregs[REG_EIP] += 6; + context->uc_mcontext.gregs[REG_EAX] = XBEGIN_UD; +# endif + +#elif defined(__FreeBSD__) +# ifdef __x86_64__ + context->uc_mcontext.mc_rip += 6; + context->uc_mcontext.mc_rax = XBEGIN_UD; +# else + context->uc_mcontext.mc_eip += 6; + context->uc_mcontext.mc_eax = XBEGIN_UD; +# endif + +#elif defined(__NetBSD__) +# ifdef __x86_64__ + context->uc_mcontext.__gregs[_REG_RIP] += 6; + context->uc_mcontext.__gregs[_REG_RAX] = XBEGIN_UD; +# else + context->uc_mcontext.__gregs[_REG_EIP] += 6; + context->uc_mcontext.__gregs[_REG_EAX] = XBEGIN_UD; +# endif + +#else +# error Unknown environment - please adjust +#endif + } + else + { + /* + * Not the SIGILL we're looking for... Restore the old handler and + * try again. Will likely coredump as a result. + */ + sigaction(SIGILL, &old_sigill, NULL); + } +} + +static void test_rtm_behaviour(void) +{ + struct sigaction new_sigill = { + .sa_flags = SA_SIGINFO, + .sa_sigaction = sigill_handler, + }; + const char *str; + + printf("Testing RTM behaviour\n"); + + /* + * Install a custom SIGILL handler while probing for RTM behaviour, as the + * XBEGIN instruction might suffer #UD. + */ + sigaction(SIGILL, &new_sigill, &old_sigill); + rtm_behaviour = probe_rtm_behaviour(); + sigaction(SIGILL, &old_sigill, NULL); + + switch ( rtm_behaviour ) + { + case RTM_UD: str = "#UD"; break; + case RTM_OK: str = "OK"; break; + case RTM_ABORT: str = "Abort"; break; + default: str = NULL; break; + } + + if ( str ) + printf(" Got %s\n", str); + else + return fail(" Got unexpected behaviour %d\n", rtm_behaviour); + + if ( host.cpuid.feat.rtm ) + { + if ( rtm_behaviour == RTM_UD ) + fail(" Host reports RTM, but appears unavailable\n"); + } + else + { + if ( rtm_behaviour != RTM_UD ) + fail(" Host reports no RTM, but appears available\n"); + } +} + +static void dump_tsx_details(const struct xc_cpu_policy *p, const char *pref) +{ + printf(" %s RTM %u, HLE %u, TSX_FORCE_ABORT %u, RTM_ALWAYS_ABORT %u, TSX_CTRL %u\n", + pref, + p->cpuid.feat.rtm, + p->cpuid.feat.hle, + p->cpuid.feat.tsx_force_abort, + p->cpuid.feat.rtm_always_abort, + p->msr.arch_caps.tsx_ctrl); +} + +/* Sanity test various invariants we expect in the default/max policies. */ +static void test_guest_policies(const struct xc_cpu_policy *max, + const struct xc_cpu_policy *def) +{ + const struct cpuid_policy *cm = &max->cpuid; + const struct cpuid_policy *cd = &def->cpuid; + const struct msr_policy *mm = &max->msr; + const struct msr_policy *md = &def->msr; + + dump_tsx_details(max, "Max:"); + dump_tsx_details(def, "Def:"); + + if ( ((cm->feat.raw[0].d | cd->feat.raw[0].d) & + (bitmaskof(X86_FEATURE_TSX_FORCE_ABORT) | + bitmaskof(X86_FEATURE_RTM_ALWAYS_ABORT))) || + ((mm->arch_caps.raw | md->arch_caps.raw) & ARCH_CAPS_TSX_CTRL) ) + fail(" Xen-only TSX controls offered to guest\n"); + + switch ( rtm_behaviour ) + { + case RTM_UD: + if ( (cm->feat.raw[0].b | cd->feat.raw[0].b) & + (bitmaskof(X86_FEATURE_HLE) | bitmaskof(X86_FEATURE_RTM)) ) + fail(" HLE/RTM offered to guests despite not being available\n"); + break; + + case RTM_ABORT: + if ( cd->feat.raw[0].b & + (bitmaskof(X86_FEATURE_HLE) | bitmaskof(X86_FEATURE_RTM)) ) + fail(" HLE/RTM offered to guests by default despite not being usable\n"); + break; + + case RTM_OK: + if ( !cm->feat.rtm || !cd->feat.rtm ) + fail(" RTM not offered to guests despite being available\n"); + break; + } + + if ( cd->feat.hle ) + fail(" Fail: HLE offered in default policy\n"); +} + +static void test_def_max_policies(void) +{ + if ( xen_has_pv ) + { + printf("Testing PV default/max policies\n"); + test_guest_policies(&pv_max, &pv_default); + } + + if ( xen_has_hvm ) + { + printf("Testing HVM default/max policies\n"); + test_guest_policies(&hvm_max, &hvm_default); + } +} + +static void test_guest(struct xen_domctl_createdomain *c) +{ + uint32_t domid = 0; + int rc; + + rc = xc_domain_create(xch, &domid, c); + if ( rc ) + return fail(" Domain create failure: %d - %s\n", + errno, strerror(errno)); + + printf(" Created d%u\n", domid); + + rc = xc_cpu_policy_get_domain(xch, domid, &guest_policy); + if ( rc ) + { + fail(" Failed to obtain domain policy: %d - %s\n", + errno, strerror(errno)); + goto out; + } + + dump_tsx_details(&guest_policy, "Cur:"); + + /* + * Check defaults given to the guest. + */ + if ( guest_policy.cpuid.feat.rtm != (rtm_behaviour == RTM_OK) ) + fail(" RTM %u in guest, despite rtm behaviour\n", + guest_policy.cpuid.feat.rtm); + + if ( guest_policy.cpuid.feat.hle || + guest_policy.cpuid.feat.tsx_force_abort || + guest_policy.cpuid.feat.rtm_always_abort || + guest_policy.msr.arch_caps.tsx_ctrl ) + fail(" Unexpected features advertised\n"); + + if ( host.cpuid.feat.rtm ) + { + unsigned int _7b0; + + /* + * If host RTM is available, all combinations of guest flags should be + * possible. Flip both HLE/RTM to check non-default settings. + */ + _7b0 = (guest_policy.cpuid.feat.raw[0].b ^= + (bitmaskof(X86_FEATURE_HLE) | bitmaskof(X86_FEATURE_RTM))); + + /* Set the new policy. */ + rc = xc_cpu_policy_set_domain(xch, domid, &guest_policy); + if ( rc ) + { + fail(" Failed to set domain policy: %d - %s\n", + errno, strerror(errno)); + goto out; + } + + /* Re-get the new policy. */ + rc = xc_cpu_policy_get_domain(xch, domid, &guest_policy); + if ( rc ) + { + fail(" Failed to obtain domain policy: %d - %s\n", + errno, strerror(errno)); + goto out; + } + + dump_tsx_details(&guest_policy, "Cur:"); + + if ( guest_policy.cpuid.feat.raw[0].b != _7b0 ) + { + fail(" Expected CPUID.7[1].b 0x%08x differs from actual 0x%08x\n", + _7b0, guest_policy.cpuid.feat.raw[0].b); + goto out; + } + } + + out: + rc = xc_domain_destroy(xch, domid); + if ( rc ) + fail(" Failed to destroy domain: %d - %s\n", + errno, strerror(errno)); +} + +static void test_guests(void) +{ + if ( xen_has_pv ) + { + struct xen_domctl_createdomain c = { + .max_vcpus = 1, + .max_grant_frames = 1, + }; + + printf("Testing PV guest\n"); + test_guest(&c); + } + + if ( xen_has_hvm ) + { + struct xen_domctl_createdomain c = { + .flags = XEN_DOMCTL_CDF_hvm, + .max_vcpus = 1, + .max_grant_frames = 1, + .arch = { + .emulation_flags = XEN_X86_EMU_LAPIC, + }, + }; + + if ( physinfo.capabilities & XEN_SYSCTL_PHYSCAP_hap ) + c.flags |= XEN_DOMCTL_CDF_hap; + else if ( !(physinfo.capabilities & XEN_SYSCTL_PHYSCAP_shadow) ) + return fail(" HVM available, but neither HAP nor Shadow\n"); + + printf("Testing HVM guest\n"); + test_guest(&c); + } +} + +/* Obtain some general data, then run the tests. */ +static void test_tsx(void) +{ + int rc; + + /* Read all policies except raw. */ + for ( unsigned int i = XEN_SYSCTL_cpu_policy_host; + i <= XEN_SYSCTL_cpu_policy_hvm_default; ++i ) + { + rc = xc_cpu_policy_get_system(xch, i, &policies[i]); + + if ( rc == -1 && errno == EOPNOTSUPP ) + { + /* + * Use EOPNOTSUPP to spot Xen missing CONFIG_{PV,HVM}, and adjust + * later testing accordingly. + */ + switch ( i ) + { + case XEN_SYSCTL_cpu_policy_pv_max: + case XEN_SYSCTL_cpu_policy_pv_default: + if ( xen_has_pv ) + printf(" Xen doesn't support PV\n"); + xen_has_pv = false; + continue; + + case XEN_SYSCTL_cpu_policy_hvm_max: + case XEN_SYSCTL_cpu_policy_hvm_default: + if ( xen_has_hvm ) + printf(" Xen doesn't support HVM\n"); + xen_has_hvm = false; + continue; + } + } + if ( rc ) + return fail("Failed to obtain policy[%u]: %d - %s\n", + i, errno, strerror(errno)); + } + + rc = xc_physinfo(xch, &physinfo); + if ( rc ) + return fail("Failed to obtain physinfo: %d - %s\n", + errno, strerror(errno)); + + printf(" Got %u CPUs\n", physinfo.max_cpu_id + 1); + + test_tsx_msrs(); + test_rtm_behaviour(); + test_def_max_policies(); + test_guests(); +} + +int main(int argc, char **argv) +{ + printf("TSX tests\n"); + + xch = xc_interface_open(NULL, NULL, 0); + + if ( !xch ) + err(1, "xc_interface_open"); + + test_tsx(); + + return !!nr_failures; +} -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Thu Jun 17 16:11:12 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 17 Jun 2021 16:11:12 +0000 Received: from list by lists.xenproject.org with outflank-mailman.144094.265277 (Exim 4.92) (envelope-from ) id 1ltubh-0006sX-S5; Thu, 17 Jun 2021 16:11:05 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 144094.265277; Thu, 17 Jun 2021 16:11:05 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1ltubh-0006sP-P5; Thu, 17 Jun 2021 16:11:05 +0000 Received: by outflank-mailman (input) for mailman id 144094; Thu, 17 Jun 2021 16:11:04 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1ltubg-0006sJ-JP for xen-changelog@lists.xenproject.org; Thu, 17 Jun 2021 16:11:04 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1ltubg-0004wv-II for xen-changelog@lists.xenproject.org; Thu, 17 Jun 2021 16:11:04 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1ltubg-0002S2-H7 for xen-changelog@lists.xenproject.org; Thu, 17 Jun 2021 16:11:04 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=fT4RFp5zYmGw+LJcSYFOpaOBywYTFdMdAbwF40XBJro=; b=J7LRemFAZ3SecYChFo1k9BGNL6 nnVJe+XY0i9cUW7QbaBEXetqO/NHDYli4+929hdZLWCKntujWI6laW5cUkd/Y1zjJhu7C3SRurdw5 4+8rcYCkB2nUFKhHxqx4ATX5weME17a1mgRNDJ+XKw2rEsH6jcK9rOfreDSAcTRZxfJ4=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] x86/mtrr: move epte_get_entry_emt to p2m-ept.c Message-Id: Date: Thu, 17 Jun 2021 16:11:04 +0000 commit ed464d4e8a9a49559307c96ee6aa59e97820f692 Author: Roger Pau Monné AuthorDate: Thu Jun 17 17:58:11 2021 +0200 Commit: Jan Beulich CommitDate: Thu Jun 17 17:58:11 2021 +0200 x86/mtrr: move epte_get_entry_emt to p2m-ept.c This is an EPT specific function, so it shouldn't live in the generic mtrr file. Such movement is also needed for future work that will require passing a p2m_type_t parameter to epte_get_entry_emt, and making that type visible to the mtrr users is cumbersome and unneeded. Moving epte_get_entry_emt out of mtrr.c requires making the helper to get the MTRR type of an address from the mtrr state public. While there rename the function to start with the mtrr prefix, like other mtrr related functions. While there fix some of the types of the function parameters. No functional change intended. Signed-off-by: Roger Pau Monné Reviewed-by: Jan Beulich Reviewed-by: Kevin Tian --- xen/arch/x86/hvm/mtrr.c | 107 +---------------------------------- xen/arch/x86/hvm/vmx/vmx.c | 4 +- xen/arch/x86/mm/p2m-ept.c | 114 ++++++++++++++++++++++++++++++++++++-- xen/include/asm-x86/hvm/vmx/vmx.h | 2 + xen/include/asm-x86/mtrr.h | 5 +- 5 files changed, 118 insertions(+), 114 deletions(-) diff --git a/xen/arch/x86/hvm/mtrr.c b/xen/arch/x86/hvm/mtrr.c index 82ded1635c..4a9f3177ed 100644 --- a/xen/arch/x86/hvm/mtrr.c +++ b/xen/arch/x86/hvm/mtrr.c @@ -194,8 +194,7 @@ void hvm_vcpu_cacheattr_destroy(struct vcpu *v) * May return a negative value when order > 0, indicating to the caller * that the respective mapping needs splitting. */ -static int get_mtrr_type(const struct mtrr_state *m, - paddr_t pa, unsigned int order) +int mtrr_get_type(const struct mtrr_state *m, paddr_t pa, unsigned int order) { uint8_t overlap_mtrr = 0; uint8_t overlap_mtrr_pos = 0; @@ -323,7 +322,7 @@ static uint8_t effective_mm_type(struct mtrr_state *m, * just use it */ if ( gmtrr_mtype == NO_HARDCODE_MEM_TYPE ) - mtrr_mtype = get_mtrr_type(m, gpa, 0); + mtrr_mtype = mtrr_get_type(m, gpa, 0); else mtrr_mtype = gmtrr_mtype; @@ -350,7 +349,7 @@ uint32_t get_pat_flags(struct vcpu *v, guest_eff_mm_type = effective_mm_type(g, pat, gpaddr, gl1e_flags, gmtrr_mtype); /* 2. Get the memory type of host physical address, with MTRR */ - shadow_mtrr_type = get_mtrr_type(&mtrr_state, spaddr, 0); + shadow_mtrr_type = mtrr_get_type(&mtrr_state, spaddr, 0); /* 3. Find the memory type in PAT, with host MTRR memory type * and guest effective memory type. @@ -789,106 +788,6 @@ void memory_type_changed(struct domain *d) } } -int epte_get_entry_emt(struct domain *d, unsigned long gfn, mfn_t mfn, - unsigned int order, uint8_t *ipat, bool_t direct_mmio) -{ - int gmtrr_mtype, hmtrr_mtype; - struct vcpu *v = current; - unsigned long i; - - *ipat = 0; - - if ( v->domain != d ) - v = d->vcpu ? d->vcpu[0] : NULL; - - /* Mask, not add, for order so it works with INVALID_MFN on unmapping */ - if ( rangeset_overlaps_range(mmio_ro_ranges, mfn_x(mfn), - mfn_x(mfn) | ((1UL << order) - 1)) ) - { - if ( !order || rangeset_contains_range(mmio_ro_ranges, mfn_x(mfn), - mfn_x(mfn) | ((1UL << order) - 1)) ) - { - *ipat = 1; - return MTRR_TYPE_UNCACHABLE; - } - /* Force invalid memory type so resolve_misconfig() will split it */ - return -1; - } - - if ( !mfn_valid(mfn) ) - { - *ipat = 1; - return MTRR_TYPE_UNCACHABLE; - } - - if ( !direct_mmio && !is_iommu_enabled(d) && !cache_flush_permitted(d) ) - { - *ipat = 1; - return MTRR_TYPE_WRBACK; - } - - for ( i = 0; i < (1ul << order); i++ ) - { - if ( is_special_page(mfn_to_page(mfn_add(mfn, i))) ) - { - if ( order ) - return -1; - *ipat = 1; - return MTRR_TYPE_WRBACK; - } - } - - if ( direct_mmio ) - return MTRR_TYPE_UNCACHABLE; - - gmtrr_mtype = hvm_get_mem_pinned_cacheattr(d, _gfn(gfn), order); - if ( gmtrr_mtype >= 0 ) - { - *ipat = 1; - return gmtrr_mtype != PAT_TYPE_UC_MINUS ? gmtrr_mtype - : MTRR_TYPE_UNCACHABLE; - } - if ( gmtrr_mtype == -EADDRNOTAVAIL ) - return -1; - - gmtrr_mtype = v ? get_mtrr_type(&v->arch.hvm.mtrr, gfn << PAGE_SHIFT, order) - : MTRR_TYPE_WRBACK; - hmtrr_mtype = get_mtrr_type(&mtrr_state, mfn_x(mfn) << PAGE_SHIFT, order); - if ( gmtrr_mtype < 0 || hmtrr_mtype < 0 ) - return -1; - - /* If both types match we're fine. */ - if ( likely(gmtrr_mtype == hmtrr_mtype) ) - return hmtrr_mtype; - - /* If either type is UC, we have to go with that one. */ - if ( gmtrr_mtype == MTRR_TYPE_UNCACHABLE || - hmtrr_mtype == MTRR_TYPE_UNCACHABLE ) - return MTRR_TYPE_UNCACHABLE; - - /* If either type is WB, we have to go with the other one. */ - if ( gmtrr_mtype == MTRR_TYPE_WRBACK ) - return hmtrr_mtype; - if ( hmtrr_mtype == MTRR_TYPE_WRBACK ) - return gmtrr_mtype; - - /* - * At this point we have disagreeing WC, WT, or WP types. The only - * combination that can be cleanly resolved is WT:WP. The ones involving - * WC need to be converted to UC, both due to the memory ordering - * differences and because WC disallows reads to be cached (WT and WP - * permit this), while WT and WP require writes to go straight to memory - * (WC can buffer them). - */ - if ( (gmtrr_mtype == MTRR_TYPE_WRTHROUGH && - hmtrr_mtype == MTRR_TYPE_WRPROT) || - (gmtrr_mtype == MTRR_TYPE_WRPROT && - hmtrr_mtype == MTRR_TYPE_WRTHROUGH) ) - return MTRR_TYPE_WRPROT; - - return MTRR_TYPE_UNCACHABLE; -} - /* * Local variables: * mode: C diff --git a/xen/arch/x86/hvm/vmx/vmx.c b/xen/arch/x86/hvm/vmx/vmx.c index 7e3e67fdc3..0d4b47681b 100644 --- a/xen/arch/x86/hvm/vmx/vmx.c +++ b/xen/arch/x86/hvm/vmx/vmx.c @@ -417,12 +417,12 @@ static int vmx_domain_initialise(struct domain *d) static void domain_creation_finished(struct domain *d) { gfn_t gfn = gaddr_to_gfn(APIC_DEFAULT_PHYS_BASE); - uint8_t ipat; + bool ipat; if ( !has_vlapic(d) || mfn_eq(apic_access_mfn, INVALID_MFN) ) return; - ASSERT(epte_get_entry_emt(d, gfn_x(gfn), apic_access_mfn, 0, &ipat, + ASSERT(epte_get_entry_emt(d, gfn, apic_access_mfn, 0, &ipat, true) == MTRR_TYPE_WRBACK); ASSERT(ipat); diff --git a/xen/arch/x86/mm/p2m-ept.c b/xen/arch/x86/mm/p2m-ept.c index a3beaf91e2..f1d1d07e92 100644 --- a/xen/arch/x86/mm/p2m-ept.c +++ b/xen/arch/x86/mm/p2m-ept.c @@ -20,6 +20,7 @@ #include #include #include +#include #include #include #include @@ -485,6 +486,108 @@ static int ept_invalidate_emt_range(struct p2m_domain *p2m, return rc; } +int epte_get_entry_emt(struct domain *d, gfn_t gfn, mfn_t mfn, + unsigned int order, bool *ipat, bool direct_mmio) +{ + int gmtrr_mtype, hmtrr_mtype; + struct vcpu *v = current; + unsigned long i; + + *ipat = false; + + if ( v->domain != d ) + v = d->vcpu ? d->vcpu[0] : NULL; + + /* Mask, not add, for order so it works with INVALID_MFN on unmapping */ + if ( rangeset_overlaps_range(mmio_ro_ranges, mfn_x(mfn), + mfn_x(mfn) | ((1UL << order) - 1)) ) + { + if ( !order || rangeset_contains_range(mmio_ro_ranges, mfn_x(mfn), + mfn_x(mfn) | ((1UL << order) - 1)) ) + { + *ipat = true; + return MTRR_TYPE_UNCACHABLE; + } + /* Force invalid memory type so resolve_misconfig() will split it */ + return -1; + } + + if ( !mfn_valid(mfn) ) + { + *ipat = true; + return MTRR_TYPE_UNCACHABLE; + } + + if ( !direct_mmio && !is_iommu_enabled(d) && !cache_flush_permitted(d) ) + { + *ipat = true; + return MTRR_TYPE_WRBACK; + } + + for ( i = 0; i < (1ul << order); i++ ) + { + if ( is_special_page(mfn_to_page(mfn_add(mfn, i))) ) + { + if ( order ) + return -1; + *ipat = true; + return MTRR_TYPE_WRBACK; + } + } + + if ( direct_mmio ) + return MTRR_TYPE_UNCACHABLE; + + gmtrr_mtype = hvm_get_mem_pinned_cacheattr(d, gfn, order); + if ( gmtrr_mtype >= 0 ) + { + *ipat = true; + return gmtrr_mtype != PAT_TYPE_UC_MINUS ? gmtrr_mtype + : MTRR_TYPE_UNCACHABLE; + } + if ( gmtrr_mtype == -EADDRNOTAVAIL ) + return -1; + + gmtrr_mtype = v ? mtrr_get_type(&v->arch.hvm.mtrr, + gfn_x(gfn) << PAGE_SHIFT, order) + : MTRR_TYPE_WRBACK; + hmtrr_mtype = mtrr_get_type(&mtrr_state, mfn_x(mfn) << PAGE_SHIFT, + order); + if ( gmtrr_mtype < 0 || hmtrr_mtype < 0 ) + return -1; + + /* If both types match we're fine. */ + if ( likely(gmtrr_mtype == hmtrr_mtype) ) + return hmtrr_mtype; + + /* If either type is UC, we have to go with that one. */ + if ( gmtrr_mtype == MTRR_TYPE_UNCACHABLE || + hmtrr_mtype == MTRR_TYPE_UNCACHABLE ) + return MTRR_TYPE_UNCACHABLE; + + /* If either type is WB, we have to go with the other one. */ + if ( gmtrr_mtype == MTRR_TYPE_WRBACK ) + return hmtrr_mtype; + if ( hmtrr_mtype == MTRR_TYPE_WRBACK ) + return gmtrr_mtype; + + /* + * At this point we have disagreeing WC, WT, or WP types. The only + * combination that can be cleanly resolved is WT:WP. The ones involving + * WC need to be converted to UC, both due to the memory ordering + * differences and because WC disallows reads to be cached (WT and WP + * permit this), while WT and WP require writes to go straight to memory + * (WC can buffer them). + */ + if ( (gmtrr_mtype == MTRR_TYPE_WRTHROUGH && + hmtrr_mtype == MTRR_TYPE_WRPROT) || + (gmtrr_mtype == MTRR_TYPE_WRPROT && + hmtrr_mtype == MTRR_TYPE_WRTHROUGH) ) + return MTRR_TYPE_WRPROT; + + return MTRR_TYPE_UNCACHABLE; +} + /* * Resolve deliberately mis-configured (EMT field set to an invalid value) * entries in the page table hierarchy for the given GFN: @@ -519,7 +622,7 @@ static int resolve_misconfig(struct p2m_domain *p2m, unsigned long gfn) if ( level == 0 || is_epte_superpage(&e) ) { - uint8_t ipat = 0; + bool ipat; if ( e.emt != MTRR_NUM_TYPES ) break; @@ -535,7 +638,7 @@ static int resolve_misconfig(struct p2m_domain *p2m, unsigned long gfn) e.emt = 0; if ( !is_epte_valid(&e) || !is_epte_present(&e) ) continue; - e.emt = epte_get_entry_emt(p2m->domain, gfn + i, + e.emt = epte_get_entry_emt(p2m->domain, _gfn(gfn + i), _mfn(e.mfn), 0, &ipat, e.sa_p2mt == p2m_mmio_direct); e.ipat = ipat; @@ -553,7 +656,8 @@ static int resolve_misconfig(struct p2m_domain *p2m, unsigned long gfn) } else { - int emt = epte_get_entry_emt(p2m->domain, gfn, _mfn(e.mfn), + int emt = epte_get_entry_emt(p2m->domain, _gfn(gfn), + _mfn(e.mfn), level * EPT_TABLE_ORDER, &ipat, e.sa_p2mt == p2m_mmio_direct); bool_t recalc = e.recalc; @@ -788,8 +892,8 @@ ept_set_entry(struct p2m_domain *p2m, gfn_t gfn_, mfn_t mfn, if ( mfn_valid(mfn) || p2m_allows_invalid_mfn(p2mt) ) { - uint8_t ipat = 0; - int emt = epte_get_entry_emt(p2m->domain, gfn, mfn, + bool ipat; + int emt = epte_get_entry_emt(p2m->domain, _gfn(gfn), mfn, i * EPT_TABLE_ORDER, &ipat, p2mt == p2m_mmio_direct); diff --git a/xen/include/asm-x86/hvm/vmx/vmx.h b/xen/include/asm-x86/hvm/vmx/vmx.h index 534e9fc221..f668ee1f09 100644 --- a/xen/include/asm-x86/hvm/vmx/vmx.h +++ b/xen/include/asm-x86/hvm/vmx/vmx.h @@ -599,6 +599,8 @@ void ept_p2m_uninit(struct p2m_domain *p2m); void ept_walk_table(struct domain *d, unsigned long gfn); bool_t ept_handle_misconfig(uint64_t gpa); +int epte_get_entry_emt(struct domain *d, gfn_t gfn, mfn_t mfn, + unsigned int order, bool *ipat, bool direct_mmio); void setup_ept_dump(void); void p2m_init_altp2m_ept(struct domain *d, unsigned int i); /* Locate an alternate p2m by its EPTP */ diff --git a/xen/include/asm-x86/mtrr.h b/xen/include/asm-x86/mtrr.h index 24e5de5c22..e0fd1005ce 100644 --- a/xen/include/asm-x86/mtrr.h +++ b/xen/include/asm-x86/mtrr.h @@ -72,12 +72,11 @@ extern int mtrr_add_page(unsigned long base, unsigned long size, unsigned int type, char increment); extern int mtrr_del(int reg, unsigned long base, unsigned long size); extern int mtrr_del_page(int reg, unsigned long base, unsigned long size); +extern int mtrr_get_type(const struct mtrr_state *m, paddr_t pa, + unsigned int order); extern void mtrr_centaur_report_mcr(int mcr, u32 lo, u32 hi); extern u32 get_pat_flags(struct vcpu *v, u32 gl1e_flags, paddr_t gpaddr, paddr_t spaddr, uint8_t gmtrr_mtype); -extern int epte_get_entry_emt(struct domain *, unsigned long gfn, mfn_t mfn, - unsigned int order, uint8_t *ipat, - bool_t direct_mmio); extern unsigned char pat_type_2_pte_flags(unsigned char pat_type); extern int hold_mtrr_updates_on_aps; extern void mtrr_aps_sync_begin(void); -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Thu Jun 17 16:11:15 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 17 Jun 2021 16:11:15 +0000 Received: from list by lists.xenproject.org with outflank-mailman.144095.265281 (Exim 4.92) (envelope-from ) id 1ltubr-0006uA-TW; Thu, 17 Jun 2021 16:11:15 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 144095.265281; Thu, 17 Jun 2021 16:11:15 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1ltubr-0006u3-Qi; Thu, 17 Jun 2021 16:11:15 +0000 Received: by outflank-mailman (input) for mailman id 144095; Thu, 17 Jun 2021 16:11:14 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1ltubq-0006tv-Mb for xen-changelog@lists.xenproject.org; Thu, 17 Jun 2021 16:11:14 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1ltubq-0004wz-Lh for xen-changelog@lists.xenproject.org; Thu, 17 Jun 2021 16:11:14 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1ltubq-0002St-Kk for xen-changelog@lists.xenproject.org; Thu, 17 Jun 2021 16:11:14 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=uNjf0jG0YAqTamp7Zlw4qLVY97atbEVrx7tRzmhj7c8=; b=1aiyV1/a6tgngdMIM9oaspekcn 4mkfkym3p9JKsOzHwwgESLw8eet6WTHWaJZq6O29u9W4rl/RwpSEXwd4/qPWS5dvGf+hP73BvkoZu Rqx3AF3aTj6eW0b9MUAVfLcWSvthNUa7ZXH6/0LTHf2d5bFaPLGPT6Y0wsdxWXbqkER0=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] x86/ept: force WB cache attributes for grant and foreign maps Message-Id: Date: Thu, 17 Jun 2021 16:11:14 +0000 commit 8af4b47f061edf6450f2b0a0a8134fdf1c13b3e5 Author: Roger Pau Monné AuthorDate: Thu Jun 17 18:00:57 2021 +0200 Commit: Jan Beulich CommitDate: Thu Jun 17 18:00:57 2021 +0200 x86/ept: force WB cache attributes for grant and foreign maps Force WB type for grants and foreign pages. Those are usually mapped over unpopulated physical ranges in the p2m, and those ranges would usually be UC in the MTRR state, which is unlikely to be the correct cache attribute. It's also cumbersome (or even impossible) for the guest to be setting the MTRR type for all those mappings as WB, as MTRR ranges are finite. Note that this is not an issue on AMD because WB cache attribute is already set on grants and foreign mappings in the p2m and MTRR types are ignored. Also on AMD Xen cannot force a cache attribute because of the lack of ignore PAT equivalent, so the behavior here slightly diverges between AMD and Intel (or EPT vs NPT/shadow). Signed-off-by: Roger Pau Monné Reviewed-by: Jan Beulich Reviewed-by: Kevin Tian --- xen/arch/x86/hvm/vmx/vmx.c | 2 +- xen/arch/x86/mm/p2m-ept.c | 37 +++++++++++++++++++++++++++++++------ xen/include/asm-x86/hvm/vmx/vmx.h | 2 +- 3 files changed, 33 insertions(+), 8 deletions(-) diff --git a/xen/arch/x86/hvm/vmx/vmx.c b/xen/arch/x86/hvm/vmx/vmx.c index 0d4b47681b..e09b7e3af9 100644 --- a/xen/arch/x86/hvm/vmx/vmx.c +++ b/xen/arch/x86/hvm/vmx/vmx.c @@ -423,7 +423,7 @@ static void domain_creation_finished(struct domain *d) return; ASSERT(epte_get_entry_emt(d, gfn, apic_access_mfn, 0, &ipat, - true) == MTRR_TYPE_WRBACK); + p2m_mmio_direct) == MTRR_TYPE_WRBACK); ASSERT(ipat); if ( set_mmio_p2m_entry(d, gfn, apic_access_mfn, PAGE_ORDER_4K) ) diff --git a/xen/arch/x86/mm/p2m-ept.c b/xen/arch/x86/mm/p2m-ept.c index f1d1d07e92..289001b66b 100644 --- a/xen/arch/x86/mm/p2m-ept.c +++ b/xen/arch/x86/mm/p2m-ept.c @@ -487,7 +487,7 @@ static int ept_invalidate_emt_range(struct p2m_domain *p2m, } int epte_get_entry_emt(struct domain *d, gfn_t gfn, mfn_t mfn, - unsigned int order, bool *ipat, bool direct_mmio) + unsigned int order, bool *ipat, p2m_type_t type) { int gmtrr_mtype, hmtrr_mtype; struct vcpu *v = current; @@ -518,7 +518,8 @@ int epte_get_entry_emt(struct domain *d, gfn_t gfn, mfn_t mfn, return MTRR_TYPE_UNCACHABLE; } - if ( !direct_mmio && !is_iommu_enabled(d) && !cache_flush_permitted(d) ) + if ( type != p2m_mmio_direct && !is_iommu_enabled(d) && + !cache_flush_permitted(d) ) { *ipat = true; return MTRR_TYPE_WRBACK; @@ -535,9 +536,33 @@ int epte_get_entry_emt(struct domain *d, gfn_t gfn, mfn_t mfn, } } - if ( direct_mmio ) + switch ( type ) + { + case p2m_mmio_direct: return MTRR_TYPE_UNCACHABLE; + case p2m_grant_map_ro: + case p2m_grant_map_rw: + case p2m_map_foreign: + /* + * Force WB type for grants and foreign pages. Those are usually mapped + * over unpopulated physical ranges in the p2m, and those would usually + * be UC in the MTRR state, which is unlikely to be the correct cache + * attribute. It's also cumbersome (or even impossible) for the guest + * to be setting the MTRR type for all those mappings as WB, as MTRR + * ranges are finite. + * + * Note that on AMD we cannot force a cache attribute because of the + * lack of ignore PAT equivalent, so the behavior here slightly + * diverges. See p2m_type_to_flags for the AMD attributes. + */ + *ipat = true; + return MTRR_TYPE_WRBACK; + + default: + break; + } + gmtrr_mtype = hvm_get_mem_pinned_cacheattr(d, gfn, order); if ( gmtrr_mtype >= 0 ) { @@ -640,7 +665,7 @@ static int resolve_misconfig(struct p2m_domain *p2m, unsigned long gfn) continue; e.emt = epte_get_entry_emt(p2m->domain, _gfn(gfn + i), _mfn(e.mfn), 0, &ipat, - e.sa_p2mt == p2m_mmio_direct); + e.sa_p2mt); e.ipat = ipat; nt = p2m_recalc_type(e.recalc, e.sa_p2mt, p2m, gfn + i); @@ -659,7 +684,7 @@ static int resolve_misconfig(struct p2m_domain *p2m, unsigned long gfn) int emt = epte_get_entry_emt(p2m->domain, _gfn(gfn), _mfn(e.mfn), level * EPT_TABLE_ORDER, &ipat, - e.sa_p2mt == p2m_mmio_direct); + e.sa_p2mt); bool_t recalc = e.recalc; if ( recalc && p2m_is_changeable(e.sa_p2mt) ) @@ -895,7 +920,7 @@ ept_set_entry(struct p2m_domain *p2m, gfn_t gfn_, mfn_t mfn, bool ipat; int emt = epte_get_entry_emt(p2m->domain, _gfn(gfn), mfn, i * EPT_TABLE_ORDER, &ipat, - p2mt == p2m_mmio_direct); + p2mt); if ( emt >= 0 ) new_entry.emt = emt; diff --git a/xen/include/asm-x86/hvm/vmx/vmx.h b/xen/include/asm-x86/hvm/vmx/vmx.h index f668ee1f09..0deb507490 100644 --- a/xen/include/asm-x86/hvm/vmx/vmx.h +++ b/xen/include/asm-x86/hvm/vmx/vmx.h @@ -600,7 +600,7 @@ void ept_p2m_uninit(struct p2m_domain *p2m); void ept_walk_table(struct domain *d, unsigned long gfn); bool_t ept_handle_misconfig(uint64_t gpa); int epte_get_entry_emt(struct domain *d, gfn_t gfn, mfn_t mfn, - unsigned int order, bool *ipat, bool direct_mmio); + unsigned int order, bool *ipat, p2m_type_t type); void setup_ept_dump(void); void p2m_init_altp2m_ept(struct domain *d, unsigned int i); /* Locate an alternate p2m by its EPTP */ -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Thu Jun 17 18:55:15 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 17 Jun 2021 18:55:15 +0000 Received: from list by lists.xenproject.org with outflank-mailman.144132.265342 (Exim 4.92) (envelope-from ) id 1ltxAR-0006zc-51; Thu, 17 Jun 2021 18:55:07 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 144132.265342; Thu, 17 Jun 2021 18:55:07 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1ltxAR-0006zU-27; Thu, 17 Jun 2021 18:55:07 +0000 Received: by outflank-mailman (input) for mailman id 144132; Thu, 17 Jun 2021 18:55:05 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1ltxAP-0006zO-9p for xen-changelog@lists.xenproject.org; Thu, 17 Jun 2021 18:55:05 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1ltxAP-0007lw-6O for xen-changelog@lists.xenproject.org; Thu, 17 Jun 2021 18:55:05 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1ltxAP-0003OS-4w for xen-changelog@lists.xenproject.org; Thu, 17 Jun 2021 18:55:05 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=OB7npx3IV6hrvXGd3CIIc2fFdCIUqvw9et1Hr3dRtzc=; b=rdGpROpKwi7Kc9ldfMoFftgpb6 M8p2sHV0VoCUlMCpghEECV5ewXDMQ+upPGa/zKEzMGPmSkNSvoU7mYz/IonGtIVofZGEgiI6pHqIS 93LTpfSFYOQDY9uOgOYSocb5wmu7pGpZFHWWvtyQyQHzvpvoKJVDLIK3PBAF2AKNTCHs=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.15] x86/tsx: Minor cleanup and improvements Message-Id: Date: Thu, 17 Jun 2021 18:55:05 +0000 commit 0aabeb9293b6a3bcd98d692de2cd3d16f4f41062 Author: Andrew Cooper AuthorDate: Tue May 18 14:53:56 2021 +0100 Commit: Andrew Cooper CommitDate: Thu Jun 17 14:36:15 2021 +0100 x86/tsx: Minor cleanup and improvements * Introduce cpu_has_arch_caps and replace boot_cpu_has(X86_FEATURE_ARCH_CAPS) * Read CPUID data into the appropriate boot_cpu_data.x86_capability[] element, as subsequent changes are going to need more cpu_has_* logic. * Use the hi/lo MSR helpers, which substantially improves code generation. No practical change. Signed-off-by: Andrew Cooper Acked-by: Jan Beulich Reviewed-by: Roger Pau Monné (cherry picked from commit 3670abcaf0324f2aedba0c4dc7939072b27efa1d) --- xen/arch/x86/cpuid.c | 2 +- xen/arch/x86/hvm/vmx/vmx.c | 2 +- xen/arch/x86/msr.c | 2 +- xen/arch/x86/spec_ctrl.c | 2 +- xen/arch/x86/tsx.c | 21 ++++++++++++--------- xen/include/asm-x86/cpufeature.h | 1 + 6 files changed, 17 insertions(+), 13 deletions(-) diff --git a/xen/arch/x86/cpuid.c b/xen/arch/x86/cpuid.c index 7ac6636efc..431f6fb173 100644 --- a/xen/arch/x86/cpuid.c +++ b/xen/arch/x86/cpuid.c @@ -713,7 +713,7 @@ int init_domain_cpuid_policy(struct domain *d) * so dom0 can turn off workarounds as appropriate. Temporary, until the * domain policy logic gains a better understanding of MSRs. */ - if ( is_hardware_domain(d) && boot_cpu_has(X86_FEATURE_ARCH_CAPS) ) + if ( is_hardware_domain(d) && cpu_has_arch_caps ) p->feat.arch_caps = true; d->arch.cpuid = p; diff --git a/xen/arch/x86/hvm/vmx/vmx.c b/xen/arch/x86/hvm/vmx/vmx.c index dde4f3b70d..fd7b193058 100644 --- a/xen/arch/x86/hvm/vmx/vmx.c +++ b/xen/arch/x86/hvm/vmx/vmx.c @@ -2573,7 +2573,7 @@ static bool __init has_if_pschange_mc(void) if ( cpu_has_hypervisor ) return false; - if ( boot_cpu_has(X86_FEATURE_ARCH_CAPS) ) + if ( cpu_has_arch_caps ) rdmsrl(MSR_ARCH_CAPABILITIES, caps); if ( caps & ARCH_CAPS_IF_PSCHANGE_MC_NO ) diff --git a/xen/arch/x86/msr.c b/xen/arch/x86/msr.c index c3a988bd11..374f92b2c5 100644 --- a/xen/arch/x86/msr.c +++ b/xen/arch/x86/msr.c @@ -136,7 +136,7 @@ int init_domain_msr_policy(struct domain *d) * so dom0 can turn off workarounds as appropriate. Temporary, until the * domain policy logic gains a better understanding of MSRs. */ - if ( is_hardware_domain(d) && boot_cpu_has(X86_FEATURE_ARCH_CAPS) ) + if ( is_hardware_domain(d) && cpu_has_arch_caps ) { uint64_t val; diff --git a/xen/arch/x86/spec_ctrl.c b/xen/arch/x86/spec_ctrl.c index bea4c3e8b4..aa85a0be46 100644 --- a/xen/arch/x86/spec_ctrl.c +++ b/xen/arch/x86/spec_ctrl.c @@ -885,7 +885,7 @@ void __init init_speculation_mitigations(void) bool cpu_has_bug_taa; uint64_t caps = 0; - if ( boot_cpu_has(X86_FEATURE_ARCH_CAPS) ) + if ( cpu_has_arch_caps ) rdmsrl(MSR_ARCH_CAPABILITIES, caps); hw_smt_enabled = check_smt_enabled(); diff --git a/xen/arch/x86/tsx.c b/xen/arch/x86/tsx.c index e09e819dce..98ecb71a4a 100644 --- a/xen/arch/x86/tsx.c +++ b/xen/arch/x86/tsx.c @@ -34,15 +34,18 @@ void tsx_init(void) { /* * This function is first called between microcode being loaded, and CPUID - * being scanned generally. Calculate from raw data whether MSR_TSX_CTRL - * is available. + * being scanned generally. Read into boot_cpu_data.x86_capability[] for + * the cpu_has_* bits we care about using here. */ if ( unlikely(cpu_has_tsx_ctrl < 0) ) { uint64_t caps = 0; - if ( boot_cpu_data.cpuid_level >= 7 && - (cpuid_count_edx(7, 0) & cpufeat_mask(X86_FEATURE_ARCH_CAPS)) ) + if ( boot_cpu_data.cpuid_level >= 7 ) + boot_cpu_data.x86_capability[cpufeat_word(X86_FEATURE_ARCH_CAPS)] + = cpuid_count_edx(7, 0); + + if ( cpu_has_arch_caps ) rdmsrl(MSR_ARCH_CAPABILITIES, caps); cpu_has_tsx_ctrl = !!(caps & ARCH_CAPS_TSX_CTRL); @@ -74,18 +77,18 @@ void tsx_init(void) if ( cpu_has_tsx_ctrl ) { - uint64_t val; + uint32_t hi, lo; - rdmsrl(MSR_TSX_CTRL, val); + rdmsr(MSR_TSX_CTRL, lo, hi); /* Check bottom bit only. Higher bits are various sentinels. */ rtm_disabled = !(opt_tsx & 1); - val &= ~(TSX_CTRL_RTM_DISABLE | TSX_CTRL_CPUID_CLEAR); + lo &= ~(TSX_CTRL_RTM_DISABLE | TSX_CTRL_CPUID_CLEAR); if ( rtm_disabled ) - val |= TSX_CTRL_RTM_DISABLE | TSX_CTRL_CPUID_CLEAR; + lo |= TSX_CTRL_RTM_DISABLE | TSX_CTRL_CPUID_CLEAR; - wrmsrl(MSR_TSX_CTRL, val); + wrmsr(MSR_TSX_CTRL, lo, hi); } else if ( opt_tsx >= 0 ) printk_once(XENLOG_WARNING diff --git a/xen/include/asm-x86/cpufeature.h b/xen/include/asm-x86/cpufeature.h index 33b2257888..9f5ae3aa0d 100644 --- a/xen/include/asm-x86/cpufeature.h +++ b/xen/include/asm-x86/cpufeature.h @@ -133,6 +133,7 @@ #define cpu_has_avx512_vp2intersect boot_cpu_has(X86_FEATURE_AVX512_VP2INTERSECT) #define cpu_has_tsx_force_abort boot_cpu_has(X86_FEATURE_TSX_FORCE_ABORT) #define cpu_has_serialize boot_cpu_has(X86_FEATURE_SERIALIZE) +#define cpu_has_arch_caps boot_cpu_has(X86_FEATURE_ARCH_CAPS) /* CPUID level 0x00000007:1.eax */ #define cpu_has_avx_vnni boot_cpu_has(X86_FEATURE_AVX_VNNI) -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.15 From xen-changelog-bounces@lists.xenproject.org Thu Jun 17 18:55:17 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 17 Jun 2021 18:55:17 +0000 Received: from list by lists.xenproject.org with outflank-mailman.144133.265347 (Exim 4.92) (envelope-from ) id 1ltxAb-00071l-8F; Thu, 17 Jun 2021 18:55:17 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 144133.265347; Thu, 17 Jun 2021 18:55:17 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1ltxAb-00071d-50; Thu, 17 Jun 2021 18:55:17 +0000 Received: by outflank-mailman (input) for mailman id 144133; Thu, 17 Jun 2021 18:55:15 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1ltxAZ-00071P-Ai for xen-changelog@lists.xenproject.org; Thu, 17 Jun 2021 18:55:15 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1ltxAZ-0007m0-9q for xen-changelog@lists.xenproject.org; Thu, 17 Jun 2021 18:55:15 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1ltxAZ-0003PW-8u for xen-changelog@lists.xenproject.org; Thu, 17 Jun 2021 18:55:15 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=sMF9cC9qYVJMLZZZOEnYePXIX0BtZ+2A4br0WvDN4CI=; b=EdxO4X0lC7+VGz2Ox4wmyIT+3R s3eH9EV6cZWFVCMsGZwXUMgFlTc4eAI21IFmg9w6H4QtDo5ft/iy7vsbR7h4TtxEnqbwGoMJiQCqL nLiPkbaphYd8GbejXbXh3eOglocQtnbC5mF/w9Sy6QUgQI6FpzRbc9zqHEtFB/t4cscc=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.15] x86/tsx: Deprecate vpmu=rtm-abort and use tsx= instead Message-Id: Date: Thu, 17 Jun 2021 18:55:15 +0000 commit 796d4056691f2b6d6f71b799173e8352eb6d73ea Author: Andrew Cooper AuthorDate: Thu May 20 19:16:11 2021 +0100 Commit: Andrew Cooper CommitDate: Thu Jun 17 14:36:15 2021 +0100 x86/tsx: Deprecate vpmu=rtm-abort and use tsx= instead This reuses the rtm_disable infrastructure, so CPUID derivation works properly when TSX is disabled in favour of working PCR3. vpmu= is not a supported feature, and having this functionality under tsx= centralises all TSX handling. Signed-off-by: Andrew Cooper Acked-by: Jan Beulich Reviewed-by: Roger Pau Monné (cherry picked from commit 9fdcf851689cb2a9501d3947cb5d767d9c7797e8) --- docs/misc/xen-command-line.pandoc | 40 +++++++++++++++--------------- xen/arch/x86/cpu/intel.c | 3 --- xen/arch/x86/cpu/vpmu.c | 4 +-- xen/arch/x86/tsx.c | 51 +++++++++++++++++++++++++++++++++++++-- xen/include/asm-x86/vpmu.h | 1 - 5 files changed, 70 insertions(+), 29 deletions(-) diff --git a/docs/misc/xen-command-line.pandoc b/docs/misc/xen-command-line.pandoc index c32a397a12..1fae872626 100644 --- a/docs/misc/xen-command-line.pandoc +++ b/docs/misc/xen-command-line.pandoc @@ -2296,14 +2296,21 @@ pages) must also be specified via the tbuf_size parameter. Controls for the use of Transactional Synchronization eXtensions. -On Intel parts released in Q3 2019 (with updated microcode), and future parts, -a control has been introduced which allows TSX to be turned off. +Several microcode updates are relevant: -On systems with the ability to turn TSX off, this boolean offers system wide -control of whether TSX is enabled or disabled. + * March 2019, fixing the TSX memory ordering errata on all TSX-enabled CPUs + to date. Introduced MSR_TSX_FORCE_ABORT on SKL/SKX/KBL/WHL/CFL parts. The + errata workaround uses Performance Counter 3, so the user can select + between working TSX and working perfcounters. + + * November 2019, fixing the TSX Async Abort speculative vulnerability. + Introduced MSR_TSX_CTRL on all TSX-enabled MDS_NO parts to date, + CLX/WHL-R/CFL-R, with the controls becoming architectural moving forward + and formally retiring HLE from the architecture. The user can disable TSX + to mitigate TAA, and elect to hide the HLE/RTM CPUID bits. -On parts vulnerable to CVE-2019-11135 / TSX Asynchronous Abort, the following -logic applies: +On systems with the ability to configure TSX, this boolean offers system wide +control of whether TSX is enabled or disabled. * An explicit `tsx=` choice is honoured, even if it is `true` and would result in a vulnerable system. @@ -2311,10 +2318,14 @@ logic applies: * When no explicit `tsx=` choice is given, parts vulnerable to TAA will be mitigated by disabling TSX, as this is the lowest overhead option. - * If the use of TSX is important, the more expensive TAA mitigations can be + If the use of TSX is important, the more expensive TAA mitigations can be opted in to with `smt=0 spec-ctrl=md-clear`, at which point TSX will remain active by default. + * When no explicit `tsx=` option is given, parts susceptible to the memory + ordering errata default to `true` to enable working TSX. Alternatively, + selecting `tsx=0` will disable TSX and restore PCR3 to a working state. + ### ucode > `= List of [ | scan=, nmi=, allow-same= ]` @@ -2456,20 +2467,7 @@ provide access to a wealth of low level processor information. * The `arch` option allows access to the pre-defined architectural events. -* The `rtm-abort` boolean controls a trade-off between working Restricted - Transactional Memory, and working performance counters. - - All processors released to date (Q1 2019) supporting Transactional Memory - Extensions suffer an erratum which has been addressed in microcode. - - Processors based on the Skylake microarchitecture with up-to-date - microcode internally use performance counter 3 to work around the erratum. - A consequence is that the counter gets reprogrammed whenever an `XBEGIN` - instruction is executed. - - An alternative mode exists where PCR3 behaves as before, at the cost of - `XBEGIN` unconditionally aborting. Enabling `rtm-abort` mode will - activate this alternative mode. +* The `rtm-abort` boolean has been superseded. Use `tsx=0` instead. *Warning:* As the virtualisation is not 100% safe, don't use the vpmu flag on diff --git a/xen/arch/x86/cpu/intel.c b/xen/arch/x86/cpu/intel.c index 37439071d9..abf8e206d7 100644 --- a/xen/arch/x86/cpu/intel.c +++ b/xen/arch/x86/cpu/intel.c @@ -356,9 +356,6 @@ static void Intel_errata_workarounds(struct cpuinfo_x86 *c) (c->x86_model == 29 || c->x86_model == 46 || c->x86_model == 47)) __set_bit(X86_FEATURE_CLFLUSH_MONITOR, c->x86_capability); - if (cpu_has_tsx_force_abort && opt_rtm_abort) - wrmsrl(MSR_TSX_FORCE_ABORT, TSX_FORCE_ABORT_RTM); - probe_c3_errata(c); } diff --git a/xen/arch/x86/cpu/vpmu.c b/xen/arch/x86/cpu/vpmu.c index ab667361d3..fb1b296a6c 100644 --- a/xen/arch/x86/cpu/vpmu.c +++ b/xen/arch/x86/cpu/vpmu.c @@ -47,7 +47,6 @@ CHECK_pmu_params; static unsigned int __read_mostly opt_vpmu_enabled; unsigned int __read_mostly vpmu_mode = XENPMU_MODE_OFF; unsigned int __read_mostly vpmu_features = 0; -bool __read_mostly opt_rtm_abort; static DEFINE_SPINLOCK(vpmu_lock); static unsigned vpmu_count; @@ -77,7 +76,8 @@ static int __init parse_vpmu_params(const char *s) else if ( !cmdline_strcmp(s, "arch") ) vpmu_features |= XENPMU_FEATURE_ARCH_ONLY; else if ( (val = parse_boolean("rtm-abort", s, ss)) >= 0 ) - opt_rtm_abort = val; + printk(XENLOG_WARNING + "'rtm-abort=' superseded. Use 'tsx=' instead\n"); else rc = -EINVAL; diff --git a/xen/arch/x86/tsx.c b/xen/arch/x86/tsx.c index 98ecb71a4a..338191df7f 100644 --- a/xen/arch/x86/tsx.c +++ b/xen/arch/x86/tsx.c @@ -6,7 +6,9 @@ * Valid values: * 1 => Explicit tsx=1 * 0 => Explicit tsx=0 - * -1 => Default, implicit tsx=1, may change to 0 to mitigate TAA + * -1 => Default, altered to 0/1 (if unspecified) by: + * - TAA heuristics/settings for speculative safety + * - "TSX vs PCR3" select for TSX memory ordering safety * -3 => Implicit tsx=1 (feed-through from spec-ctrl=0) * * This is arranged such that the bottom bit encodes whether TSX is actually @@ -50,6 +52,26 @@ void tsx_init(void) cpu_has_tsx_ctrl = !!(caps & ARCH_CAPS_TSX_CTRL); + if ( cpu_has_tsx_force_abort ) + { + /* + * On an early TSX-enable Skylake part subject to the memory + * ordering erratum, with at least the March 2019 microcode. + */ + + /* + * If no explicit tsx= option is provided, pick a default. + * + * This deliberately overrides the implicit opt_tsx=-3 from + * `spec-ctrl=0` because: + * - parse_spec_ctrl() ran before any CPU details where know. + * - We now know we're running on a CPU not affected by TAA (as + * TSX_FORCE_ABORT is enumerated). + */ + if ( opt_tsx < 0 ) + opt_tsx = 1; + } + /* * The TSX features (HLE/RTM) are handled specially. They both * enumerate features but, on certain parts, have mechanisms to be @@ -75,6 +97,12 @@ void tsx_init(void) } } + /* + * Note: MSR_TSX_CTRL is enumerated on TSX-enabled MDS_NO and later parts. + * MSR_TSX_FORCE_ABORT is enumerated on TSX-enabled pre-MDS_NO Skylake + * parts only. The two features are on a disjoint set of CPUs, and not + * offered to guests by hypervisors. + */ if ( cpu_has_tsx_ctrl ) { uint32_t hi, lo; @@ -90,9 +118,28 @@ void tsx_init(void) wrmsr(MSR_TSX_CTRL, lo, hi); } + else if ( cpu_has_tsx_force_abort ) + { + /* + * On an early TSX-enable Skylake part subject to the memory ordering + * erratum, with at least the March 2019 microcode. + */ + uint32_t hi, lo; + + rdmsr(MSR_TSX_FORCE_ABORT, lo, hi); + + /* Check bottom bit only. Higher bits are various sentinels. */ + rtm_disabled = !(opt_tsx & 1); + + lo &= ~TSX_FORCE_ABORT_RTM; + if ( rtm_disabled ) + lo |= TSX_FORCE_ABORT_RTM; + + wrmsr(MSR_TSX_FORCE_ABORT, lo, hi); + } else if ( opt_tsx >= 0 ) printk_once(XENLOG_WARNING - "MSR_TSX_CTRL not available - Ignoring tsx= setting\n"); + "TSX controls not available - Ignoring tsx= setting\n"); } /* diff --git a/xen/include/asm-x86/vpmu.h b/xen/include/asm-x86/vpmu.h index 55f85ba00f..4b0a6ba3da 100644 --- a/xen/include/asm-x86/vpmu.h +++ b/xen/include/asm-x86/vpmu.h @@ -126,7 +126,6 @@ static inline int vpmu_do_rdmsr(unsigned int msr, uint64_t *msr_content) extern unsigned int vpmu_mode; extern unsigned int vpmu_features; -extern bool opt_rtm_abort; /* Context switch */ static inline void vpmu_switch_from(struct vcpu *prev) -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.15 From xen-changelog-bounces@lists.xenproject.org Thu Jun 17 18:55:27 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 17 Jun 2021 18:55:27 +0000 Received: from list by lists.xenproject.org with outflank-mailman.144134.265350 (Exim 4.92) (envelope-from ) id 1ltxAl-00074t-9K; Thu, 17 Jun 2021 18:55:27 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 144134.265350; Thu, 17 Jun 2021 18:55:27 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1ltxAl-00074m-6U; Thu, 17 Jun 2021 18:55:27 +0000 Received: by outflank-mailman (input) for mailman id 144134; Thu, 17 Jun 2021 18:55:25 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1ltxAj-00074U-Dj for xen-changelog@lists.xenproject.org; Thu, 17 Jun 2021 18:55:25 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1ltxAj-0007mU-Cz for xen-changelog@lists.xenproject.org; Thu, 17 Jun 2021 18:55:25 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1ltxAj-0003QU-C3 for xen-changelog@lists.xenproject.org; Thu, 17 Jun 2021 18:55:25 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=F17I93tKdHT9tTMOSV0LmJ9NGv/WYAzRZ13JTBrraJg=; b=2eHJ0a6HlmyYgG//NiKzhBFn9+ 7qSuZYIxB5Lq3JowP9RlvObXGZVU3G5nXAdUTfV7mHfcG+xbjR/EtA2S1p+1MfUl1PHHaCWKa36rj 4Cm8m6LHaOrQg7AeJ0hhYVCiz4WdY0WNF2uTKcfUdzQGJaoKERdn4Ki+ZvKwRY0734u0=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.15] x86/cpuid: Fix HLE and RTM handling (again) Message-Id: Date: Thu, 17 Jun 2021 18:55:25 +0000 commit 4586e6443a8b963283bdec1d72d42fae061ebc49 Author: Andrew Cooper AuthorDate: Mon Jun 7 13:25:09 2021 +0100 Commit: Andrew Cooper CommitDate: Thu Jun 17 14:36:15 2021 +0100 x86/cpuid: Fix HLE and RTM handling (again) For reasons which are my fault, but I don't recall why, the FDP_EXCP_ONLY/NO_FPU_SEL adjustment uses the whole special_features[] array element, not the two relevant bits. HLE and RTM were recently added to the list of special features, causing them to be always set in guest view, irrespective of the toolstacks choice on the matter. Rewrite the logic to refer to the features specifically, rather than relying on the contents of the special_features[] array. Fixes: 8fe24090d9 ("x86/cpuid: Rework HLE and RTM handling") Reported-by: Edwin Török Signed-off-by: Andrew Cooper Reviewed-by: Jan Beulich (cherry picked from commit 60fa12dbf1d4d2c4ffe1ef34b495b24aa7e41aa0) --- xen/arch/x86/cpuid.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/xen/arch/x86/cpuid.c b/xen/arch/x86/cpuid.c index 431f6fb173..eefcde902e 100644 --- a/xen/arch/x86/cpuid.c +++ b/xen/arch/x86/cpuid.c @@ -640,9 +640,11 @@ void recalculate_cpuid_policy(struct domain *d) sanitise_featureset(fs); /* Fold host's FDP_EXCP_ONLY and NO_FPU_SEL into guest's view. */ - fs[FEATURESET_7b0] &= ~special_features[FEATURESET_7b0]; + fs[FEATURESET_7b0] &= ~(cpufeat_mask(X86_FEATURE_FDP_EXCP_ONLY) | + cpufeat_mask(X86_FEATURE_NO_FPU_SEL)); fs[FEATURESET_7b0] |= (host_cpuid_policy.feat._7b0 & - special_features[FEATURESET_7b0]); + (cpufeat_mask(X86_FEATURE_FDP_EXCP_ONLY) | + cpufeat_mask(X86_FEATURE_NO_FPU_SEL))); cpuid_featureset_to_policy(fs, p); -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.15 From xen-changelog-bounces@lists.xenproject.org Thu Jun 17 18:55:37 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 17 Jun 2021 18:55:37 +0000 Received: from list by lists.xenproject.org with outflank-mailman.144135.265354 (Exim 4.92) (envelope-from ) id 1ltxAv-00078s-B1; Thu, 17 Jun 2021 18:55:37 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 144135.265354; Thu, 17 Jun 2021 18:55:37 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1ltxAv-00078k-85; Thu, 17 Jun 2021 18:55:37 +0000 Received: by outflank-mailman (input) for mailman id 144135; Thu, 17 Jun 2021 18:55:35 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1ltxAt-00078K-Ht for xen-changelog@lists.xenproject.org; Thu, 17 Jun 2021 18:55:35 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1ltxAt-0007mp-H3 for xen-changelog@lists.xenproject.org; Thu, 17 Jun 2021 18:55:35 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1ltxAt-0003RF-Fb for xen-changelog@lists.xenproject.org; Thu, 17 Jun 2021 18:55:35 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=ZW1irFZkIc89LlKcjERYXYjt1seUMoh6TzsUWWBGkfk=; b=AyI9Qv/GMdsvM6nanmVbMhXxsW pYSrIg8o5xtFUFfi0qd0FWbrkakHJZlutDWc9HhapxQ6ilxin70IrS4JdszOK+cU3ReXnVeLn2cgK oSg2xVpBtbCDJuwfS/kf29N4iizdHVRCV9M5UG/bWUPMbxxbvX5lBl3dpCWN7Deu/T8M=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.15] x86/tsx: Cope with TSX deprecation on SKL/KBL/CFL/WHL Message-Id: Date: Thu, 17 Jun 2021 18:55:35 +0000 commit ec457ac2a29279e8cd91745c410b0f49d5e8f1ff Author: Andrew Cooper AuthorDate: Wed Sep 16 16:15:52 2020 +0100 Commit: Andrew Cooper CommitDate: Thu Jun 17 14:36:15 2021 +0100 x86/tsx: Cope with TSX deprecation on SKL/KBL/CFL/WHL The June 2021 microcode is formally de-featuring TSX on the older Skylake client CPUs. The workaround from the March 2019 microcode is being dropped, and replaced with additions to MSR_TSX_FORCE_ABORT to hide the HLE/RTM CPUID bits. With this microcode in place, TSX is disabled by default on these CPUs. Backwards compatibility is provided in the same way as for TAA - RTM force aborts, rather than suffering #UD, and the CPUID bits can be hidden to recover performance. Signed-off-by: Andrew Cooper Reviewed-by: Jan Beulich (cherry picked from commit 3e09045991cde360432bc7437103f8f8a6699359) --- docs/misc/xen-command-line.pandoc | 13 +++++ tools/misc/xen-cpuid.c | 2 +- xen/arch/x86/tsx.c | 76 +++++++++++++++++++++++++++-- xen/include/asm-x86/cpufeature.h | 1 + xen/include/asm-x86/msr-index.h | 2 + xen/include/public/arch-x86/cpufeatureset.h | 1 + 6 files changed, 89 insertions(+), 6 deletions(-) diff --git a/docs/misc/xen-command-line.pandoc b/docs/misc/xen-command-line.pandoc index 1fae872626..3ece83a427 100644 --- a/docs/misc/xen-command-line.pandoc +++ b/docs/misc/xen-command-line.pandoc @@ -2309,6 +2309,12 @@ Several microcode updates are relevant: and formally retiring HLE from the architecture. The user can disable TSX to mitigate TAA, and elect to hide the HLE/RTM CPUID bits. + * June 2021, removing the workaround for March 2019 on client CPUs and + formally de-featured TSX on SKL/KBL/WHL/CFL (Note: SKX still retains the + March 2019 fix). Introduced the ability to hide the HLE/RTM CPUID bits. + PCR3 works fine, and TSX is disabled by default, but the user can re-enable + TSX at their own risk, accepting that the memory order erratum is unfixed. + On systems with the ability to configure TSX, this boolean offers system wide control of whether TSX is enabled or disabled. @@ -2326,6 +2332,13 @@ control of whether TSX is enabled or disabled. ordering errata default to `true` to enable working TSX. Alternatively, selecting `tsx=0` will disable TSX and restore PCR3 to a working state. + SKX and SKL/KBL/WHL/CFL on pre-June 2021 microcode default to `true`. + Alternatively, selecting `tsx=0` will disable TSX and restore PCR3 to a + working state. + + SKL/KBL/WHL/CFL on the June 2021 microcode or later default to `false`. + Alternatively, selecting `tsx=1` will re-enable TSX at the users own risk. + ### ucode > `= List of [ | scan=, nmi=, allow-same= ]` diff --git a/tools/misc/xen-cpuid.c b/tools/misc/xen-cpuid.c index 2d04162d8d..531d56d9c9 100644 --- a/tools/misc/xen-cpuid.c +++ b/tools/misc/xen-cpuid.c @@ -161,7 +161,7 @@ static const char *const str_7d0[32] = [ 4] = "fsrm", [ 8] = "avx512-vp2intersect", [ 9] = "srbds-ctrl", - [10] = "md-clear", + [10] = "md-clear", [11] = "rtm-always-abort", /* 12 */ [13] = "tsx-force-abort", [14] = "serialize", diff --git a/xen/arch/x86/tsx.c b/xen/arch/x86/tsx.c index 338191df7f..88adf08c49 100644 --- a/xen/arch/x86/tsx.c +++ b/xen/arch/x86/tsx.c @@ -59,6 +59,38 @@ void tsx_init(void) * ordering erratum, with at least the March 2019 microcode. */ + /* + * Probe for the June 2021 microcode which de-features TSX on + * client parts. (Note - this is a subset of parts impacted by + * the memory ordering errata.) + * + * RTM_ALWAYS_ABORT enumerates the new functionality, but is also + * read as zero if TSX_FORCE_ABORT.ENABLE_RTM has been set before + * we run. + * + * Undo this behaviour in Xen's view of the world. + */ + bool has_rtm_always_abort = cpu_has_rtm_always_abort; + + if ( !has_rtm_always_abort ) + { + uint64_t val; + + rdmsrl(MSR_TSX_FORCE_ABORT, val); + + if ( val & TSX_ENABLE_RTM ) + has_rtm_always_abort = true; + } + + /* + * Always force RTM_ALWAYS_ABORT, even if it currently visible. + * If the user explicitly opts to enable TSX, we'll set + * TSX_FORCE_ABORT.ENABLE_RTM and cause RTM_ALWAYS_ABORT to be + * hidden from the general CPUID scan later. + */ + if ( has_rtm_always_abort ) + setup_force_cpu_cap(X86_FEATURE_RTM_ALWAYS_ABORT); + /* * If no explicit tsx= option is provided, pick a default. * @@ -67,9 +99,16 @@ void tsx_init(void) * - parse_spec_ctrl() ran before any CPU details where know. * - We now know we're running on a CPU not affected by TAA (as * TSX_FORCE_ABORT is enumerated). + * - When RTM_ALWAYS_ABORT is enumerated, TSX malfunctions, so we + * only ever want it enabled by explicit user choice. + * + * Without RTM_ALWAYS_ABORT, leave TSX active. In particular, + * this includes SKX where TSX is still supported. + * + * With RTM_ALWAYS_ABORT, disable TSX. */ if ( opt_tsx < 0 ) - opt_tsx = 1; + opt_tsx = !cpu_has_rtm_always_abort; } /* @@ -90,7 +129,7 @@ void tsx_init(void) * Force the features to be visible in Xen's view if we see any of the * infrastructure capable of hiding them. */ - if ( cpu_has_tsx_ctrl ) + if ( cpu_has_tsx_ctrl || cpu_has_tsx_force_abort ) { setup_force_cpu_cap(X86_FEATURE_HLE); setup_force_cpu_cap(X86_FEATURE_RTM); @@ -131,9 +170,36 @@ void tsx_init(void) /* Check bottom bit only. Higher bits are various sentinels. */ rtm_disabled = !(opt_tsx & 1); - lo &= ~TSX_FORCE_ABORT_RTM; - if ( rtm_disabled ) - lo |= TSX_FORCE_ABORT_RTM; + lo &= ~(TSX_FORCE_ABORT_RTM | TSX_CPUID_CLEAR | TSX_ENABLE_RTM); + + if ( cpu_has_rtm_always_abort ) + { + /* + * June 2021 microcode, on a client part with TSX de-featured: + * - There are no mitigations for the TSX memory ordering errata. + * - Performance counter 3 works. (I.e. it isn't being used by + * microcode to work around the memory ordering errata.) + * - TSX_FORCE_ABORT.FORCE_ABORT_RTM is fixed read1/write-discard. + * - TSX_FORCE_ABORT.TSX_CPUID_CLEAR can be used to hide the + * HLE/RTM CPUID bits. + * - TSX_FORCE_ABORT.ENABLE_RTM may be used to opt in to + * re-enabling RTM, at the users own risk. + */ + lo |= rtm_disabled ? TSX_CPUID_CLEAR : TSX_ENABLE_RTM; + } + else + { + /* + * Either a server part where TSX isn't de-featured, or pre-June + * 2021 microcode: + * - By default, the TSX memory ordering errata is worked around + * in microcode at the cost of Performance Counter 3. + * - "Working TSX" vs "Working PCR3" can be selected by way of + * setting TSX_FORCE_ABORT.FORCE_ABORT_RTM. + */ + if ( rtm_disabled ) + lo |= TSX_FORCE_ABORT_RTM; + } wrmsr(MSR_TSX_FORCE_ABORT, lo, hi); } diff --git a/xen/include/asm-x86/cpufeature.h b/xen/include/asm-x86/cpufeature.h index 9f5ae3aa0d..a539a4bacd 100644 --- a/xen/include/asm-x86/cpufeature.h +++ b/xen/include/asm-x86/cpufeature.h @@ -131,6 +131,7 @@ #define cpu_has_avx512_4vnniw boot_cpu_has(X86_FEATURE_AVX512_4VNNIW) #define cpu_has_avx512_4fmaps boot_cpu_has(X86_FEATURE_AVX512_4FMAPS) #define cpu_has_avx512_vp2intersect boot_cpu_has(X86_FEATURE_AVX512_VP2INTERSECT) +#define cpu_has_rtm_always_abort boot_cpu_has(X86_FEATURE_RTM_ALWAYS_ABORT) #define cpu_has_tsx_force_abort boot_cpu_has(X86_FEATURE_TSX_FORCE_ABORT) #define cpu_has_serialize boot_cpu_has(X86_FEATURE_SERIALIZE) #define cpu_has_arch_caps boot_cpu_has(X86_FEATURE_ARCH_CAPS) diff --git a/xen/include/asm-x86/msr-index.h b/xen/include/asm-x86/msr-index.h index bd3a3a1e7f..9a772c12b8 100644 --- a/xen/include/asm-x86/msr-index.h +++ b/xen/include/asm-x86/msr-index.h @@ -61,6 +61,8 @@ #define MSR_TSX_FORCE_ABORT 0x0000010f #define TSX_FORCE_ABORT_RTM (_AC(1, ULL) << 0) +#define TSX_CPUID_CLEAR (_AC(1, ULL) << 1) +#define TSX_ENABLE_RTM (_AC(1, ULL) << 2) #define MSR_TSX_CTRL 0x00000122 #define TSX_CTRL_RTM_DISABLE (_AC(1, ULL) << 0) diff --git a/xen/include/public/arch-x86/cpufeatureset.h b/xen/include/public/arch-x86/cpufeatureset.h index 6477f75505..b4f3d622a0 100644 --- a/xen/include/public/arch-x86/cpufeatureset.h +++ b/xen/include/public/arch-x86/cpufeatureset.h @@ -263,6 +263,7 @@ XEN_CPUFEATURE(AVX512_4FMAPS, 9*32+ 3) /*A AVX512 Multiply Accumulation Single XEN_CPUFEATURE(AVX512_VP2INTERSECT, 9*32+8) /*a VP2INTERSECT{D,Q} insns */ XEN_CPUFEATURE(SRBDS_CTRL, 9*32+ 9) /* MSR_MCU_OPT_CTRL and RNGDS_MITG_DIS. */ XEN_CPUFEATURE(MD_CLEAR, 9*32+10) /*A VERW clears microarchitectural buffers */ +XEN_CPUFEATURE(RTM_ALWAYS_ABORT, 9*32+11) /*! June 2021 TSX defeaturing in microcode. */ XEN_CPUFEATURE(TSX_FORCE_ABORT, 9*32+13) /* MSR_TSX_FORCE_ABORT.RTM_ABORT */ XEN_CPUFEATURE(SERIALIZE, 9*32+14) /*a SERIALIZE insn */ XEN_CPUFEATURE(CET_IBT, 9*32+20) /* CET - Indirect Branch Tracking */ -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.15 From xen-changelog-bounces@lists.xenproject.org Thu Jun 17 18:55:47 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 17 Jun 2021 18:55:47 +0000 Received: from list by lists.xenproject.org with outflank-mailman.144136.265359 (Exim 4.92) (envelope-from ) id 1ltxB5-0007Br-Cs; Thu, 17 Jun 2021 18:55:47 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 144136.265359; Thu, 17 Jun 2021 18:55:47 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1ltxB5-0007Bj-9h; Thu, 17 Jun 2021 18:55:47 +0000 Received: by outflank-mailman (input) for mailman id 144136; Thu, 17 Jun 2021 18:55:46 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1ltxB4-0007BX-Br for xen-changelog@lists.xenproject.org; Thu, 17 Jun 2021 18:55:46 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1ltxB4-0007n1-B1 for xen-changelog@lists.xenproject.org; Thu, 17 Jun 2021 18:55:46 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1ltxB4-0003Sd-AC for xen-changelog@lists.xenproject.org; Thu, 17 Jun 2021 18:55:46 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=iIK3IcxwdolmTu9F8WYqxWePNYKz+hNJQWIY493EZTM=; b=CoxTZ+5MKf8yFzv4ja0c+s34J5 yPS3NzwubLLvzMKntFWq8ZjpE98ynET5imJ+bfdM5Gu4jYMzT7ItU/Wbui7XZFKqZnmpA1BcEGv2s QDHQVCxN7k+y8jGyZE7EWSZhCe47GI7g4r9PH30Bo+Bkn2y9snoCEohQYjmdkjQaB6Ak=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.14] x86/tsx: Minor cleanup and improvements Message-Id: Date: Thu, 17 Jun 2021 18:55:46 +0000 commit 768138cde899685efb5ec5a1c831b1d4e8794f54 Author: Andrew Cooper AuthorDate: Tue May 18 14:53:56 2021 +0100 Commit: Andrew Cooper CommitDate: Thu Jun 17 14:45:37 2021 +0100 x86/tsx: Minor cleanup and improvements * Introduce cpu_has_arch_caps and replace boot_cpu_has(X86_FEATURE_ARCH_CAPS) * Read CPUID data into the appropriate boot_cpu_data.x86_capability[] element, as subsequent changes are going to need more cpu_has_* logic. * Use the hi/lo MSR helpers, which substantially improves code generation. No practical change. Signed-off-by: Andrew Cooper Acked-by: Jan Beulich Reviewed-by: Roger Pau Monné (cherry picked from commit 3670abcaf0324f2aedba0c4dc7939072b27efa1d) --- xen/arch/x86/cpuid.c | 2 +- xen/arch/x86/hvm/vmx/vmx.c | 2 +- xen/arch/x86/msr.c | 2 +- xen/arch/x86/spec_ctrl.c | 2 +- xen/arch/x86/tsx.c | 21 ++++++++++++--------- xen/include/asm-x86/cpufeature.h | 1 + 6 files changed, 17 insertions(+), 13 deletions(-) diff --git a/xen/arch/x86/cpuid.c b/xen/arch/x86/cpuid.c index 425dce7bc7..deae62705d 100644 --- a/xen/arch/x86/cpuid.c +++ b/xen/arch/x86/cpuid.c @@ -720,7 +720,7 @@ int init_domain_cpuid_policy(struct domain *d) * so dom0 can turn off workarounds as appropriate. Temporary, until the * domain policy logic gains a better understanding of MSRs. */ - if ( is_hardware_domain(d) && boot_cpu_has(X86_FEATURE_ARCH_CAPS) ) + if ( is_hardware_domain(d) && cpu_has_arch_caps ) p->feat.arch_caps = true; d->arch.cpuid = p; diff --git a/xen/arch/x86/hvm/vmx/vmx.c b/xen/arch/x86/hvm/vmx/vmx.c index 6972d4ab23..2a29675c1a 100644 --- a/xen/arch/x86/hvm/vmx/vmx.c +++ b/xen/arch/x86/hvm/vmx/vmx.c @@ -2397,7 +2397,7 @@ static bool __init has_if_pschange_mc(void) if ( cpu_has_hypervisor ) return false; - if ( boot_cpu_has(X86_FEATURE_ARCH_CAPS) ) + if ( cpu_has_arch_caps ) rdmsrl(MSR_ARCH_CAPABILITIES, caps); if ( caps & ARCH_CAPS_IF_PSCHANGE_MC_NO ) diff --git a/xen/arch/x86/msr.c b/xen/arch/x86/msr.c index aa107823ac..caef0c6ede 100644 --- a/xen/arch/x86/msr.c +++ b/xen/arch/x86/msr.c @@ -135,7 +135,7 @@ int init_domain_msr_policy(struct domain *d) * so dom0 can turn off workarounds as appropriate. Temporary, until the * domain policy logic gains a better understanding of MSRs. */ - if ( is_hardware_domain(d) && boot_cpu_has(X86_FEATURE_ARCH_CAPS) ) + if ( is_hardware_domain(d) && cpu_has_arch_caps ) { uint64_t val; diff --git a/xen/arch/x86/spec_ctrl.c b/xen/arch/x86/spec_ctrl.c index bea4c3e8b4..aa85a0be46 100644 --- a/xen/arch/x86/spec_ctrl.c +++ b/xen/arch/x86/spec_ctrl.c @@ -885,7 +885,7 @@ void __init init_speculation_mitigations(void) bool cpu_has_bug_taa; uint64_t caps = 0; - if ( boot_cpu_has(X86_FEATURE_ARCH_CAPS) ) + if ( cpu_has_arch_caps ) rdmsrl(MSR_ARCH_CAPABILITIES, caps); hw_smt_enabled = check_smt_enabled(); diff --git a/xen/arch/x86/tsx.c b/xen/arch/x86/tsx.c index e09e819dce..98ecb71a4a 100644 --- a/xen/arch/x86/tsx.c +++ b/xen/arch/x86/tsx.c @@ -34,15 +34,18 @@ void tsx_init(void) { /* * This function is first called between microcode being loaded, and CPUID - * being scanned generally. Calculate from raw data whether MSR_TSX_CTRL - * is available. + * being scanned generally. Read into boot_cpu_data.x86_capability[] for + * the cpu_has_* bits we care about using here. */ if ( unlikely(cpu_has_tsx_ctrl < 0) ) { uint64_t caps = 0; - if ( boot_cpu_data.cpuid_level >= 7 && - (cpuid_count_edx(7, 0) & cpufeat_mask(X86_FEATURE_ARCH_CAPS)) ) + if ( boot_cpu_data.cpuid_level >= 7 ) + boot_cpu_data.x86_capability[cpufeat_word(X86_FEATURE_ARCH_CAPS)] + = cpuid_count_edx(7, 0); + + if ( cpu_has_arch_caps ) rdmsrl(MSR_ARCH_CAPABILITIES, caps); cpu_has_tsx_ctrl = !!(caps & ARCH_CAPS_TSX_CTRL); @@ -74,18 +77,18 @@ void tsx_init(void) if ( cpu_has_tsx_ctrl ) { - uint64_t val; + uint32_t hi, lo; - rdmsrl(MSR_TSX_CTRL, val); + rdmsr(MSR_TSX_CTRL, lo, hi); /* Check bottom bit only. Higher bits are various sentinels. */ rtm_disabled = !(opt_tsx & 1); - val &= ~(TSX_CTRL_RTM_DISABLE | TSX_CTRL_CPUID_CLEAR); + lo &= ~(TSX_CTRL_RTM_DISABLE | TSX_CTRL_CPUID_CLEAR); if ( rtm_disabled ) - val |= TSX_CTRL_RTM_DISABLE | TSX_CTRL_CPUID_CLEAR; + lo |= TSX_CTRL_RTM_DISABLE | TSX_CTRL_CPUID_CLEAR; - wrmsrl(MSR_TSX_CTRL, val); + wrmsr(MSR_TSX_CTRL, lo, hi); } else if ( opt_tsx >= 0 ) printk_once(XENLOG_WARNING diff --git a/xen/include/asm-x86/cpufeature.h b/xen/include/asm-x86/cpufeature.h index f790d5c1f8..9c31032436 100644 --- a/xen/include/asm-x86/cpufeature.h +++ b/xen/include/asm-x86/cpufeature.h @@ -130,6 +130,7 @@ #define cpu_has_avx512_4fmaps boot_cpu_has(X86_FEATURE_AVX512_4FMAPS) #define cpu_has_tsx_force_abort boot_cpu_has(X86_FEATURE_TSX_FORCE_ABORT) #define cpu_has_serialize boot_cpu_has(X86_FEATURE_SERIALIZE) +#define cpu_has_arch_caps boot_cpu_has(X86_FEATURE_ARCH_CAPS) /* CPUID level 0x00000007:1.eax */ #define cpu_has_avx512_bf16 boot_cpu_has(X86_FEATURE_AVX512_BF16) -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.14 From xen-changelog-bounces@lists.xenproject.org Thu Jun 17 18:55:57 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 17 Jun 2021 18:55:57 +0000 Received: from list by lists.xenproject.org with outflank-mailman.144137.265362 (Exim 4.92) (envelope-from ) id 1ltxBF-0007F9-Fj; Thu, 17 Jun 2021 18:55:57 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 144137.265362; Thu, 17 Jun 2021 18:55:57 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1ltxBF-0007F1-Ci; Thu, 17 Jun 2021 18:55:57 +0000 Received: by outflank-mailman (input) for mailman id 144137; Thu, 17 Jun 2021 18:55:56 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1ltxBE-0007Em-FI for xen-changelog@lists.xenproject.org; Thu, 17 Jun 2021 18:55:56 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1ltxBE-0007nE-EP for xen-changelog@lists.xenproject.org; Thu, 17 Jun 2021 18:55:56 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1ltxBE-0003TY-DZ for xen-changelog@lists.xenproject.org; Thu, 17 Jun 2021 18:55:56 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=qs8fYojsPZzRjb3oluhJFyRaByUHN1YJhcuOz135zGs=; b=lSc1QWzUVZv3PYEtLoacjzZTUP JpIeYtZUrCrvJLNbG1HQ1olnYlnUM8qIPvj15L+3pWf91UvQR/ySbdveCdEtVzx5FAfYHUmI+RVqf 4i6Bg/VAeCfzaTK2vT8KoPTvOcM8Q5SOQGnAT1dp9LKLQIYFllY+z7YrGk52ZK1M82aA=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.14] x86/tsx: Deprecate vpmu=rtm-abort and use tsx= instead Message-Id: Date: Thu, 17 Jun 2021 18:55:56 +0000 commit 4b4ee05a70d6193fa7c18650743c72b609cd272b Author: Andrew Cooper AuthorDate: Thu May 20 19:16:11 2021 +0100 Commit: Andrew Cooper CommitDate: Thu Jun 17 14:45:37 2021 +0100 x86/tsx: Deprecate vpmu=rtm-abort and use tsx= instead This reuses the rtm_disable infrastructure, so CPUID derivation works properly when TSX is disabled in favour of working PCR3. vpmu= is not a supported feature, and having this functionality under tsx= centralises all TSX handling. Signed-off-by: Andrew Cooper Acked-by: Jan Beulich Reviewed-by: Roger Pau Monné (cherry picked from commit 9fdcf851689cb2a9501d3947cb5d767d9c7797e8) --- docs/misc/xen-command-line.pandoc | 40 +++++++++++++++--------------- xen/arch/x86/cpu/intel.c | 3 --- xen/arch/x86/cpu/vpmu.c | 4 +-- xen/arch/x86/tsx.c | 51 +++++++++++++++++++++++++++++++++++++-- xen/include/asm-x86/vpmu.h | 1 - 5 files changed, 70 insertions(+), 29 deletions(-) diff --git a/docs/misc/xen-command-line.pandoc b/docs/misc/xen-command-line.pandoc index b9824d443f..6c4e340824 100644 --- a/docs/misc/xen-command-line.pandoc +++ b/docs/misc/xen-command-line.pandoc @@ -2231,14 +2231,21 @@ pages) must also be specified via the tbuf_size parameter. Controls for the use of Transactional Synchronization eXtensions. -On Intel parts released in Q3 2019 (with updated microcode), and future parts, -a control has been introduced which allows TSX to be turned off. +Several microcode updates are relevant: -On systems with the ability to turn TSX off, this boolean offers system wide -control of whether TSX is enabled or disabled. + * March 2019, fixing the TSX memory ordering errata on all TSX-enabled CPUs + to date. Introduced MSR_TSX_FORCE_ABORT on SKL/SKX/KBL/WHL/CFL parts. The + errata workaround uses Performance Counter 3, so the user can select + between working TSX and working perfcounters. + + * November 2019, fixing the TSX Async Abort speculative vulnerability. + Introduced MSR_TSX_CTRL on all TSX-enabled MDS_NO parts to date, + CLX/WHL-R/CFL-R, with the controls becoming architectural moving forward + and formally retiring HLE from the architecture. The user can disable TSX + to mitigate TAA, and elect to hide the HLE/RTM CPUID bits. -On parts vulnerable to CVE-2019-11135 / TSX Asynchronous Abort, the following -logic applies: +On systems with the ability to configure TSX, this boolean offers system wide +control of whether TSX is enabled or disabled. * An explicit `tsx=` choice is honoured, even if it is `true` and would result in a vulnerable system. @@ -2246,10 +2253,14 @@ logic applies: * When no explicit `tsx=` choice is given, parts vulnerable to TAA will be mitigated by disabling TSX, as this is the lowest overhead option. - * If the use of TSX is important, the more expensive TAA mitigations can be + If the use of TSX is important, the more expensive TAA mitigations can be opted in to with `smt=0 spec-ctrl=md-clear`, at which point TSX will remain active by default. + * When no explicit `tsx=` option is given, parts susceptible to the memory + ordering errata default to `true` to enable working TSX. Alternatively, + selecting `tsx=0` will disable TSX and restore PCR3 to a working state. + ### ucode > `= List of [ | scan=, nmi= ]` @@ -2386,20 +2397,7 @@ provide access to a wealth of low level processor information. * The `arch` option allows access to the pre-defined architectural events. -* The `rtm-abort` boolean controls a trade-off between working Restricted - Transactional Memory, and working performance counters. - - All processors released to date (Q1 2019) supporting Transactional Memory - Extensions suffer an erratum which has been addressed in microcode. - - Processors based on the Skylake microarchitecture with up-to-date - microcode internally use performance counter 3 to work around the erratum. - A consequence is that the counter gets reprogrammed whenever an `XBEGIN` - instruction is executed. - - An alternative mode exists where PCR3 behaves as before, at the cost of - `XBEGIN` unconditionally aborting. Enabling `rtm-abort` mode will - activate this alternative mode. +* The `rtm-abort` boolean has been superseded. Use `tsx=0` instead. *Warning:* As the virtualisation is not 100% safe, don't use the vpmu flag on diff --git a/xen/arch/x86/cpu/intel.c b/xen/arch/x86/cpu/intel.c index 37439071d9..abf8e206d7 100644 --- a/xen/arch/x86/cpu/intel.c +++ b/xen/arch/x86/cpu/intel.c @@ -356,9 +356,6 @@ static void Intel_errata_workarounds(struct cpuinfo_x86 *c) (c->x86_model == 29 || c->x86_model == 46 || c->x86_model == 47)) __set_bit(X86_FEATURE_CLFLUSH_MONITOR, c->x86_capability); - if (cpu_has_tsx_force_abort && opt_rtm_abort) - wrmsrl(MSR_TSX_FORCE_ABORT, TSX_FORCE_ABORT_RTM); - probe_c3_errata(c); } diff --git a/xen/arch/x86/cpu/vpmu.c b/xen/arch/x86/cpu/vpmu.c index 1ed39ef03f..0c3f378a02 100644 --- a/xen/arch/x86/cpu/vpmu.c +++ b/xen/arch/x86/cpu/vpmu.c @@ -47,7 +47,6 @@ CHECK_pmu_params; static unsigned int __read_mostly opt_vpmu_enabled; unsigned int __read_mostly vpmu_mode = XENPMU_MODE_OFF; unsigned int __read_mostly vpmu_features = 0; -bool __read_mostly opt_rtm_abort; static DEFINE_SPINLOCK(vpmu_lock); static unsigned vpmu_count; @@ -77,7 +76,8 @@ static int __init parse_vpmu_params(const char *s) else if ( !cmdline_strcmp(s, "arch") ) vpmu_features |= XENPMU_FEATURE_ARCH_ONLY; else if ( (val = parse_boolean("rtm-abort", s, ss)) >= 0 ) - opt_rtm_abort = val; + printk(XENLOG_WARNING + "'rtm-abort=' superseded. Use 'tsx=' instead\n"); else rc = -EINVAL; diff --git a/xen/arch/x86/tsx.c b/xen/arch/x86/tsx.c index 98ecb71a4a..338191df7f 100644 --- a/xen/arch/x86/tsx.c +++ b/xen/arch/x86/tsx.c @@ -6,7 +6,9 @@ * Valid values: * 1 => Explicit tsx=1 * 0 => Explicit tsx=0 - * -1 => Default, implicit tsx=1, may change to 0 to mitigate TAA + * -1 => Default, altered to 0/1 (if unspecified) by: + * - TAA heuristics/settings for speculative safety + * - "TSX vs PCR3" select for TSX memory ordering safety * -3 => Implicit tsx=1 (feed-through from spec-ctrl=0) * * This is arranged such that the bottom bit encodes whether TSX is actually @@ -50,6 +52,26 @@ void tsx_init(void) cpu_has_tsx_ctrl = !!(caps & ARCH_CAPS_TSX_CTRL); + if ( cpu_has_tsx_force_abort ) + { + /* + * On an early TSX-enable Skylake part subject to the memory + * ordering erratum, with at least the March 2019 microcode. + */ + + /* + * If no explicit tsx= option is provided, pick a default. + * + * This deliberately overrides the implicit opt_tsx=-3 from + * `spec-ctrl=0` because: + * - parse_spec_ctrl() ran before any CPU details where know. + * - We now know we're running on a CPU not affected by TAA (as + * TSX_FORCE_ABORT is enumerated). + */ + if ( opt_tsx < 0 ) + opt_tsx = 1; + } + /* * The TSX features (HLE/RTM) are handled specially. They both * enumerate features but, on certain parts, have mechanisms to be @@ -75,6 +97,12 @@ void tsx_init(void) } } + /* + * Note: MSR_TSX_CTRL is enumerated on TSX-enabled MDS_NO and later parts. + * MSR_TSX_FORCE_ABORT is enumerated on TSX-enabled pre-MDS_NO Skylake + * parts only. The two features are on a disjoint set of CPUs, and not + * offered to guests by hypervisors. + */ if ( cpu_has_tsx_ctrl ) { uint32_t hi, lo; @@ -90,9 +118,28 @@ void tsx_init(void) wrmsr(MSR_TSX_CTRL, lo, hi); } + else if ( cpu_has_tsx_force_abort ) + { + /* + * On an early TSX-enable Skylake part subject to the memory ordering + * erratum, with at least the March 2019 microcode. + */ + uint32_t hi, lo; + + rdmsr(MSR_TSX_FORCE_ABORT, lo, hi); + + /* Check bottom bit only. Higher bits are various sentinels. */ + rtm_disabled = !(opt_tsx & 1); + + lo &= ~TSX_FORCE_ABORT_RTM; + if ( rtm_disabled ) + lo |= TSX_FORCE_ABORT_RTM; + + wrmsr(MSR_TSX_FORCE_ABORT, lo, hi); + } else if ( opt_tsx >= 0 ) printk_once(XENLOG_WARNING - "MSR_TSX_CTRL not available - Ignoring tsx= setting\n"); + "TSX controls not available - Ignoring tsx= setting\n"); } /* diff --git a/xen/include/asm-x86/vpmu.h b/xen/include/asm-x86/vpmu.h index 55f85ba00f..4b0a6ba3da 100644 --- a/xen/include/asm-x86/vpmu.h +++ b/xen/include/asm-x86/vpmu.h @@ -126,7 +126,6 @@ static inline int vpmu_do_rdmsr(unsigned int msr, uint64_t *msr_content) extern unsigned int vpmu_mode; extern unsigned int vpmu_features; -extern bool opt_rtm_abort; /* Context switch */ static inline void vpmu_switch_from(struct vcpu *prev) -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.14 From xen-changelog-bounces@lists.xenproject.org Thu Jun 17 18:56:07 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 17 Jun 2021 18:56:07 +0000 Received: from list by lists.xenproject.org with outflank-mailman.144138.265369 (Exim 4.92) (envelope-from ) id 1ltxBP-0007IC-Ik; Thu, 17 Jun 2021 18:56:07 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 144138.265369; Thu, 17 Jun 2021 18:56:07 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1ltxBP-0007I2-EF; Thu, 17 Jun 2021 18:56:07 +0000 Received: by outflank-mailman (input) for mailman id 144138; Thu, 17 Jun 2021 18:56:06 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1ltxBO-0007Ht-IE for xen-changelog@lists.xenproject.org; Thu, 17 Jun 2021 18:56:06 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1ltxBO-0007nf-HQ for xen-changelog@lists.xenproject.org; Thu, 17 Jun 2021 18:56:06 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1ltxBO-0003Uq-Gc for xen-changelog@lists.xenproject.org; Thu, 17 Jun 2021 18:56:06 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=yvI0RgPlEYbABQiKlJ5rgff5rv2mSkzP1Zzmmr3xfLk=; b=OYFPFS/06XHAuc2cGX0WFrmSSK LUiGeggOe6M9h8kX/C+Ejs557LqfFgsOnq7II0YPr0nWuNRMXl0GnFIcCUIm1TJDh/ooNL/kWxy70 SaaWv6MgzLopsYdqbIPDQPqtcw0vyTTyGuwxsDAcMFK5DyMPwdi2WwBNkXV305XB6Og0=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.14] x86/cpuid: Fix HLE and RTM handling (again) Message-Id: Date: Thu, 17 Jun 2021 18:56:06 +0000 commit ee5425c0d532a773093cc4ff91c4c06caf2ff447 Author: Andrew Cooper AuthorDate: Mon Jun 7 13:25:09 2021 +0100 Commit: Andrew Cooper CommitDate: Thu Jun 17 14:45:37 2021 +0100 x86/cpuid: Fix HLE and RTM handling (again) For reasons which are my fault, but I don't recall why, the FDP_EXCP_ONLY/NO_FPU_SEL adjustment uses the whole special_features[] array element, not the two relevant bits. HLE and RTM were recently added to the list of special features, causing them to be always set in guest view, irrespective of the toolstacks choice on the matter. Rewrite the logic to refer to the features specifically, rather than relying on the contents of the special_features[] array. Fixes: 8fe24090d9 ("x86/cpuid: Rework HLE and RTM handling") Reported-by: Edwin Török Signed-off-by: Andrew Cooper Reviewed-by: Jan Beulich (cherry picked from commit 60fa12dbf1d4d2c4ffe1ef34b495b24aa7e41aa0) --- xen/arch/x86/cpuid.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/xen/arch/x86/cpuid.c b/xen/arch/x86/cpuid.c index deae62705d..5229eba595 100644 --- a/xen/arch/x86/cpuid.c +++ b/xen/arch/x86/cpuid.c @@ -648,9 +648,11 @@ void recalculate_cpuid_policy(struct domain *d) sanitise_featureset(fs); /* Fold host's FDP_EXCP_ONLY and NO_FPU_SEL into guest's view. */ - fs[FEATURESET_7b0] &= ~special_features[FEATURESET_7b0]; + fs[FEATURESET_7b0] &= ~(cpufeat_mask(X86_FEATURE_FDP_EXCP_ONLY) | + cpufeat_mask(X86_FEATURE_NO_FPU_SEL)); fs[FEATURESET_7b0] |= (host_cpuid_policy.feat._7b0 & - special_features[FEATURESET_7b0]); + (cpufeat_mask(X86_FEATURE_FDP_EXCP_ONLY) | + cpufeat_mask(X86_FEATURE_NO_FPU_SEL))); cpuid_featureset_to_policy(fs, p); -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.14 From xen-changelog-bounces@lists.xenproject.org Thu Jun 17 18:56:17 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 17 Jun 2021 18:56:17 +0000 Received: from list by lists.xenproject.org with outflank-mailman.144139.265371 (Exim 4.92) (envelope-from ) id 1ltxBZ-0007LC-Iv; Thu, 17 Jun 2021 18:56:17 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 144139.265371; Thu, 17 Jun 2021 18:56:17 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1ltxBZ-0007L4-Fp; Thu, 17 Jun 2021 18:56:17 +0000 Received: by outflank-mailman (input) for mailman id 144139; Thu, 17 Jun 2021 18:56:16 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1ltxBY-0007Kt-Lv for xen-changelog@lists.xenproject.org; Thu, 17 Jun 2021 18:56:16 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1ltxBY-0007np-LE for xen-changelog@lists.xenproject.org; Thu, 17 Jun 2021 18:56:16 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1ltxBY-0003Vx-K2 for xen-changelog@lists.xenproject.org; Thu, 17 Jun 2021 18:56:16 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=KmXQOdbacOX5KmH+n2VMmrcn7fWb0r4dVz30R7z2XUU=; b=rLEYtoJuJoyTl3brNkj/5jTez2 +WegObxJMc3FbAnjmWvjhUhlhzXYmTlZGQW0J/GSTGkEpxQ47qmdYgGvlP4vIpIPiGc0AgbAuFZIr k9oIPnI1Tosxmr23ARFwGwJug0vJC19mSfezYyYD3FHIbY6L8gaHJw/RtTKmPf+lubJY=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.14] x86/tsx: Cope with TSX deprecation on SKL/KBL/CFL/WHL Message-Id: Date: Thu, 17 Jun 2021 18:56:16 +0000 commit 45710c02563602343261b690787bb1c76a676bef Author: Andrew Cooper AuthorDate: Wed Sep 16 16:15:52 2020 +0100 Commit: Andrew Cooper CommitDate: Thu Jun 17 14:49:05 2021 +0100 x86/tsx: Cope with TSX deprecation on SKL/KBL/CFL/WHL The June 2021 microcode is formally de-featuring TSX on the older Skylake client CPUs. The workaround from the March 2019 microcode is being dropped, and replaced with additions to MSR_TSX_FORCE_ABORT to hide the HLE/RTM CPUID bits. With this microcode in place, TSX is disabled by default on these CPUs. Backwards compatibility is provided in the same way as for TAA - RTM force aborts, rather than suffering #UD, and the CPUID bits can be hidden to recover performance. Signed-off-by: Andrew Cooper Reviewed-by: Jan Beulich (cherry picked from commit 3e09045991cde360432bc7437103f8f8a6699359) --- docs/misc/xen-command-line.pandoc | 13 +++++ tools/misc/xen-cpuid.c | 2 +- xen/arch/x86/tsx.c | 76 +++++++++++++++++++++++++++-- xen/include/asm-x86/cpufeature.h | 1 + xen/include/asm-x86/msr-index.h | 2 + xen/include/public/arch-x86/cpufeatureset.h | 1 + 6 files changed, 89 insertions(+), 6 deletions(-) diff --git a/docs/misc/xen-command-line.pandoc b/docs/misc/xen-command-line.pandoc index 6c4e340824..a889c23c64 100644 --- a/docs/misc/xen-command-line.pandoc +++ b/docs/misc/xen-command-line.pandoc @@ -2244,6 +2244,12 @@ Several microcode updates are relevant: and formally retiring HLE from the architecture. The user can disable TSX to mitigate TAA, and elect to hide the HLE/RTM CPUID bits. + * June 2021, removing the workaround for March 2019 on client CPUs and + formally de-featured TSX on SKL/KBL/WHL/CFL (Note: SKX still retains the + March 2019 fix). Introduced the ability to hide the HLE/RTM CPUID bits. + PCR3 works fine, and TSX is disabled by default, but the user can re-enable + TSX at their own risk, accepting that the memory order erratum is unfixed. + On systems with the ability to configure TSX, this boolean offers system wide control of whether TSX is enabled or disabled. @@ -2261,6 +2267,13 @@ control of whether TSX is enabled or disabled. ordering errata default to `true` to enable working TSX. Alternatively, selecting `tsx=0` will disable TSX and restore PCR3 to a working state. + SKX and SKL/KBL/WHL/CFL on pre-June 2021 microcode default to `true`. + Alternatively, selecting `tsx=0` will disable TSX and restore PCR3 to a + working state. + + SKL/KBL/WHL/CFL on the June 2021 microcode or later default to `false`. + Alternatively, selecting `tsx=1` will re-enable TSX at the users own risk. + ### ucode > `= List of [ | scan=, nmi= ]` diff --git a/tools/misc/xen-cpuid.c b/tools/misc/xen-cpuid.c index 8d8f3532a2..a231d96e13 100644 --- a/tools/misc/xen-cpuid.c +++ b/tools/misc/xen-cpuid.c @@ -161,7 +161,7 @@ static const char *const str_7d0[32] = [ 4] = "fsrm", /* 8 */ [ 9] = "srbds-ctrl", - [10] = "md-clear", + [10] = "md-clear", [11] = "rtm-always-abort", /* 12 */ [13] = "tsx-force-abort", [14] = "serialize", diff --git a/xen/arch/x86/tsx.c b/xen/arch/x86/tsx.c index 338191df7f..88adf08c49 100644 --- a/xen/arch/x86/tsx.c +++ b/xen/arch/x86/tsx.c @@ -59,6 +59,38 @@ void tsx_init(void) * ordering erratum, with at least the March 2019 microcode. */ + /* + * Probe for the June 2021 microcode which de-features TSX on + * client parts. (Note - this is a subset of parts impacted by + * the memory ordering errata.) + * + * RTM_ALWAYS_ABORT enumerates the new functionality, but is also + * read as zero if TSX_FORCE_ABORT.ENABLE_RTM has been set before + * we run. + * + * Undo this behaviour in Xen's view of the world. + */ + bool has_rtm_always_abort = cpu_has_rtm_always_abort; + + if ( !has_rtm_always_abort ) + { + uint64_t val; + + rdmsrl(MSR_TSX_FORCE_ABORT, val); + + if ( val & TSX_ENABLE_RTM ) + has_rtm_always_abort = true; + } + + /* + * Always force RTM_ALWAYS_ABORT, even if it currently visible. + * If the user explicitly opts to enable TSX, we'll set + * TSX_FORCE_ABORT.ENABLE_RTM and cause RTM_ALWAYS_ABORT to be + * hidden from the general CPUID scan later. + */ + if ( has_rtm_always_abort ) + setup_force_cpu_cap(X86_FEATURE_RTM_ALWAYS_ABORT); + /* * If no explicit tsx= option is provided, pick a default. * @@ -67,9 +99,16 @@ void tsx_init(void) * - parse_spec_ctrl() ran before any CPU details where know. * - We now know we're running on a CPU not affected by TAA (as * TSX_FORCE_ABORT is enumerated). + * - When RTM_ALWAYS_ABORT is enumerated, TSX malfunctions, so we + * only ever want it enabled by explicit user choice. + * + * Without RTM_ALWAYS_ABORT, leave TSX active. In particular, + * this includes SKX where TSX is still supported. + * + * With RTM_ALWAYS_ABORT, disable TSX. */ if ( opt_tsx < 0 ) - opt_tsx = 1; + opt_tsx = !cpu_has_rtm_always_abort; } /* @@ -90,7 +129,7 @@ void tsx_init(void) * Force the features to be visible in Xen's view if we see any of the * infrastructure capable of hiding them. */ - if ( cpu_has_tsx_ctrl ) + if ( cpu_has_tsx_ctrl || cpu_has_tsx_force_abort ) { setup_force_cpu_cap(X86_FEATURE_HLE); setup_force_cpu_cap(X86_FEATURE_RTM); @@ -131,9 +170,36 @@ void tsx_init(void) /* Check bottom bit only. Higher bits are various sentinels. */ rtm_disabled = !(opt_tsx & 1); - lo &= ~TSX_FORCE_ABORT_RTM; - if ( rtm_disabled ) - lo |= TSX_FORCE_ABORT_RTM; + lo &= ~(TSX_FORCE_ABORT_RTM | TSX_CPUID_CLEAR | TSX_ENABLE_RTM); + + if ( cpu_has_rtm_always_abort ) + { + /* + * June 2021 microcode, on a client part with TSX de-featured: + * - There are no mitigations for the TSX memory ordering errata. + * - Performance counter 3 works. (I.e. it isn't being used by + * microcode to work around the memory ordering errata.) + * - TSX_FORCE_ABORT.FORCE_ABORT_RTM is fixed read1/write-discard. + * - TSX_FORCE_ABORT.TSX_CPUID_CLEAR can be used to hide the + * HLE/RTM CPUID bits. + * - TSX_FORCE_ABORT.ENABLE_RTM may be used to opt in to + * re-enabling RTM, at the users own risk. + */ + lo |= rtm_disabled ? TSX_CPUID_CLEAR : TSX_ENABLE_RTM; + } + else + { + /* + * Either a server part where TSX isn't de-featured, or pre-June + * 2021 microcode: + * - By default, the TSX memory ordering errata is worked around + * in microcode at the cost of Performance Counter 3. + * - "Working TSX" vs "Working PCR3" can be selected by way of + * setting TSX_FORCE_ABORT.FORCE_ABORT_RTM. + */ + if ( rtm_disabled ) + lo |= TSX_FORCE_ABORT_RTM; + } wrmsr(MSR_TSX_FORCE_ABORT, lo, hi); } diff --git a/xen/include/asm-x86/cpufeature.h b/xen/include/asm-x86/cpufeature.h index 9c31032436..80eff688d4 100644 --- a/xen/include/asm-x86/cpufeature.h +++ b/xen/include/asm-x86/cpufeature.h @@ -128,6 +128,7 @@ /* CPUID level 0x00000007:0.edx */ #define cpu_has_avx512_4vnniw boot_cpu_has(X86_FEATURE_AVX512_4VNNIW) #define cpu_has_avx512_4fmaps boot_cpu_has(X86_FEATURE_AVX512_4FMAPS) +#define cpu_has_rtm_always_abort boot_cpu_has(X86_FEATURE_RTM_ALWAYS_ABORT) #define cpu_has_tsx_force_abort boot_cpu_has(X86_FEATURE_TSX_FORCE_ABORT) #define cpu_has_serialize boot_cpu_has(X86_FEATURE_SERIALIZE) #define cpu_has_arch_caps boot_cpu_has(X86_FEATURE_ARCH_CAPS) diff --git a/xen/include/asm-x86/msr-index.h b/xen/include/asm-x86/msr-index.h index d66c92b946..80764266bf 100644 --- a/xen/include/asm-x86/msr-index.h +++ b/xen/include/asm-x86/msr-index.h @@ -61,6 +61,8 @@ #define MSR_TSX_FORCE_ABORT 0x0000010f #define TSX_FORCE_ABORT_RTM (_AC(1, ULL) << 0) +#define TSX_CPUID_CLEAR (_AC(1, ULL) << 1) +#define TSX_ENABLE_RTM (_AC(1, ULL) << 2) #define MSR_TSX_CTRL 0x00000122 #define TSX_CTRL_RTM_DISABLE (_AC(1, ULL) << 0) diff --git a/xen/include/public/arch-x86/cpufeatureset.h b/xen/include/public/arch-x86/cpufeatureset.h index 02b3a2a265..d2422386c0 100644 --- a/xen/include/public/arch-x86/cpufeatureset.h +++ b/xen/include/public/arch-x86/cpufeatureset.h @@ -261,6 +261,7 @@ XEN_CPUFEATURE(AVX512_4VNNIW, 9*32+ 2) /*A AVX512 Neural Network Instructions * XEN_CPUFEATURE(AVX512_4FMAPS, 9*32+ 3) /*A AVX512 Multiply Accumulation Single Precision */ XEN_CPUFEATURE(SRBDS_CTRL, 9*32+ 9) /* MSR_MCU_OPT_CTRL and RNGDS_MITG_DIS. */ XEN_CPUFEATURE(MD_CLEAR, 9*32+10) /*A VERW clears microarchitectural buffers */ +XEN_CPUFEATURE(RTM_ALWAYS_ABORT, 9*32+11) /*! June 2021 TSX defeaturing in microcode. */ XEN_CPUFEATURE(TSX_FORCE_ABORT, 9*32+13) /* MSR_TSX_FORCE_ABORT.RTM_ABORT */ XEN_CPUFEATURE(SERIALIZE, 9*32+14) /*a SERIALIZE insn */ XEN_CPUFEATURE(CET_IBT, 9*32+20) /* CET - Indirect Branch Tracking */ -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.14 From xen-changelog-bounces@lists.xenproject.org Thu Jun 17 22:55:10 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 17 Jun 2021 22:55:10 +0000 Received: from list by lists.xenproject.org with outflank-mailman.144190.265450 (Exim 4.92) (envelope-from ) id 1lu0uf-000697-Kj; Thu, 17 Jun 2021 22:55:05 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 144190.265450; Thu, 17 Jun 2021 22:55:05 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lu0uf-00068z-Hc; Thu, 17 Jun 2021 22:55:05 +0000 Received: by outflank-mailman (input) for mailman id 144190; Thu, 17 Jun 2021 22:55:04 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lu0ue-00068t-Ks for xen-changelog@lists.xenproject.org; Thu, 17 Jun 2021 22:55:04 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lu0ue-0003PM-Gz for xen-changelog@lists.xenproject.org; Thu, 17 Jun 2021 22:55:04 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lu0ue-0000JW-Fn for xen-changelog@lists.xenproject.org; Thu, 17 Jun 2021 22:55:04 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=wd0yG2dINYOcA1BwjDW85s0aPnyIVzXe8279CJ2aafg=; b=UOsTndfpo6XbBfOlABP5CQU1hB 58VFV7SFUjZTIL5JAoJCIb6eCil35KlAa1KjvRG301rIz62KYxr8CH7QtNdaQuGMQNU3BbtXHB9x3 TggUAgkmA776FaPYn+WRst/AnPCtJM1vbEO7aXcfodnCPid0uTEzttn6pAa1Qu7DBP/c=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging-4.14] golang/xenlight: fix code generation for python 2.6 Message-Id: Date: Thu, 17 Jun 2021 22:55:04 +0000 commit b6a8c4f72def4d1135ff42660a86276ce2565c8c Author: Nick Rosbrook AuthorDate: Mon Jul 20 19:54:40 2020 -0400 Commit: Andrew Cooper CommitDate: Thu Jun 17 19:45:39 2021 +0100 golang/xenlight: fix code generation for python 2.6 Before python 2.7, str.format() calls required that the format fields were explicitly enumerated, e.g.: '{0} {1}'.format(foo, bar) vs. '{} {}'.format(foo, bar) Currently, gengotypes.py uses the latter pattern everywhere, which means the Go bindings do not build on python 2.6. Use the 2.6 syntax for format() in order to support python 2.6 for now. Signed-off-by: Nick Rosbrook Acked-by: Wei Liu (cherry picked from commit 6d49fbdeab3e687a6818f809ca3d98ac7ced2c8d) --- tools/golang/xenlight/gengotypes.py | 204 ++++++++++++++++++------------------ 1 file changed, 102 insertions(+), 102 deletions(-) diff --git a/tools/golang/xenlight/gengotypes.py b/tools/golang/xenlight/gengotypes.py index 557fecd07b..ebec938224 100644 --- a/tools/golang/xenlight/gengotypes.py +++ b/tools/golang/xenlight/gengotypes.py @@ -3,7 +3,7 @@ import os import sys -sys.path.append('{}/tools/libxl'.format(os.environ['XEN_ROOT'])) +sys.path.append('{0}/tools/libxl'.format(os.environ['XEN_ROOT'])) import idl # Go versions of some builtin types. @@ -73,14 +73,14 @@ def xenlight_golang_define_enum(ty = None): if ty.typename is not None: typename = xenlight_golang_fmt_name(ty.typename) - s += 'type {} int\n'.format(typename) + s += 'type {0} int\n'.format(typename) # Start const block s += 'const(\n' for v in ty.values: name = xenlight_golang_fmt_name(v.name) - s += '{} {} = {}\n'.format(name, typename, v.value) + s += '{0} {1} = {2}\n'.format(name, typename, v.value) # End const block s += ')\n' @@ -99,9 +99,9 @@ def xenlight_golang_define_struct(ty = None, typename = None, nested = False): # Begin struct definition if nested: - s += '{} struct {{\n'.format(name) + s += '{0} struct {{\n'.format(name) else: - s += 'type {} struct {{\n'.format(name) + s += 'type {0} struct {{\n'.format(name) # Write struct fields for f in ty.fields: @@ -111,13 +111,13 @@ def xenlight_golang_define_struct(ty = None, typename = None, nested = False): typename = xenlight_golang_fmt_name(typename) name = xenlight_golang_fmt_name(f.name) - s += '{} []{}\n'.format(name, typename) + s += '{0} []{1}\n'.format(name, typename) else: typename = f.type.typename typename = xenlight_golang_fmt_name(typename) name = xenlight_golang_fmt_name(f.name) - s += '{} {}\n'.format(name, typename) + s += '{0} {1}\n'.format(name, typename) elif isinstance(f.type, idl.Struct): r = xenlight_golang_define_struct(f.type, typename=f.name, nested=True) @@ -132,7 +132,7 @@ def xenlight_golang_define_struct(ty = None, typename = None, nested = False): extras.extend(r[1]) else: - raise Exception('type {} not supported'.format(f.type)) + raise Exception('type {0} not supported'.format(f.type)) # End struct definition s += '}\n' @@ -151,11 +151,11 @@ def xenlight_golang_define_union(ty = None, struct_name = '', union_name = ''): s = '' extras = [] - interface_name = '{}_{}_union'.format(struct_name, ty.keyvar.name) + interface_name = '{0}_{1}_union'.format(struct_name, ty.keyvar.name) interface_name = xenlight_golang_fmt_name(interface_name, exported=False) - s += 'type {} interface {{\n'.format(interface_name) - s += 'is{}()\n'.format(interface_name) + s += 'type {0} interface {{\n'.format(interface_name) + s += 'is{0}()\n'.format(interface_name) s += '}\n' extras.append(s) @@ -165,7 +165,7 @@ def xenlight_golang_define_union(ty = None, struct_name = '', union_name = ''): continue # Define struct - name = '{}_{}_union_{}'.format(struct_name, ty.keyvar.name, f.name) + name = '{0}_{1}_union_{2}'.format(struct_name, ty.keyvar.name, f.name) r = xenlight_golang_define_struct(f.type, typename=name) extras.append(r[0]) extras.extend(r[1]) @@ -173,21 +173,21 @@ def xenlight_golang_define_union(ty = None, struct_name = '', union_name = ''): # This typeof trick ensures that the fields used in the cgo struct # used for marshaling are the same as the fields of the union in the # actual C type, and avoids re-defining all of those fields. - s = 'typedef typeof(((struct {} *)NULL)->{}.{}){};' + s = 'typedef typeof(((struct {0} *)NULL)->{1}.{2}){3};' s = s.format(struct_name, union_name, f.name, name) cgo_helpers_preamble.append(s) # Define function to implement 'union' interface name = xenlight_golang_fmt_name(name) - s = 'func (x {}) is{}(){{}}\n'.format(name, interface_name) + s = 'func (x {0}) is{1}(){{}}\n'.format(name, interface_name) extras.append(s) fname = xenlight_golang_fmt_name(ty.keyvar.name) ftype = xenlight_golang_fmt_name(ty.keyvar.type.typename) - s = '{} {}\n'.format(fname, ftype) + s = '{0} {1}\n'.format(fname, ftype) - fname = xenlight_golang_fmt_name('{}_union'.format(ty.keyvar.name)) - s += '{} {}\n'.format(fname, interface_name) + fname = xenlight_golang_fmt_name('{0}_union'.format(ty.keyvar.name)) + s += '{0} {1}\n'.format(fname, interface_name) return (s,extras) @@ -243,7 +243,7 @@ def xenlight_golang_define_from_C(ty = None): Define the fromC marshaling function for the type represented by ty. """ - func = 'func (x *{}) fromC(xc *C.{}) error {{\n {}\n return nil}}\n' + func = 'func (x *{0}) fromC(xc *C.{1}) error {{\n {2}\n return nil}}\n' goname = xenlight_golang_fmt_name(ty.typename) cname = ty.typename @@ -271,7 +271,7 @@ def xenlight_golang_define_from_C(ty = None): extras.extend(r[1]) else: - raise Exception('type {} not supported'.format(f.type)) + raise Exception('type {0} not supported'.format(f.type)) return (func.format(goname, cname, body), extras) @@ -300,8 +300,8 @@ def xenlight_golang_convert_from_C(ty = None, outer_name = None, cvarname = None # If outer_name is set, treat this as nested. if outer_name is not None: - goname = '{}.{}'.format(xenlight_golang_fmt_name(outer_name), goname) - cname = '{}.{}'.format(outer_name, cname) + goname = '{0}.{1}'.format(xenlight_golang_fmt_name(outer_name), goname) + cname = '{0}.{1}'.format(outer_name, cname) # Types that satisfy this condition can be easily casted or # converted to a Go builtin type. @@ -312,15 +312,15 @@ def xenlight_golang_convert_from_C(ty = None, outer_name = None, cvarname = None if not is_castable: # If the type is not castable, we need to call its fromC # function. - s += 'if err := x.{}.fromC(&{}.{});'.format(goname,cvarname,cname) - s += 'err != nil {{\nreturn fmt.Errorf("converting field {}: %v", err)\n}}\n'.format(goname) + s += 'if err := x.{0}.fromC(&{1}.{2});'.format(goname,cvarname,cname) + s += 'err != nil {{\nreturn fmt.Errorf("converting field {0}: %v", err)\n}}\n'.format(goname) elif gotypename == 'string': # Use the cgo helper for converting C strings. - s += 'x.{} = C.GoString({}.{})\n'.format(goname,cvarname,cname) + s += 'x.{0} = C.GoString({1}.{2})\n'.format(goname,cvarname,cname) else: - s += 'x.{} = {}({}.{})\n'.format(goname,gotypename,cvarname,cname) + s += 'x.{0} = {1}({2}.{3})\n'.format(goname,gotypename,cvarname,cname) return s @@ -331,9 +331,9 @@ def xenlight_golang_union_from_C(ty = None, union_name = '', struct_name = ''): gokeyname = xenlight_golang_fmt_name(keyname) keytype = ty.keyvar.type.typename gokeytype = xenlight_golang_fmt_name(keytype) - field_name = xenlight_golang_fmt_name('{}_union'.format(keyname)) + field_name = xenlight_golang_fmt_name('{0}_union'.format(keyname)) - interface_name = '{}_{}_union'.format(struct_name, keyname) + interface_name = '{0}_{1}_union'.format(struct_name, keyname) interface_name = xenlight_golang_fmt_name(interface_name, exported=False) cgo_keyname = keyname @@ -343,7 +343,7 @@ def xenlight_golang_union_from_C(ty = None, union_name = '', struct_name = ''): cases = {} for f in ty.fields: - val = '{}_{}'.format(keytype, f.name) + val = '{0}_{1}'.format(keytype, f.name) val = xenlight_golang_fmt_name(val) # Add to list of cases to make for the switch @@ -354,17 +354,17 @@ def xenlight_golang_union_from_C(ty = None, union_name = '', struct_name = ''): continue # Define fromC func for 'union' struct. - typename = '{}_{}_union_{}'.format(struct_name,keyname,f.name) + typename = '{0}_{1}_union_{2}'.format(struct_name,keyname,f.name) gotypename = xenlight_golang_fmt_name(typename) # Define the function here. The cases for keyed unions are a little # different. - s = 'func (x *{}) fromC(xc *C.{}) error {{\n'.format(gotypename,struct_name) - s += 'if {}(xc.{}) != {} {{\n'.format(gokeytype,cgo_keyname,val) - err_string = '"expected union key {}"'.format(val) - s += 'return errors.New({})\n'.format(err_string) + s = 'func (x *{0}) fromC(xc *C.{1}) error {{\n'.format(gotypename,struct_name) + s += 'if {0}(xc.{1}) != {2} {{\n'.format(gokeytype,cgo_keyname,val) + err_string = '"expected union key {0}"'.format(val) + s += 'return errors.New({0})\n'.format(err_string) s += '}\n\n' - s += 'tmp := (*C.{})(unsafe.Pointer(&xc.{}[0]))\n'.format(typename,union_name) + s += 'tmp := (*C.{0})(unsafe.Pointer(&xc.{1}[0]))\n'.format(typename,union_name) for nf in f.type.fields: s += xenlight_golang_convert_from_C(nf,cvarname='tmp') @@ -374,35 +374,35 @@ def xenlight_golang_union_from_C(ty = None, union_name = '', struct_name = ''): extras.append(s) - s = 'x.{} = {}(xc.{})\n'.format(gokeyname,gokeytype,cgo_keyname) - s += 'switch x.{}{{\n'.format(gokeyname) + s = 'x.{0} = {1}(xc.{2})\n'.format(gokeyname,gokeytype,cgo_keyname) + s += 'switch x.{0}{{\n'.format(gokeyname) # Create switch statement to determine which 'union element' # to populate in the Go struct. for case_name, case_tuple in sorted(cases.items()): (case_val, case_type) = case_tuple - s += 'case {}:\n'.format(case_val) + s += 'case {0}:\n'.format(case_val) if case_type is None: - s += "x.{} = nil\n".format(field_name) + s += "x.{0} = nil\n".format(field_name) continue - gotype = '{}_{}_union_{}'.format(struct_name,keyname,case_name) + gotype = '{0}_{1}_union_{2}'.format(struct_name,keyname,case_name) gotype = xenlight_golang_fmt_name(gotype) - goname = '{}_{}'.format(keyname,case_name) + goname = '{0}_{1}'.format(keyname,case_name) goname = xenlight_golang_fmt_name(goname,exported=False) - s += 'var {} {}\n'.format(goname, gotype) - s += 'if err := {}.fromC(xc);'.format(goname) - s += 'err != nil {{\n return fmt.Errorf("converting field {}: %v", err)\n}}\n'.format(goname) + s += 'var {0} {1}\n'.format(goname, gotype) + s += 'if err := {0}.fromC(xc);'.format(goname) + s += 'err != nil {{\n return fmt.Errorf("converting field {0}: %v", err)\n}}\n'.format(goname) - s += 'x.{} = {}\n'.format(field_name, goname) + s += 'x.{0} = {1}\n'.format(field_name, goname) # End switch statement s += 'default:\n' - err_string = '"invalid union key \'%v\'", x.{}'.format(gokeyname) - s += 'return fmt.Errorf({})'.format(err_string) + err_string = '"invalid union key \'%v\'", x.{0}'.format(gokeyname) + s += 'return fmt.Errorf({0})'.format(err_string) s += '}\n' return (s,extras) @@ -420,22 +420,22 @@ def xenlight_golang_array_from_C(ty = None): goname = xenlight_golang_fmt_name(ty.name) ctypename = ty.type.elem_type.typename cname = ty.name - cslice = 'c{}'.format(goname) + cslice = 'c{0}'.format(goname) clenvar = ty.type.lenvar.name - s += 'x.{} = nil\n'.format(goname) - s += 'if n := int(xc.{}); n > 0 {{\n'.format(clenvar) - s += '{} := '.format(cslice) - s +='(*[1<<28]C.{})(unsafe.Pointer(xc.{}))[:n:n]\n'.format(ctypename, cname) - s += 'x.{} = make([]{}, n)\n'.format(goname, gotypename) - s += 'for i, v := range {} {{\n'.format(cslice) + s += 'x.{0} = nil\n'.format(goname) + s += 'if n := int(xc.{0}); n > 0 {{\n'.format(clenvar) + s += '{0} := '.format(cslice) + s +='(*[1<<28]C.{0})(unsafe.Pointer(xc.{1}))[:n:n]\n'.format(ctypename, cname) + s += 'x.{0} = make([]{1}, n)\n'.format(goname, gotypename) + s += 'for i, v := range {0} {{\n'.format(cslice) is_enum = isinstance(ty.type.elem_type,idl.Enumeration) if gotypename in go_builtin_types or is_enum: - s += 'x.{}[i] = {}(v)\n'.format(goname, gotypename) + s += 'x.{0}[i] = {1}(v)\n'.format(goname, gotypename) else: - s += 'if err := x.{}[i].fromC(&v); err != nil {{\n'.format(goname) - s += 'return fmt.Errorf("converting field {}: %v", err) }}\n'.format(goname) + s += 'if err := x.{0}[i].fromC(&v); err != nil {{\n'.format(goname) + s += 'return fmt.Errorf("converting field {0}: %v", err) }}\n'.format(goname) s += '}\n}\n' @@ -446,11 +446,11 @@ def xenlight_golang_define_to_C(ty = None, typename = None, nested = False): Define the toC marshaling function for the type represented by ty. """ - func = 'func (x *{}) toC(xc *C.{}) (err error){{{}\n return nil\n }}\n' + func = 'func (x *{0}) toC(xc *C.{1}) (err error){{{2}\n return nil\n }}\n' body = '' if ty.dispose_fn is not None: - body += 'defer func(){{\nif err != nil{{\nC.{}(xc)}}\n}}()\n\n'.format(ty.dispose_fn) + body += 'defer func(){{\nif err != nil{{\nC.{0}(xc)}}\n}}()\n\n'.format(ty.dispose_fn) goname = xenlight_golang_fmt_name(ty.typename) cname = ty.typename @@ -471,7 +471,7 @@ def xenlight_golang_define_to_C(ty = None, typename = None, nested = False): body += xenlight_golang_union_to_C(f.type, f.name, ty.typename) else: - raise Exception('type {} not supported'.format(f.type)) + raise Exception('type {0} not supported'.format(f.type)) return func.format(goname, cname, body) @@ -506,26 +506,26 @@ def xenlight_golang_convert_to_C(ty = None, outer_name = None, # If outer_name is set, treat this as nested. if outer_name is not None: - goname = '{}.{}'.format(xenlight_golang_fmt_name(outer_name), goname) - cname = '{}.{}'.format(outer_name, cname) + goname = '{0}.{1}'.format(xenlight_golang_fmt_name(outer_name), goname) + cname = '{0}.{1}'.format(outer_name, cname) is_castable = (ty.type.json_parse_type == 'JSON_INTEGER' or isinstance(ty.type, idl.Enumeration) or gotypename in go_builtin_types) if not is_castable: - s += 'if err := {}.{}.toC(&{}.{}); err != nil {{\n'.format(govarname,goname, + s += 'if err := {0}.{1}.toC(&{2}.{3}); err != nil {{\n'.format(govarname,goname, cvarname,cname) - s += 'return fmt.Errorf("converting field {}: %v", err)\n}}\n'.format(goname) + s += 'return fmt.Errorf("converting field {0}: %v", err)\n}}\n'.format(goname) elif gotypename == 'string': # Use the cgo helper for converting C strings. - s += 'if {}.{} != "" {{\n'.format(govarname,goname) - s += '{}.{} = C.CString({}.{})}}\n'.format(cvarname,cname, + s += 'if {0}.{1} != "" {{\n'.format(govarname,goname) + s += '{0}.{1} = C.CString({2}.{3})}}\n'.format(cvarname,cname, govarname,goname) else: - s += '{}.{} = C.{}({}.{})\n'.format(cvarname,cname,ctypename, + s += '{0}.{1} = C.{2}({3}.{4})\n'.format(cvarname,cname,ctypename, govarname,goname) return s @@ -537,7 +537,7 @@ def xenlight_golang_union_to_C(ty = None, union_name = '', keytype = ty.keyvar.type.typename gokeytype = xenlight_golang_fmt_name(keytype) - interface_name = '{}_{}_union'.format(struct_name, keyname) + interface_name = '{0}_{1}_union'.format(struct_name, keyname) interface_name = xenlight_golang_fmt_name(interface_name, exported=False) cgo_keyname = keyname @@ -545,44 +545,44 @@ def xenlight_golang_union_to_C(ty = None, union_name = '', cgo_keyname = '_' + cgo_keyname - s = 'xc.{} = C.{}(x.{})\n'.format(cgo_keyname,keytype,gokeyname) - s += 'switch x.{}{{\n'.format(gokeyname) + s = 'xc.{0} = C.{1}(x.{2})\n'.format(cgo_keyname,keytype,gokeyname) + s += 'switch x.{0}{{\n'.format(gokeyname) # Create switch statement to determine how to populate the C union. for f in ty.fields: - key_val = '{}_{}'.format(keytype, f.name) + key_val = '{0}_{1}'.format(keytype, f.name) key_val = xenlight_golang_fmt_name(key_val) - s += 'case {}:\n'.format(key_val) + s += 'case {0}:\n'.format(key_val) if f.type is None: s += "break\n" continue - cgotype = '{}_{}_union_{}'.format(struct_name,keyname,f.name) + cgotype = '{0}_{1}_union_{2}'.format(struct_name,keyname,f.name) gotype = xenlight_golang_fmt_name(cgotype) - field_name = xenlight_golang_fmt_name('{}_union'.format(keyname)) - s += 'tmp, ok := x.{}.({})\n'.format(field_name,gotype) + field_name = xenlight_golang_fmt_name('{0}_union'.format(keyname)) + s += 'tmp, ok := x.{0}.({1})\n'.format(field_name,gotype) s += 'if !ok {\n' - s += 'return errors.New("wrong type for union key {}")\n'.format(keyname) + s += 'return errors.New("wrong type for union key {0}")\n'.format(keyname) s += '}\n' - s += 'var {} C.{}\n'.format(f.name,cgotype) + s += 'var {0} C.{1}\n'.format(f.name,cgotype) for uf in f.type.fields: s += xenlight_golang_convert_to_C(uf,cvarname=f.name, govarname='tmp') # The union is still represented as Go []byte. - s += '{}Bytes := C.GoBytes(unsafe.Pointer(&{}),C.sizeof_{})\n'.format(f.name, + s += '{0}Bytes := C.GoBytes(unsafe.Pointer(&{1}),C.sizeof_{2})\n'.format(f.name, f.name, cgotype) - s += 'copy(xc.{}[:],{}Bytes)\n'.format(union_name,f.name) + s += 'copy(xc.{0}[:],{1}Bytes)\n'.format(union_name,f.name) # End switch statement s += 'default:\n' - err_string = '"invalid union key \'%v\'", x.{}'.format(gokeyname) - s += 'return fmt.Errorf({})'.format(err_string) + err_string = '"invalid union key \'%v\'", x.{0}'.format(gokeyname) + s += 'return fmt.Errorf({0})'.format(err_string) s += '}\n' return s @@ -599,29 +599,29 @@ def xenlight_golang_array_to_C(ty = None): is_enum = isinstance(ty.type.elem_type,idl.Enumeration) if gotypename in go_builtin_types or is_enum: - s += 'if {} := len(x.{}); {} > 0 {{\n'.format(golenvar,goname,golenvar) - s += 'xc.{} = (*C.{})(C.malloc(C.size_t({}*{})))\n'.format(cname,ctypename, + s += 'if {0} := len(x.{1}); {2} > 0 {{\n'.format(golenvar,goname,golenvar) + s += 'xc.{0} = (*C.{1})(C.malloc(C.size_t({2}*{3})))\n'.format(cname,ctypename, golenvar,golenvar) - s += 'xc.{} = C.int({})\n'.format(clenvar,golenvar) - s += 'c{} := (*[1<<28]C.{})(unsafe.Pointer(xc.{}))[:{}:{}]\n'.format(goname, + s += 'xc.{0} = C.int({1})\n'.format(clenvar,golenvar) + s += 'c{0} := (*[1<<28]C.{1})(unsafe.Pointer(xc.{2}))[:{3}:{4}]\n'.format(goname, ctypename,cname, golenvar,golenvar) - s += 'for i,v := range x.{} {{\n'.format(goname) - s += 'c{}[i] = C.{}(v)\n'.format(goname,ctypename) + s += 'for i,v := range x.{0} {{\n'.format(goname) + s += 'c{0}[i] = C.{1}(v)\n'.format(goname,ctypename) s += '}\n}\n' return s - s += 'if {} := len(x.{}); {} > 0 {{\n'.format(golenvar,goname,golenvar) - s += 'xc.{} = (*C.{})(C.malloc(C.ulong({})*C.sizeof_{}))\n'.format(cname,ctypename, + s += 'if {0} := len(x.{1}); {2} > 0 {{\n'.format(golenvar,goname,golenvar) + s += 'xc.{0} = (*C.{1})(C.malloc(C.ulong({2})*C.sizeof_{3}))\n'.format(cname,ctypename, golenvar,ctypename) - s += 'xc.{} = C.int({})\n'.format(clenvar,golenvar) - s += 'c{} := (*[1<<28]C.{})(unsafe.Pointer(xc.{}))[:{}:{}]\n'.format(goname, + s += 'xc.{0} = C.int({1})\n'.format(clenvar,golenvar) + s += 'c{0} := (*[1<<28]C.{1})(unsafe.Pointer(xc.{2}))[:{3}:{4}]\n'.format(goname, ctypename,cname, golenvar,golenvar) - s += 'for i,v := range x.{} {{\n'.format(goname) - s += 'if err := v.toC(&c{}[i]); err != nil {{\n'.format(goname) - s += 'return fmt.Errorf("converting field {}: %v", err)\n'.format(goname) + s += 'for i,v := range x.{0} {{\n'.format(goname) + s += 'if err := v.toC(&c{0}[i]); err != nil {{\n'.format(goname) + s += 'return fmt.Errorf("converting field {0}: %v", err)\n'.format(goname) s += '}\n}\n}\n' return s @@ -633,7 +633,7 @@ def xenlight_golang_define_constructor(ty = None): gotypename = xenlight_golang_fmt_name(ctypename) # Since this func is exported, add a comment as per Go conventions. - s += '// New{} returns an instance of {}'.format(gotypename,gotypename) + s += '// New{0} returns an instance of {1}'.format(gotypename,gotypename) s += ' initialized with defaults.\n' # If a struct has a keyed union, an extra argument is @@ -643,7 +643,7 @@ def xenlight_golang_define_constructor(ty = None): init_fns = [] # Add call to parent init_fn first. - init_fns.append('C.{}(&xc)'.format(ty.init_fn)) + init_fns.append('C.{0}(&xc)'.format(ty.init_fn)) for f in ty.fields: if not isinstance(f.type, idl.KeyedUnion): @@ -658,24 +658,24 @@ def xenlight_golang_define_constructor(ty = None): # Serveral keyed unions use 'type' as the key variable name. In # that case, prepend the first letter of the Go type name. if param_goname == 'type': - param_goname = '{}type'.format(param_gotype.lower()[0]) + param_goname = '{0}type'.format(param_gotype.lower()[0]) # Add call to keyed union's init_fn. - init_fns.append('C.{}_{}(&xc, C.{}({}))'.format(ty.init_fn, + init_fns.append('C.{0}_{1}(&xc, C.{2}({3}))'.format(ty.init_fn, param.name, param_ctype, param_goname)) # Add to params list. - params.append('{} {}'.format(param_goname, param_gotype)) + params.append('{0} {1}'.format(param_goname, param_gotype)) # Define function - s += 'func New{}({}) (*{}, error) {{\n'.format(gotypename, + s += 'func New{0}({1}) (*{2}, error) {{\n'.format(gotypename, ','.join(params), gotypename) # Declare variables. - s += 'var (\nx {}\nxc C.{})\n\n'.format(gotypename, ctypename) + s += 'var (\nx {0}\nxc C.{1})\n\n'.format(gotypename, ctypename) # Write init_fn calls. s += '\n'.join(init_fns) @@ -684,7 +684,7 @@ def xenlight_golang_define_constructor(ty = None): # Make sure dispose_fn get's called when constructor # returns. if ty.dispose_fn is not None: - s += 'defer C.{}(&xc)\n'.format(ty.dispose_fn) + s += 'defer C.{0}(&xc)\n'.format(ty.dispose_fn) s += '\n' @@ -727,7 +727,7 @@ if __name__ == '__main__': header_comment="""// DO NOT EDIT. // // This file is generated by: -// {} +// {0} // """.format(' '.join(sys.argv)) -- generated by git-patchbot for /home/xen/git/xen.git#staging-4.14 From xen-changelog-bounces@lists.xenproject.org Fri Jun 18 07:22:09 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Fri, 18 Jun 2021 07:22:09 +0000 Received: from list by lists.xenproject.org with outflank-mailman.144342.265683 (Exim 4.92) (envelope-from ) id 1lu8pH-0001h6-V4; Fri, 18 Jun 2021 07:22:03 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 144342.265683; Fri, 18 Jun 2021 07:22:03 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lu8pH-0001gy-S3; Fri, 18 Jun 2021 07:22:03 +0000 Received: by outflank-mailman (input) for mailman id 144342; Fri, 18 Jun 2021 07:22:02 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lu8pG-0001ge-Fz for xen-changelog@lists.xenproject.org; Fri, 18 Jun 2021 07:22:02 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lu8pG-00022a-F5 for xen-changelog@lists.xenproject.org; Fri, 18 Jun 2021 07:22:02 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lu8pG-00074c-EB for xen-changelog@lists.xenproject.org; Fri, 18 Jun 2021 07:22:02 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=5XlHXj3VvZc2juMLXI7aw63ntCsbVYX90kXeor7dGps=; b=yR/h+ruluhuLt6AeZYg2i4c+8a 0bjJ+4xLPhBSEWIzgX4ziEiZwAU5qU/x0c0LcuLaGAuexXTiBb2HLPxpf62aue/JvUHijoOcPUhe1 yyPaWb7EDTua9wXYlK7MuEIJeXCy6Lc9ILpziMM6ys0om5859IyCLpcdsqBHEjAp7uy8=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.15] x86/tsx: Minor cleanup and improvements Message-Id: Date: Fri, 18 Jun 2021 07:22:02 +0000 commit 0aabeb9293b6a3bcd98d692de2cd3d16f4f41062 Author: Andrew Cooper AuthorDate: Tue May 18 14:53:56 2021 +0100 Commit: Andrew Cooper CommitDate: Thu Jun 17 14:36:15 2021 +0100 x86/tsx: Minor cleanup and improvements * Introduce cpu_has_arch_caps and replace boot_cpu_has(X86_FEATURE_ARCH_CAPS) * Read CPUID data into the appropriate boot_cpu_data.x86_capability[] element, as subsequent changes are going to need more cpu_has_* logic. * Use the hi/lo MSR helpers, which substantially improves code generation. No practical change. Signed-off-by: Andrew Cooper Acked-by: Jan Beulich Reviewed-by: Roger Pau Monné (cherry picked from commit 3670abcaf0324f2aedba0c4dc7939072b27efa1d) --- xen/arch/x86/cpuid.c | 2 +- xen/arch/x86/hvm/vmx/vmx.c | 2 +- xen/arch/x86/msr.c | 2 +- xen/arch/x86/spec_ctrl.c | 2 +- xen/arch/x86/tsx.c | 21 ++++++++++++--------- xen/include/asm-x86/cpufeature.h | 1 + 6 files changed, 17 insertions(+), 13 deletions(-) diff --git a/xen/arch/x86/cpuid.c b/xen/arch/x86/cpuid.c index 7ac6636efc..431f6fb173 100644 --- a/xen/arch/x86/cpuid.c +++ b/xen/arch/x86/cpuid.c @@ -713,7 +713,7 @@ int init_domain_cpuid_policy(struct domain *d) * so dom0 can turn off workarounds as appropriate. Temporary, until the * domain policy logic gains a better understanding of MSRs. */ - if ( is_hardware_domain(d) && boot_cpu_has(X86_FEATURE_ARCH_CAPS) ) + if ( is_hardware_domain(d) && cpu_has_arch_caps ) p->feat.arch_caps = true; d->arch.cpuid = p; diff --git a/xen/arch/x86/hvm/vmx/vmx.c b/xen/arch/x86/hvm/vmx/vmx.c index dde4f3b70d..fd7b193058 100644 --- a/xen/arch/x86/hvm/vmx/vmx.c +++ b/xen/arch/x86/hvm/vmx/vmx.c @@ -2573,7 +2573,7 @@ static bool __init has_if_pschange_mc(void) if ( cpu_has_hypervisor ) return false; - if ( boot_cpu_has(X86_FEATURE_ARCH_CAPS) ) + if ( cpu_has_arch_caps ) rdmsrl(MSR_ARCH_CAPABILITIES, caps); if ( caps & ARCH_CAPS_IF_PSCHANGE_MC_NO ) diff --git a/xen/arch/x86/msr.c b/xen/arch/x86/msr.c index c3a988bd11..374f92b2c5 100644 --- a/xen/arch/x86/msr.c +++ b/xen/arch/x86/msr.c @@ -136,7 +136,7 @@ int init_domain_msr_policy(struct domain *d) * so dom0 can turn off workarounds as appropriate. Temporary, until the * domain policy logic gains a better understanding of MSRs. */ - if ( is_hardware_domain(d) && boot_cpu_has(X86_FEATURE_ARCH_CAPS) ) + if ( is_hardware_domain(d) && cpu_has_arch_caps ) { uint64_t val; diff --git a/xen/arch/x86/spec_ctrl.c b/xen/arch/x86/spec_ctrl.c index bea4c3e8b4..aa85a0be46 100644 --- a/xen/arch/x86/spec_ctrl.c +++ b/xen/arch/x86/spec_ctrl.c @@ -885,7 +885,7 @@ void __init init_speculation_mitigations(void) bool cpu_has_bug_taa; uint64_t caps = 0; - if ( boot_cpu_has(X86_FEATURE_ARCH_CAPS) ) + if ( cpu_has_arch_caps ) rdmsrl(MSR_ARCH_CAPABILITIES, caps); hw_smt_enabled = check_smt_enabled(); diff --git a/xen/arch/x86/tsx.c b/xen/arch/x86/tsx.c index e09e819dce..98ecb71a4a 100644 --- a/xen/arch/x86/tsx.c +++ b/xen/arch/x86/tsx.c @@ -34,15 +34,18 @@ void tsx_init(void) { /* * This function is first called between microcode being loaded, and CPUID - * being scanned generally. Calculate from raw data whether MSR_TSX_CTRL - * is available. + * being scanned generally. Read into boot_cpu_data.x86_capability[] for + * the cpu_has_* bits we care about using here. */ if ( unlikely(cpu_has_tsx_ctrl < 0) ) { uint64_t caps = 0; - if ( boot_cpu_data.cpuid_level >= 7 && - (cpuid_count_edx(7, 0) & cpufeat_mask(X86_FEATURE_ARCH_CAPS)) ) + if ( boot_cpu_data.cpuid_level >= 7 ) + boot_cpu_data.x86_capability[cpufeat_word(X86_FEATURE_ARCH_CAPS)] + = cpuid_count_edx(7, 0); + + if ( cpu_has_arch_caps ) rdmsrl(MSR_ARCH_CAPABILITIES, caps); cpu_has_tsx_ctrl = !!(caps & ARCH_CAPS_TSX_CTRL); @@ -74,18 +77,18 @@ void tsx_init(void) if ( cpu_has_tsx_ctrl ) { - uint64_t val; + uint32_t hi, lo; - rdmsrl(MSR_TSX_CTRL, val); + rdmsr(MSR_TSX_CTRL, lo, hi); /* Check bottom bit only. Higher bits are various sentinels. */ rtm_disabled = !(opt_tsx & 1); - val &= ~(TSX_CTRL_RTM_DISABLE | TSX_CTRL_CPUID_CLEAR); + lo &= ~(TSX_CTRL_RTM_DISABLE | TSX_CTRL_CPUID_CLEAR); if ( rtm_disabled ) - val |= TSX_CTRL_RTM_DISABLE | TSX_CTRL_CPUID_CLEAR; + lo |= TSX_CTRL_RTM_DISABLE | TSX_CTRL_CPUID_CLEAR; - wrmsrl(MSR_TSX_CTRL, val); + wrmsr(MSR_TSX_CTRL, lo, hi); } else if ( opt_tsx >= 0 ) printk_once(XENLOG_WARNING diff --git a/xen/include/asm-x86/cpufeature.h b/xen/include/asm-x86/cpufeature.h index 33b2257888..9f5ae3aa0d 100644 --- a/xen/include/asm-x86/cpufeature.h +++ b/xen/include/asm-x86/cpufeature.h @@ -133,6 +133,7 @@ #define cpu_has_avx512_vp2intersect boot_cpu_has(X86_FEATURE_AVX512_VP2INTERSECT) #define cpu_has_tsx_force_abort boot_cpu_has(X86_FEATURE_TSX_FORCE_ABORT) #define cpu_has_serialize boot_cpu_has(X86_FEATURE_SERIALIZE) +#define cpu_has_arch_caps boot_cpu_has(X86_FEATURE_ARCH_CAPS) /* CPUID level 0x00000007:1.eax */ #define cpu_has_avx_vnni boot_cpu_has(X86_FEATURE_AVX_VNNI) -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.15 From xen-changelog-bounces@lists.xenproject.org Fri Jun 18 07:22:14 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Fri, 18 Jun 2021 07:22:14 +0000 Received: from list by lists.xenproject.org with outflank-mailman.144345.265686 (Exim 4.92) (envelope-from ) id 1lu8pS-0001l1-0H; Fri, 18 Jun 2021 07:22:14 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 144345.265686; Fri, 18 Jun 2021 07:22:13 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lu8pR-0001kt-TW; Fri, 18 Jun 2021 07:22:13 +0000 Received: by outflank-mailman (input) for mailman id 144345; Fri, 18 Jun 2021 07:22:12 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lu8pQ-0001kQ-JS for xen-changelog@lists.xenproject.org; Fri, 18 Jun 2021 07:22:12 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lu8pQ-00022h-Ia for xen-changelog@lists.xenproject.org; Fri, 18 Jun 2021 07:22:12 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lu8pQ-00077N-Hc for xen-changelog@lists.xenproject.org; Fri, 18 Jun 2021 07:22:12 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=nrFEgers14tTRYekSXXM1VrkjXWOW03N0uef0r1JXeg=; b=fV0M8EIGJjNdD4FEi3VrDkT1Sx lXVWJmVFU/RjItVH4mSM+qJFrYmss6E3nn9UwV2REdm12KipEnmuJTrlYW6uQOGYZKKnO4M4at3NY d8TlnXvKjnUqMHgmROJu7lWufQZCjBWBbxCvjMNxV/TPtycUf72qLDS8PZ7Qqb55soZU=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.15] x86/tsx: Deprecate vpmu=rtm-abort and use tsx= instead Message-Id: Date: Fri, 18 Jun 2021 07:22:12 +0000 commit 796d4056691f2b6d6f71b799173e8352eb6d73ea Author: Andrew Cooper AuthorDate: Thu May 20 19:16:11 2021 +0100 Commit: Andrew Cooper CommitDate: Thu Jun 17 14:36:15 2021 +0100 x86/tsx: Deprecate vpmu=rtm-abort and use tsx= instead This reuses the rtm_disable infrastructure, so CPUID derivation works properly when TSX is disabled in favour of working PCR3. vpmu= is not a supported feature, and having this functionality under tsx= centralises all TSX handling. Signed-off-by: Andrew Cooper Acked-by: Jan Beulich Reviewed-by: Roger Pau Monné (cherry picked from commit 9fdcf851689cb2a9501d3947cb5d767d9c7797e8) --- docs/misc/xen-command-line.pandoc | 40 +++++++++++++++--------------- xen/arch/x86/cpu/intel.c | 3 --- xen/arch/x86/cpu/vpmu.c | 4 +-- xen/arch/x86/tsx.c | 51 +++++++++++++++++++++++++++++++++++++-- xen/include/asm-x86/vpmu.h | 1 - 5 files changed, 70 insertions(+), 29 deletions(-) diff --git a/docs/misc/xen-command-line.pandoc b/docs/misc/xen-command-line.pandoc index c32a397a12..1fae872626 100644 --- a/docs/misc/xen-command-line.pandoc +++ b/docs/misc/xen-command-line.pandoc @@ -2296,14 +2296,21 @@ pages) must also be specified via the tbuf_size parameter. Controls for the use of Transactional Synchronization eXtensions. -On Intel parts released in Q3 2019 (with updated microcode), and future parts, -a control has been introduced which allows TSX to be turned off. +Several microcode updates are relevant: -On systems with the ability to turn TSX off, this boolean offers system wide -control of whether TSX is enabled or disabled. + * March 2019, fixing the TSX memory ordering errata on all TSX-enabled CPUs + to date. Introduced MSR_TSX_FORCE_ABORT on SKL/SKX/KBL/WHL/CFL parts. The + errata workaround uses Performance Counter 3, so the user can select + between working TSX and working perfcounters. + + * November 2019, fixing the TSX Async Abort speculative vulnerability. + Introduced MSR_TSX_CTRL on all TSX-enabled MDS_NO parts to date, + CLX/WHL-R/CFL-R, with the controls becoming architectural moving forward + and formally retiring HLE from the architecture. The user can disable TSX + to mitigate TAA, and elect to hide the HLE/RTM CPUID bits. -On parts vulnerable to CVE-2019-11135 / TSX Asynchronous Abort, the following -logic applies: +On systems with the ability to configure TSX, this boolean offers system wide +control of whether TSX is enabled or disabled. * An explicit `tsx=` choice is honoured, even if it is `true` and would result in a vulnerable system. @@ -2311,10 +2318,14 @@ logic applies: * When no explicit `tsx=` choice is given, parts vulnerable to TAA will be mitigated by disabling TSX, as this is the lowest overhead option. - * If the use of TSX is important, the more expensive TAA mitigations can be + If the use of TSX is important, the more expensive TAA mitigations can be opted in to with `smt=0 spec-ctrl=md-clear`, at which point TSX will remain active by default. + * When no explicit `tsx=` option is given, parts susceptible to the memory + ordering errata default to `true` to enable working TSX. Alternatively, + selecting `tsx=0` will disable TSX and restore PCR3 to a working state. + ### ucode > `= List of [ | scan=, nmi=, allow-same= ]` @@ -2456,20 +2467,7 @@ provide access to a wealth of low level processor information. * The `arch` option allows access to the pre-defined architectural events. -* The `rtm-abort` boolean controls a trade-off between working Restricted - Transactional Memory, and working performance counters. - - All processors released to date (Q1 2019) supporting Transactional Memory - Extensions suffer an erratum which has been addressed in microcode. - - Processors based on the Skylake microarchitecture with up-to-date - microcode internally use performance counter 3 to work around the erratum. - A consequence is that the counter gets reprogrammed whenever an `XBEGIN` - instruction is executed. - - An alternative mode exists where PCR3 behaves as before, at the cost of - `XBEGIN` unconditionally aborting. Enabling `rtm-abort` mode will - activate this alternative mode. +* The `rtm-abort` boolean has been superseded. Use `tsx=0` instead. *Warning:* As the virtualisation is not 100% safe, don't use the vpmu flag on diff --git a/xen/arch/x86/cpu/intel.c b/xen/arch/x86/cpu/intel.c index 37439071d9..abf8e206d7 100644 --- a/xen/arch/x86/cpu/intel.c +++ b/xen/arch/x86/cpu/intel.c @@ -356,9 +356,6 @@ static void Intel_errata_workarounds(struct cpuinfo_x86 *c) (c->x86_model == 29 || c->x86_model == 46 || c->x86_model == 47)) __set_bit(X86_FEATURE_CLFLUSH_MONITOR, c->x86_capability); - if (cpu_has_tsx_force_abort && opt_rtm_abort) - wrmsrl(MSR_TSX_FORCE_ABORT, TSX_FORCE_ABORT_RTM); - probe_c3_errata(c); } diff --git a/xen/arch/x86/cpu/vpmu.c b/xen/arch/x86/cpu/vpmu.c index ab667361d3..fb1b296a6c 100644 --- a/xen/arch/x86/cpu/vpmu.c +++ b/xen/arch/x86/cpu/vpmu.c @@ -47,7 +47,6 @@ CHECK_pmu_params; static unsigned int __read_mostly opt_vpmu_enabled; unsigned int __read_mostly vpmu_mode = XENPMU_MODE_OFF; unsigned int __read_mostly vpmu_features = 0; -bool __read_mostly opt_rtm_abort; static DEFINE_SPINLOCK(vpmu_lock); static unsigned vpmu_count; @@ -77,7 +76,8 @@ static int __init parse_vpmu_params(const char *s) else if ( !cmdline_strcmp(s, "arch") ) vpmu_features |= XENPMU_FEATURE_ARCH_ONLY; else if ( (val = parse_boolean("rtm-abort", s, ss)) >= 0 ) - opt_rtm_abort = val; + printk(XENLOG_WARNING + "'rtm-abort=' superseded. Use 'tsx=' instead\n"); else rc = -EINVAL; diff --git a/xen/arch/x86/tsx.c b/xen/arch/x86/tsx.c index 98ecb71a4a..338191df7f 100644 --- a/xen/arch/x86/tsx.c +++ b/xen/arch/x86/tsx.c @@ -6,7 +6,9 @@ * Valid values: * 1 => Explicit tsx=1 * 0 => Explicit tsx=0 - * -1 => Default, implicit tsx=1, may change to 0 to mitigate TAA + * -1 => Default, altered to 0/1 (if unspecified) by: + * - TAA heuristics/settings for speculative safety + * - "TSX vs PCR3" select for TSX memory ordering safety * -3 => Implicit tsx=1 (feed-through from spec-ctrl=0) * * This is arranged such that the bottom bit encodes whether TSX is actually @@ -50,6 +52,26 @@ void tsx_init(void) cpu_has_tsx_ctrl = !!(caps & ARCH_CAPS_TSX_CTRL); + if ( cpu_has_tsx_force_abort ) + { + /* + * On an early TSX-enable Skylake part subject to the memory + * ordering erratum, with at least the March 2019 microcode. + */ + + /* + * If no explicit tsx= option is provided, pick a default. + * + * This deliberately overrides the implicit opt_tsx=-3 from + * `spec-ctrl=0` because: + * - parse_spec_ctrl() ran before any CPU details where know. + * - We now know we're running on a CPU not affected by TAA (as + * TSX_FORCE_ABORT is enumerated). + */ + if ( opt_tsx < 0 ) + opt_tsx = 1; + } + /* * The TSX features (HLE/RTM) are handled specially. They both * enumerate features but, on certain parts, have mechanisms to be @@ -75,6 +97,12 @@ void tsx_init(void) } } + /* + * Note: MSR_TSX_CTRL is enumerated on TSX-enabled MDS_NO and later parts. + * MSR_TSX_FORCE_ABORT is enumerated on TSX-enabled pre-MDS_NO Skylake + * parts only. The two features are on a disjoint set of CPUs, and not + * offered to guests by hypervisors. + */ if ( cpu_has_tsx_ctrl ) { uint32_t hi, lo; @@ -90,9 +118,28 @@ void tsx_init(void) wrmsr(MSR_TSX_CTRL, lo, hi); } + else if ( cpu_has_tsx_force_abort ) + { + /* + * On an early TSX-enable Skylake part subject to the memory ordering + * erratum, with at least the March 2019 microcode. + */ + uint32_t hi, lo; + + rdmsr(MSR_TSX_FORCE_ABORT, lo, hi); + + /* Check bottom bit only. Higher bits are various sentinels. */ + rtm_disabled = !(opt_tsx & 1); + + lo &= ~TSX_FORCE_ABORT_RTM; + if ( rtm_disabled ) + lo |= TSX_FORCE_ABORT_RTM; + + wrmsr(MSR_TSX_FORCE_ABORT, lo, hi); + } else if ( opt_tsx >= 0 ) printk_once(XENLOG_WARNING - "MSR_TSX_CTRL not available - Ignoring tsx= setting\n"); + "TSX controls not available - Ignoring tsx= setting\n"); } /* diff --git a/xen/include/asm-x86/vpmu.h b/xen/include/asm-x86/vpmu.h index 55f85ba00f..4b0a6ba3da 100644 --- a/xen/include/asm-x86/vpmu.h +++ b/xen/include/asm-x86/vpmu.h @@ -126,7 +126,6 @@ static inline int vpmu_do_rdmsr(unsigned int msr, uint64_t *msr_content) extern unsigned int vpmu_mode; extern unsigned int vpmu_features; -extern bool opt_rtm_abort; /* Context switch */ static inline void vpmu_switch_from(struct vcpu *prev) -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.15 From xen-changelog-bounces@lists.xenproject.org Fri Jun 18 07:22:24 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Fri, 18 Jun 2021 07:22:24 +0000 Received: from list by lists.xenproject.org with outflank-mailman.144348.265690 (Exim 4.92) (envelope-from ) id 1lu8pc-0001qV-2R; Fri, 18 Jun 2021 07:22:24 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 144348.265690; Fri, 18 Jun 2021 07:22:24 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lu8pb-0001qL-Uz; Fri, 18 Jun 2021 07:22:23 +0000 Received: by outflank-mailman (input) for mailman id 144348; Fri, 18 Jun 2021 07:22:22 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lu8pa-0001pF-MX for xen-changelog@lists.xenproject.org; Fri, 18 Jun 2021 07:22:22 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lu8pa-00022x-Lf for xen-changelog@lists.xenproject.org; Fri, 18 Jun 2021 07:22:22 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lu8pa-00078e-Kk for xen-changelog@lists.xenproject.org; Fri, 18 Jun 2021 07:22:22 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=v9PpOoDYGFl9MCkNq/Z2YflodjaE487unFiCWFakKVQ=; b=1V4g6gg8JbgKQTc320HThh4lLU j4blY3gbB+t9TcazWmULKEffwY4GGZ0esF4GdCKA5ATjab0GzZ1yLt7h4rgZramXoHIf4SzS0xwTy U6XVoKdSBKh3avosUWlpZlhYnB8lPRoy/KU954wVrqKZNIfoQjP7PvAy746nVDdl4f/c=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.15] x86/cpuid: Fix HLE and RTM handling (again) Message-Id: Date: Fri, 18 Jun 2021 07:22:22 +0000 commit 4586e6443a8b963283bdec1d72d42fae061ebc49 Author: Andrew Cooper AuthorDate: Mon Jun 7 13:25:09 2021 +0100 Commit: Andrew Cooper CommitDate: Thu Jun 17 14:36:15 2021 +0100 x86/cpuid: Fix HLE and RTM handling (again) For reasons which are my fault, but I don't recall why, the FDP_EXCP_ONLY/NO_FPU_SEL adjustment uses the whole special_features[] array element, not the two relevant bits. HLE and RTM were recently added to the list of special features, causing them to be always set in guest view, irrespective of the toolstacks choice on the matter. Rewrite the logic to refer to the features specifically, rather than relying on the contents of the special_features[] array. Fixes: 8fe24090d9 ("x86/cpuid: Rework HLE and RTM handling") Reported-by: Edwin Török Signed-off-by: Andrew Cooper Reviewed-by: Jan Beulich (cherry picked from commit 60fa12dbf1d4d2c4ffe1ef34b495b24aa7e41aa0) --- xen/arch/x86/cpuid.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/xen/arch/x86/cpuid.c b/xen/arch/x86/cpuid.c index 431f6fb173..eefcde902e 100644 --- a/xen/arch/x86/cpuid.c +++ b/xen/arch/x86/cpuid.c @@ -640,9 +640,11 @@ void recalculate_cpuid_policy(struct domain *d) sanitise_featureset(fs); /* Fold host's FDP_EXCP_ONLY and NO_FPU_SEL into guest's view. */ - fs[FEATURESET_7b0] &= ~special_features[FEATURESET_7b0]; + fs[FEATURESET_7b0] &= ~(cpufeat_mask(X86_FEATURE_FDP_EXCP_ONLY) | + cpufeat_mask(X86_FEATURE_NO_FPU_SEL)); fs[FEATURESET_7b0] |= (host_cpuid_policy.feat._7b0 & - special_features[FEATURESET_7b0]); + (cpufeat_mask(X86_FEATURE_FDP_EXCP_ONLY) | + cpufeat_mask(X86_FEATURE_NO_FPU_SEL))); cpuid_featureset_to_policy(fs, p); -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.15 From xen-changelog-bounces@lists.xenproject.org Fri Jun 18 07:22:34 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Fri, 18 Jun 2021 07:22:34 +0000 Received: from list by lists.xenproject.org with outflank-mailman.144349.265695 (Exim 4.92) (envelope-from ) id 1lu8pm-0001y1-4O; Fri, 18 Jun 2021 07:22:34 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 144349.265695; Fri, 18 Jun 2021 07:22:34 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lu8pm-0001xu-0W; Fri, 18 Jun 2021 07:22:34 +0000 Received: by outflank-mailman (input) for mailman id 144349; Fri, 18 Jun 2021 07:22:32 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lu8pk-0001xN-QD for xen-changelog@lists.xenproject.org; Fri, 18 Jun 2021 07:22:32 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lu8pk-00023C-PK for xen-changelog@lists.xenproject.org; Fri, 18 Jun 2021 07:22:32 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lu8pk-00079p-O1 for xen-changelog@lists.xenproject.org; Fri, 18 Jun 2021 07:22:32 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=5uKMHwUHS6B/7TuC5xH94xRbHo8TLMcngWxf1vQhHkk=; b=khZ2Oc9FXUC2uaQU5R/23wkOG5 rp4Ot3C4bDPqxYP4Znc/mxuNrJM+cF0UXtS/cUr5gqk8HWenZiVqOsjJvnJhjGvBJlh8AFdCvso5K cc4EjK6KFSKLA4avbt7tJ+0+sZPlwzPnksAQYlwlB6kTqOWmG+/UCj6GWy7gmhisqlxw=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.15] x86/tsx: Cope with TSX deprecation on SKL/KBL/CFL/WHL Message-Id: Date: Fri, 18 Jun 2021 07:22:32 +0000 commit ec457ac2a29279e8cd91745c410b0f49d5e8f1ff Author: Andrew Cooper AuthorDate: Wed Sep 16 16:15:52 2020 +0100 Commit: Andrew Cooper CommitDate: Thu Jun 17 14:36:15 2021 +0100 x86/tsx: Cope with TSX deprecation on SKL/KBL/CFL/WHL The June 2021 microcode is formally de-featuring TSX on the older Skylake client CPUs. The workaround from the March 2019 microcode is being dropped, and replaced with additions to MSR_TSX_FORCE_ABORT to hide the HLE/RTM CPUID bits. With this microcode in place, TSX is disabled by default on these CPUs. Backwards compatibility is provided in the same way as for TAA - RTM force aborts, rather than suffering #UD, and the CPUID bits can be hidden to recover performance. Signed-off-by: Andrew Cooper Reviewed-by: Jan Beulich (cherry picked from commit 3e09045991cde360432bc7437103f8f8a6699359) --- docs/misc/xen-command-line.pandoc | 13 +++++ tools/misc/xen-cpuid.c | 2 +- xen/arch/x86/tsx.c | 76 +++++++++++++++++++++++++++-- xen/include/asm-x86/cpufeature.h | 1 + xen/include/asm-x86/msr-index.h | 2 + xen/include/public/arch-x86/cpufeatureset.h | 1 + 6 files changed, 89 insertions(+), 6 deletions(-) diff --git a/docs/misc/xen-command-line.pandoc b/docs/misc/xen-command-line.pandoc index 1fae872626..3ece83a427 100644 --- a/docs/misc/xen-command-line.pandoc +++ b/docs/misc/xen-command-line.pandoc @@ -2309,6 +2309,12 @@ Several microcode updates are relevant: and formally retiring HLE from the architecture. The user can disable TSX to mitigate TAA, and elect to hide the HLE/RTM CPUID bits. + * June 2021, removing the workaround for March 2019 on client CPUs and + formally de-featured TSX on SKL/KBL/WHL/CFL (Note: SKX still retains the + March 2019 fix). Introduced the ability to hide the HLE/RTM CPUID bits. + PCR3 works fine, and TSX is disabled by default, but the user can re-enable + TSX at their own risk, accepting that the memory order erratum is unfixed. + On systems with the ability to configure TSX, this boolean offers system wide control of whether TSX is enabled or disabled. @@ -2326,6 +2332,13 @@ control of whether TSX is enabled or disabled. ordering errata default to `true` to enable working TSX. Alternatively, selecting `tsx=0` will disable TSX and restore PCR3 to a working state. + SKX and SKL/KBL/WHL/CFL on pre-June 2021 microcode default to `true`. + Alternatively, selecting `tsx=0` will disable TSX and restore PCR3 to a + working state. + + SKL/KBL/WHL/CFL on the June 2021 microcode or later default to `false`. + Alternatively, selecting `tsx=1` will re-enable TSX at the users own risk. + ### ucode > `= List of [ | scan=, nmi=, allow-same= ]` diff --git a/tools/misc/xen-cpuid.c b/tools/misc/xen-cpuid.c index 2d04162d8d..531d56d9c9 100644 --- a/tools/misc/xen-cpuid.c +++ b/tools/misc/xen-cpuid.c @@ -161,7 +161,7 @@ static const char *const str_7d0[32] = [ 4] = "fsrm", [ 8] = "avx512-vp2intersect", [ 9] = "srbds-ctrl", - [10] = "md-clear", + [10] = "md-clear", [11] = "rtm-always-abort", /* 12 */ [13] = "tsx-force-abort", [14] = "serialize", diff --git a/xen/arch/x86/tsx.c b/xen/arch/x86/tsx.c index 338191df7f..88adf08c49 100644 --- a/xen/arch/x86/tsx.c +++ b/xen/arch/x86/tsx.c @@ -59,6 +59,38 @@ void tsx_init(void) * ordering erratum, with at least the March 2019 microcode. */ + /* + * Probe for the June 2021 microcode which de-features TSX on + * client parts. (Note - this is a subset of parts impacted by + * the memory ordering errata.) + * + * RTM_ALWAYS_ABORT enumerates the new functionality, but is also + * read as zero if TSX_FORCE_ABORT.ENABLE_RTM has been set before + * we run. + * + * Undo this behaviour in Xen's view of the world. + */ + bool has_rtm_always_abort = cpu_has_rtm_always_abort; + + if ( !has_rtm_always_abort ) + { + uint64_t val; + + rdmsrl(MSR_TSX_FORCE_ABORT, val); + + if ( val & TSX_ENABLE_RTM ) + has_rtm_always_abort = true; + } + + /* + * Always force RTM_ALWAYS_ABORT, even if it currently visible. + * If the user explicitly opts to enable TSX, we'll set + * TSX_FORCE_ABORT.ENABLE_RTM and cause RTM_ALWAYS_ABORT to be + * hidden from the general CPUID scan later. + */ + if ( has_rtm_always_abort ) + setup_force_cpu_cap(X86_FEATURE_RTM_ALWAYS_ABORT); + /* * If no explicit tsx= option is provided, pick a default. * @@ -67,9 +99,16 @@ void tsx_init(void) * - parse_spec_ctrl() ran before any CPU details where know. * - We now know we're running on a CPU not affected by TAA (as * TSX_FORCE_ABORT is enumerated). + * - When RTM_ALWAYS_ABORT is enumerated, TSX malfunctions, so we + * only ever want it enabled by explicit user choice. + * + * Without RTM_ALWAYS_ABORT, leave TSX active. In particular, + * this includes SKX where TSX is still supported. + * + * With RTM_ALWAYS_ABORT, disable TSX. */ if ( opt_tsx < 0 ) - opt_tsx = 1; + opt_tsx = !cpu_has_rtm_always_abort; } /* @@ -90,7 +129,7 @@ void tsx_init(void) * Force the features to be visible in Xen's view if we see any of the * infrastructure capable of hiding them. */ - if ( cpu_has_tsx_ctrl ) + if ( cpu_has_tsx_ctrl || cpu_has_tsx_force_abort ) { setup_force_cpu_cap(X86_FEATURE_HLE); setup_force_cpu_cap(X86_FEATURE_RTM); @@ -131,9 +170,36 @@ void tsx_init(void) /* Check bottom bit only. Higher bits are various sentinels. */ rtm_disabled = !(opt_tsx & 1); - lo &= ~TSX_FORCE_ABORT_RTM; - if ( rtm_disabled ) - lo |= TSX_FORCE_ABORT_RTM; + lo &= ~(TSX_FORCE_ABORT_RTM | TSX_CPUID_CLEAR | TSX_ENABLE_RTM); + + if ( cpu_has_rtm_always_abort ) + { + /* + * June 2021 microcode, on a client part with TSX de-featured: + * - There are no mitigations for the TSX memory ordering errata. + * - Performance counter 3 works. (I.e. it isn't being used by + * microcode to work around the memory ordering errata.) + * - TSX_FORCE_ABORT.FORCE_ABORT_RTM is fixed read1/write-discard. + * - TSX_FORCE_ABORT.TSX_CPUID_CLEAR can be used to hide the + * HLE/RTM CPUID bits. + * - TSX_FORCE_ABORT.ENABLE_RTM may be used to opt in to + * re-enabling RTM, at the users own risk. + */ + lo |= rtm_disabled ? TSX_CPUID_CLEAR : TSX_ENABLE_RTM; + } + else + { + /* + * Either a server part where TSX isn't de-featured, or pre-June + * 2021 microcode: + * - By default, the TSX memory ordering errata is worked around + * in microcode at the cost of Performance Counter 3. + * - "Working TSX" vs "Working PCR3" can be selected by way of + * setting TSX_FORCE_ABORT.FORCE_ABORT_RTM. + */ + if ( rtm_disabled ) + lo |= TSX_FORCE_ABORT_RTM; + } wrmsr(MSR_TSX_FORCE_ABORT, lo, hi); } diff --git a/xen/include/asm-x86/cpufeature.h b/xen/include/asm-x86/cpufeature.h index 9f5ae3aa0d..a539a4bacd 100644 --- a/xen/include/asm-x86/cpufeature.h +++ b/xen/include/asm-x86/cpufeature.h @@ -131,6 +131,7 @@ #define cpu_has_avx512_4vnniw boot_cpu_has(X86_FEATURE_AVX512_4VNNIW) #define cpu_has_avx512_4fmaps boot_cpu_has(X86_FEATURE_AVX512_4FMAPS) #define cpu_has_avx512_vp2intersect boot_cpu_has(X86_FEATURE_AVX512_VP2INTERSECT) +#define cpu_has_rtm_always_abort boot_cpu_has(X86_FEATURE_RTM_ALWAYS_ABORT) #define cpu_has_tsx_force_abort boot_cpu_has(X86_FEATURE_TSX_FORCE_ABORT) #define cpu_has_serialize boot_cpu_has(X86_FEATURE_SERIALIZE) #define cpu_has_arch_caps boot_cpu_has(X86_FEATURE_ARCH_CAPS) diff --git a/xen/include/asm-x86/msr-index.h b/xen/include/asm-x86/msr-index.h index bd3a3a1e7f..9a772c12b8 100644 --- a/xen/include/asm-x86/msr-index.h +++ b/xen/include/asm-x86/msr-index.h @@ -61,6 +61,8 @@ #define MSR_TSX_FORCE_ABORT 0x0000010f #define TSX_FORCE_ABORT_RTM (_AC(1, ULL) << 0) +#define TSX_CPUID_CLEAR (_AC(1, ULL) << 1) +#define TSX_ENABLE_RTM (_AC(1, ULL) << 2) #define MSR_TSX_CTRL 0x00000122 #define TSX_CTRL_RTM_DISABLE (_AC(1, ULL) << 0) diff --git a/xen/include/public/arch-x86/cpufeatureset.h b/xen/include/public/arch-x86/cpufeatureset.h index 6477f75505..b4f3d622a0 100644 --- a/xen/include/public/arch-x86/cpufeatureset.h +++ b/xen/include/public/arch-x86/cpufeatureset.h @@ -263,6 +263,7 @@ XEN_CPUFEATURE(AVX512_4FMAPS, 9*32+ 3) /*A AVX512 Multiply Accumulation Single XEN_CPUFEATURE(AVX512_VP2INTERSECT, 9*32+8) /*a VP2INTERSECT{D,Q} insns */ XEN_CPUFEATURE(SRBDS_CTRL, 9*32+ 9) /* MSR_MCU_OPT_CTRL and RNGDS_MITG_DIS. */ XEN_CPUFEATURE(MD_CLEAR, 9*32+10) /*A VERW clears microarchitectural buffers */ +XEN_CPUFEATURE(RTM_ALWAYS_ABORT, 9*32+11) /*! June 2021 TSX defeaturing in microcode. */ XEN_CPUFEATURE(TSX_FORCE_ABORT, 9*32+13) /* MSR_TSX_FORCE_ABORT.RTM_ABORT */ XEN_CPUFEATURE(SERIALIZE, 9*32+14) /*a SERIALIZE insn */ XEN_CPUFEATURE(CET_IBT, 9*32+20) /* CET - Indirect Branch Tracking */ -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.15 From xen-changelog-bounces@lists.xenproject.org Fri Jun 18 11:00:11 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Fri, 18 Jun 2021 11:00:11 +0000 Received: from list by lists.xenproject.org with outflank-mailman.144463.265882 (Exim 4.92) (envelope-from ) id 1luCEG-0005xQ-Lu; Fri, 18 Jun 2021 11:00:04 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 144463.265882; Fri, 18 Jun 2021 11:00:04 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1luCEG-0005x0-IC; Fri, 18 Jun 2021 11:00:04 +0000 Received: by outflank-mailman (input) for mailman id 144463; Fri, 18 Jun 2021 11:00:02 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1luCEE-0005jH-QF for xen-changelog@lists.xenproject.org; Fri, 18 Jun 2021 11:00:02 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1luCEE-0006Np-M7 for xen-changelog@lists.xenproject.org; Fri, 18 Jun 2021 11:00:02 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1luCEE-0008RF-Kw for xen-changelog@lists.xenproject.org; Fri, 18 Jun 2021 11:00:02 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=qTP001CqeEEB/Kda5NIx1pFZfbMD82/BfGYdiCO39MA=; b=eQYbcub3T28f7wH2jbGpY2hYPv kLlTEZyTkUWQBp13D3TjOdoUOo/l+QBQbrvT6Fm+TEA0cBPYRJc38oRyC1PbV1jgtcYJ0aZggv/50 num7s+8JUcuAy2nc2y0DrfpRPkENdiYuCP8tFPmgjOvyIYV76BYLsE5uaTKZ1sulFWuQ=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.14] x86/tsx: Minor cleanup and improvements Message-Id: Date: Fri, 18 Jun 2021 11:00:02 +0000 commit 768138cde899685efb5ec5a1c831b1d4e8794f54 Author: Andrew Cooper AuthorDate: Tue May 18 14:53:56 2021 +0100 Commit: Andrew Cooper CommitDate: Thu Jun 17 14:45:37 2021 +0100 x86/tsx: Minor cleanup and improvements * Introduce cpu_has_arch_caps and replace boot_cpu_has(X86_FEATURE_ARCH_CAPS) * Read CPUID data into the appropriate boot_cpu_data.x86_capability[] element, as subsequent changes are going to need more cpu_has_* logic. * Use the hi/lo MSR helpers, which substantially improves code generation. No practical change. Signed-off-by: Andrew Cooper Acked-by: Jan Beulich Reviewed-by: Roger Pau Monné (cherry picked from commit 3670abcaf0324f2aedba0c4dc7939072b27efa1d) --- xen/arch/x86/cpuid.c | 2 +- xen/arch/x86/hvm/vmx/vmx.c | 2 +- xen/arch/x86/msr.c | 2 +- xen/arch/x86/spec_ctrl.c | 2 +- xen/arch/x86/tsx.c | 21 ++++++++++++--------- xen/include/asm-x86/cpufeature.h | 1 + 6 files changed, 17 insertions(+), 13 deletions(-) diff --git a/xen/arch/x86/cpuid.c b/xen/arch/x86/cpuid.c index 425dce7bc7..deae62705d 100644 --- a/xen/arch/x86/cpuid.c +++ b/xen/arch/x86/cpuid.c @@ -720,7 +720,7 @@ int init_domain_cpuid_policy(struct domain *d) * so dom0 can turn off workarounds as appropriate. Temporary, until the * domain policy logic gains a better understanding of MSRs. */ - if ( is_hardware_domain(d) && boot_cpu_has(X86_FEATURE_ARCH_CAPS) ) + if ( is_hardware_domain(d) && cpu_has_arch_caps ) p->feat.arch_caps = true; d->arch.cpuid = p; diff --git a/xen/arch/x86/hvm/vmx/vmx.c b/xen/arch/x86/hvm/vmx/vmx.c index 6972d4ab23..2a29675c1a 100644 --- a/xen/arch/x86/hvm/vmx/vmx.c +++ b/xen/arch/x86/hvm/vmx/vmx.c @@ -2397,7 +2397,7 @@ static bool __init has_if_pschange_mc(void) if ( cpu_has_hypervisor ) return false; - if ( boot_cpu_has(X86_FEATURE_ARCH_CAPS) ) + if ( cpu_has_arch_caps ) rdmsrl(MSR_ARCH_CAPABILITIES, caps); if ( caps & ARCH_CAPS_IF_PSCHANGE_MC_NO ) diff --git a/xen/arch/x86/msr.c b/xen/arch/x86/msr.c index aa107823ac..caef0c6ede 100644 --- a/xen/arch/x86/msr.c +++ b/xen/arch/x86/msr.c @@ -135,7 +135,7 @@ int init_domain_msr_policy(struct domain *d) * so dom0 can turn off workarounds as appropriate. Temporary, until the * domain policy logic gains a better understanding of MSRs. */ - if ( is_hardware_domain(d) && boot_cpu_has(X86_FEATURE_ARCH_CAPS) ) + if ( is_hardware_domain(d) && cpu_has_arch_caps ) { uint64_t val; diff --git a/xen/arch/x86/spec_ctrl.c b/xen/arch/x86/spec_ctrl.c index bea4c3e8b4..aa85a0be46 100644 --- a/xen/arch/x86/spec_ctrl.c +++ b/xen/arch/x86/spec_ctrl.c @@ -885,7 +885,7 @@ void __init init_speculation_mitigations(void) bool cpu_has_bug_taa; uint64_t caps = 0; - if ( boot_cpu_has(X86_FEATURE_ARCH_CAPS) ) + if ( cpu_has_arch_caps ) rdmsrl(MSR_ARCH_CAPABILITIES, caps); hw_smt_enabled = check_smt_enabled(); diff --git a/xen/arch/x86/tsx.c b/xen/arch/x86/tsx.c index e09e819dce..98ecb71a4a 100644 --- a/xen/arch/x86/tsx.c +++ b/xen/arch/x86/tsx.c @@ -34,15 +34,18 @@ void tsx_init(void) { /* * This function is first called between microcode being loaded, and CPUID - * being scanned generally. Calculate from raw data whether MSR_TSX_CTRL - * is available. + * being scanned generally. Read into boot_cpu_data.x86_capability[] for + * the cpu_has_* bits we care about using here. */ if ( unlikely(cpu_has_tsx_ctrl < 0) ) { uint64_t caps = 0; - if ( boot_cpu_data.cpuid_level >= 7 && - (cpuid_count_edx(7, 0) & cpufeat_mask(X86_FEATURE_ARCH_CAPS)) ) + if ( boot_cpu_data.cpuid_level >= 7 ) + boot_cpu_data.x86_capability[cpufeat_word(X86_FEATURE_ARCH_CAPS)] + = cpuid_count_edx(7, 0); + + if ( cpu_has_arch_caps ) rdmsrl(MSR_ARCH_CAPABILITIES, caps); cpu_has_tsx_ctrl = !!(caps & ARCH_CAPS_TSX_CTRL); @@ -74,18 +77,18 @@ void tsx_init(void) if ( cpu_has_tsx_ctrl ) { - uint64_t val; + uint32_t hi, lo; - rdmsrl(MSR_TSX_CTRL, val); + rdmsr(MSR_TSX_CTRL, lo, hi); /* Check bottom bit only. Higher bits are various sentinels. */ rtm_disabled = !(opt_tsx & 1); - val &= ~(TSX_CTRL_RTM_DISABLE | TSX_CTRL_CPUID_CLEAR); + lo &= ~(TSX_CTRL_RTM_DISABLE | TSX_CTRL_CPUID_CLEAR); if ( rtm_disabled ) - val |= TSX_CTRL_RTM_DISABLE | TSX_CTRL_CPUID_CLEAR; + lo |= TSX_CTRL_RTM_DISABLE | TSX_CTRL_CPUID_CLEAR; - wrmsrl(MSR_TSX_CTRL, val); + wrmsr(MSR_TSX_CTRL, lo, hi); } else if ( opt_tsx >= 0 ) printk_once(XENLOG_WARNING diff --git a/xen/include/asm-x86/cpufeature.h b/xen/include/asm-x86/cpufeature.h index f790d5c1f8..9c31032436 100644 --- a/xen/include/asm-x86/cpufeature.h +++ b/xen/include/asm-x86/cpufeature.h @@ -130,6 +130,7 @@ #define cpu_has_avx512_4fmaps boot_cpu_has(X86_FEATURE_AVX512_4FMAPS) #define cpu_has_tsx_force_abort boot_cpu_has(X86_FEATURE_TSX_FORCE_ABORT) #define cpu_has_serialize boot_cpu_has(X86_FEATURE_SERIALIZE) +#define cpu_has_arch_caps boot_cpu_has(X86_FEATURE_ARCH_CAPS) /* CPUID level 0x00000007:1.eax */ #define cpu_has_avx512_bf16 boot_cpu_has(X86_FEATURE_AVX512_BF16) -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.14 From xen-changelog-bounces@lists.xenproject.org Fri Jun 18 11:00:13 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Fri, 18 Jun 2021 11:00:13 +0000 Received: from list by lists.xenproject.org with outflank-mailman.144464.265886 (Exim 4.92) (envelope-from ) id 1luCEP-0006K4-Ng; Fri, 18 Jun 2021 11:00:13 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 144464.265886; Fri, 18 Jun 2021 11:00:13 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1luCEP-0006Jw-KP; Fri, 18 Jun 2021 11:00:13 +0000 Received: by outflank-mailman (input) for mailman id 144464; Fri, 18 Jun 2021 11:00:12 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1luCEO-0006Jq-RF for xen-changelog@lists.xenproject.org; Fri, 18 Jun 2021 11:00:12 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1luCEO-0006Nt-QU for xen-changelog@lists.xenproject.org; Fri, 18 Jun 2021 11:00:12 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1luCEO-0008Sk-Og for xen-changelog@lists.xenproject.org; Fri, 18 Jun 2021 11:00:12 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=jk/szYL48Q8p30uu4aniPSAP1OXlaQyNScvTAMc11NQ=; b=w/cVOeC/39BTNCYV+BSnPI5IVv MZb1ETqzE2Eyx62HQto5Wvi0zNBt7bIXdCfXvRbaQf41DKnbaJHFm2cTvD6o9+L8h32ljwzPrMaNC hFkN/w8cfsKTBrvKapuwCJLJCnFuJYQ+zjgvFzh/z+hOwuPPFjiYfQjsFGERDwwzwXZY=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.14] x86/tsx: Deprecate vpmu=rtm-abort and use tsx= instead Message-Id: Date: Fri, 18 Jun 2021 11:00:12 +0000 commit 4b4ee05a70d6193fa7c18650743c72b609cd272b Author: Andrew Cooper AuthorDate: Thu May 20 19:16:11 2021 +0100 Commit: Andrew Cooper CommitDate: Thu Jun 17 14:45:37 2021 +0100 x86/tsx: Deprecate vpmu=rtm-abort and use tsx= instead This reuses the rtm_disable infrastructure, so CPUID derivation works properly when TSX is disabled in favour of working PCR3. vpmu= is not a supported feature, and having this functionality under tsx= centralises all TSX handling. Signed-off-by: Andrew Cooper Acked-by: Jan Beulich Reviewed-by: Roger Pau Monné (cherry picked from commit 9fdcf851689cb2a9501d3947cb5d767d9c7797e8) --- docs/misc/xen-command-line.pandoc | 40 +++++++++++++++--------------- xen/arch/x86/cpu/intel.c | 3 --- xen/arch/x86/cpu/vpmu.c | 4 +-- xen/arch/x86/tsx.c | 51 +++++++++++++++++++++++++++++++++++++-- xen/include/asm-x86/vpmu.h | 1 - 5 files changed, 70 insertions(+), 29 deletions(-) diff --git a/docs/misc/xen-command-line.pandoc b/docs/misc/xen-command-line.pandoc index b9824d443f..6c4e340824 100644 --- a/docs/misc/xen-command-line.pandoc +++ b/docs/misc/xen-command-line.pandoc @@ -2231,14 +2231,21 @@ pages) must also be specified via the tbuf_size parameter. Controls for the use of Transactional Synchronization eXtensions. -On Intel parts released in Q3 2019 (with updated microcode), and future parts, -a control has been introduced which allows TSX to be turned off. +Several microcode updates are relevant: -On systems with the ability to turn TSX off, this boolean offers system wide -control of whether TSX is enabled or disabled. + * March 2019, fixing the TSX memory ordering errata on all TSX-enabled CPUs + to date. Introduced MSR_TSX_FORCE_ABORT on SKL/SKX/KBL/WHL/CFL parts. The + errata workaround uses Performance Counter 3, so the user can select + between working TSX and working perfcounters. + + * November 2019, fixing the TSX Async Abort speculative vulnerability. + Introduced MSR_TSX_CTRL on all TSX-enabled MDS_NO parts to date, + CLX/WHL-R/CFL-R, with the controls becoming architectural moving forward + and formally retiring HLE from the architecture. The user can disable TSX + to mitigate TAA, and elect to hide the HLE/RTM CPUID bits. -On parts vulnerable to CVE-2019-11135 / TSX Asynchronous Abort, the following -logic applies: +On systems with the ability to configure TSX, this boolean offers system wide +control of whether TSX is enabled or disabled. * An explicit `tsx=` choice is honoured, even if it is `true` and would result in a vulnerable system. @@ -2246,10 +2253,14 @@ logic applies: * When no explicit `tsx=` choice is given, parts vulnerable to TAA will be mitigated by disabling TSX, as this is the lowest overhead option. - * If the use of TSX is important, the more expensive TAA mitigations can be + If the use of TSX is important, the more expensive TAA mitigations can be opted in to with `smt=0 spec-ctrl=md-clear`, at which point TSX will remain active by default. + * When no explicit `tsx=` option is given, parts susceptible to the memory + ordering errata default to `true` to enable working TSX. Alternatively, + selecting `tsx=0` will disable TSX and restore PCR3 to a working state. + ### ucode > `= List of [ | scan=, nmi= ]` @@ -2386,20 +2397,7 @@ provide access to a wealth of low level processor information. * The `arch` option allows access to the pre-defined architectural events. -* The `rtm-abort` boolean controls a trade-off between working Restricted - Transactional Memory, and working performance counters. - - All processors released to date (Q1 2019) supporting Transactional Memory - Extensions suffer an erratum which has been addressed in microcode. - - Processors based on the Skylake microarchitecture with up-to-date - microcode internally use performance counter 3 to work around the erratum. - A consequence is that the counter gets reprogrammed whenever an `XBEGIN` - instruction is executed. - - An alternative mode exists where PCR3 behaves as before, at the cost of - `XBEGIN` unconditionally aborting. Enabling `rtm-abort` mode will - activate this alternative mode. +* The `rtm-abort` boolean has been superseded. Use `tsx=0` instead. *Warning:* As the virtualisation is not 100% safe, don't use the vpmu flag on diff --git a/xen/arch/x86/cpu/intel.c b/xen/arch/x86/cpu/intel.c index 37439071d9..abf8e206d7 100644 --- a/xen/arch/x86/cpu/intel.c +++ b/xen/arch/x86/cpu/intel.c @@ -356,9 +356,6 @@ static void Intel_errata_workarounds(struct cpuinfo_x86 *c) (c->x86_model == 29 || c->x86_model == 46 || c->x86_model == 47)) __set_bit(X86_FEATURE_CLFLUSH_MONITOR, c->x86_capability); - if (cpu_has_tsx_force_abort && opt_rtm_abort) - wrmsrl(MSR_TSX_FORCE_ABORT, TSX_FORCE_ABORT_RTM); - probe_c3_errata(c); } diff --git a/xen/arch/x86/cpu/vpmu.c b/xen/arch/x86/cpu/vpmu.c index 1ed39ef03f..0c3f378a02 100644 --- a/xen/arch/x86/cpu/vpmu.c +++ b/xen/arch/x86/cpu/vpmu.c @@ -47,7 +47,6 @@ CHECK_pmu_params; static unsigned int __read_mostly opt_vpmu_enabled; unsigned int __read_mostly vpmu_mode = XENPMU_MODE_OFF; unsigned int __read_mostly vpmu_features = 0; -bool __read_mostly opt_rtm_abort; static DEFINE_SPINLOCK(vpmu_lock); static unsigned vpmu_count; @@ -77,7 +76,8 @@ static int __init parse_vpmu_params(const char *s) else if ( !cmdline_strcmp(s, "arch") ) vpmu_features |= XENPMU_FEATURE_ARCH_ONLY; else if ( (val = parse_boolean("rtm-abort", s, ss)) >= 0 ) - opt_rtm_abort = val; + printk(XENLOG_WARNING + "'rtm-abort=' superseded. Use 'tsx=' instead\n"); else rc = -EINVAL; diff --git a/xen/arch/x86/tsx.c b/xen/arch/x86/tsx.c index 98ecb71a4a..338191df7f 100644 --- a/xen/arch/x86/tsx.c +++ b/xen/arch/x86/tsx.c @@ -6,7 +6,9 @@ * Valid values: * 1 => Explicit tsx=1 * 0 => Explicit tsx=0 - * -1 => Default, implicit tsx=1, may change to 0 to mitigate TAA + * -1 => Default, altered to 0/1 (if unspecified) by: + * - TAA heuristics/settings for speculative safety + * - "TSX vs PCR3" select for TSX memory ordering safety * -3 => Implicit tsx=1 (feed-through from spec-ctrl=0) * * This is arranged such that the bottom bit encodes whether TSX is actually @@ -50,6 +52,26 @@ void tsx_init(void) cpu_has_tsx_ctrl = !!(caps & ARCH_CAPS_TSX_CTRL); + if ( cpu_has_tsx_force_abort ) + { + /* + * On an early TSX-enable Skylake part subject to the memory + * ordering erratum, with at least the March 2019 microcode. + */ + + /* + * If no explicit tsx= option is provided, pick a default. + * + * This deliberately overrides the implicit opt_tsx=-3 from + * `spec-ctrl=0` because: + * - parse_spec_ctrl() ran before any CPU details where know. + * - We now know we're running on a CPU not affected by TAA (as + * TSX_FORCE_ABORT is enumerated). + */ + if ( opt_tsx < 0 ) + opt_tsx = 1; + } + /* * The TSX features (HLE/RTM) are handled specially. They both * enumerate features but, on certain parts, have mechanisms to be @@ -75,6 +97,12 @@ void tsx_init(void) } } + /* + * Note: MSR_TSX_CTRL is enumerated on TSX-enabled MDS_NO and later parts. + * MSR_TSX_FORCE_ABORT is enumerated on TSX-enabled pre-MDS_NO Skylake + * parts only. The two features are on a disjoint set of CPUs, and not + * offered to guests by hypervisors. + */ if ( cpu_has_tsx_ctrl ) { uint32_t hi, lo; @@ -90,9 +118,28 @@ void tsx_init(void) wrmsr(MSR_TSX_CTRL, lo, hi); } + else if ( cpu_has_tsx_force_abort ) + { + /* + * On an early TSX-enable Skylake part subject to the memory ordering + * erratum, with at least the March 2019 microcode. + */ + uint32_t hi, lo; + + rdmsr(MSR_TSX_FORCE_ABORT, lo, hi); + + /* Check bottom bit only. Higher bits are various sentinels. */ + rtm_disabled = !(opt_tsx & 1); + + lo &= ~TSX_FORCE_ABORT_RTM; + if ( rtm_disabled ) + lo |= TSX_FORCE_ABORT_RTM; + + wrmsr(MSR_TSX_FORCE_ABORT, lo, hi); + } else if ( opt_tsx >= 0 ) printk_once(XENLOG_WARNING - "MSR_TSX_CTRL not available - Ignoring tsx= setting\n"); + "TSX controls not available - Ignoring tsx= setting\n"); } /* diff --git a/xen/include/asm-x86/vpmu.h b/xen/include/asm-x86/vpmu.h index 55f85ba00f..4b0a6ba3da 100644 --- a/xen/include/asm-x86/vpmu.h +++ b/xen/include/asm-x86/vpmu.h @@ -126,7 +126,6 @@ static inline int vpmu_do_rdmsr(unsigned int msr, uint64_t *msr_content) extern unsigned int vpmu_mode; extern unsigned int vpmu_features; -extern bool opt_rtm_abort; /* Context switch */ static inline void vpmu_switch_from(struct vcpu *prev) -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.14 From xen-changelog-bounces@lists.xenproject.org Fri Jun 18 11:00:23 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Fri, 18 Jun 2021 11:00:23 +0000 Received: from list by lists.xenproject.org with outflank-mailman.144465.265890 (Exim 4.92) (envelope-from ) id 1luCEZ-0006Mv-P8; Fri, 18 Jun 2021 11:00:23 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 144465.265890; Fri, 18 Jun 2021 11:00:23 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1luCEZ-0006Mn-M9; Fri, 18 Jun 2021 11:00:23 +0000 Received: by outflank-mailman (input) for mailman id 144465; Fri, 18 Jun 2021 11:00:23 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1luCEY-0006Md-Ub for xen-changelog@lists.xenproject.org; Fri, 18 Jun 2021 11:00:22 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1luCEY-0006OB-Tk for xen-changelog@lists.xenproject.org; Fri, 18 Jun 2021 11:00:22 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1luCEY-0008Tk-Sv for xen-changelog@lists.xenproject.org; Fri, 18 Jun 2021 11:00:22 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=lW2kA1DNwhHeXkPqL66FxHhXM8o2A6p960RIuH1cDUY=; b=GSixsW7/EOMRnvNIxb2c488xiQ 7pCer/ZoxqUkaFCvZcbTAznWUzSw7gaqMklJyDGiMpUYismwkZ2qLNKECpR35LoirveB270C6a4it NnBPH4nRVZFnwO63GRBn1ZuTFsY8r1V6gb/aXfSYmm5X00zKXSe1STG7oYXw9T0eQPKU=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.14] x86/cpuid: Fix HLE and RTM handling (again) Message-Id: Date: Fri, 18 Jun 2021 11:00:22 +0000 commit ee5425c0d532a773093cc4ff91c4c06caf2ff447 Author: Andrew Cooper AuthorDate: Mon Jun 7 13:25:09 2021 +0100 Commit: Andrew Cooper CommitDate: Thu Jun 17 14:45:37 2021 +0100 x86/cpuid: Fix HLE and RTM handling (again) For reasons which are my fault, but I don't recall why, the FDP_EXCP_ONLY/NO_FPU_SEL adjustment uses the whole special_features[] array element, not the two relevant bits. HLE and RTM were recently added to the list of special features, causing them to be always set in guest view, irrespective of the toolstacks choice on the matter. Rewrite the logic to refer to the features specifically, rather than relying on the contents of the special_features[] array. Fixes: 8fe24090d9 ("x86/cpuid: Rework HLE and RTM handling") Reported-by: Edwin Török Signed-off-by: Andrew Cooper Reviewed-by: Jan Beulich (cherry picked from commit 60fa12dbf1d4d2c4ffe1ef34b495b24aa7e41aa0) --- xen/arch/x86/cpuid.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/xen/arch/x86/cpuid.c b/xen/arch/x86/cpuid.c index deae62705d..5229eba595 100644 --- a/xen/arch/x86/cpuid.c +++ b/xen/arch/x86/cpuid.c @@ -648,9 +648,11 @@ void recalculate_cpuid_policy(struct domain *d) sanitise_featureset(fs); /* Fold host's FDP_EXCP_ONLY and NO_FPU_SEL into guest's view. */ - fs[FEATURESET_7b0] &= ~special_features[FEATURESET_7b0]; + fs[FEATURESET_7b0] &= ~(cpufeat_mask(X86_FEATURE_FDP_EXCP_ONLY) | + cpufeat_mask(X86_FEATURE_NO_FPU_SEL)); fs[FEATURESET_7b0] |= (host_cpuid_policy.feat._7b0 & - special_features[FEATURESET_7b0]); + (cpufeat_mask(X86_FEATURE_FDP_EXCP_ONLY) | + cpufeat_mask(X86_FEATURE_NO_FPU_SEL))); cpuid_featureset_to_policy(fs, p); -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.14 From xen-changelog-bounces@lists.xenproject.org Fri Jun 18 11:00:33 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Fri, 18 Jun 2021 11:00:33 +0000 Received: from list by lists.xenproject.org with outflank-mailman.144466.265894 (Exim 4.92) (envelope-from ) id 1luCEj-0006Qy-Qv; Fri, 18 Jun 2021 11:00:33 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 144466.265894; Fri, 18 Jun 2021 11:00:33 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1luCEj-0006Qp-Ng; Fri, 18 Jun 2021 11:00:33 +0000 Received: by outflank-mailman (input) for mailman id 144466; Fri, 18 Jun 2021 11:00:33 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1luCEj-0006Qd-1a for xen-changelog@lists.xenproject.org; Fri, 18 Jun 2021 11:00:33 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1luCEj-0006OS-0l for xen-changelog@lists.xenproject.org; Fri, 18 Jun 2021 11:00:33 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1luCEj-0008Up-01 for xen-changelog@lists.xenproject.org; Fri, 18 Jun 2021 11:00:33 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=uXIGZp7wxxXvIxkkypUYEdNTv95ZEgVlSLWCq/d8gwY=; b=nSEndROh45ZNL0G/25D4nul8ky cbDtzVMmkUw63WPkUF2UjE7IZXoZCUiCk8Wd98PzoP00sYOsnNqofRirHN2oFbz8qZNMWEOW8SPRG xFiuyJg7AfP3dtUy1HJepd8nQMbKuRzL43zxbIVV+ZnaXcXac/5ojnNnCLmA7//wVaDE=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.14] x86/tsx: Cope with TSX deprecation on SKL/KBL/CFL/WHL Message-Id: Date: Fri, 18 Jun 2021 11:00:33 +0000 commit 45710c02563602343261b690787bb1c76a676bef Author: Andrew Cooper AuthorDate: Wed Sep 16 16:15:52 2020 +0100 Commit: Andrew Cooper CommitDate: Thu Jun 17 14:49:05 2021 +0100 x86/tsx: Cope with TSX deprecation on SKL/KBL/CFL/WHL The June 2021 microcode is formally de-featuring TSX on the older Skylake client CPUs. The workaround from the March 2019 microcode is being dropped, and replaced with additions to MSR_TSX_FORCE_ABORT to hide the HLE/RTM CPUID bits. With this microcode in place, TSX is disabled by default on these CPUs. Backwards compatibility is provided in the same way as for TAA - RTM force aborts, rather than suffering #UD, and the CPUID bits can be hidden to recover performance. Signed-off-by: Andrew Cooper Reviewed-by: Jan Beulich (cherry picked from commit 3e09045991cde360432bc7437103f8f8a6699359) --- docs/misc/xen-command-line.pandoc | 13 +++++ tools/misc/xen-cpuid.c | 2 +- xen/arch/x86/tsx.c | 76 +++++++++++++++++++++++++++-- xen/include/asm-x86/cpufeature.h | 1 + xen/include/asm-x86/msr-index.h | 2 + xen/include/public/arch-x86/cpufeatureset.h | 1 + 6 files changed, 89 insertions(+), 6 deletions(-) diff --git a/docs/misc/xen-command-line.pandoc b/docs/misc/xen-command-line.pandoc index 6c4e340824..a889c23c64 100644 --- a/docs/misc/xen-command-line.pandoc +++ b/docs/misc/xen-command-line.pandoc @@ -2244,6 +2244,12 @@ Several microcode updates are relevant: and formally retiring HLE from the architecture. The user can disable TSX to mitigate TAA, and elect to hide the HLE/RTM CPUID bits. + * June 2021, removing the workaround for March 2019 on client CPUs and + formally de-featured TSX on SKL/KBL/WHL/CFL (Note: SKX still retains the + March 2019 fix). Introduced the ability to hide the HLE/RTM CPUID bits. + PCR3 works fine, and TSX is disabled by default, but the user can re-enable + TSX at their own risk, accepting that the memory order erratum is unfixed. + On systems with the ability to configure TSX, this boolean offers system wide control of whether TSX is enabled or disabled. @@ -2261,6 +2267,13 @@ control of whether TSX is enabled or disabled. ordering errata default to `true` to enable working TSX. Alternatively, selecting `tsx=0` will disable TSX and restore PCR3 to a working state. + SKX and SKL/KBL/WHL/CFL on pre-June 2021 microcode default to `true`. + Alternatively, selecting `tsx=0` will disable TSX and restore PCR3 to a + working state. + + SKL/KBL/WHL/CFL on the June 2021 microcode or later default to `false`. + Alternatively, selecting `tsx=1` will re-enable TSX at the users own risk. + ### ucode > `= List of [ | scan=, nmi= ]` diff --git a/tools/misc/xen-cpuid.c b/tools/misc/xen-cpuid.c index 8d8f3532a2..a231d96e13 100644 --- a/tools/misc/xen-cpuid.c +++ b/tools/misc/xen-cpuid.c @@ -161,7 +161,7 @@ static const char *const str_7d0[32] = [ 4] = "fsrm", /* 8 */ [ 9] = "srbds-ctrl", - [10] = "md-clear", + [10] = "md-clear", [11] = "rtm-always-abort", /* 12 */ [13] = "tsx-force-abort", [14] = "serialize", diff --git a/xen/arch/x86/tsx.c b/xen/arch/x86/tsx.c index 338191df7f..88adf08c49 100644 --- a/xen/arch/x86/tsx.c +++ b/xen/arch/x86/tsx.c @@ -59,6 +59,38 @@ void tsx_init(void) * ordering erratum, with at least the March 2019 microcode. */ + /* + * Probe for the June 2021 microcode which de-features TSX on + * client parts. (Note - this is a subset of parts impacted by + * the memory ordering errata.) + * + * RTM_ALWAYS_ABORT enumerates the new functionality, but is also + * read as zero if TSX_FORCE_ABORT.ENABLE_RTM has been set before + * we run. + * + * Undo this behaviour in Xen's view of the world. + */ + bool has_rtm_always_abort = cpu_has_rtm_always_abort; + + if ( !has_rtm_always_abort ) + { + uint64_t val; + + rdmsrl(MSR_TSX_FORCE_ABORT, val); + + if ( val & TSX_ENABLE_RTM ) + has_rtm_always_abort = true; + } + + /* + * Always force RTM_ALWAYS_ABORT, even if it currently visible. + * If the user explicitly opts to enable TSX, we'll set + * TSX_FORCE_ABORT.ENABLE_RTM and cause RTM_ALWAYS_ABORT to be + * hidden from the general CPUID scan later. + */ + if ( has_rtm_always_abort ) + setup_force_cpu_cap(X86_FEATURE_RTM_ALWAYS_ABORT); + /* * If no explicit tsx= option is provided, pick a default. * @@ -67,9 +99,16 @@ void tsx_init(void) * - parse_spec_ctrl() ran before any CPU details where know. * - We now know we're running on a CPU not affected by TAA (as * TSX_FORCE_ABORT is enumerated). + * - When RTM_ALWAYS_ABORT is enumerated, TSX malfunctions, so we + * only ever want it enabled by explicit user choice. + * + * Without RTM_ALWAYS_ABORT, leave TSX active. In particular, + * this includes SKX where TSX is still supported. + * + * With RTM_ALWAYS_ABORT, disable TSX. */ if ( opt_tsx < 0 ) - opt_tsx = 1; + opt_tsx = !cpu_has_rtm_always_abort; } /* @@ -90,7 +129,7 @@ void tsx_init(void) * Force the features to be visible in Xen's view if we see any of the * infrastructure capable of hiding them. */ - if ( cpu_has_tsx_ctrl ) + if ( cpu_has_tsx_ctrl || cpu_has_tsx_force_abort ) { setup_force_cpu_cap(X86_FEATURE_HLE); setup_force_cpu_cap(X86_FEATURE_RTM); @@ -131,9 +170,36 @@ void tsx_init(void) /* Check bottom bit only. Higher bits are various sentinels. */ rtm_disabled = !(opt_tsx & 1); - lo &= ~TSX_FORCE_ABORT_RTM; - if ( rtm_disabled ) - lo |= TSX_FORCE_ABORT_RTM; + lo &= ~(TSX_FORCE_ABORT_RTM | TSX_CPUID_CLEAR | TSX_ENABLE_RTM); + + if ( cpu_has_rtm_always_abort ) + { + /* + * June 2021 microcode, on a client part with TSX de-featured: + * - There are no mitigations for the TSX memory ordering errata. + * - Performance counter 3 works. (I.e. it isn't being used by + * microcode to work around the memory ordering errata.) + * - TSX_FORCE_ABORT.FORCE_ABORT_RTM is fixed read1/write-discard. + * - TSX_FORCE_ABORT.TSX_CPUID_CLEAR can be used to hide the + * HLE/RTM CPUID bits. + * - TSX_FORCE_ABORT.ENABLE_RTM may be used to opt in to + * re-enabling RTM, at the users own risk. + */ + lo |= rtm_disabled ? TSX_CPUID_CLEAR : TSX_ENABLE_RTM; + } + else + { + /* + * Either a server part where TSX isn't de-featured, or pre-June + * 2021 microcode: + * - By default, the TSX memory ordering errata is worked around + * in microcode at the cost of Performance Counter 3. + * - "Working TSX" vs "Working PCR3" can be selected by way of + * setting TSX_FORCE_ABORT.FORCE_ABORT_RTM. + */ + if ( rtm_disabled ) + lo |= TSX_FORCE_ABORT_RTM; + } wrmsr(MSR_TSX_FORCE_ABORT, lo, hi); } diff --git a/xen/include/asm-x86/cpufeature.h b/xen/include/asm-x86/cpufeature.h index 9c31032436..80eff688d4 100644 --- a/xen/include/asm-x86/cpufeature.h +++ b/xen/include/asm-x86/cpufeature.h @@ -128,6 +128,7 @@ /* CPUID level 0x00000007:0.edx */ #define cpu_has_avx512_4vnniw boot_cpu_has(X86_FEATURE_AVX512_4VNNIW) #define cpu_has_avx512_4fmaps boot_cpu_has(X86_FEATURE_AVX512_4FMAPS) +#define cpu_has_rtm_always_abort boot_cpu_has(X86_FEATURE_RTM_ALWAYS_ABORT) #define cpu_has_tsx_force_abort boot_cpu_has(X86_FEATURE_TSX_FORCE_ABORT) #define cpu_has_serialize boot_cpu_has(X86_FEATURE_SERIALIZE) #define cpu_has_arch_caps boot_cpu_has(X86_FEATURE_ARCH_CAPS) diff --git a/xen/include/asm-x86/msr-index.h b/xen/include/asm-x86/msr-index.h index d66c92b946..80764266bf 100644 --- a/xen/include/asm-x86/msr-index.h +++ b/xen/include/asm-x86/msr-index.h @@ -61,6 +61,8 @@ #define MSR_TSX_FORCE_ABORT 0x0000010f #define TSX_FORCE_ABORT_RTM (_AC(1, ULL) << 0) +#define TSX_CPUID_CLEAR (_AC(1, ULL) << 1) +#define TSX_ENABLE_RTM (_AC(1, ULL) << 2) #define MSR_TSX_CTRL 0x00000122 #define TSX_CTRL_RTM_DISABLE (_AC(1, ULL) << 0) diff --git a/xen/include/public/arch-x86/cpufeatureset.h b/xen/include/public/arch-x86/cpufeatureset.h index 02b3a2a265..d2422386c0 100644 --- a/xen/include/public/arch-x86/cpufeatureset.h +++ b/xen/include/public/arch-x86/cpufeatureset.h @@ -261,6 +261,7 @@ XEN_CPUFEATURE(AVX512_4VNNIW, 9*32+ 2) /*A AVX512 Neural Network Instructions * XEN_CPUFEATURE(AVX512_4FMAPS, 9*32+ 3) /*A AVX512 Multiply Accumulation Single Precision */ XEN_CPUFEATURE(SRBDS_CTRL, 9*32+ 9) /* MSR_MCU_OPT_CTRL and RNGDS_MITG_DIS. */ XEN_CPUFEATURE(MD_CLEAR, 9*32+10) /*A VERW clears microarchitectural buffers */ +XEN_CPUFEATURE(RTM_ALWAYS_ABORT, 9*32+11) /*! June 2021 TSX defeaturing in microcode. */ XEN_CPUFEATURE(TSX_FORCE_ABORT, 9*32+13) /* MSR_TSX_FORCE_ABORT.RTM_ABORT */ XEN_CPUFEATURE(SERIALIZE, 9*32+14) /*a SERIALIZE insn */ XEN_CPUFEATURE(CET_IBT, 9*32+20) /* CET - Indirect Branch Tracking */ -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.14 From xen-changelog-bounces@lists.xenproject.org Sat Jun 19 07:33:08 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Sat, 19 Jun 2021 07:33:08 +0000 Received: from list by lists.xenproject.org with outflank-mailman.145128.267035 (Exim 4.92) (envelope-from ) id 1luVTT-0005zO-8v; Sat, 19 Jun 2021 07:33:03 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 145128.267035; Sat, 19 Jun 2021 07:33:03 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1luVTT-0005zG-5b; Sat, 19 Jun 2021 07:33:03 +0000 Received: by outflank-mailman (input) for mailman id 145128; Sat, 19 Jun 2021 07:33:02 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1luVTR-0005z9-Vx for xen-changelog@lists.xenproject.org; Sat, 19 Jun 2021 07:33:02 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1luVTR-0001Bk-Ri for xen-changelog@lists.xenproject.org; Sat, 19 Jun 2021 07:33:01 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1luVTR-0003q4-Qf for xen-changelog@lists.xenproject.org; Sat, 19 Jun 2021 07:33:01 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=F6Y6StVNmmFkigp0yA+nIxkXmCYl5WN+e1sFindzbCQ=; b=7DJKmpUHDufPvWHAN0+XOF3HtU juRbOY73FuAelBTfFvl8oWdzUPlBza4EqQ5G6rA1QLlvhDohhDj7XV0+lk2TGTwz3mBlEDwNCMjOx /t1SSM01jr1LkmhheeahbWHSfSmxCfEtM3YzvAq1yBBktwSezZLhguYDsdqhQH2tAQBA=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen stable-4.14] golang/xenlight: fix code generation for python 2.6 Message-Id: Date: Sat, 19 Jun 2021 07:33:01 +0000 commit b6a8c4f72def4d1135ff42660a86276ce2565c8c Author: Nick Rosbrook AuthorDate: Mon Jul 20 19:54:40 2020 -0400 Commit: Andrew Cooper CommitDate: Thu Jun 17 19:45:39 2021 +0100 golang/xenlight: fix code generation for python 2.6 Before python 2.7, str.format() calls required that the format fields were explicitly enumerated, e.g.: '{0} {1}'.format(foo, bar) vs. '{} {}'.format(foo, bar) Currently, gengotypes.py uses the latter pattern everywhere, which means the Go bindings do not build on python 2.6. Use the 2.6 syntax for format() in order to support python 2.6 for now. Signed-off-by: Nick Rosbrook Acked-by: Wei Liu (cherry picked from commit 6d49fbdeab3e687a6818f809ca3d98ac7ced2c8d) --- tools/golang/xenlight/gengotypes.py | 204 ++++++++++++++++++------------------ 1 file changed, 102 insertions(+), 102 deletions(-) diff --git a/tools/golang/xenlight/gengotypes.py b/tools/golang/xenlight/gengotypes.py index 557fecd07b..ebec938224 100644 --- a/tools/golang/xenlight/gengotypes.py +++ b/tools/golang/xenlight/gengotypes.py @@ -3,7 +3,7 @@ import os import sys -sys.path.append('{}/tools/libxl'.format(os.environ['XEN_ROOT'])) +sys.path.append('{0}/tools/libxl'.format(os.environ['XEN_ROOT'])) import idl # Go versions of some builtin types. @@ -73,14 +73,14 @@ def xenlight_golang_define_enum(ty = None): if ty.typename is not None: typename = xenlight_golang_fmt_name(ty.typename) - s += 'type {} int\n'.format(typename) + s += 'type {0} int\n'.format(typename) # Start const block s += 'const(\n' for v in ty.values: name = xenlight_golang_fmt_name(v.name) - s += '{} {} = {}\n'.format(name, typename, v.value) + s += '{0} {1} = {2}\n'.format(name, typename, v.value) # End const block s += ')\n' @@ -99,9 +99,9 @@ def xenlight_golang_define_struct(ty = None, typename = None, nested = False): # Begin struct definition if nested: - s += '{} struct {{\n'.format(name) + s += '{0} struct {{\n'.format(name) else: - s += 'type {} struct {{\n'.format(name) + s += 'type {0} struct {{\n'.format(name) # Write struct fields for f in ty.fields: @@ -111,13 +111,13 @@ def xenlight_golang_define_struct(ty = None, typename = None, nested = False): typename = xenlight_golang_fmt_name(typename) name = xenlight_golang_fmt_name(f.name) - s += '{} []{}\n'.format(name, typename) + s += '{0} []{1}\n'.format(name, typename) else: typename = f.type.typename typename = xenlight_golang_fmt_name(typename) name = xenlight_golang_fmt_name(f.name) - s += '{} {}\n'.format(name, typename) + s += '{0} {1}\n'.format(name, typename) elif isinstance(f.type, idl.Struct): r = xenlight_golang_define_struct(f.type, typename=f.name, nested=True) @@ -132,7 +132,7 @@ def xenlight_golang_define_struct(ty = None, typename = None, nested = False): extras.extend(r[1]) else: - raise Exception('type {} not supported'.format(f.type)) + raise Exception('type {0} not supported'.format(f.type)) # End struct definition s += '}\n' @@ -151,11 +151,11 @@ def xenlight_golang_define_union(ty = None, struct_name = '', union_name = ''): s = '' extras = [] - interface_name = '{}_{}_union'.format(struct_name, ty.keyvar.name) + interface_name = '{0}_{1}_union'.format(struct_name, ty.keyvar.name) interface_name = xenlight_golang_fmt_name(interface_name, exported=False) - s += 'type {} interface {{\n'.format(interface_name) - s += 'is{}()\n'.format(interface_name) + s += 'type {0} interface {{\n'.format(interface_name) + s += 'is{0}()\n'.format(interface_name) s += '}\n' extras.append(s) @@ -165,7 +165,7 @@ def xenlight_golang_define_union(ty = None, struct_name = '', union_name = ''): continue # Define struct - name = '{}_{}_union_{}'.format(struct_name, ty.keyvar.name, f.name) + name = '{0}_{1}_union_{2}'.format(struct_name, ty.keyvar.name, f.name) r = xenlight_golang_define_struct(f.type, typename=name) extras.append(r[0]) extras.extend(r[1]) @@ -173,21 +173,21 @@ def xenlight_golang_define_union(ty = None, struct_name = '', union_name = ''): # This typeof trick ensures that the fields used in the cgo struct # used for marshaling are the same as the fields of the union in the # actual C type, and avoids re-defining all of those fields. - s = 'typedef typeof(((struct {} *)NULL)->{}.{}){};' + s = 'typedef typeof(((struct {0} *)NULL)->{1}.{2}){3};' s = s.format(struct_name, union_name, f.name, name) cgo_helpers_preamble.append(s) # Define function to implement 'union' interface name = xenlight_golang_fmt_name(name) - s = 'func (x {}) is{}(){{}}\n'.format(name, interface_name) + s = 'func (x {0}) is{1}(){{}}\n'.format(name, interface_name) extras.append(s) fname = xenlight_golang_fmt_name(ty.keyvar.name) ftype = xenlight_golang_fmt_name(ty.keyvar.type.typename) - s = '{} {}\n'.format(fname, ftype) + s = '{0} {1}\n'.format(fname, ftype) - fname = xenlight_golang_fmt_name('{}_union'.format(ty.keyvar.name)) - s += '{} {}\n'.format(fname, interface_name) + fname = xenlight_golang_fmt_name('{0}_union'.format(ty.keyvar.name)) + s += '{0} {1}\n'.format(fname, interface_name) return (s,extras) @@ -243,7 +243,7 @@ def xenlight_golang_define_from_C(ty = None): Define the fromC marshaling function for the type represented by ty. """ - func = 'func (x *{}) fromC(xc *C.{}) error {{\n {}\n return nil}}\n' + func = 'func (x *{0}) fromC(xc *C.{1}) error {{\n {2}\n return nil}}\n' goname = xenlight_golang_fmt_name(ty.typename) cname = ty.typename @@ -271,7 +271,7 @@ def xenlight_golang_define_from_C(ty = None): extras.extend(r[1]) else: - raise Exception('type {} not supported'.format(f.type)) + raise Exception('type {0} not supported'.format(f.type)) return (func.format(goname, cname, body), extras) @@ -300,8 +300,8 @@ def xenlight_golang_convert_from_C(ty = None, outer_name = None, cvarname = None # If outer_name is set, treat this as nested. if outer_name is not None: - goname = '{}.{}'.format(xenlight_golang_fmt_name(outer_name), goname) - cname = '{}.{}'.format(outer_name, cname) + goname = '{0}.{1}'.format(xenlight_golang_fmt_name(outer_name), goname) + cname = '{0}.{1}'.format(outer_name, cname) # Types that satisfy this condition can be easily casted or # converted to a Go builtin type. @@ -312,15 +312,15 @@ def xenlight_golang_convert_from_C(ty = None, outer_name = None, cvarname = None if not is_castable: # If the type is not castable, we need to call its fromC # function. - s += 'if err := x.{}.fromC(&{}.{});'.format(goname,cvarname,cname) - s += 'err != nil {{\nreturn fmt.Errorf("converting field {}: %v", err)\n}}\n'.format(goname) + s += 'if err := x.{0}.fromC(&{1}.{2});'.format(goname,cvarname,cname) + s += 'err != nil {{\nreturn fmt.Errorf("converting field {0}: %v", err)\n}}\n'.format(goname) elif gotypename == 'string': # Use the cgo helper for converting C strings. - s += 'x.{} = C.GoString({}.{})\n'.format(goname,cvarname,cname) + s += 'x.{0} = C.GoString({1}.{2})\n'.format(goname,cvarname,cname) else: - s += 'x.{} = {}({}.{})\n'.format(goname,gotypename,cvarname,cname) + s += 'x.{0} = {1}({2}.{3})\n'.format(goname,gotypename,cvarname,cname) return s @@ -331,9 +331,9 @@ def xenlight_golang_union_from_C(ty = None, union_name = '', struct_name = ''): gokeyname = xenlight_golang_fmt_name(keyname) keytype = ty.keyvar.type.typename gokeytype = xenlight_golang_fmt_name(keytype) - field_name = xenlight_golang_fmt_name('{}_union'.format(keyname)) + field_name = xenlight_golang_fmt_name('{0}_union'.format(keyname)) - interface_name = '{}_{}_union'.format(struct_name, keyname) + interface_name = '{0}_{1}_union'.format(struct_name, keyname) interface_name = xenlight_golang_fmt_name(interface_name, exported=False) cgo_keyname = keyname @@ -343,7 +343,7 @@ def xenlight_golang_union_from_C(ty = None, union_name = '', struct_name = ''): cases = {} for f in ty.fields: - val = '{}_{}'.format(keytype, f.name) + val = '{0}_{1}'.format(keytype, f.name) val = xenlight_golang_fmt_name(val) # Add to list of cases to make for the switch @@ -354,17 +354,17 @@ def xenlight_golang_union_from_C(ty = None, union_name = '', struct_name = ''): continue # Define fromC func for 'union' struct. - typename = '{}_{}_union_{}'.format(struct_name,keyname,f.name) + typename = '{0}_{1}_union_{2}'.format(struct_name,keyname,f.name) gotypename = xenlight_golang_fmt_name(typename) # Define the function here. The cases for keyed unions are a little # different. - s = 'func (x *{}) fromC(xc *C.{}) error {{\n'.format(gotypename,struct_name) - s += 'if {}(xc.{}) != {} {{\n'.format(gokeytype,cgo_keyname,val) - err_string = '"expected union key {}"'.format(val) - s += 'return errors.New({})\n'.format(err_string) + s = 'func (x *{0}) fromC(xc *C.{1}) error {{\n'.format(gotypename,struct_name) + s += 'if {0}(xc.{1}) != {2} {{\n'.format(gokeytype,cgo_keyname,val) + err_string = '"expected union key {0}"'.format(val) + s += 'return errors.New({0})\n'.format(err_string) s += '}\n\n' - s += 'tmp := (*C.{})(unsafe.Pointer(&xc.{}[0]))\n'.format(typename,union_name) + s += 'tmp := (*C.{0})(unsafe.Pointer(&xc.{1}[0]))\n'.format(typename,union_name) for nf in f.type.fields: s += xenlight_golang_convert_from_C(nf,cvarname='tmp') @@ -374,35 +374,35 @@ def xenlight_golang_union_from_C(ty = None, union_name = '', struct_name = ''): extras.append(s) - s = 'x.{} = {}(xc.{})\n'.format(gokeyname,gokeytype,cgo_keyname) - s += 'switch x.{}{{\n'.format(gokeyname) + s = 'x.{0} = {1}(xc.{2})\n'.format(gokeyname,gokeytype,cgo_keyname) + s += 'switch x.{0}{{\n'.format(gokeyname) # Create switch statement to determine which 'union element' # to populate in the Go struct. for case_name, case_tuple in sorted(cases.items()): (case_val, case_type) = case_tuple - s += 'case {}:\n'.format(case_val) + s += 'case {0}:\n'.format(case_val) if case_type is None: - s += "x.{} = nil\n".format(field_name) + s += "x.{0} = nil\n".format(field_name) continue - gotype = '{}_{}_union_{}'.format(struct_name,keyname,case_name) + gotype = '{0}_{1}_union_{2}'.format(struct_name,keyname,case_name) gotype = xenlight_golang_fmt_name(gotype) - goname = '{}_{}'.format(keyname,case_name) + goname = '{0}_{1}'.format(keyname,case_name) goname = xenlight_golang_fmt_name(goname,exported=False) - s += 'var {} {}\n'.format(goname, gotype) - s += 'if err := {}.fromC(xc);'.format(goname) - s += 'err != nil {{\n return fmt.Errorf("converting field {}: %v", err)\n}}\n'.format(goname) + s += 'var {0} {1}\n'.format(goname, gotype) + s += 'if err := {0}.fromC(xc);'.format(goname) + s += 'err != nil {{\n return fmt.Errorf("converting field {0}: %v", err)\n}}\n'.format(goname) - s += 'x.{} = {}\n'.format(field_name, goname) + s += 'x.{0} = {1}\n'.format(field_name, goname) # End switch statement s += 'default:\n' - err_string = '"invalid union key \'%v\'", x.{}'.format(gokeyname) - s += 'return fmt.Errorf({})'.format(err_string) + err_string = '"invalid union key \'%v\'", x.{0}'.format(gokeyname) + s += 'return fmt.Errorf({0})'.format(err_string) s += '}\n' return (s,extras) @@ -420,22 +420,22 @@ def xenlight_golang_array_from_C(ty = None): goname = xenlight_golang_fmt_name(ty.name) ctypename = ty.type.elem_type.typename cname = ty.name - cslice = 'c{}'.format(goname) + cslice = 'c{0}'.format(goname) clenvar = ty.type.lenvar.name - s += 'x.{} = nil\n'.format(goname) - s += 'if n := int(xc.{}); n > 0 {{\n'.format(clenvar) - s += '{} := '.format(cslice) - s +='(*[1<<28]C.{})(unsafe.Pointer(xc.{}))[:n:n]\n'.format(ctypename, cname) - s += 'x.{} = make([]{}, n)\n'.format(goname, gotypename) - s += 'for i, v := range {} {{\n'.format(cslice) + s += 'x.{0} = nil\n'.format(goname) + s += 'if n := int(xc.{0}); n > 0 {{\n'.format(clenvar) + s += '{0} := '.format(cslice) + s +='(*[1<<28]C.{0})(unsafe.Pointer(xc.{1}))[:n:n]\n'.format(ctypename, cname) + s += 'x.{0} = make([]{1}, n)\n'.format(goname, gotypename) + s += 'for i, v := range {0} {{\n'.format(cslice) is_enum = isinstance(ty.type.elem_type,idl.Enumeration) if gotypename in go_builtin_types or is_enum: - s += 'x.{}[i] = {}(v)\n'.format(goname, gotypename) + s += 'x.{0}[i] = {1}(v)\n'.format(goname, gotypename) else: - s += 'if err := x.{}[i].fromC(&v); err != nil {{\n'.format(goname) - s += 'return fmt.Errorf("converting field {}: %v", err) }}\n'.format(goname) + s += 'if err := x.{0}[i].fromC(&v); err != nil {{\n'.format(goname) + s += 'return fmt.Errorf("converting field {0}: %v", err) }}\n'.format(goname) s += '}\n}\n' @@ -446,11 +446,11 @@ def xenlight_golang_define_to_C(ty = None, typename = None, nested = False): Define the toC marshaling function for the type represented by ty. """ - func = 'func (x *{}) toC(xc *C.{}) (err error){{{}\n return nil\n }}\n' + func = 'func (x *{0}) toC(xc *C.{1}) (err error){{{2}\n return nil\n }}\n' body = '' if ty.dispose_fn is not None: - body += 'defer func(){{\nif err != nil{{\nC.{}(xc)}}\n}}()\n\n'.format(ty.dispose_fn) + body += 'defer func(){{\nif err != nil{{\nC.{0}(xc)}}\n}}()\n\n'.format(ty.dispose_fn) goname = xenlight_golang_fmt_name(ty.typename) cname = ty.typename @@ -471,7 +471,7 @@ def xenlight_golang_define_to_C(ty = None, typename = None, nested = False): body += xenlight_golang_union_to_C(f.type, f.name, ty.typename) else: - raise Exception('type {} not supported'.format(f.type)) + raise Exception('type {0} not supported'.format(f.type)) return func.format(goname, cname, body) @@ -506,26 +506,26 @@ def xenlight_golang_convert_to_C(ty = None, outer_name = None, # If outer_name is set, treat this as nested. if outer_name is not None: - goname = '{}.{}'.format(xenlight_golang_fmt_name(outer_name), goname) - cname = '{}.{}'.format(outer_name, cname) + goname = '{0}.{1}'.format(xenlight_golang_fmt_name(outer_name), goname) + cname = '{0}.{1}'.format(outer_name, cname) is_castable = (ty.type.json_parse_type == 'JSON_INTEGER' or isinstance(ty.type, idl.Enumeration) or gotypename in go_builtin_types) if not is_castable: - s += 'if err := {}.{}.toC(&{}.{}); err != nil {{\n'.format(govarname,goname, + s += 'if err := {0}.{1}.toC(&{2}.{3}); err != nil {{\n'.format(govarname,goname, cvarname,cname) - s += 'return fmt.Errorf("converting field {}: %v", err)\n}}\n'.format(goname) + s += 'return fmt.Errorf("converting field {0}: %v", err)\n}}\n'.format(goname) elif gotypename == 'string': # Use the cgo helper for converting C strings. - s += 'if {}.{} != "" {{\n'.format(govarname,goname) - s += '{}.{} = C.CString({}.{})}}\n'.format(cvarname,cname, + s += 'if {0}.{1} != "" {{\n'.format(govarname,goname) + s += '{0}.{1} = C.CString({2}.{3})}}\n'.format(cvarname,cname, govarname,goname) else: - s += '{}.{} = C.{}({}.{})\n'.format(cvarname,cname,ctypename, + s += '{0}.{1} = C.{2}({3}.{4})\n'.format(cvarname,cname,ctypename, govarname,goname) return s @@ -537,7 +537,7 @@ def xenlight_golang_union_to_C(ty = None, union_name = '', keytype = ty.keyvar.type.typename gokeytype = xenlight_golang_fmt_name(keytype) - interface_name = '{}_{}_union'.format(struct_name, keyname) + interface_name = '{0}_{1}_union'.format(struct_name, keyname) interface_name = xenlight_golang_fmt_name(interface_name, exported=False) cgo_keyname = keyname @@ -545,44 +545,44 @@ def xenlight_golang_union_to_C(ty = None, union_name = '', cgo_keyname = '_' + cgo_keyname - s = 'xc.{} = C.{}(x.{})\n'.format(cgo_keyname,keytype,gokeyname) - s += 'switch x.{}{{\n'.format(gokeyname) + s = 'xc.{0} = C.{1}(x.{2})\n'.format(cgo_keyname,keytype,gokeyname) + s += 'switch x.{0}{{\n'.format(gokeyname) # Create switch statement to determine how to populate the C union. for f in ty.fields: - key_val = '{}_{}'.format(keytype, f.name) + key_val = '{0}_{1}'.format(keytype, f.name) key_val = xenlight_golang_fmt_name(key_val) - s += 'case {}:\n'.format(key_val) + s += 'case {0}:\n'.format(key_val) if f.type is None: s += "break\n" continue - cgotype = '{}_{}_union_{}'.format(struct_name,keyname,f.name) + cgotype = '{0}_{1}_union_{2}'.format(struct_name,keyname,f.name) gotype = xenlight_golang_fmt_name(cgotype) - field_name = xenlight_golang_fmt_name('{}_union'.format(keyname)) - s += 'tmp, ok := x.{}.({})\n'.format(field_name,gotype) + field_name = xenlight_golang_fmt_name('{0}_union'.format(keyname)) + s += 'tmp, ok := x.{0}.({1})\n'.format(field_name,gotype) s += 'if !ok {\n' - s += 'return errors.New("wrong type for union key {}")\n'.format(keyname) + s += 'return errors.New("wrong type for union key {0}")\n'.format(keyname) s += '}\n' - s += 'var {} C.{}\n'.format(f.name,cgotype) + s += 'var {0} C.{1}\n'.format(f.name,cgotype) for uf in f.type.fields: s += xenlight_golang_convert_to_C(uf,cvarname=f.name, govarname='tmp') # The union is still represented as Go []byte. - s += '{}Bytes := C.GoBytes(unsafe.Pointer(&{}),C.sizeof_{})\n'.format(f.name, + s += '{0}Bytes := C.GoBytes(unsafe.Pointer(&{1}),C.sizeof_{2})\n'.format(f.name, f.name, cgotype) - s += 'copy(xc.{}[:],{}Bytes)\n'.format(union_name,f.name) + s += 'copy(xc.{0}[:],{1}Bytes)\n'.format(union_name,f.name) # End switch statement s += 'default:\n' - err_string = '"invalid union key \'%v\'", x.{}'.format(gokeyname) - s += 'return fmt.Errorf({})'.format(err_string) + err_string = '"invalid union key \'%v\'", x.{0}'.format(gokeyname) + s += 'return fmt.Errorf({0})'.format(err_string) s += '}\n' return s @@ -599,29 +599,29 @@ def xenlight_golang_array_to_C(ty = None): is_enum = isinstance(ty.type.elem_type,idl.Enumeration) if gotypename in go_builtin_types or is_enum: - s += 'if {} := len(x.{}); {} > 0 {{\n'.format(golenvar,goname,golenvar) - s += 'xc.{} = (*C.{})(C.malloc(C.size_t({}*{})))\n'.format(cname,ctypename, + s += 'if {0} := len(x.{1}); {2} > 0 {{\n'.format(golenvar,goname,golenvar) + s += 'xc.{0} = (*C.{1})(C.malloc(C.size_t({2}*{3})))\n'.format(cname,ctypename, golenvar,golenvar) - s += 'xc.{} = C.int({})\n'.format(clenvar,golenvar) - s += 'c{} := (*[1<<28]C.{})(unsafe.Pointer(xc.{}))[:{}:{}]\n'.format(goname, + s += 'xc.{0} = C.int({1})\n'.format(clenvar,golenvar) + s += 'c{0} := (*[1<<28]C.{1})(unsafe.Pointer(xc.{2}))[:{3}:{4}]\n'.format(goname, ctypename,cname, golenvar,golenvar) - s += 'for i,v := range x.{} {{\n'.format(goname) - s += 'c{}[i] = C.{}(v)\n'.format(goname,ctypename) + s += 'for i,v := range x.{0} {{\n'.format(goname) + s += 'c{0}[i] = C.{1}(v)\n'.format(goname,ctypename) s += '}\n}\n' return s - s += 'if {} := len(x.{}); {} > 0 {{\n'.format(golenvar,goname,golenvar) - s += 'xc.{} = (*C.{})(C.malloc(C.ulong({})*C.sizeof_{}))\n'.format(cname,ctypename, + s += 'if {0} := len(x.{1}); {2} > 0 {{\n'.format(golenvar,goname,golenvar) + s += 'xc.{0} = (*C.{1})(C.malloc(C.ulong({2})*C.sizeof_{3}))\n'.format(cname,ctypename, golenvar,ctypename) - s += 'xc.{} = C.int({})\n'.format(clenvar,golenvar) - s += 'c{} := (*[1<<28]C.{})(unsafe.Pointer(xc.{}))[:{}:{}]\n'.format(goname, + s += 'xc.{0} = C.int({1})\n'.format(clenvar,golenvar) + s += 'c{0} := (*[1<<28]C.{1})(unsafe.Pointer(xc.{2}))[:{3}:{4}]\n'.format(goname, ctypename,cname, golenvar,golenvar) - s += 'for i,v := range x.{} {{\n'.format(goname) - s += 'if err := v.toC(&c{}[i]); err != nil {{\n'.format(goname) - s += 'return fmt.Errorf("converting field {}: %v", err)\n'.format(goname) + s += 'for i,v := range x.{0} {{\n'.format(goname) + s += 'if err := v.toC(&c{0}[i]); err != nil {{\n'.format(goname) + s += 'return fmt.Errorf("converting field {0}: %v", err)\n'.format(goname) s += '}\n}\n}\n' return s @@ -633,7 +633,7 @@ def xenlight_golang_define_constructor(ty = None): gotypename = xenlight_golang_fmt_name(ctypename) # Since this func is exported, add a comment as per Go conventions. - s += '// New{} returns an instance of {}'.format(gotypename,gotypename) + s += '// New{0} returns an instance of {1}'.format(gotypename,gotypename) s += ' initialized with defaults.\n' # If a struct has a keyed union, an extra argument is @@ -643,7 +643,7 @@ def xenlight_golang_define_constructor(ty = None): init_fns = [] # Add call to parent init_fn first. - init_fns.append('C.{}(&xc)'.format(ty.init_fn)) + init_fns.append('C.{0}(&xc)'.format(ty.init_fn)) for f in ty.fields: if not isinstance(f.type, idl.KeyedUnion): @@ -658,24 +658,24 @@ def xenlight_golang_define_constructor(ty = None): # Serveral keyed unions use 'type' as the key variable name. In # that case, prepend the first letter of the Go type name. if param_goname == 'type': - param_goname = '{}type'.format(param_gotype.lower()[0]) + param_goname = '{0}type'.format(param_gotype.lower()[0]) # Add call to keyed union's init_fn. - init_fns.append('C.{}_{}(&xc, C.{}({}))'.format(ty.init_fn, + init_fns.append('C.{0}_{1}(&xc, C.{2}({3}))'.format(ty.init_fn, param.name, param_ctype, param_goname)) # Add to params list. - params.append('{} {}'.format(param_goname, param_gotype)) + params.append('{0} {1}'.format(param_goname, param_gotype)) # Define function - s += 'func New{}({}) (*{}, error) {{\n'.format(gotypename, + s += 'func New{0}({1}) (*{2}, error) {{\n'.format(gotypename, ','.join(params), gotypename) # Declare variables. - s += 'var (\nx {}\nxc C.{})\n\n'.format(gotypename, ctypename) + s += 'var (\nx {0}\nxc C.{1})\n\n'.format(gotypename, ctypename) # Write init_fn calls. s += '\n'.join(init_fns) @@ -684,7 +684,7 @@ def xenlight_golang_define_constructor(ty = None): # Make sure dispose_fn get's called when constructor # returns. if ty.dispose_fn is not None: - s += 'defer C.{}(&xc)\n'.format(ty.dispose_fn) + s += 'defer C.{0}(&xc)\n'.format(ty.dispose_fn) s += '\n' @@ -727,7 +727,7 @@ if __name__ == '__main__': header_comment="""// DO NOT EDIT. // // This file is generated by: -// {} +// {0} // """.format(' '.join(sys.argv)) -- generated by git-patchbot for /home/xen/git/xen.git#stable-4.14 From xen-changelog-bounces@lists.xenproject.org Mon Jun 21 15:55:11 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Mon, 21 Jun 2021 15:55:11 +0000 Received: from list by lists.xenproject.org with outflank-mailman.145598.267737 (Exim 4.92) (envelope-from ) id 1lvMGQ-0001OA-Bz; Mon, 21 Jun 2021 15:55:06 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 145598.267737; Mon, 21 Jun 2021 15:55:06 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lvMGQ-0001O2-96; Mon, 21 Jun 2021 15:55:06 +0000 Received: by outflank-mailman (input) for mailman id 145598; Mon, 21 Jun 2021 15:55:04 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lvMGO-0001Nq-Hr for xen-changelog@lists.xenproject.org; Mon, 21 Jun 2021 15:55:04 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lvMGO-0003zJ-GF for xen-changelog@lists.xenproject.org; Mon, 21 Jun 2021 15:55:04 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lvMGO-0000Y9-F8 for xen-changelog@lists.xenproject.org; Mon, 21 Jun 2021 15:55:04 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=TmRLFzk/tzgDRWe7RJiCqZrsYOq6wkbCOIRMHyji+G0=; b=JS2ZxmBD9CRJ8B1CpqtYfa5A3L gid51ABVg9RdASH3vhJA8bThmDUTVuWxWecaOxlWXOb2NBUq90e0AeluO/3WBUuCJ247QxyLJkLcK Sn8OpWqK3E6ZdEQymi1QrG/RcjMuj4Zspo6xUCmrN2/aaK8TjNjWcdyVx2IUOkno/eEk=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] golang/xenlight: update generated code Message-Id: Date: Mon, 21 Jun 2021 15:55:04 +0000 commit b291ce703b9cebef0800267446334e867588354a Author: Nick Rosbrook AuthorDate: Mon May 24 16:36:42 2021 -0400 Commit: George Dunlap CommitDate: Mon Jun 21 16:49:08 2021 +0100 golang/xenlight: update generated code Re-generate code to reflect changes to libxl_types.idl from the following commits: 0570d7f276 x86/msr: introduce an option for compatible MSR behavior selection 7e5cffcd1e viridian: allow vCPU hotplug for Windows VMs 9835246710 viridian: remove implicit limit of 64 VPs per partition Signed-off-by: Nick Rosbrook Reviewed-by: George Dunlap --- tools/golang/xenlight/helpers.gen.go | 6 ++++++ tools/golang/xenlight/types.gen.go | 5 +++++ 2 files changed, 11 insertions(+) diff --git a/tools/golang/xenlight/helpers.gen.go b/tools/golang/xenlight/helpers.gen.go index 4c60d27a9c..b454b12d52 100644 --- a/tools/golang/xenlight/helpers.gen.go +++ b/tools/golang/xenlight/helpers.gen.go @@ -1113,6 +1113,9 @@ default: return fmt.Errorf("invalid union key '%v'", x.Type)} x.ArchArm.GicVersion = GicVersion(xc.arch_arm.gic_version) x.ArchArm.Vuart = VuartType(xc.arch_arm.vuart) +if err := x.ArchX86.MsrRelaxed.fromC(&xc.arch_x86.msr_relaxed);err != nil { +return fmt.Errorf("converting field ArchX86.MsrRelaxed: %v", err) +} x.Altp2M = Altp2MMode(xc.altp2m) x.VmtraceBufKb = int(xc.vmtrace_buf_kb) @@ -1589,6 +1592,9 @@ default: return fmt.Errorf("invalid union key '%v'", x.Type)} xc.arch_arm.gic_version = C.libxl_gic_version(x.ArchArm.GicVersion) xc.arch_arm.vuart = C.libxl_vuart_type(x.ArchArm.Vuart) +if err := x.ArchX86.MsrRelaxed.toC(&xc.arch_x86.msr_relaxed); err != nil { +return fmt.Errorf("converting field ArchX86.MsrRelaxed: %v", err) +} xc.altp2m = C.libxl_altp2m_mode(x.Altp2M) xc.vmtrace_buf_kb = C.int(x.VmtraceBufKb) diff --git a/tools/golang/xenlight/types.gen.go b/tools/golang/xenlight/types.gen.go index cb13002fdb..f2ceceb61c 100644 --- a/tools/golang/xenlight/types.gen.go +++ b/tools/golang/xenlight/types.gen.go @@ -211,6 +211,8 @@ ViridianEnlightenmentSynic ViridianEnlightenment = 7 ViridianEnlightenmentStimer ViridianEnlightenment = 8 ViridianEnlightenmentHcallIpi ViridianEnlightenment = 9 ViridianEnlightenmentExProcessorMasks ViridianEnlightenment = 10 +ViridianEnlightenmentNoVpLimit ViridianEnlightenment = 11 +ViridianEnlightenmentCpuHotplug ViridianEnlightenment = 12 ) type Hdtype int @@ -513,6 +515,9 @@ ArchArm struct { GicVersion GicVersion Vuart VuartType } +ArchX86 struct { +MsrRelaxed Defbool +} Altp2M Altp2MMode VmtraceBufKb int } -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Mon Jun 21 15:55:16 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Mon, 21 Jun 2021 15:55:16 +0000 Received: from list by lists.xenproject.org with outflank-mailman.145599.267741 (Exim 4.92) (envelope-from ) id 1lvMGa-0001Q4-Db; Mon, 21 Jun 2021 15:55:16 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 145599.267741; Mon, 21 Jun 2021 15:55:16 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lvMGa-0001Pw-AZ; Mon, 21 Jun 2021 15:55:16 +0000 Received: by outflank-mailman (input) for mailman id 145599; Mon, 21 Jun 2021 15:55:14 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lvMGY-0001Ph-K7 for xen-changelog@lists.xenproject.org; Mon, 21 Jun 2021 15:55:14 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lvMGY-0003za-JM for xen-changelog@lists.xenproject.org; Mon, 21 Jun 2021 15:55:14 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lvMGY-0000Yw-IO for xen-changelog@lists.xenproject.org; Mon, 21 Jun 2021 15:55:14 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=Z7YOXLxdIm8qJgacggOQsqQMIkIg6mrMv3iQcW/JEZE=; b=pahlmCBG/TmDpt1G26IKY2RVZX Y4NNzpTVeAh/PbJeqq2enRuhDXFEQBW8yXo1mn3eS8jACbmZ7PChgGB2tfDGV3DReoauDSRttm6qk qgyZv2WaUFls9MWdBmUFZMbouqfe3PTW5HP3bjNiRYqhJ7GCZiPHKrLEGrvXONbgTef8=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] golang/xenlight: fix StringList toC conversion Message-Id: Date: Mon, 21 Jun 2021 15:55:14 +0000 commit 1422d8db1b3dfdf7d9179944e594876e5e356a4b Author: Nick Rosbrook AuthorDate: Mon May 24 16:36:43 2021 -0400 Commit: George Dunlap CommitDate: Mon Jun 21 16:49:08 2021 +0100 golang/xenlight: fix StringList toC conversion The current implementation of StringList.toC does not correctly account for how libxl_string_list is expected to be laid out in C, which is clear when one looks at libxl_string_list_length in libxl.c. In particular, StringList.toC does not account for the extra memory that should be allocated for the "sentinel" entry. And, when using the "slice trick" to create a slice that can address C memory, the unsafe.Pointer conversion should be on a C.libxl_string_list, not *C.libxl_string_list. Fix these problems by (1) allocating an extra slot in the slice used to address the C memory, and explicity set the last entry to nil so the C memory will be zeroed out, and (2) dereferencing csl in the unsafe.Pointer conversion. Signed-off-by: Nick Rosbrook Reviewed-by: George Dunlap --- tools/golang/xenlight/xenlight.go | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/tools/golang/xenlight/xenlight.go b/tools/golang/xenlight/xenlight.go index b9189dec5c..13171d0ad1 100644 --- a/tools/golang/xenlight/xenlight.go +++ b/tools/golang/xenlight/xenlight.go @@ -491,13 +491,14 @@ func (sl *StringList) fromC(csl *C.libxl_string_list) error { func (sl StringList) toC(csl *C.libxl_string_list) error { var char *C.char - size := len(sl) + size := len(sl) + 1 *csl = (C.libxl_string_list)(C.malloc(C.ulong(size) * C.ulong(unsafe.Sizeof(char)))) - clist := (*[1 << 30]*C.char)(unsafe.Pointer(csl))[:size:size] + clist := (*[1 << 30]*C.char)(unsafe.Pointer(*csl))[:size:size] for i, v := range sl { clist[i] = C.CString(v) } + clist[len(clist)-1] = nil return nil } -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Mon Jun 21 15:55:26 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Mon, 21 Jun 2021 15:55:26 +0000 Received: from list by lists.xenproject.org with outflank-mailman.145600.267746 (Exim 4.92) (envelope-from ) id 1lvMGk-0001TW-Ff; Mon, 21 Jun 2021 15:55:26 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 145600.267746; Mon, 21 Jun 2021 15:55:26 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lvMGk-0001TM-CD; Mon, 21 Jun 2021 15:55:26 +0000 Received: by outflank-mailman (input) for mailman id 145600; Mon, 21 Jun 2021 15:55:24 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lvMGi-0001Sv-N6 for xen-changelog@lists.xenproject.org; Mon, 21 Jun 2021 15:55:24 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lvMGi-0003zo-MM for xen-changelog@lists.xenproject.org; Mon, 21 Jun 2021 15:55:24 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lvMGi-0000Ze-LW for xen-changelog@lists.xenproject.org; Mon, 21 Jun 2021 15:55:24 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=jBLjuvyis+raT1wBPpgyICv6OWS6GIv2dByjb5BuiWg=; b=s947elvsbAfCanhC7fpcXTwvaF 2CP9r4FJv8v9jkbDQcc5wpLNJk23x1uFYbVVZzwF8qwBRkMJP/DLDdWE2NijYiFD2EG/FysL7XLL7 cv81AXIgnlV3oDIC4Myx6VdAwUC8ov+PlpIN1/9z6xFI0BJOpY31SFhVbhOmlJeYt3Hc=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] golang/xenlight: export keyed union interface types Message-Id: Date: Mon, 21 Jun 2021 15:55:24 +0000 commit bc9f632e31ee66be3f1860fc7303fe91a42e56a6 Author: Nick Rosbrook AuthorDate: Mon May 24 16:36:45 2021 -0400 Commit: George Dunlap CommitDate: Mon Jun 21 16:49:08 2021 +0100 golang/xenlight: export keyed union interface types For structs that have a keyed union, e.g. DomainBuildInfo, the TypeUnion field must be exported so that package users can get/set the fields within. This means that users are aware of the existence of the interface type used in those fields (see [1]), so it is awkward that the interface itself is not exported. However, the single method within the interface must remain unexported so that users cannot mistakenly "implement" those interfaces. Since there seems to be no reason to do otherwise, export the keyed union interface types. [1] https://pkg.go.dev/xenbits.xenproject.org/git-http/xen.git/tools/golang/xenlight?tab=doc#DeviceUsbdev Signed-off-by: Nick Rosbrook Reviewed-by: George Dunlap --- tools/golang/xenlight/gengotypes.py | 6 ++-- tools/golang/xenlight/types.gen.go | 58 ++++++++++++++++++------------------- 2 files changed, 32 insertions(+), 32 deletions(-) diff --git a/tools/golang/xenlight/gengotypes.py b/tools/golang/xenlight/gengotypes.py index 3e40c3d5dc..ef3279a910 100644 --- a/tools/golang/xenlight/gengotypes.py +++ b/tools/golang/xenlight/gengotypes.py @@ -159,7 +159,7 @@ def xenlight_golang_define_union(ty = None, struct_name = '', union_name = ''): extras = [] interface_name = '{0}_{1}_union'.format(struct_name, ty.keyvar.name) - interface_name = xenlight_golang_fmt_name(interface_name, exported=False) + interface_name = xenlight_golang_fmt_name(interface_name) s += 'type {0} interface {{\n'.format(interface_name) s += 'is{0}()\n'.format(interface_name) @@ -341,7 +341,7 @@ def xenlight_golang_union_from_C(ty = None, union_name = '', struct_name = ''): field_name = xenlight_golang_fmt_name('{0}_union'.format(keyname)) interface_name = '{0}_{1}_union'.format(struct_name, keyname) - interface_name = xenlight_golang_fmt_name(interface_name, exported=False) + interface_name = xenlight_golang_fmt_name(interface_name) cgo_keyname = keyname if cgo_keyname in go_keywords: @@ -545,7 +545,7 @@ def xenlight_golang_union_to_C(ty = None, union_name = '', gokeytype = xenlight_golang_fmt_name(keytype) interface_name = '{0}_{1}_union'.format(struct_name, keyname) - interface_name = xenlight_golang_fmt_name(interface_name, exported=False) + interface_name = xenlight_golang_fmt_name(interface_name) cgo_keyname = keyname if cgo_keyname in go_keywords: diff --git a/tools/golang/xenlight/types.gen.go b/tools/golang/xenlight/types.gen.go index f2ceceb61c..a214dd9df6 100644 --- a/tools/golang/xenlight/types.gen.go +++ b/tools/golang/xenlight/types.gen.go @@ -337,18 +337,18 @@ State int Evtch int Rref int Connection ChannelConnection -ConnectionUnion channelinfoConnectionUnion +ConnectionUnion ChannelinfoConnectionUnion } -type channelinfoConnectionUnion interface { -ischannelinfoConnectionUnion() +type ChannelinfoConnectionUnion interface { +isChannelinfoConnectionUnion() } type ChannelinfoConnectionUnionPty struct { Path string } -func (x ChannelinfoConnectionUnionPty) ischannelinfoConnectionUnion(){} +func (x ChannelinfoConnectionUnionPty) isChannelinfoConnectionUnion(){} type Vminfo struct { Uuid Uuid @@ -510,7 +510,7 @@ Apic Defbool DmRestrict Defbool Tee TeeType Type DomainType -TypeUnion domainBuildInfoTypeUnion +TypeUnion DomainBuildInfoTypeUnion ArchArm struct { GicVersion GicVersion Vuart VuartType @@ -522,8 +522,8 @@ Altp2M Altp2MMode VmtraceBufKb int } -type domainBuildInfoTypeUnion interface { -isdomainBuildInfoTypeUnion() +type DomainBuildInfoTypeUnion interface { +isDomainBuildInfoTypeUnion() } type DomainBuildInfoTypeUnionHvm struct { @@ -575,7 +575,7 @@ RdmMemBoundaryMemkb uint64 McaCaps uint64 } -func (x DomainBuildInfoTypeUnionHvm) isdomainBuildInfoTypeUnion(){} +func (x DomainBuildInfoTypeUnionHvm) isDomainBuildInfoTypeUnion(){} type DomainBuildInfoTypeUnionPv struct { Kernel string @@ -588,7 +588,7 @@ Features string E820Host Defbool } -func (x DomainBuildInfoTypeUnionPv) isdomainBuildInfoTypeUnion(){} +func (x DomainBuildInfoTypeUnionPv) isDomainBuildInfoTypeUnion(){} type DomainBuildInfoTypeUnionPvh struct { Pvshim Defbool @@ -597,7 +597,7 @@ PvshimCmdline string PvshimExtra string } -func (x DomainBuildInfoTypeUnionPvh) isdomainBuildInfoTypeUnion(){} +func (x DomainBuildInfoTypeUnionPvh) isDomainBuildInfoTypeUnion(){} type DeviceVfb struct { BackendDomid Domid @@ -761,11 +761,11 @@ type DeviceUsbdev struct { Ctrl Devid Port int Type UsbdevType -TypeUnion deviceUsbdevTypeUnion +TypeUnion DeviceUsbdevTypeUnion } -type deviceUsbdevTypeUnion interface { -isdeviceUsbdevTypeUnion() +type DeviceUsbdevTypeUnion interface { +isDeviceUsbdevTypeUnion() } type DeviceUsbdevTypeUnionHostdev struct { @@ -773,7 +773,7 @@ Hostbus byte Hostaddr byte } -func (x DeviceUsbdevTypeUnionHostdev) isdeviceUsbdevTypeUnion(){} +func (x DeviceUsbdevTypeUnionHostdev) isDeviceUsbdevTypeUnion(){} type DeviceDtdev struct { Path string @@ -807,18 +807,18 @@ BackendDomname string Devid Devid Name string Connection ChannelConnection -ConnectionUnion deviceChannelConnectionUnion +ConnectionUnion DeviceChannelConnectionUnion } -type deviceChannelConnectionUnion interface { -isdeviceChannelConnectionUnion() +type DeviceChannelConnectionUnion interface { +isDeviceChannelConnectionUnion() } type DeviceChannelConnectionUnionSocket struct { Path string } -func (x DeviceChannelConnectionUnionSocket) isdeviceChannelConnectionUnion(){} +func (x DeviceChannelConnectionUnionSocket) isDeviceChannelConnectionUnion(){} type ConnectorParam struct { UniqueId string @@ -1116,31 +1116,31 @@ Domid Domid Domuuid Uuid ForUser uint64 Type EventType -TypeUnion eventTypeUnion +TypeUnion EventTypeUnion } -type eventTypeUnion interface { -iseventTypeUnion() +type EventTypeUnion interface { +isEventTypeUnion() } type EventTypeUnionDomainShutdown struct { ShutdownReason byte } -func (x EventTypeUnionDomainShutdown) iseventTypeUnion(){} +func (x EventTypeUnionDomainShutdown) isEventTypeUnion(){} type EventTypeUnionDiskEject struct { Vdev string Disk DeviceDisk } -func (x EventTypeUnionDiskEject) iseventTypeUnion(){} +func (x EventTypeUnionDiskEject) isEventTypeUnion(){} type EventTypeUnionOperationComplete struct { Rc int } -func (x EventTypeUnionOperationComplete) iseventTypeUnion(){} +func (x EventTypeUnionOperationComplete) isEventTypeUnion(){} type PsrCmtType int const( @@ -1175,11 +1175,11 @@ PsrFeatTypeMba PsrFeatType = 2 type PsrHwInfo struct { Id uint32 Type PsrFeatType -TypeUnion psrHwInfoTypeUnion +TypeUnion PsrHwInfoTypeUnion } -type psrHwInfoTypeUnion interface { -ispsrHwInfoTypeUnion() +type PsrHwInfoTypeUnion interface { +isPsrHwInfoTypeUnion() } type PsrHwInfoTypeUnionCat struct { @@ -1188,7 +1188,7 @@ CbmLen uint32 CdpEnabled bool } -func (x PsrHwInfoTypeUnionCat) ispsrHwInfoTypeUnion(){} +func (x PsrHwInfoTypeUnionCat) isPsrHwInfoTypeUnion(){} type PsrHwInfoTypeUnionMba struct { CosMax uint32 @@ -1196,5 +1196,5 @@ ThrtlMax uint32 Linear bool } -func (x PsrHwInfoTypeUnionMba) ispsrHwInfoTypeUnion(){} +func (x PsrHwInfoTypeUnionMba) isPsrHwInfoTypeUnion(){} -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Mon Jun 21 15:55:36 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Mon, 21 Jun 2021 15:55:36 +0000 Received: from list by lists.xenproject.org with outflank-mailman.145601.267749 (Exim 4.92) (envelope-from ) id 1lvMGu-0001Xn-HG; Mon, 21 Jun 2021 15:55:36 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 145601.267749; Mon, 21 Jun 2021 15:55:36 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lvMGu-0001Xf-E4; Mon, 21 Jun 2021 15:55:36 +0000 Received: by outflank-mailman (input) for mailman id 145601; Mon, 21 Jun 2021 15:55:34 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lvMGs-0001XJ-QT for xen-changelog@lists.xenproject.org; Mon, 21 Jun 2021 15:55:34 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lvMGs-00040Q-Pe for xen-changelog@lists.xenproject.org; Mon, 21 Jun 2021 15:55:34 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lvMGs-0000aI-Oi for xen-changelog@lists.xenproject.org; Mon, 21 Jun 2021 15:55:34 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=IBMsUcETC+XpABkTki8xN+Msdr3kfRhbxAS82lGDMCg=; b=hyaI2JtpdSYVhu7FOESsLekxF4 Qre+KE2XqMM6uLGp/zZ3TL25ynuv/o+eQKLu+5wG6ulUwUv4zXosRR/nBBFUCTF/hcDR4epyAiXOC 2H+q0PBpz/7DqgVN0qQ3fwumyz1GSRMk5RzMfl3XH9UHJgUSJObE8+u43DJS3tIbJxoY=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] golang/xenlight: use struct pointers in keyed union fields Message-Id: Date: Mon, 21 Jun 2021 15:55:34 +0000 commit 1997940ad25e3566d1ab38496b8c7b07a086695a Author: Nick Rosbrook AuthorDate: Mon May 24 16:36:46 2021 -0400 Commit: George Dunlap CommitDate: Mon Jun 21 16:49:08 2021 +0100 golang/xenlight: use struct pointers in keyed union fields Currently, when marshalig Go types with keyed union fields, we assign the value of the struct (e.g. DomainBuildInfoTypeUnionHvm) which implements the interface of the keyed union field (e.g. DomainBuildInfoTypeUnion). As-is, this means that if a populated DomainBuildInfo is marshaled to e.g. JSON, unmarshaling back to DomainBuildInfo will fail. When the encoding/json is unmarshaling data into a Go type, and encounters a JSON object, it basically can either marshal the data into an empty interface, a map, or a struct. It cannot, however, marshal data into an interface with at least one method defined on it (e.g. DomainBuildInfoTypeUnion). Before this check is done, however, the decoder will check if the Go type is a pointer, and dereference it if so. It will then use the type of this value as the "target" type. This means that if the TypeUnion field is populated with a DomainBuildInfoTypeUnion, the decoder will see a non-empty interface and fail. If the TypeUnion field is populated with a *DomainBuildInfoTypeUnionHvm, it dereferences the pointer and sees a struct instead, allowing decoding to continue normally. Since there does not appear to be a strict need for NOT using pointers in these fields, update code generation to set keyed union fields to pointers of their implementing structs. Signed-off-by: Nick Rosbrook Reviewed-by: George Dunlap --- tools/golang/xenlight/gengotypes.py | 4 ++-- tools/golang/xenlight/helpers.gen.go | 44 ++++++++++++++++++------------------ 2 files changed, 24 insertions(+), 24 deletions(-) diff --git a/tools/golang/xenlight/gengotypes.py b/tools/golang/xenlight/gengotypes.py index ef3279a910..ac1cf060dd 100644 --- a/tools/golang/xenlight/gengotypes.py +++ b/tools/golang/xenlight/gengotypes.py @@ -404,7 +404,7 @@ def xenlight_golang_union_from_C(ty = None, union_name = '', struct_name = ''): s += 'if err := {0}.fromC(xc);'.format(goname) s += 'err != nil {{\n return fmt.Errorf("converting field {0}: %v", err)\n}}\n'.format(goname) - s += 'x.{0} = {1}\n'.format(field_name, goname) + s += 'x.{0} = &{1}\n'.format(field_name, goname) # End switch statement s += 'default:\n' @@ -570,7 +570,7 @@ def xenlight_golang_union_to_C(ty = None, union_name = '', gotype = xenlight_golang_fmt_name(cgotype) field_name = xenlight_golang_fmt_name('{0}_union'.format(keyname)) - s += 'tmp, ok := x.{0}.({1})\n'.format(field_name,gotype) + s += 'tmp, ok := x.{0}.(*{1})\n'.format(field_name,gotype) s += 'if !ok {\n' s += 'return errors.New("wrong type for union key {0}")\n'.format(keyname) s += '}\n' diff --git a/tools/golang/xenlight/helpers.gen.go b/tools/golang/xenlight/helpers.gen.go index b454b12d52..db82537b42 100644 --- a/tools/golang/xenlight/helpers.gen.go +++ b/tools/golang/xenlight/helpers.gen.go @@ -433,7 +433,7 @@ var connectionPty ChannelinfoConnectionUnionPty if err := connectionPty.fromC(xc);err != nil { return fmt.Errorf("converting field connectionPty: %v", err) } -x.ConnectionUnion = connectionPty +x.ConnectionUnion = &connectionPty case ChannelConnectionSocket: x.ConnectionUnion = nil case ChannelConnectionUnknown: @@ -473,7 +473,7 @@ switch x.Connection{ case ChannelConnectionUnknown: break case ChannelConnectionPty: -tmp, ok := x.ConnectionUnion.(ChannelinfoConnectionUnionPty) +tmp, ok := x.ConnectionUnion.(*ChannelinfoConnectionUnionPty) if !ok { return errors.New("wrong type for union key connection") } @@ -1094,7 +1094,7 @@ var typeHvm DomainBuildInfoTypeUnionHvm if err := typeHvm.fromC(xc);err != nil { return fmt.Errorf("converting field typeHvm: %v", err) } -x.TypeUnion = typeHvm +x.TypeUnion = &typeHvm case DomainTypeInvalid: x.TypeUnion = nil case DomainTypePv: @@ -1102,13 +1102,13 @@ var typePv DomainBuildInfoTypeUnionPv if err := typePv.fromC(xc);err != nil { return fmt.Errorf("converting field typePv: %v", err) } -x.TypeUnion = typePv +x.TypeUnion = &typePv case DomainTypePvh: var typePvh DomainBuildInfoTypeUnionPvh if err := typePvh.fromC(xc);err != nil { return fmt.Errorf("converting field typePvh: %v", err) } -x.TypeUnion = typePvh +x.TypeUnion = &typePvh default: return fmt.Errorf("invalid union key '%v'", x.Type)} x.ArchArm.GicVersion = GicVersion(xc.arch_arm.gic_version) @@ -1427,7 +1427,7 @@ xc.tee = C.libxl_tee_type(x.Tee) xc._type = C.libxl_domain_type(x.Type) switch x.Type{ case DomainTypeHvm: -tmp, ok := x.TypeUnion.(DomainBuildInfoTypeUnionHvm) +tmp, ok := x.TypeUnion.(*DomainBuildInfoTypeUnionHvm) if !ok { return errors.New("wrong type for union key type") } @@ -1545,7 +1545,7 @@ hvm.mca_caps = C.uint64_t(tmp.McaCaps) hvmBytes := C.GoBytes(unsafe.Pointer(&hvm),C.sizeof_libxl_domain_build_info_type_union_hvm) copy(xc.u[:],hvmBytes) case DomainTypePv: -tmp, ok := x.TypeUnion.(DomainBuildInfoTypeUnionPv) +tmp, ok := x.TypeUnion.(*DomainBuildInfoTypeUnionPv) if !ok { return errors.New("wrong type for union key type") } @@ -1570,7 +1570,7 @@ return fmt.Errorf("converting field E820Host: %v", err) pvBytes := C.GoBytes(unsafe.Pointer(&pv),C.sizeof_libxl_domain_build_info_type_union_pv) copy(xc.u[:],pvBytes) case DomainTypePvh: -tmp, ok := x.TypeUnion.(DomainBuildInfoTypeUnionPvh) +tmp, ok := x.TypeUnion.(*DomainBuildInfoTypeUnionPvh) if !ok { return errors.New("wrong type for union key type") } @@ -2157,7 +2157,7 @@ var typeHostdev DeviceUsbdevTypeUnionHostdev if err := typeHostdev.fromC(xc);err != nil { return fmt.Errorf("converting field typeHostdev: %v", err) } -x.TypeUnion = typeHostdev +x.TypeUnion = &typeHostdev default: return fmt.Errorf("invalid union key '%v'", x.Type)} @@ -2184,7 +2184,7 @@ xc.port = C.int(x.Port) xc._type = C.libxl_usbdev_type(x.Type) switch x.Type{ case UsbdevTypeHostdev: -tmp, ok := x.TypeUnion.(DeviceUsbdevTypeUnionHostdev) +tmp, ok := x.TypeUnion.(*DeviceUsbdevTypeUnionHostdev) if !ok { return errors.New("wrong type for union key type") } @@ -2375,7 +2375,7 @@ var connectionSocket DeviceChannelConnectionUnionSocket if err := connectionSocket.fromC(xc);err != nil { return fmt.Errorf("converting field connectionSocket: %v", err) } -x.ConnectionUnion = connectionSocket +x.ConnectionUnion = &connectionSocket case ChannelConnectionUnknown: x.ConnectionUnion = nil default: @@ -2411,7 +2411,7 @@ break case ChannelConnectionPty: break case ChannelConnectionSocket: -tmp, ok := x.ConnectionUnion.(DeviceChannelConnectionUnionSocket) +tmp, ok := x.ConnectionUnion.(*DeviceChannelConnectionUnionSocket) if !ok { return errors.New("wrong type for union key connection") } @@ -3948,7 +3948,7 @@ var typeDiskEject EventTypeUnionDiskEject if err := typeDiskEject.fromC(xc);err != nil { return fmt.Errorf("converting field typeDiskEject: %v", err) } -x.TypeUnion = typeDiskEject +x.TypeUnion = &typeDiskEject case EventTypeDomainCreateConsoleAvailable: x.TypeUnion = nil case EventTypeDomainDeath: @@ -3958,13 +3958,13 @@ var typeDomainShutdown EventTypeUnionDomainShutdown if err := typeDomainShutdown.fromC(xc);err != nil { return fmt.Errorf("converting field typeDomainShutdown: %v", err) } -x.TypeUnion = typeDomainShutdown +x.TypeUnion = &typeDomainShutdown case EventTypeOperationComplete: var typeOperationComplete EventTypeUnionOperationComplete if err := typeOperationComplete.fromC(xc);err != nil { return fmt.Errorf("converting field typeOperationComplete: %v", err) } -x.TypeUnion = typeOperationComplete +x.TypeUnion = &typeOperationComplete default: return fmt.Errorf("invalid union key '%v'", x.Type)} @@ -4019,7 +4019,7 @@ xc.for_user = C.uint64_t(x.ForUser) xc._type = C.libxl_event_type(x.Type) switch x.Type{ case EventTypeDomainShutdown: -tmp, ok := x.TypeUnion.(EventTypeUnionDomainShutdown) +tmp, ok := x.TypeUnion.(*EventTypeUnionDomainShutdown) if !ok { return errors.New("wrong type for union key type") } @@ -4030,7 +4030,7 @@ copy(xc.u[:],domain_shutdownBytes) case EventTypeDomainDeath: break case EventTypeDiskEject: -tmp, ok := x.TypeUnion.(EventTypeUnionDiskEject) +tmp, ok := x.TypeUnion.(*EventTypeUnionDiskEject) if !ok { return errors.New("wrong type for union key type") } @@ -4043,7 +4043,7 @@ return fmt.Errorf("converting field Disk: %v", err) disk_ejectBytes := C.GoBytes(unsafe.Pointer(&disk_eject),C.sizeof_libxl_event_type_union_disk_eject) copy(xc.u[:],disk_ejectBytes) case EventTypeOperationComplete: -tmp, ok := x.TypeUnion.(EventTypeUnionOperationComplete) +tmp, ok := x.TypeUnion.(*EventTypeUnionOperationComplete) if !ok { return errors.New("wrong type for union key type") } @@ -4118,13 +4118,13 @@ var typeCat PsrHwInfoTypeUnionCat if err := typeCat.fromC(xc);err != nil { return fmt.Errorf("converting field typeCat: %v", err) } -x.TypeUnion = typeCat +x.TypeUnion = &typeCat case PsrFeatTypeMba: var typeMba PsrHwInfoTypeUnionMba if err := typeMba.fromC(xc);err != nil { return fmt.Errorf("converting field typeMba: %v", err) } -x.TypeUnion = typeMba +x.TypeUnion = &typeMba default: return fmt.Errorf("invalid union key '%v'", x.Type)} @@ -4163,7 +4163,7 @@ xc.id = C.uint32_t(x.Id) xc._type = C.libxl_psr_feat_type(x.Type) switch x.Type{ case PsrFeatTypeCat: -tmp, ok := x.TypeUnion.(PsrHwInfoTypeUnionCat) +tmp, ok := x.TypeUnion.(*PsrHwInfoTypeUnionCat) if !ok { return errors.New("wrong type for union key type") } @@ -4174,7 +4174,7 @@ cat.cdp_enabled = C.bool(tmp.CdpEnabled) catBytes := C.GoBytes(unsafe.Pointer(&cat),C.sizeof_libxl_psr_hw_info_type_union_cat) copy(xc.u[:],catBytes) case PsrFeatTypeMba: -tmp, ok := x.TypeUnion.(PsrHwInfoTypeUnionMba) +tmp, ok := x.TypeUnion.(*PsrHwInfoTypeUnionMba) if !ok { return errors.New("wrong type for union key type") } -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Mon Jun 21 15:55:46 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Mon, 21 Jun 2021 15:55:46 +0000 Received: from list by lists.xenproject.org with outflank-mailman.145602.267753 (Exim 4.92) (envelope-from ) id 1lvMH4-0001bV-LS; Mon, 21 Jun 2021 15:55:46 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 145602.267753; Mon, 21 Jun 2021 15:55:46 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lvMH4-0001bN-IG; Mon, 21 Jun 2021 15:55:46 +0000 Received: by outflank-mailman (input) for mailman id 145602; Mon, 21 Jun 2021 15:55:44 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lvMH2-0001au-TL for xen-changelog@lists.xenproject.org; Mon, 21 Jun 2021 15:55:44 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lvMH2-00040e-Sa for xen-changelog@lists.xenproject.org; Mon, 21 Jun 2021 15:55:44 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lvMH2-0000aw-Rm for xen-changelog@lists.xenproject.org; Mon, 21 Jun 2021 15:55:44 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=AoAf0qkzMedxzTE6UivuUeVgPHvUDqc1UopMNfK4ITE=; b=H2QFQzZHB0vBuO4/xdZT1ODObF A5p2TdUVFqbLrUbzm4LRs793f+4N9QHfJGnLpdKF5wzaRPfGTPofK5DOEOA01Rgsv6fk7CNE9+RAj c1VRy70QM8ZCXaHPnxosvqgGSJAS9hkp9HlQj5RkP/TY8TLWHhKnfzJlrwUYtu7OOeBQ=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] golang/xenlight: rename Ctx receivers to ctx Message-Id: Date: Mon, 21 Jun 2021 15:55:44 +0000 commit c089de0e2fa56d846cfb658b7b5efc3426895973 Author: Nick Rosbrook AuthorDate: Mon May 24 16:36:47 2021 -0400 Commit: George Dunlap CommitDate: Mon Jun 21 16:49:08 2021 +0100 golang/xenlight: rename Ctx receivers to ctx As a matter of style, it is strange to see capitalized receiver names, due to the significance of capitalized symbols in Go (although there is in fact nothing special about a capitalized receiver name). Fix this in xenlight.go by running: gofmt -w -r 'Ctx -> ctx' xenlight.go from tools/golang/xenlight. There is no functional change. Signed-off-by: Nick Rosbrook Reviewed-by: George Dunlap --- tools/golang/xenlight/xenlight.go | 154 +++++++++++++++++++------------------- 1 file changed, 77 insertions(+), 77 deletions(-) diff --git a/tools/golang/xenlight/xenlight.go b/tools/golang/xenlight/xenlight.go index 13171d0ad1..fc3eb0bf3f 100644 --- a/tools/golang/xenlight/xenlight.go +++ b/tools/golang/xenlight/xenlight.go @@ -203,13 +203,13 @@ type Domid uint32 // NameToDomid does not guarantee that the domid associated with name at // the time NameToDomid is called is the same as the domid associated with // name at the time NameToDomid returns. -func (Ctx *Context) NameToDomid(name string) (Domid, error) { +func (ctx *Context) NameToDomid(name string) (Domid, error) { var domid C.uint32_t cname := C.CString(name) defer C.free(unsafe.Pointer(cname)) - if ret := C.libxl_name_to_domid(Ctx.ctx, cname, &domid); ret != 0 { + if ret := C.libxl_name_to_domid(ctx.ctx, cname, &domid); ret != 0 { return DomidInvalid, Error(ret) } @@ -223,8 +223,8 @@ func (Ctx *Context) NameToDomid(name string) (Domid, error) { // DomidToName does not guarantee that the name (if any) associated with domid // at the time DomidToName is called is the same as the name (if any) associated // with domid at the time DomidToName returns. -func (Ctx *Context) DomidToName(domid Domid) string { - cname := C.libxl_domid_to_name(Ctx.ctx, C.uint32_t(domid)) +func (ctx *Context) DomidToName(domid Domid) string { + cname := C.libxl_domid_to_name(ctx.ctx, C.uint32_t(domid)) defer C.free(unsafe.Pointer(cname)) return C.GoString(cname) @@ -594,10 +594,10 @@ func SchedulerFromString(name string) (s Scheduler, err error) { // libxl_cpupoolinfo * libxl_list_cpupool(libxl_ctx*, int *nb_pool_out); // void libxl_cpupoolinfo_list_free(libxl_cpupoolinfo *list, int nb_pool); -func (Ctx *Context) ListCpupool() (list []Cpupoolinfo) { +func (ctx *Context) ListCpupool() (list []Cpupoolinfo) { var nbPool C.int - c_cpupool_list := C.libxl_list_cpupool(Ctx.ctx, &nbPool) + c_cpupool_list := C.libxl_list_cpupool(ctx.ctx, &nbPool) defer C.libxl_cpupoolinfo_list_free(c_cpupool_list, nbPool) @@ -617,10 +617,10 @@ func (Ctx *Context) ListCpupool() (list []Cpupoolinfo) { } // int libxl_cpupool_info(libxl_ctx *ctx, libxl_cpupoolinfo *info, uint32_t poolid); -func (Ctx *Context) CpupoolInfo(Poolid uint32) (pool Cpupoolinfo, err error) { +func (ctx *Context) CpupoolInfo(Poolid uint32) (pool Cpupoolinfo, err error) { var c_cpupool C.libxl_cpupoolinfo - ret := C.libxl_cpupool_info(Ctx.ctx, &c_cpupool, C.uint32_t(Poolid)) + ret := C.libxl_cpupool_info(ctx.ctx, &c_cpupool, C.uint32_t(Poolid)) if ret != 0 { err = Error(-ret) return @@ -638,7 +638,7 @@ func (Ctx *Context) CpupoolInfo(Poolid uint32) (pool Cpupoolinfo, err error) { // uint32_t *poolid); // FIXME: uuid // FIXME: Setting poolid -func (Ctx *Context) CpupoolCreate(Name string, Scheduler Scheduler, Cpumap Bitmap) (err error, Poolid uint32) { +func (ctx *Context) CpupoolCreate(Name string, Scheduler Scheduler, Cpumap Bitmap) (err error, Poolid uint32) { poolid := C.uint32_t(C.LIBXL_CPUPOOL_POOLID_ANY) name := C.CString(Name) defer C.free(unsafe.Pointer(name)) @@ -653,7 +653,7 @@ func (Ctx *Context) CpupoolCreate(Name string, Scheduler Scheduler, Cpumap Bitma } defer C.libxl_bitmap_dispose(&cbm) - ret := C.libxl_cpupool_create(Ctx.ctx, name, C.libxl_scheduler(Scheduler), + ret := C.libxl_cpupool_create(ctx.ctx, name, C.libxl_scheduler(Scheduler), cbm, &uuid, &poolid) if ret != 0 { err = Error(-ret) @@ -666,8 +666,8 @@ func (Ctx *Context) CpupoolCreate(Name string, Scheduler Scheduler, Cpumap Bitma } // int libxl_cpupool_destroy(libxl_ctx *ctx, uint32_t poolid); -func (Ctx *Context) CpupoolDestroy(Poolid uint32) (err error) { - ret := C.libxl_cpupool_destroy(Ctx.ctx, C.uint32_t(Poolid)) +func (ctx *Context) CpupoolDestroy(Poolid uint32) (err error) { + ret := C.libxl_cpupool_destroy(ctx.ctx, C.uint32_t(Poolid)) if ret != 0 { err = Error(-ret) return @@ -677,8 +677,8 @@ func (Ctx *Context) CpupoolDestroy(Poolid uint32) (err error) { } // int libxl_cpupool_cpuadd(libxl_ctx *ctx, uint32_t poolid, int cpu); -func (Ctx *Context) CpupoolCpuadd(Poolid uint32, Cpu int) (err error) { - ret := C.libxl_cpupool_cpuadd(Ctx.ctx, C.uint32_t(Poolid), C.int(Cpu)) +func (ctx *Context) CpupoolCpuadd(Poolid uint32, Cpu int) (err error) { + ret := C.libxl_cpupool_cpuadd(ctx.ctx, C.uint32_t(Poolid), C.int(Cpu)) if ret != 0 { err = Error(-ret) return @@ -689,14 +689,14 @@ func (Ctx *Context) CpupoolCpuadd(Poolid uint32, Cpu int) (err error) { // int libxl_cpupool_cpuadd_cpumap(libxl_ctx *ctx, uint32_t poolid, // const libxl_bitmap *cpumap); -func (Ctx *Context) CpupoolCpuaddCpumap(Poolid uint32, Cpumap Bitmap) (err error) { +func (ctx *Context) CpupoolCpuaddCpumap(Poolid uint32, Cpumap Bitmap) (err error) { var cbm C.libxl_bitmap if err = Cpumap.toC(&cbm); err != nil { return } defer C.libxl_bitmap_dispose(&cbm) - ret := C.libxl_cpupool_cpuadd_cpumap(Ctx.ctx, C.uint32_t(Poolid), &cbm) + ret := C.libxl_cpupool_cpuadd_cpumap(ctx.ctx, C.uint32_t(Poolid), &cbm) if ret != 0 { err = Error(-ret) return @@ -706,8 +706,8 @@ func (Ctx *Context) CpupoolCpuaddCpumap(Poolid uint32, Cpumap Bitmap) (err error } // int libxl_cpupool_cpuremove(libxl_ctx *ctx, uint32_t poolid, int cpu); -func (Ctx *Context) CpupoolCpuremove(Poolid uint32, Cpu int) (err error) { - ret := C.libxl_cpupool_cpuremove(Ctx.ctx, C.uint32_t(Poolid), C.int(Cpu)) +func (ctx *Context) CpupoolCpuremove(Poolid uint32, Cpu int) (err error) { + ret := C.libxl_cpupool_cpuremove(ctx.ctx, C.uint32_t(Poolid), C.int(Cpu)) if ret != 0 { err = Error(-ret) return @@ -718,14 +718,14 @@ func (Ctx *Context) CpupoolCpuremove(Poolid uint32, Cpu int) (err error) { // int libxl_cpupool_cpuremove_cpumap(libxl_ctx *ctx, uint32_t poolid, // const libxl_bitmap *cpumap); -func (Ctx *Context) CpupoolCpuremoveCpumap(Poolid uint32, Cpumap Bitmap) (err error) { +func (ctx *Context) CpupoolCpuremoveCpumap(Poolid uint32, Cpumap Bitmap) (err error) { var cbm C.libxl_bitmap if err = Cpumap.toC(&cbm); err != nil { return } defer C.libxl_bitmap_dispose(&cbm) - ret := C.libxl_cpupool_cpuremove_cpumap(Ctx.ctx, C.uint32_t(Poolid), &cbm) + ret := C.libxl_cpupool_cpuremove_cpumap(ctx.ctx, C.uint32_t(Poolid), &cbm) if ret != 0 { err = Error(-ret) return @@ -735,11 +735,11 @@ func (Ctx *Context) CpupoolCpuremoveCpumap(Poolid uint32, Cpumap Bitmap) (err er } // int libxl_cpupool_rename(libxl_ctx *ctx, const char *name, uint32_t poolid); -func (Ctx *Context) CpupoolRename(Name string, Poolid uint32) (err error) { +func (ctx *Context) CpupoolRename(Name string, Poolid uint32) (err error) { name := C.CString(Name) defer C.free(unsafe.Pointer(name)) - ret := C.libxl_cpupool_rename(Ctx.ctx, name, C.uint32_t(Poolid)) + ret := C.libxl_cpupool_rename(ctx.ctx, name, C.uint32_t(Poolid)) if ret != 0 { err = Error(-ret) return @@ -749,10 +749,10 @@ func (Ctx *Context) CpupoolRename(Name string, Poolid uint32) (err error) { } // int libxl_cpupool_cpuadd_node(libxl_ctx *ctx, uint32_t poolid, int node, int *cpus); -func (Ctx *Context) CpupoolCpuaddNode(Poolid uint32, Node int) (Cpus int, err error) { +func (ctx *Context) CpupoolCpuaddNode(Poolid uint32, Node int) (Cpus int, err error) { ccpus := C.int(0) - ret := C.libxl_cpupool_cpuadd_node(Ctx.ctx, C.uint32_t(Poolid), C.int(Node), &ccpus) + ret := C.libxl_cpupool_cpuadd_node(ctx.ctx, C.uint32_t(Poolid), C.int(Node), &ccpus) if ret != 0 { err = Error(-ret) return @@ -764,10 +764,10 @@ func (Ctx *Context) CpupoolCpuaddNode(Poolid uint32, Node int) (Cpus int, err er } // int libxl_cpupool_cpuremove_node(libxl_ctx *ctx, uint32_t poolid, int node, int *cpus); -func (Ctx *Context) CpupoolCpuremoveNode(Poolid uint32, Node int) (Cpus int, err error) { +func (ctx *Context) CpupoolCpuremoveNode(Poolid uint32, Node int) (Cpus int, err error) { ccpus := C.int(0) - ret := C.libxl_cpupool_cpuremove_node(Ctx.ctx, C.uint32_t(Poolid), C.int(Node), &ccpus) + ret := C.libxl_cpupool_cpuremove_node(ctx.ctx, C.uint32_t(Poolid), C.int(Node), &ccpus) if ret != 0 { err = Error(-ret) return @@ -779,8 +779,8 @@ func (Ctx *Context) CpupoolCpuremoveNode(Poolid uint32, Node int) (Cpus int, err } // int libxl_cpupool_movedomain(libxl_ctx *ctx, uint32_t poolid, uint32_t domid); -func (Ctx *Context) CpupoolMovedomain(Poolid uint32, Id Domid) (err error) { - ret := C.libxl_cpupool_movedomain(Ctx.ctx, C.uint32_t(Poolid), C.uint32_t(Id)) +func (ctx *Context) CpupoolMovedomain(Poolid uint32, Id Domid) (err error) { + ret := C.libxl_cpupool_movedomain(ctx.ctx, C.uint32_t(Poolid), C.uint32_t(Id)) if ret != 0 { err = Error(-ret) return @@ -792,8 +792,8 @@ func (Ctx *Context) CpupoolMovedomain(Poolid uint32, Id Domid) (err error) { // // Utility functions // -func (Ctx *Context) CpupoolFindByName(name string) (info Cpupoolinfo, found bool) { - plist := Ctx.ListCpupool() +func (ctx *Context) CpupoolFindByName(name string) (info Cpupoolinfo, found bool) { + plist := ctx.ListCpupool() for i := range plist { if plist[i].PoolName == name { @@ -805,14 +805,14 @@ func (Ctx *Context) CpupoolFindByName(name string) (info Cpupoolinfo, found bool return } -func (Ctx *Context) CpupoolMakeFree(Cpumap Bitmap) (err error) { - plist := Ctx.ListCpupool() +func (ctx *Context) CpupoolMakeFree(Cpumap Bitmap) (err error) { + plist := ctx.ListCpupool() for i := range plist { var Intersection Bitmap Intersection = Cpumap.And(plist[i].Cpumap) if !Intersection.IsEmpty() { - err = Ctx.CpupoolCpuremoveCpumap(plist[i].Poolid, Intersection) + err = ctx.CpupoolCpuremoveCpumap(plist[i].Poolid, Intersection) if err != nil { return } @@ -940,8 +940,8 @@ func (bm Bitmap) String() (s string) { } //int libxl_get_max_cpus(libxl_ctx *ctx); -func (Ctx *Context) GetMaxCpus() (maxCpus int, err error) { - ret := C.libxl_get_max_cpus(Ctx.ctx) +func (ctx *Context) GetMaxCpus() (maxCpus int, err error) { + ret := C.libxl_get_max_cpus(ctx.ctx) if ret < 0 { err = Error(-ret) return @@ -951,8 +951,8 @@ func (Ctx *Context) GetMaxCpus() (maxCpus int, err error) { } //int libxl_get_online_cpus(libxl_ctx *ctx); -func (Ctx *Context) GetOnlineCpus() (onCpus int, err error) { - ret := C.libxl_get_online_cpus(Ctx.ctx) +func (ctx *Context) GetOnlineCpus() (onCpus int, err error) { + ret := C.libxl_get_online_cpus(ctx.ctx) if ret < 0 { err = Error(-ret) return @@ -962,8 +962,8 @@ func (Ctx *Context) GetOnlineCpus() (onCpus int, err error) { } //int libxl_get_max_nodes(libxl_ctx *ctx); -func (Ctx *Context) GetMaxNodes() (maxNodes int, err error) { - ret := C.libxl_get_max_nodes(Ctx.ctx) +func (ctx *Context) GetMaxNodes() (maxNodes int, err error) { + ret := C.libxl_get_max_nodes(ctx.ctx) if ret < 0 { err = Error(-ret) return @@ -973,9 +973,9 @@ func (Ctx *Context) GetMaxNodes() (maxNodes int, err error) { } //int libxl_get_free_memory(libxl_ctx *ctx, uint64_t *memkb); -func (Ctx *Context) GetFreeMemory() (memkb uint64, err error) { +func (ctx *Context) GetFreeMemory() (memkb uint64, err error) { var cmem C.uint64_t - ret := C.libxl_get_free_memory(Ctx.ctx, &cmem) + ret := C.libxl_get_free_memory(ctx.ctx, &cmem) if ret < 0 { err = Error(-ret) @@ -988,12 +988,12 @@ func (Ctx *Context) GetFreeMemory() (memkb uint64, err error) { } //int libxl_get_physinfo(libxl_ctx *ctx, libxl_physinfo *physinfo) -func (Ctx *Context) GetPhysinfo() (physinfo *Physinfo, err error) { +func (ctx *Context) GetPhysinfo() (physinfo *Physinfo, err error) { var cphys C.libxl_physinfo C.libxl_physinfo_init(&cphys) defer C.libxl_physinfo_dispose(&cphys) - ret := C.libxl_get_physinfo(Ctx.ctx, &cphys) + ret := C.libxl_get_physinfo(ctx.ctx, &cphys) if ret < 0 { err = Error(ret) @@ -1005,22 +1005,22 @@ func (Ctx *Context) GetPhysinfo() (physinfo *Physinfo, err error) { } //const libxl_version_info* libxl_get_version_info(libxl_ctx *ctx); -func (Ctx *Context) GetVersionInfo() (info *VersionInfo, err error) { +func (ctx *Context) GetVersionInfo() (info *VersionInfo, err error) { var cinfo *C.libxl_version_info - cinfo = C.libxl_get_version_info(Ctx.ctx) + cinfo = C.libxl_get_version_info(ctx.ctx) err = info.fromC(cinfo) return } -func (Ctx *Context) DomainInfo(Id Domid) (di *Dominfo, err error) { +func (ctx *Context) DomainInfo(Id Domid) (di *Dominfo, err error) { var cdi C.libxl_dominfo C.libxl_dominfo_init(&cdi) defer C.libxl_dominfo_dispose(&cdi) - ret := C.libxl_domain_info(Ctx.ctx, &cdi, C.uint32_t(Id)) + ret := C.libxl_domain_info(ctx.ctx, &cdi, C.uint32_t(Id)) if ret != 0 { err = Error(-ret) @@ -1032,8 +1032,8 @@ func (Ctx *Context) DomainInfo(Id Domid) (di *Dominfo, err error) { return } -func (Ctx *Context) DomainUnpause(Id Domid) (err error) { - ret := C.libxl_domain_unpause(Ctx.ctx, C.uint32_t(Id), nil) +func (ctx *Context) DomainUnpause(Id Domid) (err error) { + ret := C.libxl_domain_unpause(ctx.ctx, C.uint32_t(Id), nil) if ret != 0 { err = Error(-ret) @@ -1042,8 +1042,8 @@ func (Ctx *Context) DomainUnpause(Id Domid) (err error) { } //int libxl_domain_pause(libxl_ctx *ctx, uint32_t domain); -func (Ctx *Context) DomainPause(id Domid) (err error) { - ret := C.libxl_domain_pause(Ctx.ctx, C.uint32_t(id), nil) +func (ctx *Context) DomainPause(id Domid) (err error) { + ret := C.libxl_domain_pause(ctx.ctx, C.uint32_t(id), nil) if ret != 0 { err = Error(-ret) @@ -1052,8 +1052,8 @@ func (Ctx *Context) DomainPause(id Domid) (err error) { } //int libxl_domain_shutdown(libxl_ctx *ctx, uint32_t domid); -func (Ctx *Context) DomainShutdown(id Domid) (err error) { - ret := C.libxl_domain_shutdown(Ctx.ctx, C.uint32_t(id), nil) +func (ctx *Context) DomainShutdown(id Domid) (err error) { + ret := C.libxl_domain_shutdown(ctx.ctx, C.uint32_t(id), nil) if ret != 0 { err = Error(-ret) @@ -1062,8 +1062,8 @@ func (Ctx *Context) DomainShutdown(id Domid) (err error) { } //int libxl_domain_reboot(libxl_ctx *ctx, uint32_t domid); -func (Ctx *Context) DomainReboot(id Domid) (err error) { - ret := C.libxl_domain_reboot(Ctx.ctx, C.uint32_t(id), nil) +func (ctx *Context) DomainReboot(id Domid) (err error) { + ret := C.libxl_domain_reboot(ctx.ctx, C.uint32_t(id), nil) if ret != 0 { err = Error(-ret) @@ -1073,9 +1073,9 @@ func (Ctx *Context) DomainReboot(id Domid) (err error) { //libxl_dominfo * libxl_list_domain(libxl_ctx*, int *nb_domain_out); //void libxl_dominfo_list_free(libxl_dominfo *list, int nb_domain); -func (Ctx *Context) ListDomain() (glist []Dominfo) { +func (ctx *Context) ListDomain() (glist []Dominfo) { var nbDomain C.int - clist := C.libxl_list_domain(Ctx.ctx, &nbDomain) + clist := C.libxl_list_domain(ctx.ctx, &nbDomain) defer C.libxl_dominfo_list_free(clist, nbDomain) if int(nbDomain) == 0 { @@ -1095,11 +1095,11 @@ func (Ctx *Context) ListDomain() (glist []Dominfo) { //libxl_vcpuinfo *libxl_list_vcpu(libxl_ctx *ctx, uint32_t domid, // int *nb_vcpu, int *nr_cpus_out); //void libxl_vcpuinfo_list_free(libxl_vcpuinfo *, int nr_vcpus); -func (Ctx *Context) ListVcpu(id Domid) (glist []Vcpuinfo) { +func (ctx *Context) ListVcpu(id Domid) (glist []Vcpuinfo) { var nbVcpu C.int var nrCpu C.int - clist := C.libxl_list_vcpu(Ctx.ctx, C.uint32_t(id), &nbVcpu, &nrCpu) + clist := C.libxl_list_vcpu(ctx.ctx, C.uint32_t(id), &nbVcpu, &nrCpu) defer C.libxl_vcpuinfo_list_free(clist, nbVcpu) if int(nbVcpu) == 0 { @@ -1125,9 +1125,9 @@ func (ct ConsoleType) String() (str string) { //int libxl_console_get_tty(libxl_ctx *ctx, uint32_t domid, int cons_num, //libxl_console_type type, char **path); -func (Ctx *Context) ConsoleGetTty(id Domid, consNum int, conType ConsoleType) (path string, err error) { +func (ctx *Context) ConsoleGetTty(id Domid, consNum int, conType ConsoleType) (path string, err error) { var cpath *C.char - ret := C.libxl_console_get_tty(Ctx.ctx, C.uint32_t(id), C.int(consNum), C.libxl_console_type(conType), &cpath) + ret := C.libxl_console_get_tty(ctx.ctx, C.uint32_t(id), C.int(consNum), C.libxl_console_type(conType), &cpath) if ret != 0 { err = Error(-ret) return @@ -1140,9 +1140,9 @@ func (Ctx *Context) ConsoleGetTty(id Domid, consNum int, conType ConsoleType) (p //int libxl_primary_console_get_tty(libxl_ctx *ctx, uint32_t domid_vm, // char **path); -func (Ctx *Context) PrimaryConsoleGetTty(domid uint32) (path string, err error) { +func (ctx *Context) PrimaryConsoleGetTty(domid uint32) (path string, err error) { var cpath *C.char - ret := C.libxl_primary_console_get_tty(Ctx.ctx, C.uint32_t(domid), &cpath) + ret := C.libxl_primary_console_get_tty(ctx.ctx, C.uint32_t(domid), &cpath) if ret != 0 { err = Error(-ret) return @@ -1154,7 +1154,7 @@ func (Ctx *Context) PrimaryConsoleGetTty(domid uint32) (path string, err error) } // DeviceNicAdd adds a nic to a domain. -func (Ctx *Context) DeviceNicAdd(domid Domid, nic *DeviceNic) error { +func (ctx *Context) DeviceNicAdd(domid Domid, nic *DeviceNic) error { var cnic C.libxl_device_nic if err := nic.toC(&cnic); err != nil { @@ -1162,7 +1162,7 @@ func (Ctx *Context) DeviceNicAdd(domid Domid, nic *DeviceNic) error { } defer C.libxl_device_nic_dispose(&cnic) - ret := C.libxl_device_nic_add(Ctx.ctx, C.uint32_t(domid), &cnic, nil) + ret := C.libxl_device_nic_add(ctx.ctx, C.uint32_t(domid), &cnic, nil) if ret != 0 { return Error(ret) } @@ -1171,7 +1171,7 @@ func (Ctx *Context) DeviceNicAdd(domid Domid, nic *DeviceNic) error { } // DeviceNicRemove removes a nic from a domain. -func (Ctx *Context) DeviceNicRemove(domid Domid, nic *DeviceNic) error { +func (ctx *Context) DeviceNicRemove(domid Domid, nic *DeviceNic) error { var cnic C.libxl_device_nic if err := nic.toC(&cnic); err != nil { @@ -1179,7 +1179,7 @@ func (Ctx *Context) DeviceNicRemove(domid Domid, nic *DeviceNic) error { } defer C.libxl_device_nic_dispose(&cnic) - ret := C.libxl_device_nic_remove(Ctx.ctx, C.uint32_t(domid), &cnic, nil) + ret := C.libxl_device_nic_remove(ctx.ctx, C.uint32_t(domid), &cnic, nil) if ret != 0 { return Error(ret) } @@ -1188,7 +1188,7 @@ func (Ctx *Context) DeviceNicRemove(domid Domid, nic *DeviceNic) error { } // DevicePciAdd is used to passthrough a PCI device to a domain. -func (Ctx *Context) DevicePciAdd(domid Domid, pci *DevicePci) error { +func (ctx *Context) DevicePciAdd(domid Domid, pci *DevicePci) error { var cpci C.libxl_device_pci if err := pci.toC(&cpci); err != nil { @@ -1196,7 +1196,7 @@ func (Ctx *Context) DevicePciAdd(domid Domid, pci *DevicePci) error { } defer C.libxl_device_pci_dispose(&cpci) - ret := C.libxl_device_pci_add(Ctx.ctx, C.uint32_t(domid), &cpci, nil) + ret := C.libxl_device_pci_add(ctx.ctx, C.uint32_t(domid), &cpci, nil) if ret != 0 { return Error(ret) } @@ -1205,7 +1205,7 @@ func (Ctx *Context) DevicePciAdd(domid Domid, pci *DevicePci) error { } // DevicePciRemove removes a PCI device from a domain. -func (Ctx *Context) DevicePciRemove(domid Domid, pci *DevicePci) error { +func (ctx *Context) DevicePciRemove(domid Domid, pci *DevicePci) error { var cpci C.libxl_device_pci if err := pci.toC(&cpci); err != nil { @@ -1213,7 +1213,7 @@ func (Ctx *Context) DevicePciRemove(domid Domid, pci *DevicePci) error { } defer C.libxl_device_pci_dispose(&cpci) - ret := C.libxl_device_pci_remove(Ctx.ctx, C.uint32_t(domid), &cpci, nil) + ret := C.libxl_device_pci_remove(ctx.ctx, C.uint32_t(domid), &cpci, nil) if ret != 0 { return Error(ret) } @@ -1222,7 +1222,7 @@ func (Ctx *Context) DevicePciRemove(domid Domid, pci *DevicePci) error { } // DeviceUsbdevAdd adds a USB device to a domain. -func (Ctx *Context) DeviceUsbdevAdd(domid Domid, usbdev *DeviceUsbdev) error { +func (ctx *Context) DeviceUsbdevAdd(domid Domid, usbdev *DeviceUsbdev) error { var cusbdev C.libxl_device_usbdev if err := usbdev.toC(&cusbdev); err != nil { @@ -1230,7 +1230,7 @@ func (Ctx *Context) DeviceUsbdevAdd(domid Domid, usbdev *DeviceUsbdev) error { } defer C.libxl_device_usbdev_dispose(&cusbdev) - ret := C.libxl_device_usbdev_add(Ctx.ctx, C.uint32_t(domid), &cusbdev, nil) + ret := C.libxl_device_usbdev_add(ctx.ctx, C.uint32_t(domid), &cusbdev, nil) if ret != 0 { return Error(ret) } @@ -1239,7 +1239,7 @@ func (Ctx *Context) DeviceUsbdevAdd(domid Domid, usbdev *DeviceUsbdev) error { } // DeviceUsbdevRemove removes a USB device from a domain. -func (Ctx *Context) DeviceUsbdevRemove(domid Domid, usbdev *DeviceUsbdev) error { +func (ctx *Context) DeviceUsbdevRemove(domid Domid, usbdev *DeviceUsbdev) error { var cusbdev C.libxl_device_usbdev if err := usbdev.toC(&cusbdev); err != nil { @@ -1247,7 +1247,7 @@ func (Ctx *Context) DeviceUsbdevRemove(domid Domid, usbdev *DeviceUsbdev) error } defer C.libxl_device_usbdev_dispose(&cusbdev) - ret := C.libxl_device_usbdev_remove(Ctx.ctx, C.uint32_t(domid), &cusbdev, nil) + ret := C.libxl_device_usbdev_remove(ctx.ctx, C.uint32_t(domid), &cusbdev, nil) if ret != 0 { return Error(ret) } @@ -1256,7 +1256,7 @@ func (Ctx *Context) DeviceUsbdevRemove(domid Domid, usbdev *DeviceUsbdev) error } // DomainCreateNew creates a new domain. -func (Ctx *Context) DomainCreateNew(config *DomainConfig) (Domid, error) { +func (ctx *Context) DomainCreateNew(config *DomainConfig) (Domid, error) { var cdomid C.uint32_t var cconfig C.libxl_domain_config err := config.toC(&cconfig) @@ -1265,7 +1265,7 @@ func (Ctx *Context) DomainCreateNew(config *DomainConfig) (Domid, error) { } defer C.libxl_domain_config_dispose(&cconfig) - ret := C.libxl_domain_create_new(Ctx.ctx, &cconfig, &cdomid, nil, nil) + ret := C.libxl_domain_create_new(ctx.ctx, &cconfig, &cdomid, nil, nil) if ret != 0 { return Domid(0), Error(ret) } -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Mon Jun 21 15:55:56 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Mon, 21 Jun 2021 15:55:56 +0000 Received: from list by lists.xenproject.org with outflank-mailman.145603.267757 (Exim 4.92) (envelope-from ) id 1lvMHE-0001eI-Mz; Mon, 21 Jun 2021 15:55:56 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 145603.267757; Mon, 21 Jun 2021 15:55:56 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lvMHE-0001eA-Jv; Mon, 21 Jun 2021 15:55:56 +0000 Received: by outflank-mailman (input) for mailman id 145603; Mon, 21 Jun 2021 15:55:55 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lvMHC-0001dq-W5 for xen-changelog@lists.xenproject.org; Mon, 21 Jun 2021 15:55:54 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lvMHC-00040r-VN for xen-changelog@lists.xenproject.org; Mon, 21 Jun 2021 15:55:54 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lvMHC-0000bd-Ug for xen-changelog@lists.xenproject.org; Mon, 21 Jun 2021 15:55:54 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=BvkGh8Umr7fEBTFv95DNGHI/0wN/ppBnqIRnJHBWKkI=; b=D6H+d6RefKMBqPkW8yRUK7zQBL 8FXMG54QYGs+nsSmk1DCUPAyK5NVSod1tIK0KulbfKIbhsHe4WerCnq4okhUOGc3BvLH19nTCAZnh oTzZ1lMyKaSTlMGrLz2ibBbj+cI/VOkO9P35ZqyljqM/BkvqTO2APlQfOiFvRQfuAWvU=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] golang/xenlight: add DomainDestroy wrapper Message-Id: Date: Mon, 21 Jun 2021 15:55:54 +0000 commit 9b6d865e0af56e376740ba03b1ccdf316362a71e Author: Nick Rosbrook AuthorDate: Mon May 24 16:36:50 2021 -0400 Commit: George Dunlap CommitDate: Mon Jun 21 16:49:08 2021 +0100 golang/xenlight: add DomainDestroy wrapper Add a wrapper around libxl_domain_destroy. Signed-off-by: Nick Rosbrook Reviewed-by: George Dunlap --- tools/golang/xenlight/xenlight.go | 10 ++++++++++ 1 file changed, 10 insertions(+) diff --git a/tools/golang/xenlight/xenlight.go b/tools/golang/xenlight/xenlight.go index fc3eb0bf3f..f605f17613 100644 --- a/tools/golang/xenlight/xenlight.go +++ b/tools/golang/xenlight/xenlight.go @@ -1272,3 +1272,13 @@ func (ctx *Context) DomainCreateNew(config *DomainConfig) (Domid, error) { return Domid(cdomid), nil } + +// DomainDestroy destroys a domain given a domid. +func (ctx *Context) DomainDestroy(domid Domid) error { + ret := C.libxl_domain_destroy(ctx.ctx, C.uint32_t(domid), nil) + if ret != 0 { + return Error(ret) + } + + return nil +} -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Mon Jun 21 15:56:06 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Mon, 21 Jun 2021 15:56:06 +0000 Received: from list by lists.xenproject.org with outflank-mailman.145604.267761 (Exim 4.92) (envelope-from ) id 1lvMHO-0001hT-PK; Mon, 21 Jun 2021 15:56:06 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 145604.267761; Mon, 21 Jun 2021 15:56:06 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lvMHO-0001hL-LS; Mon, 21 Jun 2021 15:56:06 +0000 Received: by outflank-mailman (input) for mailman id 145604; Mon, 21 Jun 2021 15:56:05 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lvMHN-0001h3-2m for xen-changelog@lists.xenproject.org; Mon, 21 Jun 2021 15:56:05 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lvMHN-00041G-21 for xen-changelog@lists.xenproject.org; Mon, 21 Jun 2021 15:56:05 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lvMHN-0000cY-1B for xen-changelog@lists.xenproject.org; Mon, 21 Jun 2021 15:56:05 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=ica5mme/AMgwNvQbABf026sKYZsrxH0thH4YSrGOMMY=; b=eiVH60K44nwyATAJ3B6e4ORHWe 9PtWWpKzh/pX2Y0/d7G6TBhtY0OIgFNZaIVXHUWmhKXJ1ooGxEatBm428I2+v3mapBpm0hAqqCdc9 GAWfcIH10dIGaTl4Nr0bgWRnXZ8PdYDqLXfP4dF3UMuj+99SmozyshU76HJxEMTov8tY=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] golang/xenlight: add SendTrigger wrapper Message-Id: Date: Mon, 21 Jun 2021 15:56:05 +0000 commit 1d95fd75df18bf25cb445feb47caf62da25c00e8 Author: Nick Rosbrook AuthorDate: Mon May 24 16:36:51 2021 -0400 Commit: George Dunlap CommitDate: Mon Jun 21 16:49:08 2021 +0100 golang/xenlight: add SendTrigger wrapper Add a warpper around libxl_send_trigger. Signed-off-by: Nick Rosbrook Reviewed-by: George Dunlap --- tools/golang/xenlight/xenlight.go | 11 +++++++++++ 1 file changed, 11 insertions(+) diff --git a/tools/golang/xenlight/xenlight.go b/tools/golang/xenlight/xenlight.go index f605f17613..a651554d0c 100644 --- a/tools/golang/xenlight/xenlight.go +++ b/tools/golang/xenlight/xenlight.go @@ -1282,3 +1282,14 @@ func (ctx *Context) DomainDestroy(domid Domid) error { return nil } + +// SendTrigger sends a Trigger to the domain specified by domid. +func (ctx *Context) SendTrigger(domid Domid, trigger Trigger, vcpuid int) error { + ret := C.libxl_send_trigger(ctx.ctx, C.uint32_t(domid), + C.libxl_trigger(trigger), C.uint32_t(vcpuid), nil) + if ret != 0 { + return Error(ret) + } + + return nil +} -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Mon Jun 21 15:56:16 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Mon, 21 Jun 2021 15:56:16 +0000 Received: from list by lists.xenproject.org with outflank-mailman.145605.267765 (Exim 4.92) (envelope-from ) id 1lvMHY-0001la-SH; Mon, 21 Jun 2021 15:56:16 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 145605.267765; Mon, 21 Jun 2021 15:56:16 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lvMHY-0001lS-P8; Mon, 21 Jun 2021 15:56:16 +0000 Received: by outflank-mailman (input) for mailman id 145605; Mon, 21 Jun 2021 15:56:15 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lvMHX-0001lC-5n for xen-changelog@lists.xenproject.org; Mon, 21 Jun 2021 15:56:15 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lvMHX-00041R-50 for xen-changelog@lists.xenproject.org; Mon, 21 Jun 2021 15:56:15 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lvMHX-0000dW-45 for xen-changelog@lists.xenproject.org; Mon, 21 Jun 2021 15:56:15 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=NzvuGxbT584uZB/UT3lxtPpAVnfRmNjeWQI0avxsCGc=; b=yp+V/YMeO9rDInT/gFcxVUCHCB xLfC+zVC6/pI95/0oV22qWHU8HHHcJidMugbeWuqhH54ll89Odf7hc+aJGjxyOR+tyQANgIjoK+9x LUgrr6JMW/Hfz9CsT0CyI/Sf/bxXFUMbIsoEKAh4r5ZLNLDNjM4XL53h5+2Q7TftsUmU=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] golang/xenlight: do not negate ret when converting to Error Message-Id: Date: Mon, 21 Jun 2021 15:56:15 +0000 commit c9b59f9032d41be8bade8a25d9148cf6ed203704 Author: Nick Rosbrook AuthorDate: Mon May 24 16:36:52 2021 -0400 Commit: George Dunlap CommitDate: Mon Jun 21 16:49:08 2021 +0100 golang/xenlight: do not negate ret when converting to Error Commit 871e51d2d4 changed the sign on the xenlight error types (making the values negative, same as the C-generated constants), but failed to remove the code changing the sign before casting to Error(). This results in error strings like "libxl error: ", rather than the correct message. Fix all occurrances of this by running: gofmt -w -r 'Error(-ret) -> Error(ret)' xenlight.go from tools/golang/xenlight. Signed-off-by: Nick Rosbrook Acked-by: George Dunlap --- tools/golang/xenlight/xenlight.go | 46 +++++++++++++++++++-------------------- 1 file changed, 23 insertions(+), 23 deletions(-) diff --git a/tools/golang/xenlight/xenlight.go b/tools/golang/xenlight/xenlight.go index a651554d0c..a45c636952 100644 --- a/tools/golang/xenlight/xenlight.go +++ b/tools/golang/xenlight/xenlight.go @@ -583,7 +583,7 @@ func SchedulerFromString(name string) (s Scheduler, err error) { ret := C.libxl_scheduler_from_string(cname, &cs) if ret != 0 { - err = Error(-ret) + err = Error(ret) return } @@ -622,7 +622,7 @@ func (ctx *Context) CpupoolInfo(Poolid uint32) (pool Cpupoolinfo, err error) { ret := C.libxl_cpupool_info(ctx.ctx, &c_cpupool, C.uint32_t(Poolid)) if ret != 0 { - err = Error(-ret) + err = Error(ret) return } defer C.libxl_cpupoolinfo_dispose(&c_cpupool) @@ -656,7 +656,7 @@ func (ctx *Context) CpupoolCreate(Name string, Scheduler Scheduler, Cpumap Bitma ret := C.libxl_cpupool_create(ctx.ctx, name, C.libxl_scheduler(Scheduler), cbm, &uuid, &poolid) if ret != 0 { - err = Error(-ret) + err = Error(ret) return } @@ -669,7 +669,7 @@ func (ctx *Context) CpupoolCreate(Name string, Scheduler Scheduler, Cpumap Bitma func (ctx *Context) CpupoolDestroy(Poolid uint32) (err error) { ret := C.libxl_cpupool_destroy(ctx.ctx, C.uint32_t(Poolid)) if ret != 0 { - err = Error(-ret) + err = Error(ret) return } @@ -680,7 +680,7 @@ func (ctx *Context) CpupoolDestroy(Poolid uint32) (err error) { func (ctx *Context) CpupoolCpuadd(Poolid uint32, Cpu int) (err error) { ret := C.libxl_cpupool_cpuadd(ctx.ctx, C.uint32_t(Poolid), C.int(Cpu)) if ret != 0 { - err = Error(-ret) + err = Error(ret) return } @@ -698,7 +698,7 @@ func (ctx *Context) CpupoolCpuaddCpumap(Poolid uint32, Cpumap Bitmap) (err error ret := C.libxl_cpupool_cpuadd_cpumap(ctx.ctx, C.uint32_t(Poolid), &cbm) if ret != 0 { - err = Error(-ret) + err = Error(ret) return } @@ -709,7 +709,7 @@ func (ctx *Context) CpupoolCpuaddCpumap(Poolid uint32, Cpumap Bitmap) (err error func (ctx *Context) CpupoolCpuremove(Poolid uint32, Cpu int) (err error) { ret := C.libxl_cpupool_cpuremove(ctx.ctx, C.uint32_t(Poolid), C.int(Cpu)) if ret != 0 { - err = Error(-ret) + err = Error(ret) return } @@ -727,7 +727,7 @@ func (ctx *Context) CpupoolCpuremoveCpumap(Poolid uint32, Cpumap Bitmap) (err er ret := C.libxl_cpupool_cpuremove_cpumap(ctx.ctx, C.uint32_t(Poolid), &cbm) if ret != 0 { - err = Error(-ret) + err = Error(ret) return } @@ -741,7 +741,7 @@ func (ctx *Context) CpupoolRename(Name string, Poolid uint32) (err error) { ret := C.libxl_cpupool_rename(ctx.ctx, name, C.uint32_t(Poolid)) if ret != 0 { - err = Error(-ret) + err = Error(ret) return } @@ -754,7 +754,7 @@ func (ctx *Context) CpupoolCpuaddNode(Poolid uint32, Node int) (Cpus int, err er ret := C.libxl_cpupool_cpuadd_node(ctx.ctx, C.uint32_t(Poolid), C.int(Node), &ccpus) if ret != 0 { - err = Error(-ret) + err = Error(ret) return } @@ -769,7 +769,7 @@ func (ctx *Context) CpupoolCpuremoveNode(Poolid uint32, Node int) (Cpus int, err ret := C.libxl_cpupool_cpuremove_node(ctx.ctx, C.uint32_t(Poolid), C.int(Node), &ccpus) if ret != 0 { - err = Error(-ret) + err = Error(ret) return } @@ -782,7 +782,7 @@ func (ctx *Context) CpupoolCpuremoveNode(Poolid uint32, Node int) (Cpus int, err func (ctx *Context) CpupoolMovedomain(Poolid uint32, Id Domid) (err error) { ret := C.libxl_cpupool_movedomain(ctx.ctx, C.uint32_t(Poolid), C.uint32_t(Id)) if ret != 0 { - err = Error(-ret) + err = Error(ret) return } @@ -943,7 +943,7 @@ func (bm Bitmap) String() (s string) { func (ctx *Context) GetMaxCpus() (maxCpus int, err error) { ret := C.libxl_get_max_cpus(ctx.ctx) if ret < 0 { - err = Error(-ret) + err = Error(ret) return } maxCpus = int(ret) @@ -954,7 +954,7 @@ func (ctx *Context) GetMaxCpus() (maxCpus int, err error) { func (ctx *Context) GetOnlineCpus() (onCpus int, err error) { ret := C.libxl_get_online_cpus(ctx.ctx) if ret < 0 { - err = Error(-ret) + err = Error(ret) return } onCpus = int(ret) @@ -965,7 +965,7 @@ func (ctx *Context) GetOnlineCpus() (onCpus int, err error) { func (ctx *Context) GetMaxNodes() (maxNodes int, err error) { ret := C.libxl_get_max_nodes(ctx.ctx) if ret < 0 { - err = Error(-ret) + err = Error(ret) return } maxNodes = int(ret) @@ -978,7 +978,7 @@ func (ctx *Context) GetFreeMemory() (memkb uint64, err error) { ret := C.libxl_get_free_memory(ctx.ctx, &cmem) if ret < 0 { - err = Error(-ret) + err = Error(ret) return } @@ -1023,7 +1023,7 @@ func (ctx *Context) DomainInfo(Id Domid) (di *Dominfo, err error) { ret := C.libxl_domain_info(ctx.ctx, &cdi, C.uint32_t(Id)) if ret != 0 { - err = Error(-ret) + err = Error(ret) return } @@ -1036,7 +1036,7 @@ func (ctx *Context) DomainUnpause(Id Domid) (err error) { ret := C.libxl_domain_unpause(ctx.ctx, C.uint32_t(Id), nil) if ret != 0 { - err = Error(-ret) + err = Error(ret) } return } @@ -1046,7 +1046,7 @@ func (ctx *Context) DomainPause(id Domid) (err error) { ret := C.libxl_domain_pause(ctx.ctx, C.uint32_t(id), nil) if ret != 0 { - err = Error(-ret) + err = Error(ret) } return } @@ -1056,7 +1056,7 @@ func (ctx *Context) DomainShutdown(id Domid) (err error) { ret := C.libxl_domain_shutdown(ctx.ctx, C.uint32_t(id), nil) if ret != 0 { - err = Error(-ret) + err = Error(ret) } return } @@ -1066,7 +1066,7 @@ func (ctx *Context) DomainReboot(id Domid) (err error) { ret := C.libxl_domain_reboot(ctx.ctx, C.uint32_t(id), nil) if ret != 0 { - err = Error(-ret) + err = Error(ret) } return } @@ -1129,7 +1129,7 @@ func (ctx *Context) ConsoleGetTty(id Domid, consNum int, conType ConsoleType) (p var cpath *C.char ret := C.libxl_console_get_tty(ctx.ctx, C.uint32_t(id), C.int(consNum), C.libxl_console_type(conType), &cpath) if ret != 0 { - err = Error(-ret) + err = Error(ret) return } defer C.free(unsafe.Pointer(cpath)) @@ -1144,7 +1144,7 @@ func (ctx *Context) PrimaryConsoleGetTty(domid uint32) (path string, err error) var cpath *C.char ret := C.libxl_primary_console_get_tty(ctx.ctx, C.uint32_t(domid), &cpath) if ret != 0 { - err = Error(-ret) + err = Error(ret) return } defer C.free(unsafe.Pointer(cpath)) -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Tue Jun 22 17:11:13 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 22 Jun 2021 17:11:13 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146003.268568 (Exim 4.92) (envelope-from ) id 1lvjvW-0003d9-IF; Tue, 22 Jun 2021 17:11:06 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146003.268568; Tue, 22 Jun 2021 17:11:06 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lvjvW-0003d0-Et; Tue, 22 Jun 2021 17:11:06 +0000 Received: by outflank-mailman (input) for mailman id 146003; Tue, 22 Jun 2021 17:11:05 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lvjvV-0003cu-5q for xen-changelog@lists.xenproject.org; Tue, 22 Jun 2021 17:11:05 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lvjvV-0007oD-3Q for xen-changelog@lists.xenproject.org; Tue, 22 Jun 2021 17:11:05 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lvjvV-00071b-2N for xen-changelog@lists.xenproject.org; Tue, 22 Jun 2021 17:11:05 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=GqHwLVUXOOlzFo050SIPmg+7ZN1BZhSiO0jyvmkUbU4=; b=CdrzbbbYWZ3UdJb3YtWvJ7PtA7 nFrmuGx7Kn0Ezxa+gb+u3f7qzLAC09d+MrdB81tVyyfv6h51g3L3tO+ZD2fVs05m7fVSjyMnZveFy d+DgNiUHS1T24zs5E/ybEXn42ByEeIQ0OorH/xHnvrVoXgWKKiza+8UEsCp/smgEfzrA=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] Revert "tools/firmware/ovmf: Use OvmfXen platform file is exist" Message-Id: Date: Tue, 22 Jun 2021 17:11:05 +0000 commit c7691f5e340f3b579d40c77548f63133cdf5aac7 Author: Andrew Cooper AuthorDate: Tue Jun 22 16:14:53 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 22 18:00:08 2021 +0100 Revert "tools/firmware/ovmf: Use OvmfXen platform file is exist" This reverts commit aad7b5c11d51d57659978e04702ac970906894e8. The change from OvmfX64 to OvmfXen causes a change in behaviour, whereby OvmfXen maps its shared info page at the top of address space. When trying to migrate such a domain, XENMEM_maximum_gpfn returns a very large value. This has uncovered multiple issues: 1) The userspace hypercall wrappers truncate all return values to int on Linux and Solaris, even on 64bit. This needs fixing in libxenctrl. 2) 32bit toolstacks can't migrate any domain with RAM above the 2^40 mark, because of virtual address constraints. This needs fixing in OVMF. Fixes for both of these aren't completely trivial. Revert the change to unblock staging in the meantime. Signed-off-by: Andrew Cooper Acked-by: Anthony PERARD --- tools/firmware/ovmf-makefile | 10 ++-------- 1 file changed, 2 insertions(+), 8 deletions(-) diff --git a/tools/firmware/ovmf-makefile b/tools/firmware/ovmf-makefile index 637ee509c3..55f9992145 100644 --- a/tools/firmware/ovmf-makefile +++ b/tools/firmware/ovmf-makefile @@ -17,14 +17,8 @@ all: build .PHONY: build build: if test -e .git ; then $(GIT) submodule update --init --recursive ; fi - set -ex; \ - if test -e OvmfPkg/OvmfXen.dsc; then \ - OvmfPkg/build.sh -a X64 -b $(TARGET) -n 4 -p OvmfPkg/OvmfXen.dsc; \ - cp Build/OvmfXen/$(TARGET)_GCC*/FV/OVMF.fd ovmf.bin; \ - else \ - OvmfPkg/build.sh -a X64 -b $(TARGET) -n 4; \ - cp Build/OvmfX64/$(TARGET)_GCC*/FV/OVMF.fd ovmf.bin; \ - fi + OvmfPkg/build.sh -a X64 -b $(TARGET) -n 4 + cp Build/OvmfX64/$(TARGET)_GCC*/FV/OVMF.fd ovmf.bin .PHONY: clean clean: -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:22:07 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:22:07 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146573.269683 (Exim 4.92) (envelope-from ) id 1lwMUl-00044S-G1; Thu, 24 Jun 2021 10:22:03 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146573.269683; Thu, 24 Jun 2021 10:22:03 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMUl-00044K-DC; Thu, 24 Jun 2021 10:22:03 +0000 Received: by outflank-mailman (input) for mailman id 146573; Thu, 24 Jun 2021 10:22:02 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMUk-000442-LU for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:22:02 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMUk-0005uw-Ka for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:22:02 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwMUk-00067f-JV for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:22:02 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=goXWlMt/xWcPE6R5qb2EBwyKna85bHAnn9km0o5s174=; b=xGll12IF2KEhM08JLLFWOLJeTe /DQCOn//27CE4P4GwCkYlpKFn2xO7whrOsRHCAR3bQMb/emHu70f2mpO6hX4flbLLRDfTxvkk04+z UpvVAbDUs4JvAVf4weMK3+JNdoXcv6X+5spBJ15mK7G0JrUDccBdIOXSnlcRoc8VbFss=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] SUPPORT.md: Un-shimmed 32-bit PV guests are no longer supported Message-Id: Date: Thu, 24 Jun 2021 10:22:02 +0000 commit 1a0f2fe2297d122a08fee2b26de5de995fdeca13 Author: George Dunlap AuthorDate: Thu May 6 13:38:02 2021 +0100 Commit: George Dunlap CommitDate: Fri Jun 4 17:24:05 2021 +0100 SUPPORT.md: Un-shimmed 32-bit PV guests are no longer supported The support status of 32-bit guests doesn't seem particularly useful. With it changed to fully unsupported outside of PV-shim, adjust the PV32 Kconfig default accordingly. Reported-by: Jann Horn Signed-off-by: George Dunlap Signed-off-by: Jan Beulich --- SUPPORT.md | 9 +-------- xen/arch/x86/Kconfig | 7 +++++-- 2 files changed, 6 insertions(+), 10 deletions(-) diff --git a/SUPPORT.md b/SUPPORT.md index 7a53635c9e..317392d8f3 100644 --- a/SUPPORT.md +++ b/SUPPORT.md @@ -86,14 +86,7 @@ No hardware requirements Status, x86_64: Supported Status, x86_32, shim: Supported - Status, x86_32, without shim: Supported, with caveats - -Due to architectural limitations, -32-bit PV guests must be assumed to be able to read arbitrary host memory -using speculative execution attacks. -Advisories will continue to be issued -for new vulnerabilities related to un-shimmed 32-bit PV guests -enabling denial-of-service attacks or privilege escalation attacks. + Status, x86_32, without shim: Supported, not security supported ### x86/HVM diff --git a/xen/arch/x86/Kconfig b/xen/arch/x86/Kconfig index e55e029b79..9b164db641 100644 --- a/xen/arch/x86/Kconfig +++ b/xen/arch/x86/Kconfig @@ -55,7 +55,7 @@ config PV config PV32 bool "Support for 32bit PV guests" depends on PV - default y + default PV_SHIM select COMPAT ---help--- The 32bit PV ABI uses Ring1, an area of the x86 architecture which @@ -67,7 +67,10 @@ config PV32 reduction, or performance reasons. Backwards compatibility can be provided via the PV Shim mechanism. - If unsure, say Y. + Note that outside of PV Shim, 32-bit PV guests are not security + supported anymore. + + If unsure, use the default setting. config PV_LINEAR_PT bool "Support for PV linear pagetables" -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:22:13 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:22:13 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146575.269688 (Exim 4.92) (envelope-from ) id 1lwMUv-00046y-Hv; Thu, 24 Jun 2021 10:22:13 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146575.269688; Thu, 24 Jun 2021 10:22:13 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMUv-00046o-Ej; Thu, 24 Jun 2021 10:22:13 +0000 Received: by outflank-mailman (input) for mailman id 146575; Thu, 24 Jun 2021 10:22:12 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMUu-00046c-Om for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:22:12 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMUu-0005vQ-Nz for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:22:12 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwMUu-0006AK-N6 for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:22:12 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=kgui8yObfcbMT4pSPYFgn94JDG6m7QIv+MgIJmnK1mY=; b=RYAOzDN9eamRgEg7slngYv5vuK x9xsnwwT9eYag/AvrSisxHzHUheMrVqhCIxaNxIym29Bob91BYCh948mp24TgCoCdjZkzJKrg/Fc8 wK6hyM6QxFF4AsSCcSln7+w/I+hcl9x7fizqELo8GJhrmDV5IySvbWcQvyO8K84DRl58=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] tools/libs/guest: fix max_pfn setting in map_p2m() Message-Id: Date: Thu, 24 Jun 2021 10:22:12 +0000 commit 7bd8989ab77b6ade3b7a5f4b640a55248d1791a3 Author: Juergen Gross AuthorDate: Fri Jun 4 08:02:09 2021 +0200 Commit: Julien Grall CommitDate: Fri Jun 4 18:56:46 2021 +0100 tools/libs/guest: fix max_pfn setting in map_p2m() When setting the highest pfn used in the guest, don't subtract 1 from the value read from the shared_info data. The value read already is the correct pfn. Fixes: 91e204d37f449 ("libxc: try to find last used pfn when migrating") Signed-off-by: Juergen Gross Acked-by: Wei Liu --- tools/libs/guest/xg_sr_save_x86_pv.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/tools/libs/guest/xg_sr_save_x86_pv.c b/tools/libs/guest/xg_sr_save_x86_pv.c index 4964f1f7b8..dae7f2817f 100644 --- a/tools/libs/guest/xg_sr_save_x86_pv.c +++ b/tools/libs/guest/xg_sr_save_x86_pv.c @@ -468,7 +468,7 @@ static int map_p2m(struct xc_sr_context *ctx) ctx->x86.pv.p2m_generation = ~0ULL; ctx->x86.pv.max_pfn = GET_FIELD(ctx->x86.pv.shinfo, arch.max_pfn, - ctx->x86.pv.width) - 1; + ctx->x86.pv.width); p2m_cr3 = GET_FIELD(ctx->x86.pv.shinfo, arch.p2m_cr3, ctx->x86.pv.width); return p2m_cr3 ? map_p2m_list(ctx, p2m_cr3) : map_p2m_tree(ctx); -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:22:23 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:22:23 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146576.269693 (Exim 4.92) (envelope-from ) id 1lwMV5-0004Ak-Jk; Thu, 24 Jun 2021 10:22:23 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146576.269693; Thu, 24 Jun 2021 10:22:23 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMV5-0004Aa-GH; Thu, 24 Jun 2021 10:22:23 +0000 Received: by outflank-mailman (input) for mailman id 146576; Thu, 24 Jun 2021 10:22:22 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMV4-0004AR-SM for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:22:22 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMV4-0005vh-RV for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:22:22 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwMV4-0006BT-Qb for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:22:22 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=V72XOpdnNyNpxWTsHrNAkRvuMmX3o6s8VCf7SsN9bO8=; b=oe3RDj0hWyJyJITzhhEQmOBEv1 OtOrrLbMC3MzQpB2d0Vu/xBtbm1h36nklO/OjaUWouukWrSDDd9v1xdEyGLN5Nsg3yBExKd3c+Sd5 M4PXVoKYgGtBQFUWYKXLoZI1JQo2yStCMDaAIeg5hEH/3rq8d/i0cv5w5rqGXcb3rVAg=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] tools/libs/ctrl: fix xc_core_arch_map_p2m() to support linear p2m table Message-Id: Date: Thu, 24 Jun 2021 10:22:22 +0000 commit bd7a29c3d0b937ab542abea06ff1b575abe7247a Author: Juergen Gross AuthorDate: Fri Jun 4 08:02:10 2021 +0200 Commit: Julien Grall CommitDate: Fri Jun 4 18:56:46 2021 +0100 tools/libs/ctrl: fix xc_core_arch_map_p2m() to support linear p2m table The core of a pv linux guest produced via "xl dump-core" is nor usable as since kernel 4.14 only the linear p2m table is kept if Xen indicates it is supporting that. Unfortunately xc_core_arch_map_p2m() is still supporting the 3-level p2m tree only. Fix that by copying the functionality of map_p2m() from libxenguest to libxenctrl. Additionally the mapped p2m isn't of a fixed length now, so the interface to the mapping functions needs to be adapted. In order not to add even more parameters, expand struct domain_info_context and use a pointer to that as a parameter. Fixes: dc6d60937121 ("libxc: set flag for support of linear p2m list in domain builder") Signed-off-by: Juergen Gross Acked-by: Wei Liu --- tools/include/xenguest.h | 1 + tools/libs/ctrl/xc_core.c | 5 +- tools/libs/ctrl/xc_core.h | 8 +- tools/libs/ctrl/xc_core_arm.c | 23 ++-- tools/libs/ctrl/xc_core_x86.c | 256 ++++++++++++++++++++++++++++++++++-------- tools/libs/ctrl/xc_private.h | 1 + tools/libs/guest/xg_domain.c | 17 +-- 7 files changed, 233 insertions(+), 78 deletions(-) diff --git a/tools/include/xenguest.h b/tools/include/xenguest.h index a4492038cf..f9fb0449ad 100644 --- a/tools/include/xenguest.h +++ b/tools/include/xenguest.h @@ -700,6 +700,7 @@ struct xc_domain_meminfo { xen_pfn_t *pfn_type; xen_pfn_t *p2m_table; unsigned long p2m_size; + unsigned int p2m_frames; }; int xc_map_domain_meminfo(xc_interface *xch, uint32_t domid, diff --git a/tools/libs/ctrl/xc_core.c b/tools/libs/ctrl/xc_core.c index b47ab2f6d8..9576bec5a3 100644 --- a/tools/libs/ctrl/xc_core.c +++ b/tools/libs/ctrl/xc_core.c @@ -574,8 +574,7 @@ xc_domain_dumpcore_via_callback(xc_interface *xch, goto out; } - sts = xc_core_arch_map_p2m(xch, dinfo->guest_width, &info, live_shinfo, - &p2m, &dinfo->p2m_size); + sts = xc_core_arch_map_p2m(xch, dinfo, &info, live_shinfo, &p2m); if ( sts != 0 ) goto out; @@ -945,7 +944,7 @@ out: if ( memory_map != NULL ) free(memory_map); if ( p2m != NULL ) - munmap(p2m, PAGE_SIZE * P2M_FL_ENTRIES); + munmap(p2m, PAGE_SIZE * dinfo->p2m_frames); if ( p2m_array != NULL ) free(p2m_array); if ( pfn_array != NULL ) diff --git a/tools/libs/ctrl/xc_core.h b/tools/libs/ctrl/xc_core.h index 36fb755da2..8ea1f93a10 100644 --- a/tools/libs/ctrl/xc_core.h +++ b/tools/libs/ctrl/xc_core.h @@ -138,14 +138,14 @@ int xc_core_arch_memory_map_get(xc_interface *xch, xc_dominfo_t *info, shared_info_any_t *live_shinfo, xc_core_memory_map_t **mapp, unsigned int *nr_entries); -int xc_core_arch_map_p2m(xc_interface *xch, unsigned int guest_width, +int xc_core_arch_map_p2m(xc_interface *xch, struct domain_info_context *dinfo, xc_dominfo_t *info, shared_info_any_t *live_shinfo, - xen_pfn_t **live_p2m, unsigned long *pfnp); + xen_pfn_t **live_p2m); -int xc_core_arch_map_p2m_writable(xc_interface *xch, unsigned int guest_width, +int xc_core_arch_map_p2m_writable(xc_interface *xch, struct domain_info_context *dinfo, xc_dominfo_t *info, shared_info_any_t *live_shinfo, - xen_pfn_t **live_p2m, unsigned long *pfnp); + xen_pfn_t **live_p2m); int xc_core_arch_get_scratch_gpfn(xc_interface *xch, uint32_t domid, xen_pfn_t *gpfn); diff --git a/tools/libs/ctrl/xc_core_arm.c b/tools/libs/ctrl/xc_core_arm.c index 7b587b4cc5..93765a565f 100644 --- a/tools/libs/ctrl/xc_core_arm.c +++ b/tools/libs/ctrl/xc_core_arm.c @@ -66,33 +66,24 @@ xc_core_arch_memory_map_get(xc_interface *xch, struct xc_core_arch_context *unus static int xc_core_arch_map_p2m_rw(xc_interface *xch, struct domain_info_context *dinfo, xc_dominfo_t *info, - shared_info_any_t *live_shinfo, xen_pfn_t **live_p2m, - unsigned long *pfnp, int rw) + shared_info_any_t *live_shinfo, xen_pfn_t **live_p2m, int rw) { errno = ENOSYS; return -1; } int -xc_core_arch_map_p2m(xc_interface *xch, unsigned int guest_width, xc_dominfo_t *info, - shared_info_any_t *live_shinfo, xen_pfn_t **live_p2m, - unsigned long *pfnp) +xc_core_arch_map_p2m(xc_interface *xch, struct domain_info_context *dinfo, xc_dominfo_t *info, + shared_info_any_t *live_shinfo, xen_pfn_t **live_p2m) { - struct domain_info_context _dinfo = { .guest_width = guest_width }; - struct domain_info_context *dinfo = &_dinfo; - return xc_core_arch_map_p2m_rw(xch, dinfo, info, - live_shinfo, live_p2m, pfnp, 0); + return xc_core_arch_map_p2m_rw(xch, dinfo, info, live_shinfo, live_p2m, 0); } int -xc_core_arch_map_p2m_writable(xc_interface *xch, unsigned int guest_width, xc_dominfo_t *info, - shared_info_any_t *live_shinfo, xen_pfn_t **live_p2m, - unsigned long *pfnp) +xc_core_arch_map_p2m_writable(xc_interface *xch, struct domain_info_context *dinfo, xc_dominfo_t *info, + shared_info_any_t *live_shinfo, xen_pfn_t **live_p2m) { - struct domain_info_context _dinfo = { .guest_width = guest_width }; - struct domain_info_context *dinfo = &_dinfo; - return xc_core_arch_map_p2m_rw(xch, dinfo, info, - live_shinfo, live_p2m, pfnp, 1); + return xc_core_arch_map_p2m_rw(xch, dinfo, info, live_shinfo, live_p2m, 1); } int diff --git a/tools/libs/ctrl/xc_core_x86.c b/tools/libs/ctrl/xc_core_x86.c index cb76e6207b..c8f71d4b75 100644 --- a/tools/libs/ctrl/xc_core_x86.c +++ b/tools/libs/ctrl/xc_core_x86.c @@ -17,6 +17,7 @@ * */ +#include #include "xc_private.h" #include "xc_core.h" #include @@ -65,34 +66,169 @@ xc_core_arch_memory_map_get(xc_interface *xch, struct xc_core_arch_context *unus return 0; } -static int -xc_core_arch_map_p2m_rw(xc_interface *xch, struct domain_info_context *dinfo, xc_dominfo_t *info, - shared_info_any_t *live_shinfo, xen_pfn_t **live_p2m, - unsigned long *pfnp, int rw) +static inline bool is_canonical_address(uint64_t vaddr) { - /* Double and single indirect references to the live P2M table */ - xen_pfn_t *live_p2m_frame_list_list = NULL; - xen_pfn_t *live_p2m_frame_list = NULL; - /* Copies of the above. */ - xen_pfn_t *p2m_frame_list_list = NULL; - xen_pfn_t *p2m_frame_list = NULL; + return ((int64_t)vaddr >> 47) == ((int64_t)vaddr >> 63); +} - uint32_t dom = info->domid; - int ret = -1; - int err; - int i; +/* Virtual address ranges reserved for hypervisor. */ +#define HYPERVISOR_VIRT_START_X86_64 0xFFFF800000000000ULL +#define HYPERVISOR_VIRT_END_X86_64 0xFFFF87FFFFFFFFFFULL - if ( xc_domain_nr_gpfns(xch, info->domid, &dinfo->p2m_size) < 0 ) +#define HYPERVISOR_VIRT_START_X86_32 0x00000000F5800000ULL +#define HYPERVISOR_VIRT_END_X86_32 0x00000000FFFFFFFFULL + +static xen_pfn_t * +xc_core_arch_map_p2m_list_rw(xc_interface *xch, struct domain_info_context *dinfo, + uint32_t dom, shared_info_any_t *live_shinfo, + uint64_t p2m_cr3) +{ + uint64_t p2m_vaddr, p2m_end, mask, off; + xen_pfn_t p2m_mfn, mfn, saved_mfn, max_pfn; + uint64_t *ptes = NULL; + xen_pfn_t *mfns = NULL; + unsigned int fpp, n_pages, level, n_levels, shift, + idx_start, idx_end, idx, saved_idx; + + p2m_vaddr = GET_FIELD(live_shinfo, arch.p2m_vaddr, dinfo->guest_width); + fpp = PAGE_SIZE / dinfo->guest_width; + dinfo->p2m_frames = (dinfo->p2m_size - 1) / fpp + 1; + p2m_end = p2m_vaddr + dinfo->p2m_frames * PAGE_SIZE - 1; + + if ( dinfo->guest_width == 8 ) { - ERROR("Could not get maximum GPFN!"); - goto out; + mask = 0x0000ffffffffffffULL; + n_levels = 4; + p2m_mfn = p2m_cr3 >> 12; + if ( !is_canonical_address(p2m_vaddr) || + !is_canonical_address(p2m_end) || + p2m_end < p2m_vaddr || + (p2m_vaddr <= HYPERVISOR_VIRT_END_X86_64 && + p2m_end > HYPERVISOR_VIRT_START_X86_64) ) + { + ERROR("Bad virtual p2m address range %#" PRIx64 "-%#" PRIx64, + p2m_vaddr, p2m_end); + errno = ERANGE; + goto out; + } + } + else + { + mask = 0x00000000ffffffffULL; + n_levels = 3; + if ( p2m_cr3 & ~mask ) + p2m_mfn = ~0UL; + else + p2m_mfn = (uint32_t)((p2m_cr3 >> 12) | (p2m_cr3 << 20)); + if ( p2m_vaddr > mask || p2m_end > mask || p2m_end < p2m_vaddr || + (p2m_vaddr <= HYPERVISOR_VIRT_END_X86_32 && + p2m_end > HYPERVISOR_VIRT_START_X86_32) ) + { + ERROR("Bad virtual p2m address range %#" PRIx64 "-%#" PRIx64, + p2m_vaddr, p2m_end); + errno = ERANGE; + goto out; + } } - if ( dinfo->p2m_size < info->nr_pages ) + mfns = malloc(sizeof(*mfns)); + if ( !mfns ) { - ERROR("p2m_size < nr_pages -1 (%lx < %lx", dinfo->p2m_size, info->nr_pages - 1); + ERROR("Cannot allocate memory for array of %u mfns", 1); goto out; } + mfns[0] = p2m_mfn; + off = 0; + saved_mfn = 0; + idx_start = idx_end = saved_idx = 0; + + for ( level = n_levels; level > 0; level-- ) + { + n_pages = idx_end - idx_start + 1; + ptes = xc_map_foreign_pages(xch, dom, PROT_READ, mfns, n_pages); + if ( !ptes ) + { + PERROR("Failed to map %u page table pages for p2m list", n_pages); + goto out; + } + free(mfns); + + shift = level * 9 + 3; + idx_start = ((p2m_vaddr - off) & mask) >> shift; + idx_end = ((p2m_end - off) & mask) >> shift; + idx = idx_end - idx_start + 1; + mfns = malloc(sizeof(*mfns) * idx); + if ( !mfns ) + { + ERROR("Cannot allocate memory for array of %u mfns", idx); + goto out; + } + + for ( idx = idx_start; idx <= idx_end; idx++ ) + { + mfn = (ptes[idx] & 0x000ffffffffff000ULL) >> PAGE_SHIFT; + if ( mfn == 0 ) + { + ERROR("Bad mfn %#lx during page table walk for vaddr %#" PRIx64 " at level %d of p2m list", + mfn, off + ((uint64_t)idx << shift), level); + errno = ERANGE; + goto out; + } + mfns[idx - idx_start] = mfn; + + /* Maximum pfn check at level 2. Same reasoning as for p2m tree. */ + if ( level == 2 ) + { + if ( mfn != saved_mfn ) + { + saved_mfn = mfn; + saved_idx = idx - idx_start; + } + } + } + + if ( level == 2 ) + { + if ( saved_idx == idx_end ) + saved_idx++; + max_pfn = ((xen_pfn_t)saved_idx << 9) * fpp; + if ( max_pfn < dinfo->p2m_size ) + { + dinfo->p2m_size = max_pfn; + dinfo->p2m_frames = (dinfo->p2m_size + fpp - 1) / fpp; + p2m_end = p2m_vaddr + dinfo->p2m_frames * PAGE_SIZE - 1; + idx_end = idx_start + saved_idx; + } + } + + munmap(ptes, n_pages * PAGE_SIZE); + ptes = NULL; + off = p2m_vaddr & ((mask >> shift) << shift); + } + + return mfns; + + out: + free(mfns); + if ( ptes ) + munmap(ptes, n_pages * PAGE_SIZE); + + return NULL; +} + +static xen_pfn_t * +xc_core_arch_map_p2m_tree_rw(xc_interface *xch, struct domain_info_context *dinfo, + uint32_t dom, shared_info_any_t *live_shinfo) +{ + /* Double and single indirect references to the live P2M table */ + xen_pfn_t *live_p2m_frame_list_list; + xen_pfn_t *live_p2m_frame_list = NULL; + /* Copies of the above. */ + xen_pfn_t *p2m_frame_list_list = NULL; + xen_pfn_t *p2m_frame_list; + + int err; + int i; live_p2m_frame_list_list = xc_map_foreign_range(xch, dom, PAGE_SIZE, PROT_READ, @@ -151,10 +287,60 @@ xc_core_arch_map_p2m_rw(xc_interface *xch, struct domain_info_context *dinfo, xc for ( i = P2M_FL_ENTRIES - 1; i >= 0; i-- ) p2m_frame_list[i] = ((uint32_t *)p2m_frame_list)[i]; + dinfo->p2m_frames = P2M_FL_ENTRIES; + + return p2m_frame_list; + + out: + err = errno; + + if ( live_p2m_frame_list_list ) + munmap(live_p2m_frame_list_list, PAGE_SIZE); + + if ( live_p2m_frame_list ) + munmap(live_p2m_frame_list, P2M_FLL_ENTRIES * PAGE_SIZE); + + free(p2m_frame_list_list); + + errno = err; + + return NULL; +} + +static int +xc_core_arch_map_p2m_rw(xc_interface *xch, struct domain_info_context *dinfo, xc_dominfo_t *info, + shared_info_any_t *live_shinfo, xen_pfn_t **live_p2m, int rw) +{ + xen_pfn_t *p2m_frame_list = NULL; + uint64_t p2m_cr3; + uint32_t dom = info->domid; + int ret = -1; + int err; + + if ( xc_domain_nr_gpfns(xch, info->domid, &dinfo->p2m_size) < 0 ) + { + ERROR("Could not get maximum GPFN!"); + goto out; + } + + if ( dinfo->p2m_size < info->nr_pages ) + { + ERROR("p2m_size < nr_pages -1 (%lx < %lx", dinfo->p2m_size, info->nr_pages - 1); + goto out; + } + + p2m_cr3 = GET_FIELD(live_shinfo, arch.p2m_cr3, dinfo->guest_width); + + p2m_frame_list = p2m_cr3 ? xc_core_arch_map_p2m_list_rw(xch, dinfo, dom, live_shinfo, p2m_cr3) + : xc_core_arch_map_p2m_tree_rw(xch, dinfo, dom, live_shinfo); + + if ( !p2m_frame_list ) + goto out; + *live_p2m = xc_map_foreign_pages(xch, dom, rw ? (PROT_READ | PROT_WRITE) : PROT_READ, p2m_frame_list, - P2M_FL_ENTRIES); + dinfo->p2m_frames); if ( !*live_p2m ) { @@ -162,21 +348,11 @@ xc_core_arch_map_p2m_rw(xc_interface *xch, struct domain_info_context *dinfo, xc goto out; } - *pfnp = dinfo->p2m_size; - ret = 0; out: err = errno; - if ( live_p2m_frame_list_list ) - munmap(live_p2m_frame_list_list, PAGE_SIZE); - - if ( live_p2m_frame_list ) - munmap(live_p2m_frame_list, P2M_FLL_ENTRIES * PAGE_SIZE); - - free(p2m_frame_list_list); - free(p2m_frame_list); errno = err; @@ -184,25 +360,17 @@ out: } int -xc_core_arch_map_p2m(xc_interface *xch, unsigned int guest_width, xc_dominfo_t *info, - shared_info_any_t *live_shinfo, xen_pfn_t **live_p2m, - unsigned long *pfnp) +xc_core_arch_map_p2m(xc_interface *xch, struct domain_info_context *dinfo, xc_dominfo_t *info, + shared_info_any_t *live_shinfo, xen_pfn_t **live_p2m) { - struct domain_info_context _dinfo = { .guest_width = guest_width }; - struct domain_info_context *dinfo = &_dinfo; - return xc_core_arch_map_p2m_rw(xch, dinfo, info, - live_shinfo, live_p2m, pfnp, 0); + return xc_core_arch_map_p2m_rw(xch, dinfo, info, live_shinfo, live_p2m, 0); } int -xc_core_arch_map_p2m_writable(xc_interface *xch, unsigned int guest_width, xc_dominfo_t *info, - shared_info_any_t *live_shinfo, xen_pfn_t **live_p2m, - unsigned long *pfnp) +xc_core_arch_map_p2m_writable(xc_interface *xch, struct domain_info_context *dinfo, xc_dominfo_t *info, + shared_info_any_t *live_shinfo, xen_pfn_t **live_p2m) { - struct domain_info_context _dinfo = { .guest_width = guest_width }; - struct domain_info_context *dinfo = &_dinfo; - return xc_core_arch_map_p2m_rw(xch, dinfo, info, - live_shinfo, live_p2m, pfnp, 1); + return xc_core_arch_map_p2m_rw(xch, dinfo, info, live_shinfo, live_p2m, 1); } int diff --git a/tools/libs/ctrl/xc_private.h b/tools/libs/ctrl/xc_private.h index f0b5f83ac8..8ebc0b59da 100644 --- a/tools/libs/ctrl/xc_private.h +++ b/tools/libs/ctrl/xc_private.h @@ -79,6 +79,7 @@ struct iovec { struct domain_info_context { unsigned int guest_width; + unsigned int p2m_frames; unsigned long p2m_size; }; diff --git a/tools/libs/guest/xg_domain.c b/tools/libs/guest/xg_domain.c index 5019c84e0e..dd7db2cbd8 100644 --- a/tools/libs/guest/xg_domain.c +++ b/tools/libs/guest/xg_domain.c @@ -24,13 +24,9 @@ int xc_unmap_domain_meminfo(xc_interface *xch, struct xc_domain_meminfo *minfo) { - struct domain_info_context _di = { .guest_width = minfo->guest_width, - .p2m_size = minfo->p2m_size}; - struct domain_info_context *dinfo = &_di; - free(minfo->pfn_type); if ( minfo->p2m_table ) - munmap(minfo->p2m_table, P2M_FL_ENTRIES * PAGE_SIZE); + munmap(minfo->p2m_table, minfo->p2m_frames * PAGE_SIZE); minfo->p2m_table = NULL; return 0; @@ -40,7 +36,6 @@ int xc_map_domain_meminfo(xc_interface *xch, uint32_t domid, struct xc_domain_meminfo *minfo) { struct domain_info_context _di; - struct domain_info_context *dinfo = &_di; xc_dominfo_t info; shared_info_any_t *live_shinfo; @@ -96,16 +91,16 @@ int xc_map_domain_meminfo(xc_interface *xch, uint32_t domid, return -1; } - if ( xc_core_arch_map_p2m_writable(xch, minfo->guest_width, &info, - live_shinfo, &minfo->p2m_table, - &minfo->p2m_size) ) + if ( xc_core_arch_map_p2m_writable(xch, &_di, &info, + live_shinfo, &minfo->p2m_table) ) { PERROR("Could not map the P2M table"); munmap(live_shinfo, PAGE_SIZE); return -1; } munmap(live_shinfo, PAGE_SIZE); - _di.p2m_size = minfo->p2m_size; + minfo->p2m_size = _di.p2m_size; + minfo->p2m_frames = _di.p2m_frames; /* Make space and prepare for getting the PFN types */ minfo->pfn_type = calloc(sizeof(*minfo->pfn_type), minfo->p2m_size); @@ -141,7 +136,7 @@ failed: } if ( minfo->p2m_table ) { - munmap(minfo->p2m_table, P2M_FL_ENTRIES * PAGE_SIZE); + munmap(minfo->p2m_table, minfo->p2m_frames * PAGE_SIZE); minfo->p2m_table = NULL; } -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:22:33 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:22:33 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146577.269696 (Exim 4.92) (envelope-from ) id 1lwMVF-0004FS-Kk; Thu, 24 Jun 2021 10:22:33 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146577.269696; Thu, 24 Jun 2021 10:22:33 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMVF-0004FK-Hs; Thu, 24 Jun 2021 10:22:33 +0000 Received: by outflank-mailman (input) for mailman id 146577; Thu, 24 Jun 2021 10:22:33 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMVE-0004FC-VF for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:22:32 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMVE-0005vv-UQ for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:22:32 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwMVE-0006Ca-Tk for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:22:32 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=vW0pFW/Y77YrA44bZ5hGB+bLbghWG6AjEF866dh9kNs=; b=ciXblquxjxhrAsIOF0K3tnoMAb QfLbtK64FWlYHRei9nb55y+VkBm5iPfaPsPJ/I5xWCc9zHHIUY3+dJrG3Y62nqOwPRv6tyHZ//XcW VgeYwdW/ZRTBT+FpXet2H0GStsKz0XysHYZGNsbDh5lavg+VUgSpk+tL8fsRl963m9g4=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] tools/libs/ctrl: use common p2m mapping code in xc_domain_resume_any() Message-Id: Date: Thu, 24 Jun 2021 10:22:32 +0000 commit f183854facad996fe891c086c024bca7cbcdc1e4 Author: Juergen Gross AuthorDate: Fri Jun 4 08:02:11 2021 +0200 Commit: Julien Grall CommitDate: Fri Jun 4 18:56:46 2021 +0100 tools/libs/ctrl: use common p2m mapping code in xc_domain_resume_any() Instead of open coding the mapping of the p2m list use the already existing xc_core_arch_map_p2m() call, especially as the current code does not support guests with the linear p2m map. It should be noted that this code is needed for colo/remus only. Switching to xc_core_arch_map_p2m() drops the need to bail out for bitness of tool stack and guest differing. Signed-off-by: Juergen Gross Acked-by: Christian Lindig Acked-by: Wei Liu --- tools/libs/ctrl/xc_resume.c | 66 ++++++++++++------------------------- tools/ocaml/libs/xc/xenctrl_stubs.c | 1 + 2 files changed, 22 insertions(+), 45 deletions(-) diff --git a/tools/libs/ctrl/xc_resume.c b/tools/libs/ctrl/xc_resume.c index 94c6c9fb31..e3c8e83aa9 100644 --- a/tools/libs/ctrl/xc_resume.c +++ b/tools/libs/ctrl/xc_resume.c @@ -20,6 +20,7 @@ #include #include #include +#include "xc_core.h" static int modify_returncode(xc_interface *xch, uint32_t domid) { @@ -137,12 +138,10 @@ static int xc_domain_resume_any(xc_interface *xch, uint32_t domid) struct domain_info_context _dinfo = { .guest_width = 0, .p2m_size = 0 }; struct domain_info_context *dinfo = &_dinfo; - unsigned long mfn; + xen_pfn_t mfn, store_mfn, console_mfn; vcpu_guest_context_any_t ctxt; - start_info_t *start_info; - shared_info_t *shinfo = NULL; - xen_pfn_t *p2m_frame_list_list = NULL; - xen_pfn_t *p2m_frame_list = NULL; + start_info_any_t *start_info; + shared_info_any_t *shinfo = NULL; xen_pfn_t *p2m = NULL; #endif @@ -164,11 +163,6 @@ static int xc_domain_resume_any(xc_interface *xch, uint32_t domid) PERROR("Could not get domain width"); return rc; } - if ( dinfo->guest_width != sizeof(long) ) - { - ERROR("Cannot resume uncooperative cross-address-size guests"); - return rc; - } /* Map the shared info frame */ shinfo = xc_map_foreign_range(xch, domid, PAGE_SIZE, @@ -179,34 +173,8 @@ static int xc_domain_resume_any(xc_interface *xch, uint32_t domid) goto out; } - dinfo->p2m_size = shinfo->arch.max_pfn; - - p2m_frame_list_list = - xc_map_foreign_range(xch, domid, PAGE_SIZE, PROT_READ, - shinfo->arch.pfn_to_mfn_frame_list_list); - if ( p2m_frame_list_list == NULL ) - { - ERROR("Couldn't map p2m_frame_list_list"); - goto out; - } - - p2m_frame_list = xc_map_foreign_pages(xch, domid, PROT_READ, - p2m_frame_list_list, - P2M_FLL_ENTRIES); - if ( p2m_frame_list == NULL ) - { - ERROR("Couldn't map p2m_frame_list"); - goto out; - } - - /* Map all the frames of the pfn->mfn table. For migrate to succeed, - the guest must not change which frames are used for this purpose. - (its not clear why it would want to change them, and we'll be OK - from a safety POV anyhow. */ - p2m = xc_map_foreign_pages(xch, domid, PROT_READ, - p2m_frame_list, - P2M_FL_ENTRIES); - if ( p2m == NULL ) + /* Map the p2m list */ + if ( xc_core_arch_map_p2m(xch, dinfo, &info, shinfo, &p2m) ) { ERROR("Couldn't map p2m table"); goto out; @@ -228,8 +196,20 @@ static int xc_domain_resume_any(xc_interface *xch, uint32_t domid) goto out; } - start_info->store_mfn = p2m[start_info->store_mfn]; - start_info->console.domU.mfn = p2m[start_info->console.domU.mfn]; + store_mfn = GET_FIELD(start_info, store_mfn, dinfo->guest_width); + console_mfn = GET_FIELD(start_info, console.domU.mfn, dinfo->guest_width); + if ( dinfo->guest_width == 4 ) + { + store_mfn = ((uint32_t *)p2m)[store_mfn]; + console_mfn = ((uint32_t *)p2m)[console_mfn]; + } + else + { + store_mfn = ((uint64_t *)p2m)[store_mfn]; + console_mfn = ((uint64_t *)p2m)[console_mfn]; + } + SET_FIELD(start_info, store_mfn, store_mfn, dinfo->guest_width); + SET_FIELD(start_info, console.domU.mfn, console_mfn, dinfo->guest_width); munmap(start_info, PAGE_SIZE); #endif /* defined(__i386__) || defined(__x86_64__) */ @@ -250,11 +230,7 @@ static int xc_domain_resume_any(xc_interface *xch, uint32_t domid) out: #if defined(__i386__) || defined(__x86_64__) if (p2m) - munmap(p2m, P2M_FL_ENTRIES*PAGE_SIZE); - if (p2m_frame_list) - munmap(p2m_frame_list, P2M_FLL_ENTRIES*PAGE_SIZE); - if (p2m_frame_list_list) - munmap(p2m_frame_list_list, PAGE_SIZE); + munmap(p2m, dinfo->p2m_frames * PAGE_SIZE); if (shinfo) munmap(shinfo, PAGE_SIZE); #endif diff --git a/tools/ocaml/libs/xc/xenctrl_stubs.c b/tools/ocaml/libs/xc/xenctrl_stubs.c index d05d7bb30e..6e4bc567f5 100644 --- a/tools/ocaml/libs/xc/xenctrl_stubs.c +++ b/tools/ocaml/libs/xc/xenctrl_stubs.c @@ -32,6 +32,7 @@ #define XC_WANT_COMPAT_MAP_FOREIGN_API #include +#include #include #include "mmap_stubs.h" -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:22:43 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:22:43 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146578.269700 (Exim 4.92) (envelope-from ) id 1lwMVP-0004Ir-Ng; Thu, 24 Jun 2021 10:22:43 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146578.269700; Thu, 24 Jun 2021 10:22:43 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMVP-0004Ij-Ki; Thu, 24 Jun 2021 10:22:43 +0000 Received: by outflank-mailman (input) for mailman id 146578; Thu, 24 Jun 2021 10:22:43 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMVP-0004IX-2Z for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:22:43 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMVP-0005wB-1k for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:22:43 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwMVP-0006Dl-0o for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:22:43 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=8Fxrb4fNTXAdRwQbviICN+gjz7veWPoXGUguy7i9a7I=; b=aI+7PFAZX9J4e+2cCH1gpPXhSr +lTDJEsP+Ig+A1j1SiTt6QsCpNjYsgmCfqDEsqVFbhXyww2bSw6aeTuYK+4hT2i9EELdlPbTb26/z uLr8YAVIu6Rtt8dYk6FKaZ19wI7OcWsOseEC4hXB4Qm+3V87fcmiQJru0MvB6oMEKzeI=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] tools/libs: move xc_resume.c to libxenguest Message-Id: Date: Thu, 24 Jun 2021 10:22:43 +0000 commit bf1fc18901dfea05a69f661493b934c0db7d3503 Author: Juergen Gross AuthorDate: Fri Jun 4 08:02:12 2021 +0200 Commit: Julien Grall CommitDate: Fri Jun 4 18:56:46 2021 +0100 tools/libs: move xc_resume.c to libxenguest The guest suspend functionality is already part of libxenguest. Move the resume functionality from libxenctrl to libxenguest, too. Signed-off-by: Juergen Gross Acked-by: Wei Liu --- tools/include/xenctrl.h | 63 ---------- tools/include/xenguest.h | 62 ++++++++++ tools/libs/ctrl/Makefile | 1 - tools/libs/ctrl/xc_resume.c | 264 ------------------------------------------ tools/libs/guest/Makefile | 1 + tools/libs/guest/xg_resume.c | 265 +++++++++++++++++++++++++++++++++++++++++++ 6 files changed, 328 insertions(+), 328 deletions(-) diff --git a/tools/include/xenctrl.h b/tools/include/xenctrl.h index 58d3377d6a..2a7c836a02 100644 --- a/tools/include/xenctrl.h +++ b/tools/include/xenctrl.h @@ -576,69 +576,6 @@ int xc_domain_destroy(xc_interface *xch, uint32_t domid); -/** - * This function resumes a suspended domain. The domain should have - * been previously suspended. - * - * Note that there are 'xc_domain_suspend' as suspending a domain - * is quite the endeavour. - * - * For the purpose of this explanation there are three guests: - * PV (using hypercalls for privilgied operations), HVM - * (fully hardware virtualized guests using emulated devices for everything), - * and PVHVM (PV aware with hardware virtualisation). - * - * HVM guest are the simplest - they suspend via S3 / S4 and resume from - * S3 / S4. Upon resume they have to re-negotiate with the emulated devices. - * - * PV and PVHVM communicate via hypercalls for suspend (and resume). - * For suspend the toolstack initiates the process by writing an value - * in XenBus "control/shutdown" with the string "suspend". - * - * The PV guest stashes anything it deems neccessary in 'struct - * start_info' in case of failure (PVHVM may ignore this) and calls - * the SCHEDOP_shutdown::SHUTDOWN_suspend hypercall (for PV as - * argument it passes the MFN to 'struct start_info'). - * - * And then the guest is suspended. - * - * The checkpointing or notifying a guest that the suspend failed or - * cancelled (in case of checkpoint) is by having the - * SCHEDOP_shutdown::SHUTDOWN_suspend hypercall return a non-zero - * value. - * - * The PV and PVHVM resume path are similar. For PV it would be - * similar to bootup - figure out where the 'struct start_info' is (or - * if the suspend was cancelled aka checkpointed - reuse the saved - * values). - * - * From here on they differ depending whether the guest is PV or PVHVM - * in specifics but follow overall the same path: - * - PV: Bringing up the vCPUS, - * - PVHVM: Setup vector callback, - * - Bring up vCPU runstates, - * - Remap the grant tables if checkpointing or setup from scratch, - * - * - * If the resume was not checkpointing (or if suspend was succesful) we would - * setup the PV timers and the different PV events. Lastly the PV drivers - * re-negotiate with the backend. - * - * This function would return before the guest started resuming. That is - * the guest would be in non-running state and its vCPU context would be - * in the the SCHEDOP_shutdown::SHUTDOWN_suspend hypercall return path - * (for PV and PVHVM). For HVM it would be in would be in QEMU emulated - * BIOS handling S3 suspend. - * - * @parm xch a handle to an open hypervisor interface - * @parm domid the domain id to resume - * @parm fast use cooperative resume (guest must support this) - * return 0 on success, -1 on failure - */ -int xc_domain_resume(xc_interface *xch, - uint32_t domid, - int fast); - /** * This function will shutdown a domain. This is intended for use in * fully-virtualized domains where this operation is analogous to the diff --git a/tools/include/xenguest.h b/tools/include/xenguest.h index f9fb0449ad..61d0a82f48 100644 --- a/tools/include/xenguest.h +++ b/tools/include/xenguest.h @@ -689,6 +689,68 @@ int xc_query_page_offline_status(xc_interface *xch, unsigned long start, int xc_exchange_page(xc_interface *xch, uint32_t domid, xen_pfn_t mfn); +/** + * This function resumes a suspended domain. The domain should have + * been previously suspended. + * + * Note that there are 'xc_domain_suspend' as suspending a domain + * is quite the endeavour. + * + * For the purpose of this explanation there are three guests: + * PV (using hypercalls for privilgied operations), HVM + * (fully hardware virtualized guests using emulated devices for everything), + * and PVHVM (PV aware with hardware virtualisation). + * + * HVM guest are the simplest - they suspend via S3 / S4 and resume from + * S3 / S4. Upon resume they have to re-negotiate with the emulated devices. + * + * PV and PVHVM communicate via hypercalls for suspend (and resume). + * For suspend the toolstack initiates the process by writing an value + * in XenBus "control/shutdown" with the string "suspend". + * + * The PV guest stashes anything it deems neccessary in 'struct + * start_info' in case of failure (PVHVM may ignore this) and calls + * the SCHEDOP_shutdown::SHUTDOWN_suspend hypercall (for PV as + * argument it passes the MFN to 'struct start_info'). + * + * And then the guest is suspended. + * + * The checkpointing or notifying a guest that the suspend failed or + * cancelled (in case of checkpoint) is by having the + * SCHEDOP_shutdown::SHUTDOWN_suspend hypercall return a non-zero + * value. + * + * The PV and PVHVM resume path are similar. For PV it would be + * similar to bootup - figure out where the 'struct start_info' is (or + * if the suspend was cancelled aka checkpointed - reuse the saved + * values). + * + * From here on they differ depending whether the guest is PV or PVHVM + * in specifics but follow overall the same path: + * - PV: Bringing up the vCPUS, + * - PVHVM: Setup vector callback, + * - Bring up vCPU runstates, + * - Remap the grant tables if checkpointing or setup from scratch, + * + * + * If the resume was not checkpointing (or if suspend was succesful) we would + * setup the PV timers and the different PV events. Lastly the PV drivers + * re-negotiate with the backend. + * + * This function would return before the guest started resuming. That is + * the guest would be in non-running state and its vCPU context would be + * in the the SCHEDOP_shutdown::SHUTDOWN_suspend hypercall return path + * (for PV and PVHVM). For HVM it would be in would be in QEMU emulated + * BIOS handling S3 suspend. + * + * @parm xch a handle to an open hypervisor interface + * @parm domid the domain id to resume + * @parm fast use cooperative resume (guest must support this) + * return 0 on success, -1 on failure + */ +int xc_domain_resume(xc_interface *xch, + uint32_t domid, + int fast); /** * Memory related information, such as PFN types, the P2M table, diff --git a/tools/libs/ctrl/Makefile b/tools/libs/ctrl/Makefile index ce9ecae710..fbeb3a3537 100644 --- a/tools/libs/ctrl/Makefile +++ b/tools/libs/ctrl/Makefile @@ -20,7 +20,6 @@ SRCS-y += xc_rt.c SRCS-y += xc_tbuf.c SRCS-y += xc_pm.c SRCS-y += xc_cpu_hotplug.c -SRCS-y += xc_resume.c SRCS-y += xc_vm_event.c SRCS-y += xc_vmtrace.c SRCS-y += xc_monitor.c diff --git a/tools/libs/ctrl/xc_resume.c b/tools/libs/ctrl/xc_resume.c deleted file mode 100644 index e3c8e83aa9..0000000000 --- a/tools/libs/ctrl/xc_resume.c +++ /dev/null @@ -1,264 +0,0 @@ -/* - * This library is free software; you can redistribute it and/or - * modify it under the terms of the GNU Lesser General Public - * License as published by the Free Software Foundation; - * version 2.1 of the License. - * - * This library is distributed in the hope that it will be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU - * Lesser General Public License for more details. - * - * You should have received a copy of the GNU Lesser General Public - * License along with this library; If not, see . - */ - -#include "xc_private.h" - -#if defined(__i386__) || defined(__x86_64__) - -#include -#include -#include -#include "xc_core.h" - -static int modify_returncode(xc_interface *xch, uint32_t domid) -{ - vcpu_guest_context_any_t ctxt; - xc_dominfo_t info; - xen_capabilities_info_t caps; - struct domain_info_context _dinfo = {}; - struct domain_info_context *dinfo = &_dinfo; - int rc; - - if ( xc_domain_getinfo(xch, domid, 1, &info) != 1 || - info.domid != domid ) - { - PERROR("Could not get domain info"); - return -1; - } - - if ( !info.shutdown || (info.shutdown_reason != SHUTDOWN_suspend) ) - { - ERROR("Dom %d not suspended: (shutdown %d, reason %d)", domid, - info.shutdown, info.shutdown_reason); - errno = EINVAL; - return -1; - } - - if ( info.hvm ) - { - /* HVM guests without PV drivers have no return code to modify. */ - uint64_t irq = 0; - xc_hvm_param_get(xch, domid, HVM_PARAM_CALLBACK_IRQ, &irq); - if ( !irq ) - return 0; - - /* HVM guests have host address width. */ - if ( xc_version(xch, XENVER_capabilities, &caps) != 0 ) - { - PERROR("Could not get Xen capabilities"); - return -1; - } - dinfo->guest_width = strstr(caps, "x86_64") ? 8 : 4; - } - else - { - /* Probe PV guest address width. */ - if ( xc_domain_get_guest_width(xch, domid, &dinfo->guest_width) ) - return -1; - } - - if ( (rc = xc_vcpu_getcontext(xch, domid, 0, &ctxt)) != 0 ) - return rc; - - SET_FIELD(&ctxt, user_regs.eax, 1, dinfo->guest_width); - - if ( (rc = xc_vcpu_setcontext(xch, domid, 0, &ctxt)) != 0 ) - return rc; - - return 0; -} - -#else - -static int modify_returncode(xc_interface *xch, uint32_t domid) -{ - return 0; - -} - -#endif - -static int xc_domain_resume_cooperative(xc_interface *xch, uint32_t domid) -{ - DECLARE_DOMCTL; - int rc; - - /* - * Set hypercall return code to indicate that suspend is cancelled - * (rather than resuming in a new domain context). - */ - if ( (rc = modify_returncode(xch, domid)) != 0 ) - return rc; - - domctl.cmd = XEN_DOMCTL_resumedomain; - domctl.domain = domid; - return do_domctl(xch, &domctl); -} - -#if defined(__i386__) || defined(__x86_64__) -static int xc_domain_resume_hvm(xc_interface *xch, uint32_t domid) -{ - DECLARE_DOMCTL; - - /* - * The domctl XEN_DOMCTL_resumedomain unpause each vcpu. After - * the domctl, the guest will run. - * - * If it is PVHVM, the guest called the hypercall - * SCHEDOP_shutdown:SHUTDOWN_suspend - * to suspend itself. We don't modify the return code, so the PV driver - * will disconnect and reconnect. - * - * If it is a HVM, the guest will continue running. - */ - domctl.cmd = XEN_DOMCTL_resumedomain; - domctl.domain = domid; - return do_domctl(xch, &domctl); -} -#endif - -static int xc_domain_resume_any(xc_interface *xch, uint32_t domid) -{ - DECLARE_DOMCTL; - xc_dominfo_t info; - int i, rc = -1; -#if defined(__i386__) || defined(__x86_64__) - struct domain_info_context _dinfo = { .guest_width = 0, - .p2m_size = 0 }; - struct domain_info_context *dinfo = &_dinfo; - xen_pfn_t mfn, store_mfn, console_mfn; - vcpu_guest_context_any_t ctxt; - start_info_any_t *start_info; - shared_info_any_t *shinfo = NULL; - xen_pfn_t *p2m = NULL; -#endif - - if ( xc_domain_getinfo(xch, domid, 1, &info) != 1 ) - { - PERROR("Could not get domain info"); - return rc; - } - - /* - * (x86 only) Rewrite store_mfn and console_mfn back to MFN (from PFN). - */ -#if defined(__i386__) || defined(__x86_64__) - if ( info.hvm ) - return xc_domain_resume_hvm(xch, domid); - - if ( xc_domain_get_guest_width(xch, domid, &dinfo->guest_width) != 0 ) - { - PERROR("Could not get domain width"); - return rc; - } - - /* Map the shared info frame */ - shinfo = xc_map_foreign_range(xch, domid, PAGE_SIZE, - PROT_READ, info.shared_info_frame); - if ( shinfo == NULL ) - { - ERROR("Couldn't map shared info"); - goto out; - } - - /* Map the p2m list */ - if ( xc_core_arch_map_p2m(xch, dinfo, &info, shinfo, &p2m) ) - { - ERROR("Couldn't map p2m table"); - goto out; - } - - if ( xc_vcpu_getcontext(xch, domid, 0, &ctxt) ) - { - ERROR("Could not get vcpu context"); - goto out; - } - - mfn = GET_FIELD(&ctxt, user_regs.edx, dinfo->guest_width); - - start_info = xc_map_foreign_range(xch, domid, PAGE_SIZE, - PROT_READ | PROT_WRITE, mfn); - if ( start_info == NULL ) - { - ERROR("Couldn't map start_info"); - goto out; - } - - store_mfn = GET_FIELD(start_info, store_mfn, dinfo->guest_width); - console_mfn = GET_FIELD(start_info, console.domU.mfn, dinfo->guest_width); - if ( dinfo->guest_width == 4 ) - { - store_mfn = ((uint32_t *)p2m)[store_mfn]; - console_mfn = ((uint32_t *)p2m)[console_mfn]; - } - else - { - store_mfn = ((uint64_t *)p2m)[store_mfn]; - console_mfn = ((uint64_t *)p2m)[console_mfn]; - } - SET_FIELD(start_info, store_mfn, store_mfn, dinfo->guest_width); - SET_FIELD(start_info, console.domU.mfn, console_mfn, dinfo->guest_width); - - munmap(start_info, PAGE_SIZE); -#endif /* defined(__i386__) || defined(__x86_64__) */ - - /* Reset all secondary CPU states. */ - for ( i = 1; i <= info.max_vcpu_id; i++ ) - if ( xc_vcpu_setcontext(xch, domid, i, NULL) != 0 ) - { - ERROR("Couldn't reset vcpu state"); - goto out; - } - - /* Ready to resume domain execution now. */ - domctl.cmd = XEN_DOMCTL_resumedomain; - domctl.domain = domid; - rc = do_domctl(xch, &domctl); - -out: -#if defined(__i386__) || defined(__x86_64__) - if (p2m) - munmap(p2m, dinfo->p2m_frames * PAGE_SIZE); - if (shinfo) - munmap(shinfo, PAGE_SIZE); -#endif - - return rc; -} - -/* - * Resume execution of a domain after suspend shutdown. - * This can happen in one of two ways: - * 1. (fast=1) Resume the guest without resetting the domain environment. - * The guests's call to SCHEDOP_shutdown(SHUTDOWN_suspend) will return 1. - * - * 2. (fast=0) Reset guest environment so it believes it is resumed in a new - * domain context. The guests's call to SCHEDOP_shutdown(SHUTDOWN_suspend) - * will return 0. - * - * (1) should only by used for guests which can handle the special return - * code. Also note that the insertion of the return code is quite interesting - * and that the guest MUST be paused - otherwise we would be corrupting - * the guest vCPU state. - * - * (2) should be used only for guests which cannot handle the special - * new return code - and it is always safe (but slower). - */ -int xc_domain_resume(xc_interface *xch, uint32_t domid, int fast) -{ - return (fast - ? xc_domain_resume_cooperative(xch, domid) - : xc_domain_resume_any(xch, domid)); -} diff --git a/tools/libs/guest/Makefile b/tools/libs/guest/Makefile index 6d2a1d5bbc..2a2323ff09 100644 --- a/tools/libs/guest/Makefile +++ b/tools/libs/guest/Makefile @@ -9,6 +9,7 @@ endif SRCS-y += xg_private.c SRCS-y += xg_domain.c SRCS-y += xg_suspend.c +SRCS-y += xg_resume.c ifeq ($(CONFIG_MIGRATE),y) SRCS-y += xg_sr_common.c SRCS-$(CONFIG_X86) += xg_sr_common_x86.c diff --git a/tools/libs/guest/xg_resume.c b/tools/libs/guest/xg_resume.c new file mode 100644 index 0000000000..3bdefb2eef --- /dev/null +++ b/tools/libs/guest/xg_resume.c @@ -0,0 +1,265 @@ +/* + * This library is free software; you can redistribute it and/or + * modify it under the terms of the GNU Lesser General Public + * License as published by the Free Software Foundation; + * version 2.1 of the License. + * + * This library is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public + * License along with this library; If not, see . + */ + +#include "xc_private.h" +#include "xenguest.h" + +#if defined(__i386__) || defined(__x86_64__) + +#include +#include +#include +#include "xc_core.h" + +static int modify_returncode(xc_interface *xch, uint32_t domid) +{ + vcpu_guest_context_any_t ctxt; + xc_dominfo_t info; + xen_capabilities_info_t caps; + struct domain_info_context _dinfo = {}; + struct domain_info_context *dinfo = &_dinfo; + int rc; + + if ( xc_domain_getinfo(xch, domid, 1, &info) != 1 || + info.domid != domid ) + { + PERROR("Could not get domain info"); + return -1; + } + + if ( !info.shutdown || (info.shutdown_reason != SHUTDOWN_suspend) ) + { + ERROR("Dom %d not suspended: (shutdown %d, reason %d)", domid, + info.shutdown, info.shutdown_reason); + errno = EINVAL; + return -1; + } + + if ( info.hvm ) + { + /* HVM guests without PV drivers have no return code to modify. */ + uint64_t irq = 0; + xc_hvm_param_get(xch, domid, HVM_PARAM_CALLBACK_IRQ, &irq); + if ( !irq ) + return 0; + + /* HVM guests have host address width. */ + if ( xc_version(xch, XENVER_capabilities, &caps) != 0 ) + { + PERROR("Could not get Xen capabilities"); + return -1; + } + dinfo->guest_width = strstr(caps, "x86_64") ? 8 : 4; + } + else + { + /* Probe PV guest address width. */ + if ( xc_domain_get_guest_width(xch, domid, &dinfo->guest_width) ) + return -1; + } + + if ( (rc = xc_vcpu_getcontext(xch, domid, 0, &ctxt)) != 0 ) + return rc; + + SET_FIELD(&ctxt, user_regs.eax, 1, dinfo->guest_width); + + if ( (rc = xc_vcpu_setcontext(xch, domid, 0, &ctxt)) != 0 ) + return rc; + + return 0; +} + +#else + +static int modify_returncode(xc_interface *xch, uint32_t domid) +{ + return 0; + +} + +#endif + +static int xc_domain_resume_cooperative(xc_interface *xch, uint32_t domid) +{ + DECLARE_DOMCTL; + int rc; + + /* + * Set hypercall return code to indicate that suspend is cancelled + * (rather than resuming in a new domain context). + */ + if ( (rc = modify_returncode(xch, domid)) != 0 ) + return rc; + + domctl.cmd = XEN_DOMCTL_resumedomain; + domctl.domain = domid; + return do_domctl(xch, &domctl); +} + +#if defined(__i386__) || defined(__x86_64__) +static int xc_domain_resume_hvm(xc_interface *xch, uint32_t domid) +{ + DECLARE_DOMCTL; + + /* + * The domctl XEN_DOMCTL_resumedomain unpause each vcpu. After + * the domctl, the guest will run. + * + * If it is PVHVM, the guest called the hypercall + * SCHEDOP_shutdown:SHUTDOWN_suspend + * to suspend itself. We don't modify the return code, so the PV driver + * will disconnect and reconnect. + * + * If it is a HVM, the guest will continue running. + */ + domctl.cmd = XEN_DOMCTL_resumedomain; + domctl.domain = domid; + return do_domctl(xch, &domctl); +} +#endif + +static int xc_domain_resume_any(xc_interface *xch, uint32_t domid) +{ + DECLARE_DOMCTL; + xc_dominfo_t info; + int i, rc = -1; +#if defined(__i386__) || defined(__x86_64__) + struct domain_info_context _dinfo = { .guest_width = 0, + .p2m_size = 0 }; + struct domain_info_context *dinfo = &_dinfo; + xen_pfn_t mfn, store_mfn, console_mfn; + vcpu_guest_context_any_t ctxt; + start_info_any_t *start_info; + shared_info_any_t *shinfo = NULL; + xen_pfn_t *p2m = NULL; +#endif + + if ( xc_domain_getinfo(xch, domid, 1, &info) != 1 ) + { + PERROR("Could not get domain info"); + return rc; + } + + /* + * (x86 only) Rewrite store_mfn and console_mfn back to MFN (from PFN). + */ +#if defined(__i386__) || defined(__x86_64__) + if ( info.hvm ) + return xc_domain_resume_hvm(xch, domid); + + if ( xc_domain_get_guest_width(xch, domid, &dinfo->guest_width) != 0 ) + { + PERROR("Could not get domain width"); + return rc; + } + + /* Map the shared info frame */ + shinfo = xc_map_foreign_range(xch, domid, PAGE_SIZE, + PROT_READ, info.shared_info_frame); + if ( shinfo == NULL ) + { + ERROR("Couldn't map shared info"); + goto out; + } + + /* Map the p2m list */ + if ( xc_core_arch_map_p2m(xch, dinfo, &info, shinfo, &p2m) ) + { + ERROR("Couldn't map p2m table"); + goto out; + } + + if ( xc_vcpu_getcontext(xch, domid, 0, &ctxt) ) + { + ERROR("Could not get vcpu context"); + goto out; + } + + mfn = GET_FIELD(&ctxt, user_regs.edx, dinfo->guest_width); + + start_info = xc_map_foreign_range(xch, domid, PAGE_SIZE, + PROT_READ | PROT_WRITE, mfn); + if ( start_info == NULL ) + { + ERROR("Couldn't map start_info"); + goto out; + } + + store_mfn = GET_FIELD(start_info, store_mfn, dinfo->guest_width); + console_mfn = GET_FIELD(start_info, console.domU.mfn, dinfo->guest_width); + if ( dinfo->guest_width == 4 ) + { + store_mfn = ((uint32_t *)p2m)[store_mfn]; + console_mfn = ((uint32_t *)p2m)[console_mfn]; + } + else + { + store_mfn = ((uint64_t *)p2m)[store_mfn]; + console_mfn = ((uint64_t *)p2m)[console_mfn]; + } + SET_FIELD(start_info, store_mfn, store_mfn, dinfo->guest_width); + SET_FIELD(start_info, console.domU.mfn, console_mfn, dinfo->guest_width); + + munmap(start_info, PAGE_SIZE); +#endif /* defined(__i386__) || defined(__x86_64__) */ + + /* Reset all secondary CPU states. */ + for ( i = 1; i <= info.max_vcpu_id; i++ ) + if ( xc_vcpu_setcontext(xch, domid, i, NULL) != 0 ) + { + ERROR("Couldn't reset vcpu state"); + goto out; + } + + /* Ready to resume domain execution now. */ + domctl.cmd = XEN_DOMCTL_resumedomain; + domctl.domain = domid; + rc = do_domctl(xch, &domctl); + +out: +#if defined(__i386__) || defined(__x86_64__) + if (p2m) + munmap(p2m, dinfo->p2m_frames * PAGE_SIZE); + if (shinfo) + munmap(shinfo, PAGE_SIZE); +#endif + + return rc; +} + +/* + * Resume execution of a domain after suspend shutdown. + * This can happen in one of two ways: + * 1. (fast=1) Resume the guest without resetting the domain environment. + * The guests's call to SCHEDOP_shutdown(SHUTDOWN_suspend) will return 1. + * + * 2. (fast=0) Reset guest environment so it believes it is resumed in a new + * domain context. The guests's call to SCHEDOP_shutdown(SHUTDOWN_suspend) + * will return 0. + * + * (1) should only by used for guests which can handle the special return + * code. Also note that the insertion of the return code is quite interesting + * and that the guest MUST be paused - otherwise we would be corrupting + * the guest vCPU state. + * + * (2) should be used only for guests which cannot handle the special + * new return code - and it is always safe (but slower). + */ +int xc_domain_resume(xc_interface *xch, uint32_t domid, int fast) +{ + return (fast + ? xc_domain_resume_cooperative(xch, domid) + : xc_domain_resume_any(xch, domid)); +} -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:22:53 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:22:53 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146579.269704 (Exim 4.92) (envelope-from ) id 1lwMVZ-0004Lx-Pu; Thu, 24 Jun 2021 10:22:53 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146579.269704; Thu, 24 Jun 2021 10:22:53 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMVZ-0004Lq-MW; Thu, 24 Jun 2021 10:22:53 +0000 Received: by outflank-mailman (input) for mailman id 146579; Thu, 24 Jun 2021 10:22:53 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMVZ-0004Lk-6j for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:22:53 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMVZ-0005wO-5g for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:22:53 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwMVZ-0006Et-4k for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:22:53 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=mL//WOPR3CslskCSLz22EVJCFnSVWdh22+KrZklJwlk=; b=x/jhX+h8PnVpGsFMGjnCOBvUw0 /Oa9fzuD978foHVBi9ULS+B8UiGr+mz/D90uwvGAeDwzgjMT65ju+sHlnLGiRmnZ0qEMmKbwHYIci zIDRCVHFdv68lXWii2qgEZ389X6hzoi+YJPhTVBcGUl+tSGN1evfAZjnJY+Hx909+ylA=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] tools/libs: move xc_core* from libxenctrl to libxenguest Message-Id: Date: Thu, 24 Jun 2021 10:22:53 +0000 commit 455790573d3bbad6d5a1bb7e9d28b6dd71075693 Author: Juergen Gross AuthorDate: Fri Jun 4 08:02:13 2021 +0200 Commit: Julien Grall CommitDate: Fri Jun 4 18:56:46 2021 +0100 tools/libs: move xc_core* from libxenctrl to libxenguest The functionality in xc_core* should be part of libxenguest instead of libxenctrl. Users are already either in libxenguest, or in xl. There is one single exception: xc_core_arch_auto_translated_physmap() is being used by xc_domain_memory_mapping(), which is used by qemu. So leave the xc_core_arch_auto_translated_physmap() functionality in libxenctrl. This will make it easier to merge common functionality of xc_core* and xg_sr_save*. Signed-off-by: Juergen Gross Acked-by: Wei Liu --- tools/libs/ctrl/Makefile | 3 - tools/libs/ctrl/xc_core.c | 1027 ------------------------------------ tools/libs/ctrl/xc_core.h | 176 ------ tools/libs/ctrl/xc_core_arm.c | 113 ---- tools/libs/ctrl/xc_core_arm.h | 59 --- tools/libs/ctrl/xc_core_x86.c | 391 -------------- tools/libs/ctrl/xc_core_x86.h | 60 --- tools/libs/ctrl/xc_domain.c | 2 - tools/libs/ctrl/xc_private.h | 12 + tools/libs/guest/Makefile | 3 + tools/libs/guest/xg_core.c | 1027 ++++++++++++++++++++++++++++++++++++ tools/libs/guest/xg_core.h | 175 ++++++ tools/libs/guest/xg_core_arm.c | 107 ++++ tools/libs/guest/xg_core_arm.h | 59 +++ tools/libs/guest/xg_core_x86.c | 385 ++++++++++++++ tools/libs/guest/xg_core_x86.h | 60 +++ tools/libs/guest/xg_dom_boot.c | 2 +- tools/libs/guest/xg_domain.c | 2 +- tools/libs/guest/xg_offline_page.c | 2 +- tools/libs/guest/xg_resume.c | 2 +- 20 files changed, 1832 insertions(+), 1835 deletions(-) diff --git a/tools/libs/ctrl/Makefile b/tools/libs/ctrl/Makefile index fbeb3a3537..519246b0d6 100644 --- a/tools/libs/ctrl/Makefile +++ b/tools/libs/ctrl/Makefile @@ -2,9 +2,6 @@ XEN_ROOT = $(CURDIR)/../../.. include $(XEN_ROOT)/tools/Rules.mk SRCS-y += xc_altp2m.c -SRCS-y += xc_core.c -SRCS-$(CONFIG_X86) += xc_core_x86.c -SRCS-$(CONFIG_ARM) += xc_core_arm.c SRCS-y += xc_cpupool.c SRCS-y += xc_domain.c SRCS-y += xc_evtchn.c diff --git a/tools/libs/ctrl/xc_core.c b/tools/libs/ctrl/xc_core.c deleted file mode 100644 index 9576bec5a3..0000000000 --- a/tools/libs/ctrl/xc_core.c +++ /dev/null @@ -1,1027 +0,0 @@ -/* - * Elf format, (pfn, gmfn) table, IA64 support. - * Copyright (c) 2007 Isaku Yamahata - * VA Linux Systems Japan K.K. - * - * This library is free software; you can redistribute it and/or - * modify it under the terms of the GNU Lesser General Public - * License as published by the Free Software Foundation; - * version 2.1 of the License. - * - * This library is distributed in the hope that it will be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU - * Lesser General Public License for more details. - * - * You should have received a copy of the GNU Lesser General Public - * License along with this library; If not, see . - */ - -/* - * xen dump-core file format follows ELF format specification. - * Analisys tools shouldn't depends on the order of sections. - * They should follow elf header and check section names. - * - * +--------------------------------------------------------+ - * |ELF header | - * +--------------------------------------------------------+ - * |section headers | - * | null section header | - * | .shstrtab | - * | .note.Xen | - * | .xen_prstatus | - * | .xen_shared_info if present | - * | .xen_pages | - * | .xen_p2m or .xen_pfn | - * +--------------------------------------------------------+ - * |.note.Xen:note section | - * | "Xen" is used as note name, | - * | types are defined in xen/include/public/elfnote.h | - * | and descriptors are defined in xc_core.h. | - * | dumpcore none | - * | dumpcore header | - * | dumpcore xen version | - * | dumpcore format version | - * +--------------------------------------------------------+ - * |.xen_prstatus | - * | vcpu_guest_context_t[nr_vcpus] | - * +--------------------------------------------------------+ - * |.xen_shared_info if possible | - * +--------------------------------------------------------+ - * |.xen_pages | - * | page * nr_pages | - * +--------------------------------------------------------+ - * |.xen_p2m or .xen_pfn | - * | .xen_p2m: struct xen_dumpcore_p2m[nr_pages] | - * | .xen_pfn: uint64_t[nr_pages] | - * +--------------------------------------------------------+ - * |.shstrtab: section header string table | - * +--------------------------------------------------------+ - * - */ - -#include "xc_private.h" -#include "xc_core.h" -#include -#include - -#include - -/* number of pages to write at a time */ -#define DUMP_INCREMENT (4 * 1024) - -/* string table */ -struct xc_core_strtab { - char *strings; - uint16_t length; - uint16_t max; -}; - -static struct xc_core_strtab* -xc_core_strtab_init(xc_interface *xch) -{ - struct xc_core_strtab *strtab; - char *strings; - strtab = malloc(sizeof(*strtab)); - if ( strtab == NULL ) - return NULL; - - strings = malloc(PAGE_SIZE); - if ( strings == NULL ) - { - PERROR("Could not allocate string table init"); - free(strtab); - return NULL; - } - strtab->strings = strings; - strtab->max = PAGE_SIZE; - - /* index 0 represents none */ - strtab->strings[0] = '\0'; - strtab->length = 1; - - return strtab; -} - -static void -xc_core_strtab_free(struct xc_core_strtab *strtab) -{ - free(strtab->strings); - free(strtab); -} - -static uint16_t -xc_core_strtab_get(xc_interface *xch, struct xc_core_strtab *strtab, const char *name) -{ - uint16_t ret = 0; - uint16_t len = strlen(name) + 1; - - if ( strtab->length > UINT16_MAX - len ) - { - PERROR("too long string table"); - errno = E2BIG; - return ret; - } - - if ( strtab->length + len > strtab->max ) - { - char *tmp; - if ( strtab->max > UINT16_MAX / 2 ) - { - PERROR("too long string table"); - errno = ENOMEM; - return ret; - } - - tmp = realloc(strtab->strings, strtab->max * 2); - if ( tmp == NULL ) - { - PERROR("Could not allocate string table"); - return ret; - } - - strtab->strings = tmp; - strtab->max *= 2; - } - - ret = strtab->length; - strcpy(strtab->strings + strtab->length, name); - strtab->length += len; - return ret; -} - - -/* section headers */ -struct xc_core_section_headers { - uint16_t num; - uint16_t num_max; - - Elf64_Shdr *shdrs; -}; -#define SHDR_INIT ((uint16_t)16) -#define SHDR_INC ((uint16_t)4) - -static struct xc_core_section_headers* -xc_core_shdr_init(xc_interface *xch) -{ - struct xc_core_section_headers *sheaders; - sheaders = malloc(sizeof(*sheaders)); - if ( sheaders == NULL ) - return NULL; - - sheaders->num = 0; - sheaders->num_max = SHDR_INIT; - sheaders->shdrs = malloc(sizeof(sheaders->shdrs[0]) * sheaders->num_max); - if ( sheaders->shdrs == NULL ) - { - free(sheaders); - return NULL; - } - return sheaders; -} - -static void -xc_core_shdr_free(struct xc_core_section_headers *sheaders) -{ - free(sheaders->shdrs); - free(sheaders); -} - -Elf64_Shdr* -xc_core_shdr_get(xc_interface *xch, - struct xc_core_section_headers *sheaders) -{ - Elf64_Shdr *shdr; - - if ( sheaders->num == sheaders->num_max ) - { - Elf64_Shdr *shdrs; - if ( sheaders->num_max > UINT16_MAX - SHDR_INC ) - { - errno = E2BIG; - return NULL; - } - sheaders->num_max += SHDR_INC; - shdrs = realloc(sheaders->shdrs, - sizeof(sheaders->shdrs[0]) * sheaders->num_max); - if ( shdrs == NULL ) - return NULL; - sheaders->shdrs = shdrs; - } - - shdr = &sheaders->shdrs[sheaders->num]; - sheaders->num++; - memset(shdr, 0, sizeof(*shdr)); - return shdr; -} - -int -xc_core_shdr_set(xc_interface *xch, - Elf64_Shdr *shdr, - struct xc_core_strtab *strtab, - const char *name, uint32_t type, - uint64_t offset, uint64_t size, - uint64_t addralign, uint64_t entsize) -{ - uint64_t name_idx = xc_core_strtab_get(xch, strtab, name); - if ( name_idx == 0 ) - return -1; - - shdr->sh_name = name_idx; - shdr->sh_type = type; - shdr->sh_offset = offset; - shdr->sh_size = size; - shdr->sh_addralign = addralign; - shdr->sh_entsize = entsize; - return 0; -} - -static void -xc_core_ehdr_init(Elf64_Ehdr *ehdr) -{ - memset(ehdr, 0, sizeof(*ehdr)); - ehdr->e_ident[EI_MAG0] = ELFMAG0; - ehdr->e_ident[EI_MAG1] = ELFMAG1; - ehdr->e_ident[EI_MAG2] = ELFMAG2; - ehdr->e_ident[EI_MAG3] = ELFMAG3; - ehdr->e_ident[EI_CLASS] = ELFCLASS64; - ehdr->e_ident[EI_DATA] = ELF_ARCH_DATA; - ehdr->e_ident[EI_VERSION] = EV_CURRENT; - ehdr->e_ident[EI_OSABI] = ELFOSABI_SYSV; - ehdr->e_ident[EI_ABIVERSION] = EV_CURRENT; - - ehdr->e_type = ET_CORE; - /* e_machine will be filled in later */ - ehdr->e_version = EV_CURRENT; - ehdr->e_entry = 0; - ehdr->e_phoff = 0; - ehdr->e_shoff = sizeof(*ehdr); - ehdr->e_flags = ELF_CORE_EFLAGS; - ehdr->e_ehsize = sizeof(*ehdr); - ehdr->e_phentsize = sizeof(Elf64_Phdr); - ehdr->e_phnum = 0; - ehdr->e_shentsize = sizeof(Elf64_Shdr); - /* ehdr->e_shnum and ehdr->e_shstrndx aren't known here yet. - * fill it later */ -} - -static int -elfnote_fill_xen_version(xc_interface *xch, - struct xen_dumpcore_elfnote_xen_version_desc - *xen_version) -{ - int rc; - memset(xen_version, 0, sizeof(*xen_version)); - - rc = xc_version(xch, XENVER_version, NULL); - if ( rc < 0 ) - return rc; - xen_version->major_version = rc >> 16; - xen_version->minor_version = rc & ((1 << 16) - 1); - - rc = xc_version(xch, XENVER_extraversion, - &xen_version->extra_version); - if ( rc < 0 ) - return rc; - - rc = xc_version(xch, XENVER_compile_info, - &xen_version->compile_info); - if ( rc < 0 ) - return rc; - - rc = xc_version(xch, - XENVER_capabilities, &xen_version->capabilities); - if ( rc < 0 ) - return rc; - - rc = xc_version(xch, XENVER_changeset, &xen_version->changeset); - if ( rc < 0 ) - return rc; - - rc = xc_version(xch, XENVER_platform_parameters, - &xen_version->platform_parameters); - if ( rc < 0 ) - return rc; - - rc = xc_version(xch, XENVER_pagesize, NULL); - if ( rc < 0 ) - return rc; - xen_version->pagesize = rc; - - return 0; -} - -static void -elfnote_fill_format_version(struct xen_dumpcore_elfnote_format_version_desc - *format_version) -{ - format_version->version = XEN_DUMPCORE_FORMAT_VERSION_CURRENT; -} - -static void -elfnote_init(struct elfnote *elfnote) -{ - /* elf note section */ - memset(elfnote, 0, sizeof(*elfnote)); - elfnote->namesz = strlen(XEN_DUMPCORE_ELFNOTE_NAME) + 1; - strncpy(elfnote->name, XEN_DUMPCORE_ELFNOTE_NAME, sizeof(elfnote->name)); -} - -static int -elfnote_dump_none(xc_interface *xch, void *args, dumpcore_rtn_t dump_rtn) -{ - int sts; - struct elfnote elfnote; - struct xen_dumpcore_elfnote_none_desc none; - - elfnote_init(&elfnote); - /* Avoid compile warning about constant-zero-sized memset(). */ - /*memset(&none, 0, sizeof(none));*/ - - elfnote.descsz = sizeof(none); - elfnote.type = XEN_ELFNOTE_DUMPCORE_NONE; - sts = dump_rtn(xch, args, (char*)&elfnote, sizeof(elfnote)); - if ( sts != 0 ) - return sts; - return dump_rtn(xch, args, (char*)&none, sizeof(none)); -} - -static int -elfnote_dump_core_header( - xc_interface *xch, - void *args, dumpcore_rtn_t dump_rtn, const xc_dominfo_t *info, - int nr_vcpus, unsigned long nr_pages) -{ - int sts; - struct elfnote elfnote; - struct xen_dumpcore_elfnote_header_desc header; - - elfnote_init(&elfnote); - memset(&header, 0, sizeof(header)); - - elfnote.descsz = sizeof(header); - elfnote.type = XEN_ELFNOTE_DUMPCORE_HEADER; - header.xch_magic = info->hvm ? XC_CORE_MAGIC_HVM : XC_CORE_MAGIC; - header.xch_nr_vcpus = nr_vcpus; - header.xch_nr_pages = nr_pages; - header.xch_page_size = PAGE_SIZE; - sts = dump_rtn(xch, args, (char*)&elfnote, sizeof(elfnote)); - if ( sts != 0 ) - return sts; - return dump_rtn(xch, args, (char*)&header, sizeof(header)); -} - -static int -elfnote_dump_xen_version(xc_interface *xch, void *args, - dumpcore_rtn_t dump_rtn, unsigned int guest_width) -{ - int sts; - struct elfnote elfnote; - struct xen_dumpcore_elfnote_xen_version_desc xen_version; - - elfnote_init(&elfnote); - memset(&xen_version, 0, sizeof(xen_version)); - - elfnote.descsz = sizeof(xen_version); - elfnote.type = XEN_ELFNOTE_DUMPCORE_XEN_VERSION; - elfnote_fill_xen_version(xch, &xen_version); - if (guest_width < sizeof(unsigned long)) - { - // 32 bit elf file format differs in pagesize's alignment - char *p = (char *)&xen_version.pagesize; - memmove(p - 4, p, sizeof(xen_version.pagesize)); - } - sts = dump_rtn(xch, args, (char*)&elfnote, sizeof(elfnote)); - if ( sts != 0 ) - return sts; - return dump_rtn(xch, args, (char*)&xen_version, sizeof(xen_version)); -} - -static int -elfnote_dump_format_version(xc_interface *xch, - void *args, dumpcore_rtn_t dump_rtn) -{ - int sts; - struct elfnote elfnote; - struct xen_dumpcore_elfnote_format_version_desc format_version; - - elfnote_init(&elfnote); - memset(&format_version, 0, sizeof(format_version)); - - elfnote.descsz = sizeof(format_version); - elfnote.type = XEN_ELFNOTE_DUMPCORE_FORMAT_VERSION; - elfnote_fill_format_version(&format_version); - sts = dump_rtn(xch, args, (char*)&elfnote, sizeof(elfnote)); - if ( sts != 0 ) - return sts; - return dump_rtn(xch, args, (char*)&format_version, sizeof(format_version)); -} - -int -xc_domain_dumpcore_via_callback(xc_interface *xch, - uint32_t domid, - void *args, - dumpcore_rtn_t dump_rtn) -{ - xc_dominfo_t info; - shared_info_any_t *live_shinfo = NULL; - struct domain_info_context _dinfo = {}; - struct domain_info_context *dinfo = &_dinfo; - - int nr_vcpus = 0; - char *dump_mem, *dump_mem_start = NULL; - vcpu_guest_context_any_t *ctxt = NULL; - struct xc_core_arch_context arch_ctxt; - char dummy[PAGE_SIZE]; - int dummy_len; - int sts = -1; - - unsigned long i; - unsigned long j; - unsigned long nr_pages; - unsigned long max_mfn; - - xc_core_memory_map_t *memory_map = NULL; - unsigned int nr_memory_map; - unsigned int map_idx; - - int auto_translated_physmap; - xen_pfn_t *p2m = NULL; - struct xen_dumpcore_p2m *p2m_array = NULL; - - uint64_t *pfn_array = NULL; - - Elf64_Ehdr ehdr; - uint64_t filesz; - uint64_t offset; - uint64_t fixup; - - struct xc_core_strtab *strtab = NULL; - uint16_t strtab_idx; - struct xc_core_section_headers *sheaders = NULL; - Elf64_Shdr *shdr; - - xc_core_arch_context_init(&arch_ctxt); - if ( (dump_mem_start = malloc(DUMP_INCREMENT*PAGE_SIZE)) == NULL ) - { - PERROR("Could not allocate dump_mem"); - goto out; - } - - if ( xc_domain_getinfo(xch, domid, 1, &info) != 1 ) - { - PERROR("Could not get info for domain"); - goto out; - } - /* Map the shared info frame */ - live_shinfo = xc_map_foreign_range(xch, domid, PAGE_SIZE, - PROT_READ, info.shared_info_frame); - if ( !live_shinfo && !info.hvm ) - { - PERROR("Couldn't map live_shinfo"); - goto out; - } - auto_translated_physmap = xc_core_arch_auto_translated_physmap(&info); - - if ( !auto_translated_physmap ) - - { - if ( xc_domain_get_guest_width(xch, domid, &dinfo->guest_width) != 0 ) - { - PERROR("Could not get address size for domain"); - goto out; - } - } - else - { - /* - * Autotranslated guest never sets guest width in the first - * place. Force guest_width to be sizeof(unsigned long) so - * code below functions properly. - * - * Here is why this is correct. - * - * 1. Before f969bc9fc, xc_domain_get_guest_width for HVM (x86 - * and ARM) always returned hypervisor's idea of - * sizeof(unsigned long). - * - * 2. There has never been a situation in which hypervisor's - * word width is smaller than toolstack domain's (i.e. no - * 32bit hypervisor + 64bit toolstack). - * - * Predicates in code test guest_width against toolstack - * domain's sizeof(unsigned long), so setting guest_width to - * toolstack domain's idea of sizeof(unsigned long) matches - * the original behaviour for HVM guests. - */ - dinfo->guest_width = sizeof(unsigned long); - } - - if ( domid != info.domid ) - { - PERROR("Domain %d does not exist", domid); - goto out; - } - - ctxt = calloc(sizeof(*ctxt), info.max_vcpu_id + 1); - if ( !ctxt ) - { - PERROR("Could not allocate vcpu context array"); - goto out; - } - - for ( i = 0; i <= info.max_vcpu_id; i++ ) - { - if ( xc_vcpu_getcontext(xch, domid, i, &ctxt[nr_vcpus]) == 0 ) - { - if ( xc_core_arch_context_get(&arch_ctxt, &ctxt[nr_vcpus], - xch, domid) ) - continue; - nr_vcpus++; - } - } - if ( nr_vcpus == 0 ) - { - PERROR("No VCPU context could be grabbed"); - goto out; - } - - /* obtain memory map */ - sts = xc_core_arch_memory_map_get(xch, &arch_ctxt, &info, - live_shinfo, &memory_map, - &nr_memory_map); - if ( sts != 0 ) - goto out; - - /* - * Note: this is the *current* number of pages and may change under - * a live dump-core. We'll just take this value, and if more pages - * exist, we'll skip them. If there's less, then we'll just not use - * all the array... - * - * We don't want to use the total potential size of the memory map - * since that is usually much higher than info.nr_pages. - */ - nr_pages = info.nr_pages; - - if ( !auto_translated_physmap ) - { - /* obtain p2m table */ - p2m_array = malloc(nr_pages * sizeof(p2m_array[0])); - if ( p2m_array == NULL ) - { - PERROR("Could not allocate p2m array"); - goto out; - } - - sts = xc_core_arch_map_p2m(xch, dinfo, &info, live_shinfo, &p2m); - if ( sts != 0 ) - goto out; - - sts = xc_maximum_ram_page(xch, &max_mfn); - if ( sts != 0 ) - goto out; - } - else - { - pfn_array = malloc(nr_pages * sizeof(pfn_array[0])); - if ( pfn_array == NULL ) - { - PERROR("Could not allocate pfn array"); - goto out; - } - } - - /* ehdr.e_shnum and ehdr.e_shstrndx aren't known here yet. fill it later*/ - xc_core_ehdr_init(&ehdr); - - /* create section header */ - strtab = xc_core_strtab_init(xch); - if ( strtab == NULL ) - { - PERROR("Could not allocate string table"); - goto out; - } - sheaders = xc_core_shdr_init(xch); - if ( sheaders == NULL ) - { - PERROR("Could not allocate section headers"); - goto out; - } - /* null section */ - shdr = xc_core_shdr_get(xch,sheaders); - if ( shdr == NULL ) - { - PERROR("Could not get section header for null section"); - goto out; - } - - /* .shstrtab */ - shdr = xc_core_shdr_get(xch,sheaders); - if ( shdr == NULL ) - { - PERROR("Could not get section header for shstrtab"); - goto out; - } - strtab_idx = shdr - sheaders->shdrs; - /* strtab_shdr.sh_offset, strtab_shdr.sh_size aren't unknown. - * fill it later - */ - sts = xc_core_shdr_set(xch, shdr, strtab, ELF_SHSTRTAB, SHT_STRTAB, 0, 0, 0, 0); - if ( sts != 0 ) - goto out; - - /* elf note section */ - /* here the number of section header is unknown. fix up offset later. */ - offset = sizeof(ehdr); - filesz = - sizeof(struct xen_dumpcore_elfnote_none) + /* none */ - sizeof(struct xen_dumpcore_elfnote_header) + /* core header */ - sizeof(struct xen_dumpcore_elfnote_xen_version) + /* xen version */ - sizeof(struct xen_dumpcore_elfnote_format_version);/* format version */ - shdr = xc_core_shdr_get(xch,sheaders); - if ( shdr == NULL ) - { - PERROR("Could not get section header for note section"); - goto out; - } - sts = xc_core_shdr_set(xch, shdr, strtab, XEN_DUMPCORE_SEC_NOTE, SHT_NOTE, - offset, filesz, 0, 0); - if ( sts != 0 ) - goto out; - offset += filesz; - - /* prstatus */ - shdr = xc_core_shdr_get(xch,sheaders); - if ( shdr == NULL ) - { - PERROR("Could not get section header for .xen_prstatus"); - goto out; - } - filesz = sizeof(*ctxt) * nr_vcpus; - sts = xc_core_shdr_set(xch, shdr, strtab, XEN_DUMPCORE_SEC_PRSTATUS, - SHT_PROGBITS, offset, filesz, - __alignof__(*ctxt), sizeof(*ctxt)); - if ( sts != 0 ) - goto out; - offset += filesz; - - /* arch context */ - sts = xc_core_arch_context_get_shdr(xch, &arch_ctxt, sheaders, strtab, - &filesz, offset); - if ( sts != 0 ) - goto out; - offset += filesz; - - /* shared_info */ - if ( live_shinfo != NULL ) - { - shdr = xc_core_shdr_get(xch,sheaders); - if ( shdr == NULL ) - { - PERROR("Could not get section header for .xen_shared_info"); - goto out; - } - filesz = PAGE_SIZE; - sts = xc_core_shdr_set(xch, shdr, strtab, XEN_DUMPCORE_SEC_SHARED_INFO, - SHT_PROGBITS, offset, filesz, - __alignof__(*live_shinfo), PAGE_SIZE); - if ( sts != 0 ) - goto out; - offset += filesz; - } - - /* - * pages and p2m/pfn are the last section to allocate section headers - * so that we know the number of section headers here. - * 2 = pages section and p2m/pfn table section - */ - fixup = (sheaders->num + 2) * sizeof(*shdr); - /* zeroth section should have zero offset */ - for ( i = 1; i < sheaders->num; i++ ) - sheaders->shdrs[i].sh_offset += fixup; - offset += fixup; - dummy_len = ROUNDUP(offset, PAGE_SHIFT) - offset; /* padding length */ - offset += dummy_len; - - /* pages */ - shdr = xc_core_shdr_get(xch,sheaders); - if ( shdr == NULL ) - { - PERROR("could not get section headers for .xen_pages"); - goto out; - } - filesz = (uint64_t)nr_pages * PAGE_SIZE; - sts = xc_core_shdr_set(xch, shdr, strtab, XEN_DUMPCORE_SEC_PAGES, SHT_PROGBITS, - offset, filesz, PAGE_SIZE, PAGE_SIZE); - if ( sts != 0 ) - goto out; - offset += filesz; - - /* p2m/pfn table */ - shdr = xc_core_shdr_get(xch,sheaders); - if ( shdr == NULL ) - { - PERROR("Could not get section header for .xen_{p2m, pfn} table"); - goto out; - } - if ( !auto_translated_physmap ) - { - filesz = (uint64_t)nr_pages * sizeof(p2m_array[0]); - sts = xc_core_shdr_set(xch, shdr, strtab, XEN_DUMPCORE_SEC_P2M, - SHT_PROGBITS, - offset, filesz, __alignof__(p2m_array[0]), - sizeof(p2m_array[0])); - } - else - { - filesz = (uint64_t)nr_pages * sizeof(pfn_array[0]); - sts = xc_core_shdr_set(xch, shdr, strtab, XEN_DUMPCORE_SEC_PFN, - SHT_PROGBITS, - offset, filesz, __alignof__(pfn_array[0]), - sizeof(pfn_array[0])); - } - if ( sts != 0 ) - goto out; - offset += filesz; - - /* fixing up section header string table section header */ - filesz = strtab->length; - sheaders->shdrs[strtab_idx].sh_offset = offset; - sheaders->shdrs[strtab_idx].sh_size = filesz; - - /* write out elf header */ - ehdr.e_shnum = sheaders->num; - ehdr.e_shstrndx = strtab_idx; - ehdr.e_machine = ELF_ARCH_MACHINE; - sts = dump_rtn(xch, args, (char*)&ehdr, sizeof(ehdr)); - if ( sts != 0 ) - goto out; - - /* section headers */ - sts = dump_rtn(xch, args, (char*)sheaders->shdrs, - sheaders->num * sizeof(sheaders->shdrs[0])); - if ( sts != 0 ) - goto out; - - /* elf note section: xen core header */ - sts = elfnote_dump_none(xch, args, dump_rtn); - if ( sts != 0 ) - goto out; - - /* elf note section: xen core header */ - sts = elfnote_dump_core_header(xch, args, dump_rtn, &info, nr_vcpus, nr_pages); - if ( sts != 0 ) - goto out; - - /* elf note section: xen version */ - sts = elfnote_dump_xen_version(xch, args, dump_rtn, dinfo->guest_width); - if ( sts != 0 ) - goto out; - - /* elf note section: format version */ - sts = elfnote_dump_format_version(xch, args, dump_rtn); - if ( sts != 0 ) - goto out; - - /* prstatus: .xen_prstatus */ - sts = dump_rtn(xch, args, (char *)ctxt, sizeof(*ctxt) * nr_vcpus); - if ( sts != 0 ) - goto out; - - if ( live_shinfo != NULL ) - { - /* shared_info: .xen_shared_info */ - sts = dump_rtn(xch, args, (char*)live_shinfo, PAGE_SIZE); - if ( sts != 0 ) - goto out; - } - - /* arch specific context */ - sts = xc_core_arch_context_dump(xch, &arch_ctxt, args, dump_rtn); - if ( sts != 0 ) - goto out; - - /* Pad the output data to page alignment. */ - memset(dummy, 0, PAGE_SIZE); - sts = dump_rtn(xch, args, dummy, dummy_len); - if ( sts != 0 ) - goto out; - - /* dump pages: .xen_pages */ - j = 0; - dump_mem = dump_mem_start; - for ( map_idx = 0; map_idx < nr_memory_map; map_idx++ ) - { - uint64_t pfn_start; - uint64_t pfn_end; - - pfn_start = memory_map[map_idx].addr >> PAGE_SHIFT; - pfn_end = pfn_start + (memory_map[map_idx].size >> PAGE_SHIFT); - for ( i = pfn_start; i < pfn_end; i++ ) - { - uint64_t gmfn; - void *vaddr; - - if ( !auto_translated_physmap ) - { - if ( i >= dinfo->p2m_size ) - break; - - if ( dinfo->guest_width >= sizeof(unsigned long) ) - { - if ( dinfo->guest_width == sizeof(unsigned long) ) - gmfn = p2m[i]; - else - gmfn = ((uint64_t *)p2m)[i]; - if ( gmfn == INVALID_PFN ) - continue; - } - else - { - gmfn = ((uint32_t *)p2m)[i]; - if ( gmfn == (uint32_t)INVALID_PFN ) - continue; - } - if ( gmfn > max_mfn ) - continue; - - if ( j >= nr_pages ) - { - j++; - continue; - } - - p2m_array[j].pfn = i; - p2m_array[j].gmfn = gmfn; - } - else - { - if ( !xc_core_arch_gpfn_may_present(&arch_ctxt, i) ) - continue; - - if ( j >= nr_pages ) - { - j++; - continue; - } - - gmfn = i; - pfn_array[j] = i; - } - - vaddr = xc_map_foreign_range( - xch, domid, PAGE_SIZE, PROT_READ, gmfn); - if ( vaddr == NULL ) - continue; - memcpy(dump_mem, vaddr, PAGE_SIZE); - munmap(vaddr, PAGE_SIZE); - dump_mem += PAGE_SIZE; - if ( (j + 1) % DUMP_INCREMENT == 0 ) - { - sts = dump_rtn( - xch, args, dump_mem_start, dump_mem - dump_mem_start); - if ( sts != 0 ) - goto out; - dump_mem = dump_mem_start; - } - - j++; - } - } - - if ( j > nr_pages ) - { - /* - * When live dump-mode (-L option) is specified, - * guest domain may increase memory. - */ - IPRINTF("exceeded nr_pages (%ld) losing %ld pages", nr_pages, j - nr_pages); - } - - sts = dump_rtn(xch, args, dump_mem_start, dump_mem - dump_mem_start); - if ( sts != 0 ) - goto out; - if ( j < nr_pages ) - { - /* When live dump-mode (-L option) is specified, - * guest domain may reduce memory. pad with zero pages. - */ - DPRINTF("j (%ld) != nr_pages (%ld)", j, nr_pages); - memset(dump_mem_start, 0, PAGE_SIZE); - for (; j < nr_pages; j++) { - sts = dump_rtn(xch, args, dump_mem_start, PAGE_SIZE); - if ( sts != 0 ) - goto out; - if ( !auto_translated_physmap ) - { - p2m_array[j].pfn = XC_CORE_INVALID_PFN; - p2m_array[j].gmfn = XC_CORE_INVALID_GMFN; - } - else - pfn_array[j] = XC_CORE_INVALID_PFN; - } - } - - /* p2m/pfn table: .xen_p2m/.xen_pfn */ - if ( !auto_translated_physmap ) - sts = dump_rtn( - xch, args, (char *)p2m_array, sizeof(p2m_array[0]) * nr_pages); - else - sts = dump_rtn( - xch, args, (char *)pfn_array, sizeof(pfn_array[0]) * nr_pages); - if ( sts != 0 ) - goto out; - - /* elf section header string table: .shstrtab */ - sts = dump_rtn(xch, args, strtab->strings, strtab->length); - if ( sts != 0 ) - goto out; - - sts = 0; - -out: - if ( memory_map != NULL ) - free(memory_map); - if ( p2m != NULL ) - munmap(p2m, PAGE_SIZE * dinfo->p2m_frames); - if ( p2m_array != NULL ) - free(p2m_array); - if ( pfn_array != NULL ) - free(pfn_array); - if ( sheaders != NULL ) - xc_core_shdr_free(sheaders); - if ( strtab != NULL ) - xc_core_strtab_free(strtab); - if ( ctxt != NULL ) - free(ctxt); - if ( dump_mem_start != NULL ) - free(dump_mem_start); - if ( live_shinfo != NULL ) - munmap(live_shinfo, PAGE_SIZE); - xc_core_arch_context_free(&arch_ctxt); - - return sts; -} - -/* Callback args for writing to a local dump file. */ -struct dump_args { - int fd; -}; - -/* Callback routine for writing to a local dump file. */ -static int local_file_dump(xc_interface *xch, - void *args, char *buffer, unsigned int length) -{ - struct dump_args *da = args; - - if ( write_exact(da->fd, buffer, length) == -1 ) - { - PERROR("Failed to write buffer"); - return -errno; - } - - if ( length >= (DUMP_INCREMENT * PAGE_SIZE) ) - { - // Now dumping pages -- make sure we discard clean pages from - // the cache after each write - discard_file_cache(xch, da->fd, 0 /* no flush */); - } - - return 0; -} - -int -xc_domain_dumpcore(xc_interface *xch, - uint32_t domid, - const char *corename) -{ - struct dump_args da; - int sts; - - if ( (da.fd = open(corename, O_CREAT|O_RDWR|O_TRUNC, S_IWUSR|S_IRUSR)) < 0 ) - { - PERROR("Could not open corefile %s", corename); - return -errno; - } - - sts = xc_domain_dumpcore_via_callback( - xch, domid, &da, &local_file_dump); - - /* flush and discard any remaining portion of the file from cache */ - discard_file_cache(xch, da.fd, 1/* flush first*/); - - close(da.fd); - - return sts; -} - -/* - * Local variables: - * mode: C - * c-file-style: "BSD" - * c-basic-offset: 4 - * tab-width: 4 - * indent-tabs-mode: nil - * End: - */ diff --git a/tools/libs/ctrl/xc_core.h b/tools/libs/ctrl/xc_core.h deleted file mode 100644 index 8ea1f93a10..0000000000 --- a/tools/libs/ctrl/xc_core.h +++ /dev/null @@ -1,176 +0,0 @@ -/* - * Copyright (c) 2006 Isaku Yamahata - * VA Linux Systems Japan K.K. - * - * This library is free software; you can redistribute it and/or - * modify it under the terms of the GNU Lesser General Public - * License as published by the Free Software Foundation; either - * version 2.1 of the License, or (at your option) any later version. - * - * This library is distributed in the hope that it will be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU - * Lesser General Public License for more details. - * - * You should have received a copy of the GNU Lesser General Public - * License along with this library; If not, see . - * - */ - -#ifndef XC_CORE_H -#define XC_CORE_H - -#include "xen/version.h" -#include "xc_private.h" -#include "xen/libelf/elfstructs.h" - -/* section names */ -#define XEN_DUMPCORE_SEC_NOTE ".note.Xen" -#define XEN_DUMPCORE_SEC_PRSTATUS ".xen_prstatus" -#define XEN_DUMPCORE_SEC_SHARED_INFO ".xen_shared_info" -#define XEN_DUMPCORE_SEC_P2M ".xen_p2m" -#define XEN_DUMPCORE_SEC_PFN ".xen_pfn" -#define XEN_DUMPCORE_SEC_PAGES ".xen_pages" - -/* elf note name */ -#define XEN_DUMPCORE_ELFNOTE_NAME "Xen" -/* note numbers are defined in xen/elfnote.h */ - -struct elfnote { - uint32_t namesz; /* Elf_Note note; */ - uint32_t descsz; - uint32_t type; - char name[4]; /* sizeof("Xen") = 4 - * Fotunately this is 64bit aligned so that - * we can use same structore for both 32/64bit - */ -}; - -struct xen_dumpcore_elfnote_none_desc { - /* nothing */ -}; - -struct xen_dumpcore_elfnote_header_desc { - uint64_t xch_magic; - uint64_t xch_nr_vcpus; - uint64_t xch_nr_pages; - uint64_t xch_page_size; -}; - -struct xen_dumpcore_elfnote_xen_version_desc { - uint64_t major_version; - uint64_t minor_version; - xen_extraversion_t extra_version; - xen_compile_info_t compile_info; - xen_capabilities_info_t capabilities; - xen_changeset_info_t changeset; - xen_platform_parameters_t platform_parameters; - uint64_t pagesize; -}; - -#define XEN_DUMPCORE_FORMAT_VERSION(major, minor) \ - ((major) << 32) | ((minor) & 0xffffffff) -#define XEN_DUMPCORE_FORMAT_MAJOR(version) ((major) >> 32) -#define XEN_DUMPCORE_FORMAT_MINOR(version) ((minor) & 0xffffffff) - -#define XEN_DUMPCORE_FORMAT_MAJOR_CURRENT ((uint64_t)0) -#define XEN_DUMPCORE_FORMAT_MINOR_CURRENT ((uint64_t)1) -#define XEN_DUMPCORE_FORMAT_VERSION_CURRENT \ - XEN_DUMPCORE_FORMAT_VERSION(XEN_DUMPCORE_FORMAT_MAJOR_CURRENT, \ - XEN_DUMPCORE_FORMAT_MINOR_CURRENT) - -struct xen_dumpcore_elfnote_format_version_desc { - uint64_t version; -}; - - -struct xen_dumpcore_elfnote_none { - struct elfnote elfnote; - struct xen_dumpcore_elfnote_none_desc none; -}; - -struct xen_dumpcore_elfnote_header { - struct elfnote elfnote; - struct xen_dumpcore_elfnote_header_desc header; -}; - -struct xen_dumpcore_elfnote_xen_version { - struct elfnote elfnote; - struct xen_dumpcore_elfnote_xen_version_desc xen_version; -}; - -struct xen_dumpcore_elfnote_format_version { - struct elfnote elfnote; - struct xen_dumpcore_elfnote_format_version_desc format_version; -}; - -#define XC_CORE_INVALID_PFN (~(uint64_t)0) -#define XC_CORE_INVALID_GMFN (~(uint64_t)0) -struct xen_dumpcore_p2m { - uint64_t pfn; - uint64_t gmfn; -}; - - -struct xc_core_strtab; -struct xc_core_section_headers; - -Elf64_Shdr* -xc_core_shdr_get(xc_interface *xch, - struct xc_core_section_headers *sheaders); -int -xc_core_shdr_set(xc_interface *xch, - Elf64_Shdr *shdr, - struct xc_core_strtab *strtab, - const char *name, uint32_t type, - uint64_t offset, uint64_t size, - uint64_t addralign, uint64_t entsize); - -struct xc_core_memory_map { - uint64_t addr; - uint64_t size; -}; -typedef struct xc_core_memory_map xc_core_memory_map_t; -int xc_core_arch_auto_translated_physmap(const xc_dominfo_t *info); -struct xc_core_arch_context; -int xc_core_arch_memory_map_get(xc_interface *xch, - struct xc_core_arch_context *arch_ctxt, - xc_dominfo_t *info, shared_info_any_t *live_shinfo, - xc_core_memory_map_t **mapp, - unsigned int *nr_entries); -int xc_core_arch_map_p2m(xc_interface *xch, struct domain_info_context *dinfo, - xc_dominfo_t *info, shared_info_any_t *live_shinfo, - xen_pfn_t **live_p2m); - -int xc_core_arch_map_p2m_writable(xc_interface *xch, struct domain_info_context *dinfo, - xc_dominfo_t *info, - shared_info_any_t *live_shinfo, - xen_pfn_t **live_p2m); - -int xc_core_arch_get_scratch_gpfn(xc_interface *xch, uint32_t domid, - xen_pfn_t *gpfn); - - -#if defined (__i386__) || defined (__x86_64__) -# include "xc_core_x86.h" -#elif defined (__arm__) || defined(__aarch64__) -# include "xc_core_arm.h" -#else -# error "unsupported architecture" -#endif - -#ifndef ELF_CORE_EFLAGS -# define ELF_CORE_EFLAGS 0 -#endif - -#endif /* XC_CORE_H */ - -/* - * Local variables: - * mode: C - * c-file-style: "BSD" - * c-basic-offset: 4 - * tab-width: 4 - * indent-tabs-mode: nil - * End: - */ diff --git a/tools/libs/ctrl/xc_core_arm.c b/tools/libs/ctrl/xc_core_arm.c deleted file mode 100644 index 93765a565f..0000000000 --- a/tools/libs/ctrl/xc_core_arm.c +++ /dev/null @@ -1,113 +0,0 @@ -/* - * This library is free software; you can redistribute it and/or - * modify it under the terms of the GNU Lesser General Public - * License as published by the Free Software Foundation; either - * version 2.1 of the License, or (at your option) any later version. - * - * This library is distributed in the hope that it will be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU - * Lesser General Public License for more details. - * - * You should have received a copy of the GNU Lesser General Public - * License along with this library; If not, see . - * - * Copyright (c) 2011 Citrix Systems - * - */ - -#include "xc_private.h" -#include "xc_core.h" - -#include - -int -xc_core_arch_gpfn_may_present(struct xc_core_arch_context *arch_ctxt, - unsigned long pfn) -{ - /* TODO: memory from DT */ - if (pfn >= 0x80000 && pfn < 0x88000) - return 1; - return 0; -} - -int -xc_core_arch_auto_translated_physmap(const xc_dominfo_t *info) -{ - return 1; -} - -int -xc_core_arch_memory_map_get(xc_interface *xch, struct xc_core_arch_context *unused, - xc_dominfo_t *info, shared_info_any_t *live_shinfo, - xc_core_memory_map_t **mapp, - unsigned int *nr_entries) -{ - xen_pfn_t p2m_size = 0; - xc_core_memory_map_t *map; - - if ( xc_domain_nr_gpfns(xch, info->domid, &p2m_size) < 0 ) - return -1; - - map = malloc(sizeof(*map)); - if ( map == NULL ) - { - PERROR("Could not allocate memory"); - return -1; - } - - map->addr = 0; - map->size = ((uint64_t)p2m_size) << PAGE_SHIFT; - - *mapp = map; - *nr_entries = 1; - return 0; -} - -static int -xc_core_arch_map_p2m_rw(xc_interface *xch, struct domain_info_context *dinfo, xc_dominfo_t *info, - shared_info_any_t *live_shinfo, xen_pfn_t **live_p2m, int rw) -{ - errno = ENOSYS; - return -1; -} - -int -xc_core_arch_map_p2m(xc_interface *xch, struct domain_info_context *dinfo, xc_dominfo_t *info, - shared_info_any_t *live_shinfo, xen_pfn_t **live_p2m) -{ - return xc_core_arch_map_p2m_rw(xch, dinfo, info, live_shinfo, live_p2m, 0); -} - -int -xc_core_arch_map_p2m_writable(xc_interface *xch, struct domain_info_context *dinfo, xc_dominfo_t *info, - shared_info_any_t *live_shinfo, xen_pfn_t **live_p2m) -{ - return xc_core_arch_map_p2m_rw(xch, dinfo, info, live_shinfo, live_p2m, 1); -} - -int -xc_core_arch_get_scratch_gpfn(xc_interface *xch, uint32_t domid, - xen_pfn_t *gpfn) -{ - /* - * The Grant Table region space is not used until the guest is - * booting. Use the first page for the scratch pfn. - */ - BUILD_BUG_ON(GUEST_GNTTAB_SIZE < XC_PAGE_SIZE); - - *gpfn = GUEST_GNTTAB_BASE >> XC_PAGE_SHIFT; - - return 0; -} - - -/* - * Local variables: - * mode: C - * c-file-style: "BSD" - * c-basic-offset: 4 - * tab-width: 4 - * indent-tabs-mode: nil - * End: - */ diff --git a/tools/libs/ctrl/xc_core_arm.h b/tools/libs/ctrl/xc_core_arm.h deleted file mode 100644 index 162f7a7569..0000000000 --- a/tools/libs/ctrl/xc_core_arm.h +++ /dev/null @@ -1,59 +0,0 @@ -/* - * This library is free software; you can redistribute it and/or - * modify it under the terms of the GNU Lesser General Public - * License as published by the Free Software Foundation; either - * version 2.1 of the License, or (at your option) any later version. - * - * This library is distributed in the hope that it will be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU - * Lesser General Public License for more details. - * - * You should have received a copy of the GNU Lesser General Public - * License along with this library; If not, see . - * - * Copyright (c) 2012 Citrix Systems - * - */ - -#ifndef XC_CORE_ARM_H -#define XC_CORE_ARM_H - -#define ELF_ARCH_DATA ELFDATA2LSB -#define ELF_ARCH_MACHINE EM_ARM - -struct xc_core_arch_context { - /* nothing */ -}; - -#define xc_core_arch_context_init(arch_ctxt) do {} while (0) -#define xc_core_arch_context_free(arch_ctxt) do {} while (0) -#define xc_core_arch_context_get(arch_ctxt, ctxt, xch, domid) \ - (0) -#define xc_core_arch_context_dump(xch, arch_ctxt, args, dump_rtn) (0) - -int -xc_core_arch_gpfn_may_present(struct xc_core_arch_context *arch_ctxt, - unsigned long pfn); -static inline int -xc_core_arch_context_get_shdr(xc_interface *xch, - struct xc_core_arch_context *arch_ctxt, - struct xc_core_section_headers *sheaders, - struct xc_core_strtab *strtab, - uint64_t *filesz, uint64_t offset) -{ - *filesz = 0; - return 0; -} - -#endif /* XC_CORE_ARM_H */ - -/* - * Local variables: - * mode: C - * c-file-style: "BSD" - * c-basic-offset: 4 - * tab-width: 4 - * indent-tabs-mode: nil - * End: - */ diff --git a/tools/libs/ctrl/xc_core_x86.c b/tools/libs/ctrl/xc_core_x86.c deleted file mode 100644 index c8f71d4b75..0000000000 --- a/tools/libs/ctrl/xc_core_x86.c +++ /dev/null @@ -1,391 +0,0 @@ -/* - * This library is free software; you can redistribute it and/or - * modify it under the terms of the GNU Lesser General Public - * License as published by the Free Software Foundation; either - * version 2.1 of the License, or (at your option) any later version. - * - * This library is distributed in the hope that it will be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU - * Lesser General Public License for more details. - * - * You should have received a copy of the GNU Lesser General Public - * License along with this library; If not, see . - * - * Copyright (c) 2007 Isaku Yamahata - * VA Linux Systems Japan K.K. - * - */ - -#include -#include "xc_private.h" -#include "xc_core.h" -#include - -int -xc_core_arch_gpfn_may_present(struct xc_core_arch_context *arch_ctxt, - unsigned long pfn) -{ - if ((pfn >= 0xa0 && pfn < 0xc0) /* VGA hole */ - || (pfn >= (HVM_BELOW_4G_MMIO_START >> PAGE_SHIFT) - && pfn < (1ULL<<32) >> PAGE_SHIFT)) /* MMIO */ - return 0; - return 1; -} - -int -xc_core_arch_auto_translated_physmap(const xc_dominfo_t *info) -{ - return info->hvm; -} - -int -xc_core_arch_memory_map_get(xc_interface *xch, struct xc_core_arch_context *unused, - xc_dominfo_t *info, shared_info_any_t *live_shinfo, - xc_core_memory_map_t **mapp, - unsigned int *nr_entries) -{ - xen_pfn_t p2m_size = 0; - xc_core_memory_map_t *map; - - if ( xc_domain_nr_gpfns(xch, info->domid, &p2m_size) < 0 ) - return -1; - - map = malloc(sizeof(*map)); - if ( map == NULL ) - { - PERROR("Could not allocate memory"); - return -1; - } - - map->addr = 0; - map->size = ((uint64_t)p2m_size) << PAGE_SHIFT; - - *mapp = map; - *nr_entries = 1; - return 0; -} - -static inline bool is_canonical_address(uint64_t vaddr) -{ - return ((int64_t)vaddr >> 47) == ((int64_t)vaddr >> 63); -} - -/* Virtual address ranges reserved for hypervisor. */ -#define HYPERVISOR_VIRT_START_X86_64 0xFFFF800000000000ULL -#define HYPERVISOR_VIRT_END_X86_64 0xFFFF87FFFFFFFFFFULL - -#define HYPERVISOR_VIRT_START_X86_32 0x00000000F5800000ULL -#define HYPERVISOR_VIRT_END_X86_32 0x00000000FFFFFFFFULL - -static xen_pfn_t * -xc_core_arch_map_p2m_list_rw(xc_interface *xch, struct domain_info_context *dinfo, - uint32_t dom, shared_info_any_t *live_shinfo, - uint64_t p2m_cr3) -{ - uint64_t p2m_vaddr, p2m_end, mask, off; - xen_pfn_t p2m_mfn, mfn, saved_mfn, max_pfn; - uint64_t *ptes = NULL; - xen_pfn_t *mfns = NULL; - unsigned int fpp, n_pages, level, n_levels, shift, - idx_start, idx_end, idx, saved_idx; - - p2m_vaddr = GET_FIELD(live_shinfo, arch.p2m_vaddr, dinfo->guest_width); - fpp = PAGE_SIZE / dinfo->guest_width; - dinfo->p2m_frames = (dinfo->p2m_size - 1) / fpp + 1; - p2m_end = p2m_vaddr + dinfo->p2m_frames * PAGE_SIZE - 1; - - if ( dinfo->guest_width == 8 ) - { - mask = 0x0000ffffffffffffULL; - n_levels = 4; - p2m_mfn = p2m_cr3 >> 12; - if ( !is_canonical_address(p2m_vaddr) || - !is_canonical_address(p2m_end) || - p2m_end < p2m_vaddr || - (p2m_vaddr <= HYPERVISOR_VIRT_END_X86_64 && - p2m_end > HYPERVISOR_VIRT_START_X86_64) ) - { - ERROR("Bad virtual p2m address range %#" PRIx64 "-%#" PRIx64, - p2m_vaddr, p2m_end); - errno = ERANGE; - goto out; - } - } - else - { - mask = 0x00000000ffffffffULL; - n_levels = 3; - if ( p2m_cr3 & ~mask ) - p2m_mfn = ~0UL; - else - p2m_mfn = (uint32_t)((p2m_cr3 >> 12) | (p2m_cr3 << 20)); - if ( p2m_vaddr > mask || p2m_end > mask || p2m_end < p2m_vaddr || - (p2m_vaddr <= HYPERVISOR_VIRT_END_X86_32 && - p2m_end > HYPERVISOR_VIRT_START_X86_32) ) - { - ERROR("Bad virtual p2m address range %#" PRIx64 "-%#" PRIx64, - p2m_vaddr, p2m_end); - errno = ERANGE; - goto out; - } - } - - mfns = malloc(sizeof(*mfns)); - if ( !mfns ) - { - ERROR("Cannot allocate memory for array of %u mfns", 1); - goto out; - } - mfns[0] = p2m_mfn; - off = 0; - saved_mfn = 0; - idx_start = idx_end = saved_idx = 0; - - for ( level = n_levels; level > 0; level-- ) - { - n_pages = idx_end - idx_start + 1; - ptes = xc_map_foreign_pages(xch, dom, PROT_READ, mfns, n_pages); - if ( !ptes ) - { - PERROR("Failed to map %u page table pages for p2m list", n_pages); - goto out; - } - free(mfns); - - shift = level * 9 + 3; - idx_start = ((p2m_vaddr - off) & mask) >> shift; - idx_end = ((p2m_end - off) & mask) >> shift; - idx = idx_end - idx_start + 1; - mfns = malloc(sizeof(*mfns) * idx); - if ( !mfns ) - { - ERROR("Cannot allocate memory for array of %u mfns", idx); - goto out; - } - - for ( idx = idx_start; idx <= idx_end; idx++ ) - { - mfn = (ptes[idx] & 0x000ffffffffff000ULL) >> PAGE_SHIFT; - if ( mfn == 0 ) - { - ERROR("Bad mfn %#lx during page table walk for vaddr %#" PRIx64 " at level %d of p2m list", - mfn, off + ((uint64_t)idx << shift), level); - errno = ERANGE; - goto out; - } - mfns[idx - idx_start] = mfn; - - /* Maximum pfn check at level 2. Same reasoning as for p2m tree. */ - if ( level == 2 ) - { - if ( mfn != saved_mfn ) - { - saved_mfn = mfn; - saved_idx = idx - idx_start; - } - } - } - - if ( level == 2 ) - { - if ( saved_idx == idx_end ) - saved_idx++; - max_pfn = ((xen_pfn_t)saved_idx << 9) * fpp; - if ( max_pfn < dinfo->p2m_size ) - { - dinfo->p2m_size = max_pfn; - dinfo->p2m_frames = (dinfo->p2m_size + fpp - 1) / fpp; - p2m_end = p2m_vaddr + dinfo->p2m_frames * PAGE_SIZE - 1; - idx_end = idx_start + saved_idx; - } - } - - munmap(ptes, n_pages * PAGE_SIZE); - ptes = NULL; - off = p2m_vaddr & ((mask >> shift) << shift); - } - - return mfns; - - out: - free(mfns); - if ( ptes ) - munmap(ptes, n_pages * PAGE_SIZE); - - return NULL; -} - -static xen_pfn_t * -xc_core_arch_map_p2m_tree_rw(xc_interface *xch, struct domain_info_context *dinfo, - uint32_t dom, shared_info_any_t *live_shinfo) -{ - /* Double and single indirect references to the live P2M table */ - xen_pfn_t *live_p2m_frame_list_list; - xen_pfn_t *live_p2m_frame_list = NULL; - /* Copies of the above. */ - xen_pfn_t *p2m_frame_list_list = NULL; - xen_pfn_t *p2m_frame_list; - - int err; - int i; - - live_p2m_frame_list_list = - xc_map_foreign_range(xch, dom, PAGE_SIZE, PROT_READ, - GET_FIELD(live_shinfo, arch.pfn_to_mfn_frame_list_list, dinfo->guest_width)); - - if ( !live_p2m_frame_list_list ) - { - PERROR("Couldn't map p2m_frame_list_list (errno %d)", errno); - goto out; - } - - /* Get a local copy of the live_P2M_frame_list_list */ - if ( !(p2m_frame_list_list = malloc(PAGE_SIZE)) ) - { - ERROR("Couldn't allocate p2m_frame_list_list array"); - goto out; - } - memcpy(p2m_frame_list_list, live_p2m_frame_list_list, PAGE_SIZE); - - /* Canonicalize guest's unsigned long vs ours */ - if ( dinfo->guest_width > sizeof(unsigned long) ) - for ( i = 0; i < PAGE_SIZE/sizeof(unsigned long); i++ ) - if ( i < PAGE_SIZE/dinfo->guest_width ) - p2m_frame_list_list[i] = ((uint64_t *)p2m_frame_list_list)[i]; - else - p2m_frame_list_list[i] = 0; - else if ( dinfo->guest_width < sizeof(unsigned long) ) - for ( i = PAGE_SIZE/sizeof(unsigned long) - 1; i >= 0; i-- ) - p2m_frame_list_list[i] = ((uint32_t *)p2m_frame_list_list)[i]; - - live_p2m_frame_list = - xc_map_foreign_pages(xch, dom, PROT_READ, - p2m_frame_list_list, - P2M_FLL_ENTRIES); - - if ( !live_p2m_frame_list ) - { - PERROR("Couldn't map p2m_frame_list"); - goto out; - } - - /* Get a local copy of the live_P2M_frame_list */ - if ( !(p2m_frame_list = malloc(P2M_TOOLS_FL_SIZE)) ) - { - ERROR("Couldn't allocate p2m_frame_list array"); - goto out; - } - memset(p2m_frame_list, 0, P2M_TOOLS_FL_SIZE); - memcpy(p2m_frame_list, live_p2m_frame_list, P2M_GUEST_FL_SIZE); - - /* Canonicalize guest's unsigned long vs ours */ - if ( dinfo->guest_width > sizeof(unsigned long) ) - for ( i = 0; i < P2M_FL_ENTRIES; i++ ) - p2m_frame_list[i] = ((uint64_t *)p2m_frame_list)[i]; - else if ( dinfo->guest_width < sizeof(unsigned long) ) - for ( i = P2M_FL_ENTRIES - 1; i >= 0; i-- ) - p2m_frame_list[i] = ((uint32_t *)p2m_frame_list)[i]; - - dinfo->p2m_frames = P2M_FL_ENTRIES; - - return p2m_frame_list; - - out: - err = errno; - - if ( live_p2m_frame_list_list ) - munmap(live_p2m_frame_list_list, PAGE_SIZE); - - if ( live_p2m_frame_list ) - munmap(live_p2m_frame_list, P2M_FLL_ENTRIES * PAGE_SIZE); - - free(p2m_frame_list_list); - - errno = err; - - return NULL; -} - -static int -xc_core_arch_map_p2m_rw(xc_interface *xch, struct domain_info_context *dinfo, xc_dominfo_t *info, - shared_info_any_t *live_shinfo, xen_pfn_t **live_p2m, int rw) -{ - xen_pfn_t *p2m_frame_list = NULL; - uint64_t p2m_cr3; - uint32_t dom = info->domid; - int ret = -1; - int err; - - if ( xc_domain_nr_gpfns(xch, info->domid, &dinfo->p2m_size) < 0 ) - { - ERROR("Could not get maximum GPFN!"); - goto out; - } - - if ( dinfo->p2m_size < info->nr_pages ) - { - ERROR("p2m_size < nr_pages -1 (%lx < %lx", dinfo->p2m_size, info->nr_pages - 1); - goto out; - } - - p2m_cr3 = GET_FIELD(live_shinfo, arch.p2m_cr3, dinfo->guest_width); - - p2m_frame_list = p2m_cr3 ? xc_core_arch_map_p2m_list_rw(xch, dinfo, dom, live_shinfo, p2m_cr3) - : xc_core_arch_map_p2m_tree_rw(xch, dinfo, dom, live_shinfo); - - if ( !p2m_frame_list ) - goto out; - - *live_p2m = xc_map_foreign_pages(xch, dom, - rw ? (PROT_READ | PROT_WRITE) : PROT_READ, - p2m_frame_list, - dinfo->p2m_frames); - - if ( !*live_p2m ) - { - PERROR("Couldn't map p2m table"); - goto out; - } - - ret = 0; - -out: - err = errno; - - free(p2m_frame_list); - - errno = err; - return ret; -} - -int -xc_core_arch_map_p2m(xc_interface *xch, struct domain_info_context *dinfo, xc_dominfo_t *info, - shared_info_any_t *live_shinfo, xen_pfn_t **live_p2m) -{ - return xc_core_arch_map_p2m_rw(xch, dinfo, info, live_shinfo, live_p2m, 0); -} - -int -xc_core_arch_map_p2m_writable(xc_interface *xch, struct domain_info_context *dinfo, xc_dominfo_t *info, - shared_info_any_t *live_shinfo, xen_pfn_t **live_p2m) -{ - return xc_core_arch_map_p2m_rw(xch, dinfo, info, live_shinfo, live_p2m, 1); -} - -int -xc_core_arch_get_scratch_gpfn(xc_interface *xch, uint32_t domid, - xen_pfn_t *gpfn) -{ - return xc_domain_nr_gpfns(xch, domid, gpfn); -} - -/* - * Local variables: - * mode: C - * c-file-style: "BSD" - * c-basic-offset: 4 - * tab-width: 4 - * indent-tabs-mode: nil - * End: - */ diff --git a/tools/libs/ctrl/xc_core_x86.h b/tools/libs/ctrl/xc_core_x86.h deleted file mode 100644 index 867146b1d9..0000000000 --- a/tools/libs/ctrl/xc_core_x86.h +++ /dev/null @@ -1,60 +0,0 @@ -/* - * This library is free software; you can redistribute it and/or - * modify it under the terms of the GNU Lesser General Public - * License as published by the Free Software Foundation; either - * version 2.1 of the License, or (at your option) any later version. - * - * This library is distributed in the hope that it will be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU - * Lesser General Public License for more details. - * - * You should have received a copy of the GNU Lesser General Public - * License along with this library; If not, see . - * - * Copyright (c) 2007 Isaku Yamahata - * VA Linux Systems Japan K.K. - * - */ - -#ifndef XC_CORE_X86_H -#define XC_CORE_X86_H - -#define ELF_ARCH_DATA ELFDATA2LSB -#define ELF_ARCH_MACHINE (dinfo->guest_width == 8 ? EM_X86_64 : EM_386) - -struct xc_core_arch_context { - /* nothing */ -}; - -#define xc_core_arch_context_init(arch_ctxt) do {} while (0) -#define xc_core_arch_context_free(arch_ctxt) do {} while (0) -#define xc_core_arch_context_get(arch_ctxt, ctxt, xch, domid) \ - (0) -#define xc_core_arch_context_dump(xch, arch_ctxt, args, dump_rtn) (0) - -int -xc_core_arch_gpfn_may_present(struct xc_core_arch_context *arch_ctxt, - unsigned long pfn); -static inline int -xc_core_arch_context_get_shdr(xc_interface *xch, - struct xc_core_arch_context *arch_ctxt, - struct xc_core_section_headers *sheaders, - struct xc_core_strtab *strtab, - uint64_t *filesz, uint64_t offset) -{ - *filesz = 0; - return 0; -} - -#endif /* XC_CORE_X86_H */ - -/* - * Local variables: - * mode: C - * c-file-style: "BSD" - * c-basic-offset: 4 - * tab-width: 4 - * indent-tabs-mode: nil - * End: - */ diff --git a/tools/libs/ctrl/xc_domain.c b/tools/libs/ctrl/xc_domain.c index e7cea4a17d..7d118848f1 100644 --- a/tools/libs/ctrl/xc_domain.c +++ b/tools/libs/ctrl/xc_domain.c @@ -19,8 +19,6 @@ * Copyright (c) 2003, K A Fraser. */ -#include "xc_private.h" -#include "xc_core.h" #include "xc_private.h" #include #include diff --git a/tools/libs/ctrl/xc_private.h b/tools/libs/ctrl/xc_private.h index 8ebc0b59da..dff0f0289b 100644 --- a/tools/libs/ctrl/xc_private.h +++ b/tools/libs/ctrl/xc_private.h @@ -467,6 +467,18 @@ void *xc_vm_event_enable(xc_interface *xch, uint32_t domain_id, int param, int do_dm_op(xc_interface *xch, uint32_t domid, unsigned int nr_bufs, ...); +#if defined (__i386__) || defined (__x86_64__) +static inline int xc_core_arch_auto_translated_physmap(const xc_dominfo_t *info) +{ + return info->hvm; +} +#elif defined (__arm__) || defined(__aarch64__) +static inline int xc_core_arch_auto_translated_physmap(const xc_dominfo_t *info) +{ + return 1; +} +#endif + #endif /* __XC_PRIVATE_H__ */ /* diff --git a/tools/libs/guest/Makefile b/tools/libs/guest/Makefile index 2a2323ff09..2ce92d247e 100644 --- a/tools/libs/guest/Makefile +++ b/tools/libs/guest/Makefile @@ -24,6 +24,9 @@ SRCS-y += xg_offline_page.c else SRCS-y += xg_nomigrate.c endif +SRCS-y += xg_core.c +SRCS-$(CONFIG_X86) += xg_core_x86.c +SRCS-$(CONFIG_ARM) += xg_core_arm.c CFLAGS += -I$(XEN_libxenctrl) diff --git a/tools/libs/guest/xg_core.c b/tools/libs/guest/xg_core.c new file mode 100644 index 0000000000..c52f1161c1 --- /dev/null +++ b/tools/libs/guest/xg_core.c @@ -0,0 +1,1027 @@ +/* + * Elf format, (pfn, gmfn) table, IA64 support. + * Copyright (c) 2007 Isaku Yamahata + * VA Linux Systems Japan K.K. + * + * This library is free software; you can redistribute it and/or + * modify it under the terms of the GNU Lesser General Public + * License as published by the Free Software Foundation; + * version 2.1 of the License. + * + * This library is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public + * License along with this library; If not, see . + */ + +/* + * xen dump-core file format follows ELF format specification. + * Analisys tools shouldn't depends on the order of sections. + * They should follow elf header and check section names. + * + * +--------------------------------------------------------+ + * |ELF header | + * +--------------------------------------------------------+ + * |section headers | + * | null section header | + * | .shstrtab | + * | .note.Xen | + * | .xen_prstatus | + * | .xen_shared_info if present | + * | .xen_pages | + * | .xen_p2m or .xen_pfn | + * +--------------------------------------------------------+ + * |.note.Xen:note section | + * | "Xen" is used as note name, | + * | types are defined in xen/include/public/elfnote.h | + * | and descriptors are defined in xc_core.h. | + * | dumpcore none | + * | dumpcore header | + * | dumpcore xen version | + * | dumpcore format version | + * +--------------------------------------------------------+ + * |.xen_prstatus | + * | vcpu_guest_context_t[nr_vcpus] | + * +--------------------------------------------------------+ + * |.xen_shared_info if possible | + * +--------------------------------------------------------+ + * |.xen_pages | + * | page * nr_pages | + * +--------------------------------------------------------+ + * |.xen_p2m or .xen_pfn | + * | .xen_p2m: struct xen_dumpcore_p2m[nr_pages] | + * | .xen_pfn: uint64_t[nr_pages] | + * +--------------------------------------------------------+ + * |.shstrtab: section header string table | + * +--------------------------------------------------------+ + * + */ + +#include "xc_private.h" +#include "xg_core.h" +#include +#include + +#include + +/* number of pages to write at a time */ +#define DUMP_INCREMENT (4 * 1024) + +/* string table */ +struct xc_core_strtab { + char *strings; + uint16_t length; + uint16_t max; +}; + +static struct xc_core_strtab* +xc_core_strtab_init(xc_interface *xch) +{ + struct xc_core_strtab *strtab; + char *strings; + strtab = malloc(sizeof(*strtab)); + if ( strtab == NULL ) + return NULL; + + strings = malloc(PAGE_SIZE); + if ( strings == NULL ) + { + PERROR("Could not allocate string table init"); + free(strtab); + return NULL; + } + strtab->strings = strings; + strtab->max = PAGE_SIZE; + + /* index 0 represents none */ + strtab->strings[0] = '\0'; + strtab->length = 1; + + return strtab; +} + +static void +xc_core_strtab_free(struct xc_core_strtab *strtab) +{ + free(strtab->strings); + free(strtab); +} + +static uint16_t +xc_core_strtab_get(xc_interface *xch, struct xc_core_strtab *strtab, const char *name) +{ + uint16_t ret = 0; + uint16_t len = strlen(name) + 1; + + if ( strtab->length > UINT16_MAX - len ) + { + PERROR("too long string table"); + errno = E2BIG; + return ret; + } + + if ( strtab->length + len > strtab->max ) + { + char *tmp; + if ( strtab->max > UINT16_MAX / 2 ) + { + PERROR("too long string table"); + errno = ENOMEM; + return ret; + } + + tmp = realloc(strtab->strings, strtab->max * 2); + if ( tmp == NULL ) + { + PERROR("Could not allocate string table"); + return ret; + } + + strtab->strings = tmp; + strtab->max *= 2; + } + + ret = strtab->length; + strcpy(strtab->strings + strtab->length, name); + strtab->length += len; + return ret; +} + + +/* section headers */ +struct xc_core_section_headers { + uint16_t num; + uint16_t num_max; + + Elf64_Shdr *shdrs; +}; +#define SHDR_INIT ((uint16_t)16) +#define SHDR_INC ((uint16_t)4) + +static struct xc_core_section_headers* +xc_core_shdr_init(xc_interface *xch) +{ + struct xc_core_section_headers *sheaders; + sheaders = malloc(sizeof(*sheaders)); + if ( sheaders == NULL ) + return NULL; + + sheaders->num = 0; + sheaders->num_max = SHDR_INIT; + sheaders->shdrs = malloc(sizeof(sheaders->shdrs[0]) * sheaders->num_max); + if ( sheaders->shdrs == NULL ) + { + free(sheaders); + return NULL; + } + return sheaders; +} + +static void +xc_core_shdr_free(struct xc_core_section_headers *sheaders) +{ + free(sheaders->shdrs); + free(sheaders); +} + +Elf64_Shdr* +xc_core_shdr_get(xc_interface *xch, + struct xc_core_section_headers *sheaders) +{ + Elf64_Shdr *shdr; + + if ( sheaders->num == sheaders->num_max ) + { + Elf64_Shdr *shdrs; + if ( sheaders->num_max > UINT16_MAX - SHDR_INC ) + { + errno = E2BIG; + return NULL; + } + sheaders->num_max += SHDR_INC; + shdrs = realloc(sheaders->shdrs, + sizeof(sheaders->shdrs[0]) * sheaders->num_max); + if ( shdrs == NULL ) + return NULL; + sheaders->shdrs = shdrs; + } + + shdr = &sheaders->shdrs[sheaders->num]; + sheaders->num++; + memset(shdr, 0, sizeof(*shdr)); + return shdr; +} + +int +xc_core_shdr_set(xc_interface *xch, + Elf64_Shdr *shdr, + struct xc_core_strtab *strtab, + const char *name, uint32_t type, + uint64_t offset, uint64_t size, + uint64_t addralign, uint64_t entsize) +{ + uint64_t name_idx = xc_core_strtab_get(xch, strtab, name); + if ( name_idx == 0 ) + return -1; + + shdr->sh_name = name_idx; + shdr->sh_type = type; + shdr->sh_offset = offset; + shdr->sh_size = size; + shdr->sh_addralign = addralign; + shdr->sh_entsize = entsize; + return 0; +} + +static void +xc_core_ehdr_init(Elf64_Ehdr *ehdr) +{ + memset(ehdr, 0, sizeof(*ehdr)); + ehdr->e_ident[EI_MAG0] = ELFMAG0; + ehdr->e_ident[EI_MAG1] = ELFMAG1; + ehdr->e_ident[EI_MAG2] = ELFMAG2; + ehdr->e_ident[EI_MAG3] = ELFMAG3; + ehdr->e_ident[EI_CLASS] = ELFCLASS64; + ehdr->e_ident[EI_DATA] = ELF_ARCH_DATA; + ehdr->e_ident[EI_VERSION] = EV_CURRENT; + ehdr->e_ident[EI_OSABI] = ELFOSABI_SYSV; + ehdr->e_ident[EI_ABIVERSION] = EV_CURRENT; + + ehdr->e_type = ET_CORE; + /* e_machine will be filled in later */ + ehdr->e_version = EV_CURRENT; + ehdr->e_entry = 0; + ehdr->e_phoff = 0; + ehdr->e_shoff = sizeof(*ehdr); + ehdr->e_flags = ELF_CORE_EFLAGS; + ehdr->e_ehsize = sizeof(*ehdr); + ehdr->e_phentsize = sizeof(Elf64_Phdr); + ehdr->e_phnum = 0; + ehdr->e_shentsize = sizeof(Elf64_Shdr); + /* ehdr->e_shnum and ehdr->e_shstrndx aren't known here yet. + * fill it later */ +} + +static int +elfnote_fill_xen_version(xc_interface *xch, + struct xen_dumpcore_elfnote_xen_version_desc + *xen_version) +{ + int rc; + memset(xen_version, 0, sizeof(*xen_version)); + + rc = xc_version(xch, XENVER_version, NULL); + if ( rc < 0 ) + return rc; + xen_version->major_version = rc >> 16; + xen_version->minor_version = rc & ((1 << 16) - 1); + + rc = xc_version(xch, XENVER_extraversion, + &xen_version->extra_version); + if ( rc < 0 ) + return rc; + + rc = xc_version(xch, XENVER_compile_info, + &xen_version->compile_info); + if ( rc < 0 ) + return rc; + + rc = xc_version(xch, + XENVER_capabilities, &xen_version->capabilities); + if ( rc < 0 ) + return rc; + + rc = xc_version(xch, XENVER_changeset, &xen_version->changeset); + if ( rc < 0 ) + return rc; + + rc = xc_version(xch, XENVER_platform_parameters, + &xen_version->platform_parameters); + if ( rc < 0 ) + return rc; + + rc = xc_version(xch, XENVER_pagesize, NULL); + if ( rc < 0 ) + return rc; + xen_version->pagesize = rc; + + return 0; +} + +static void +elfnote_fill_format_version(struct xen_dumpcore_elfnote_format_version_desc + *format_version) +{ + format_version->version = XEN_DUMPCORE_FORMAT_VERSION_CURRENT; +} + +static void +elfnote_init(struct elfnote *elfnote) +{ + /* elf note section */ + memset(elfnote, 0, sizeof(*elfnote)); + elfnote->namesz = strlen(XEN_DUMPCORE_ELFNOTE_NAME) + 1; + strncpy(elfnote->name, XEN_DUMPCORE_ELFNOTE_NAME, sizeof(elfnote->name)); +} + +static int +elfnote_dump_none(xc_interface *xch, void *args, dumpcore_rtn_t dump_rtn) +{ + int sts; + struct elfnote elfnote; + struct xen_dumpcore_elfnote_none_desc none; + + elfnote_init(&elfnote); + /* Avoid compile warning about constant-zero-sized memset(). */ + /*memset(&none, 0, sizeof(none));*/ + + elfnote.descsz = sizeof(none); + elfnote.type = XEN_ELFNOTE_DUMPCORE_NONE; + sts = dump_rtn(xch, args, (char*)&elfnote, sizeof(elfnote)); + if ( sts != 0 ) + return sts; + return dump_rtn(xch, args, (char*)&none, sizeof(none)); +} + +static int +elfnote_dump_core_header( + xc_interface *xch, + void *args, dumpcore_rtn_t dump_rtn, const xc_dominfo_t *info, + int nr_vcpus, unsigned long nr_pages) +{ + int sts; + struct elfnote elfnote; + struct xen_dumpcore_elfnote_header_desc header; + + elfnote_init(&elfnote); + memset(&header, 0, sizeof(header)); + + elfnote.descsz = sizeof(header); + elfnote.type = XEN_ELFNOTE_DUMPCORE_HEADER; + header.xch_magic = info->hvm ? XC_CORE_MAGIC_HVM : XC_CORE_MAGIC; + header.xch_nr_vcpus = nr_vcpus; + header.xch_nr_pages = nr_pages; + header.xch_page_size = PAGE_SIZE; + sts = dump_rtn(xch, args, (char*)&elfnote, sizeof(elfnote)); + if ( sts != 0 ) + return sts; + return dump_rtn(xch, args, (char*)&header, sizeof(header)); +} + +static int +elfnote_dump_xen_version(xc_interface *xch, void *args, + dumpcore_rtn_t dump_rtn, unsigned int guest_width) +{ + int sts; + struct elfnote elfnote; + struct xen_dumpcore_elfnote_xen_version_desc xen_version; + + elfnote_init(&elfnote); + memset(&xen_version, 0, sizeof(xen_version)); + + elfnote.descsz = sizeof(xen_version); + elfnote.type = XEN_ELFNOTE_DUMPCORE_XEN_VERSION; + elfnote_fill_xen_version(xch, &xen_version); + if (guest_width < sizeof(unsigned long)) + { + // 32 bit elf file format differs in pagesize's alignment + char *p = (char *)&xen_version.pagesize; + memmove(p - 4, p, sizeof(xen_version.pagesize)); + } + sts = dump_rtn(xch, args, (char*)&elfnote, sizeof(elfnote)); + if ( sts != 0 ) + return sts; + return dump_rtn(xch, args, (char*)&xen_version, sizeof(xen_version)); +} + +static int +elfnote_dump_format_version(xc_interface *xch, + void *args, dumpcore_rtn_t dump_rtn) +{ + int sts; + struct elfnote elfnote; + struct xen_dumpcore_elfnote_format_version_desc format_version; + + elfnote_init(&elfnote); + memset(&format_version, 0, sizeof(format_version)); + + elfnote.descsz = sizeof(format_version); + elfnote.type = XEN_ELFNOTE_DUMPCORE_FORMAT_VERSION; + elfnote_fill_format_version(&format_version); + sts = dump_rtn(xch, args, (char*)&elfnote, sizeof(elfnote)); + if ( sts != 0 ) + return sts; + return dump_rtn(xch, args, (char*)&format_version, sizeof(format_version)); +} + +int +xc_domain_dumpcore_via_callback(xc_interface *xch, + uint32_t domid, + void *args, + dumpcore_rtn_t dump_rtn) +{ + xc_dominfo_t info; + shared_info_any_t *live_shinfo = NULL; + struct domain_info_context _dinfo = {}; + struct domain_info_context *dinfo = &_dinfo; + + int nr_vcpus = 0; + char *dump_mem, *dump_mem_start = NULL; + vcpu_guest_context_any_t *ctxt = NULL; + struct xc_core_arch_context arch_ctxt; + char dummy[PAGE_SIZE]; + int dummy_len; + int sts = -1; + + unsigned long i; + unsigned long j; + unsigned long nr_pages; + unsigned long max_mfn; + + xc_core_memory_map_t *memory_map = NULL; + unsigned int nr_memory_map; + unsigned int map_idx; + + int auto_translated_physmap; + xen_pfn_t *p2m = NULL; + struct xen_dumpcore_p2m *p2m_array = NULL; + + uint64_t *pfn_array = NULL; + + Elf64_Ehdr ehdr; + uint64_t filesz; + uint64_t offset; + uint64_t fixup; + + struct xc_core_strtab *strtab = NULL; + uint16_t strtab_idx; + struct xc_core_section_headers *sheaders = NULL; + Elf64_Shdr *shdr; + + xc_core_arch_context_init(&arch_ctxt); + if ( (dump_mem_start = malloc(DUMP_INCREMENT*PAGE_SIZE)) == NULL ) + { + PERROR("Could not allocate dump_mem"); + goto out; + } + + if ( xc_domain_getinfo(xch, domid, 1, &info) != 1 ) + { + PERROR("Could not get info for domain"); + goto out; + } + /* Map the shared info frame */ + live_shinfo = xc_map_foreign_range(xch, domid, PAGE_SIZE, + PROT_READ, info.shared_info_frame); + if ( !live_shinfo && !info.hvm ) + { + PERROR("Couldn't map live_shinfo"); + goto out; + } + auto_translated_physmap = xc_core_arch_auto_translated_physmap(&info); + + if ( !auto_translated_physmap ) + + { + if ( xc_domain_get_guest_width(xch, domid, &dinfo->guest_width) != 0 ) + { + PERROR("Could not get address size for domain"); + goto out; + } + } + else + { + /* + * Autotranslated guest never sets guest width in the first + * place. Force guest_width to be sizeof(unsigned long) so + * code below functions properly. + * + * Here is why this is correct. + * + * 1. Before f969bc9fc, xc_domain_get_guest_width for HVM (x86 + * and ARM) always returned hypervisor's idea of + * sizeof(unsigned long). + * + * 2. There has never been a situation in which hypervisor's + * word width is smaller than toolstack domain's (i.e. no + * 32bit hypervisor + 64bit toolstack). + * + * Predicates in code test guest_width against toolstack + * domain's sizeof(unsigned long), so setting guest_width to + * toolstack domain's idea of sizeof(unsigned long) matches + * the original behaviour for HVM guests. + */ + dinfo->guest_width = sizeof(unsigned long); + } + + if ( domid != info.domid ) + { + PERROR("Domain %d does not exist", domid); + goto out; + } + + ctxt = calloc(sizeof(*ctxt), info.max_vcpu_id + 1); + if ( !ctxt ) + { + PERROR("Could not allocate vcpu context array"); + goto out; + } + + for ( i = 0; i <= info.max_vcpu_id; i++ ) + { + if ( xc_vcpu_getcontext(xch, domid, i, &ctxt[nr_vcpus]) == 0 ) + { + if ( xc_core_arch_context_get(&arch_ctxt, &ctxt[nr_vcpus], + xch, domid) ) + continue; + nr_vcpus++; + } + } + if ( nr_vcpus == 0 ) + { + PERROR("No VCPU context could be grabbed"); + goto out; + } + + /* obtain memory map */ + sts = xc_core_arch_memory_map_get(xch, &arch_ctxt, &info, + live_shinfo, &memory_map, + &nr_memory_map); + if ( sts != 0 ) + goto out; + + /* + * Note: this is the *current* number of pages and may change under + * a live dump-core. We'll just take this value, and if more pages + * exist, we'll skip them. If there's less, then we'll just not use + * all the array... + * + * We don't want to use the total potential size of the memory map + * since that is usually much higher than info.nr_pages. + */ + nr_pages = info.nr_pages; + + if ( !auto_translated_physmap ) + { + /* obtain p2m table */ + p2m_array = malloc(nr_pages * sizeof(p2m_array[0])); + if ( p2m_array == NULL ) + { + PERROR("Could not allocate p2m array"); + goto out; + } + + sts = xc_core_arch_map_p2m(xch, dinfo, &info, live_shinfo, &p2m); + if ( sts != 0 ) + goto out; + + sts = xc_maximum_ram_page(xch, &max_mfn); + if ( sts != 0 ) + goto out; + } + else + { + pfn_array = malloc(nr_pages * sizeof(pfn_array[0])); + if ( pfn_array == NULL ) + { + PERROR("Could not allocate pfn array"); + goto out; + } + } + + /* ehdr.e_shnum and ehdr.e_shstrndx aren't known here yet. fill it later*/ + xc_core_ehdr_init(&ehdr); + + /* create section header */ + strtab = xc_core_strtab_init(xch); + if ( strtab == NULL ) + { + PERROR("Could not allocate string table"); + goto out; + } + sheaders = xc_core_shdr_init(xch); + if ( sheaders == NULL ) + { + PERROR("Could not allocate section headers"); + goto out; + } + /* null section */ + shdr = xc_core_shdr_get(xch,sheaders); + if ( shdr == NULL ) + { + PERROR("Could not get section header for null section"); + goto out; + } + + /* .shstrtab */ + shdr = xc_core_shdr_get(xch,sheaders); + if ( shdr == NULL ) + { + PERROR("Could not get section header for shstrtab"); + goto out; + } + strtab_idx = shdr - sheaders->shdrs; + /* strtab_shdr.sh_offset, strtab_shdr.sh_size aren't unknown. + * fill it later + */ + sts = xc_core_shdr_set(xch, shdr, strtab, ELF_SHSTRTAB, SHT_STRTAB, 0, 0, 0, 0); + if ( sts != 0 ) + goto out; + + /* elf note section */ + /* here the number of section header is unknown. fix up offset later. */ + offset = sizeof(ehdr); + filesz = + sizeof(struct xen_dumpcore_elfnote_none) + /* none */ + sizeof(struct xen_dumpcore_elfnote_header) + /* core header */ + sizeof(struct xen_dumpcore_elfnote_xen_version) + /* xen version */ + sizeof(struct xen_dumpcore_elfnote_format_version);/* format version */ + shdr = xc_core_shdr_get(xch,sheaders); + if ( shdr == NULL ) + { + PERROR("Could not get section header for note section"); + goto out; + } + sts = xc_core_shdr_set(xch, shdr, strtab, XEN_DUMPCORE_SEC_NOTE, SHT_NOTE, + offset, filesz, 0, 0); + if ( sts != 0 ) + goto out; + offset += filesz; + + /* prstatus */ + shdr = xc_core_shdr_get(xch,sheaders); + if ( shdr == NULL ) + { + PERROR("Could not get section header for .xen_prstatus"); + goto out; + } + filesz = sizeof(*ctxt) * nr_vcpus; + sts = xc_core_shdr_set(xch, shdr, strtab, XEN_DUMPCORE_SEC_PRSTATUS, + SHT_PROGBITS, offset, filesz, + __alignof__(*ctxt), sizeof(*ctxt)); + if ( sts != 0 ) + goto out; + offset += filesz; + + /* arch context */ + sts = xc_core_arch_context_get_shdr(xch, &arch_ctxt, sheaders, strtab, + &filesz, offset); + if ( sts != 0 ) + goto out; + offset += filesz; + + /* shared_info */ + if ( live_shinfo != NULL ) + { + shdr = xc_core_shdr_get(xch,sheaders); + if ( shdr == NULL ) + { + PERROR("Could not get section header for .xen_shared_info"); + goto out; + } + filesz = PAGE_SIZE; + sts = xc_core_shdr_set(xch, shdr, strtab, XEN_DUMPCORE_SEC_SHARED_INFO, + SHT_PROGBITS, offset, filesz, + __alignof__(*live_shinfo), PAGE_SIZE); + if ( sts != 0 ) + goto out; + offset += filesz; + } + + /* + * pages and p2m/pfn are the last section to allocate section headers + * so that we know the number of section headers here. + * 2 = pages section and p2m/pfn table section + */ + fixup = (sheaders->num + 2) * sizeof(*shdr); + /* zeroth section should have zero offset */ + for ( i = 1; i < sheaders->num; i++ ) + sheaders->shdrs[i].sh_offset += fixup; + offset += fixup; + dummy_len = ROUNDUP(offset, PAGE_SHIFT) - offset; /* padding length */ + offset += dummy_len; + + /* pages */ + shdr = xc_core_shdr_get(xch,sheaders); + if ( shdr == NULL ) + { + PERROR("could not get section headers for .xen_pages"); + goto out; + } + filesz = (uint64_t)nr_pages * PAGE_SIZE; + sts = xc_core_shdr_set(xch, shdr, strtab, XEN_DUMPCORE_SEC_PAGES, SHT_PROGBITS, + offset, filesz, PAGE_SIZE, PAGE_SIZE); + if ( sts != 0 ) + goto out; + offset += filesz; + + /* p2m/pfn table */ + shdr = xc_core_shdr_get(xch,sheaders); + if ( shdr == NULL ) + { + PERROR("Could not get section header for .xen_{p2m, pfn} table"); + goto out; + } + if ( !auto_translated_physmap ) + { + filesz = (uint64_t)nr_pages * sizeof(p2m_array[0]); + sts = xc_core_shdr_set(xch, shdr, strtab, XEN_DUMPCORE_SEC_P2M, + SHT_PROGBITS, + offset, filesz, __alignof__(p2m_array[0]), + sizeof(p2m_array[0])); + } + else + { + filesz = (uint64_t)nr_pages * sizeof(pfn_array[0]); + sts = xc_core_shdr_set(xch, shdr, strtab, XEN_DUMPCORE_SEC_PFN, + SHT_PROGBITS, + offset, filesz, __alignof__(pfn_array[0]), + sizeof(pfn_array[0])); + } + if ( sts != 0 ) + goto out; + offset += filesz; + + /* fixing up section header string table section header */ + filesz = strtab->length; + sheaders->shdrs[strtab_idx].sh_offset = offset; + sheaders->shdrs[strtab_idx].sh_size = filesz; + + /* write out elf header */ + ehdr.e_shnum = sheaders->num; + ehdr.e_shstrndx = strtab_idx; + ehdr.e_machine = ELF_ARCH_MACHINE; + sts = dump_rtn(xch, args, (char*)&ehdr, sizeof(ehdr)); + if ( sts != 0 ) + goto out; + + /* section headers */ + sts = dump_rtn(xch, args, (char*)sheaders->shdrs, + sheaders->num * sizeof(sheaders->shdrs[0])); + if ( sts != 0 ) + goto out; + + /* elf note section: xen core header */ + sts = elfnote_dump_none(xch, args, dump_rtn); + if ( sts != 0 ) + goto out; + + /* elf note section: xen core header */ + sts = elfnote_dump_core_header(xch, args, dump_rtn, &info, nr_vcpus, nr_pages); + if ( sts != 0 ) + goto out; + + /* elf note section: xen version */ + sts = elfnote_dump_xen_version(xch, args, dump_rtn, dinfo->guest_width); + if ( sts != 0 ) + goto out; + + /* elf note section: format version */ + sts = elfnote_dump_format_version(xch, args, dump_rtn); + if ( sts != 0 ) + goto out; + + /* prstatus: .xen_prstatus */ + sts = dump_rtn(xch, args, (char *)ctxt, sizeof(*ctxt) * nr_vcpus); + if ( sts != 0 ) + goto out; + + if ( live_shinfo != NULL ) + { + /* shared_info: .xen_shared_info */ + sts = dump_rtn(xch, args, (char*)live_shinfo, PAGE_SIZE); + if ( sts != 0 ) + goto out; + } + + /* arch specific context */ + sts = xc_core_arch_context_dump(xch, &arch_ctxt, args, dump_rtn); + if ( sts != 0 ) + goto out; + + /* Pad the output data to page alignment. */ + memset(dummy, 0, PAGE_SIZE); + sts = dump_rtn(xch, args, dummy, dummy_len); + if ( sts != 0 ) + goto out; + + /* dump pages: .xen_pages */ + j = 0; + dump_mem = dump_mem_start; + for ( map_idx = 0; map_idx < nr_memory_map; map_idx++ ) + { + uint64_t pfn_start; + uint64_t pfn_end; + + pfn_start = memory_map[map_idx].addr >> PAGE_SHIFT; + pfn_end = pfn_start + (memory_map[map_idx].size >> PAGE_SHIFT); + for ( i = pfn_start; i < pfn_end; i++ ) + { + uint64_t gmfn; + void *vaddr; + + if ( !auto_translated_physmap ) + { + if ( i >= dinfo->p2m_size ) + break; + + if ( dinfo->guest_width >= sizeof(unsigned long) ) + { + if ( dinfo->guest_width == sizeof(unsigned long) ) + gmfn = p2m[i]; + else + gmfn = ((uint64_t *)p2m)[i]; + if ( gmfn == INVALID_PFN ) + continue; + } + else + { + gmfn = ((uint32_t *)p2m)[i]; + if ( gmfn == (uint32_t)INVALID_PFN ) + continue; + } + if ( gmfn > max_mfn ) + continue; + + if ( j >= nr_pages ) + { + j++; + continue; + } + + p2m_array[j].pfn = i; + p2m_array[j].gmfn = gmfn; + } + else + { + if ( !xc_core_arch_gpfn_may_present(&arch_ctxt, i) ) + continue; + + if ( j >= nr_pages ) + { + j++; + continue; + } + + gmfn = i; + pfn_array[j] = i; + } + + vaddr = xc_map_foreign_range( + xch, domid, PAGE_SIZE, PROT_READ, gmfn); + if ( vaddr == NULL ) + continue; + memcpy(dump_mem, vaddr, PAGE_SIZE); + munmap(vaddr, PAGE_SIZE); + dump_mem += PAGE_SIZE; + if ( (j + 1) % DUMP_INCREMENT == 0 ) + { + sts = dump_rtn( + xch, args, dump_mem_start, dump_mem - dump_mem_start); + if ( sts != 0 ) + goto out; + dump_mem = dump_mem_start; + } + + j++; + } + } + + if ( j > nr_pages ) + { + /* + * When live dump-mode (-L option) is specified, + * guest domain may increase memory. + */ + IPRINTF("exceeded nr_pages (%ld) losing %ld pages", nr_pages, j - nr_pages); + } + + sts = dump_rtn(xch, args, dump_mem_start, dump_mem - dump_mem_start); + if ( sts != 0 ) + goto out; + if ( j < nr_pages ) + { + /* When live dump-mode (-L option) is specified, + * guest domain may reduce memory. pad with zero pages. + */ + DPRINTF("j (%ld) != nr_pages (%ld)", j, nr_pages); + memset(dump_mem_start, 0, PAGE_SIZE); + for (; j < nr_pages; j++) { + sts = dump_rtn(xch, args, dump_mem_start, PAGE_SIZE); + if ( sts != 0 ) + goto out; + if ( !auto_translated_physmap ) + { + p2m_array[j].pfn = XC_CORE_INVALID_PFN; + p2m_array[j].gmfn = XC_CORE_INVALID_GMFN; + } + else + pfn_array[j] = XC_CORE_INVALID_PFN; + } + } + + /* p2m/pfn table: .xen_p2m/.xen_pfn */ + if ( !auto_translated_physmap ) + sts = dump_rtn( + xch, args, (char *)p2m_array, sizeof(p2m_array[0]) * nr_pages); + else + sts = dump_rtn( + xch, args, (char *)pfn_array, sizeof(pfn_array[0]) * nr_pages); + if ( sts != 0 ) + goto out; + + /* elf section header string table: .shstrtab */ + sts = dump_rtn(xch, args, strtab->strings, strtab->length); + if ( sts != 0 ) + goto out; + + sts = 0; + +out: + if ( memory_map != NULL ) + free(memory_map); + if ( p2m != NULL ) + munmap(p2m, PAGE_SIZE * dinfo->p2m_frames); + if ( p2m_array != NULL ) + free(p2m_array); + if ( pfn_array != NULL ) + free(pfn_array); + if ( sheaders != NULL ) + xc_core_shdr_free(sheaders); + if ( strtab != NULL ) + xc_core_strtab_free(strtab); + if ( ctxt != NULL ) + free(ctxt); + if ( dump_mem_start != NULL ) + free(dump_mem_start); + if ( live_shinfo != NULL ) + munmap(live_shinfo, PAGE_SIZE); + xc_core_arch_context_free(&arch_ctxt); + + return sts; +} + +/* Callback args for writing to a local dump file. */ +struct dump_args { + int fd; +}; + +/* Callback routine for writing to a local dump file. */ +static int local_file_dump(xc_interface *xch, + void *args, char *buffer, unsigned int length) +{ + struct dump_args *da = args; + + if ( write_exact(da->fd, buffer, length) == -1 ) + { + PERROR("Failed to write buffer"); + return -errno; + } + + if ( length >= (DUMP_INCREMENT * PAGE_SIZE) ) + { + // Now dumping pages -- make sure we discard clean pages from + // the cache after each write + discard_file_cache(xch, da->fd, 0 /* no flush */); + } + + return 0; +} + +int +xc_domain_dumpcore(xc_interface *xch, + uint32_t domid, + const char *corename) +{ + struct dump_args da; + int sts; + + if ( (da.fd = open(corename, O_CREAT|O_RDWR|O_TRUNC, S_IWUSR|S_IRUSR)) < 0 ) + { + PERROR("Could not open corefile %s", corename); + return -errno; + } + + sts = xc_domain_dumpcore_via_callback( + xch, domid, &da, &local_file_dump); + + /* flush and discard any remaining portion of the file from cache */ + discard_file_cache(xch, da.fd, 1/* flush first*/); + + close(da.fd); + + return sts; +} + +/* + * Local variables: From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:23:03 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:23:03 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146580.269708 (Exim 4.92) (envelope-from ) id 1lwMVj-0004P5-Tq; Thu, 24 Jun 2021 10:23:03 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146580.269708; Thu, 24 Jun 2021 10:23:03 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMVj-0004Ox-Qu; Thu, 24 Jun 2021 10:23:03 +0000 Received: by outflank-mailman (input) for mailman id 146580; Thu, 24 Jun 2021 10:23:03 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMVj-0004Ol-A3 for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:23:03 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMVj-0005yD-9D for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:23:03 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwMVj-0006Fm-8J for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:23:03 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=GWnDmbo6Wk+gNHKbWFjSTKAe2vPYkUmZ1CO0rybn/gs=; b=DXCUuQulLi2Pmkzdci5uhHnKC1 tNIgg1339wEU0i8urPVIs2Kp91dPwSNRBRTWBERXSUiLlIp2/pvkKIgk8ken1SOckeIl2xHUWjsNi /hOZVPHByybnFWuJez7nuJU/QBTxBnGLdGvqXfuurYakMvuhRfMblFwkLEtj+s8BLkVA=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] tools/libs/guest: make some definitions private to libxenguest Message-Id: Date: Thu, 24 Jun 2021 10:23:03 +0000 commit 75f13e9b221e2c8603f15ee1d53318526cf56113 Author: Juergen Gross AuthorDate: Fri Jun 4 08:02:14 2021 +0200 Commit: Julien Grall CommitDate: Fri Jun 4 18:56:46 2021 +0100 tools/libs/guest: make some definitions private to libxenguest There are some definitions which are used in libxenguest only now. Move them from libxenctrl over to libxenguest. Remove an unused macro. Signed-off-by: Juergen Gross Acked-by: Wei Liu --- tools/libs/ctrl/xc_private.h | 32 -------------------------------- tools/libs/guest/xg_core.h | 2 +- tools/libs/guest/xg_core_x86.c | 16 +++++++++++++++- tools/libs/guest/xg_private.h | 16 +++++++++++++++- tools/libs/guest/xg_resume.c | 2 +- 5 files changed, 32 insertions(+), 36 deletions(-) diff --git a/tools/libs/ctrl/xc_private.h b/tools/libs/ctrl/xc_private.h index dff0f0289b..3e299b943f 100644 --- a/tools/libs/ctrl/xc_private.h +++ b/tools/libs/ctrl/xc_private.h @@ -65,38 +65,6 @@ struct iovec { #define ROUNDUP(_x,_w) (((unsigned long)(_x)+(1UL<<(_w))-1) & ~((1UL<<(_w))-1)) -#define GET_FIELD(_p, _f, _w) (((_w) == 8) ? ((_p)->x64._f) : ((_p)->x32._f)) - -#define SET_FIELD(_p, _f, _v, _w) do { \ - if ((_w) == 8) \ - (_p)->x64._f = (_v); \ - else \ - (_p)->x32._f = (_v); \ -} while (0) - -/* XXX SMH: following skanky macros rely on variable p2m_size being set */ -/* XXX TJD: also, "guest_width" should be the guest's sizeof(unsigned long) */ - -struct domain_info_context { - unsigned int guest_width; - unsigned int p2m_frames; - unsigned long p2m_size; -}; - -/* Number of xen_pfn_t in a page */ -#define FPP (PAGE_SIZE/(dinfo->guest_width)) - -/* Number of entries in the pfn_to_mfn_frame_list_list */ -#define P2M_FLL_ENTRIES (((dinfo->p2m_size)+(FPP*FPP)-1)/(FPP*FPP)) - -/* Number of entries in the pfn_to_mfn_frame_list */ -#define P2M_FL_ENTRIES (((dinfo->p2m_size)+FPP-1)/FPP) - -/* Size in bytes of the pfn_to_mfn_frame_list */ -#define P2M_GUEST_FL_SIZE ((P2M_FL_ENTRIES) * (dinfo->guest_width)) -#define P2M_TOOLS_FL_SIZE ((P2M_FL_ENTRIES) * \ - max_t(size_t, sizeof(xen_pfn_t), dinfo->guest_width)) - #define DECLARE_DOMCTL struct xen_domctl domctl #define DECLARE_SYSCTL struct xen_sysctl sysctl #define DECLARE_PHYSDEV_OP struct physdev_op physdev_op diff --git a/tools/libs/guest/xg_core.h b/tools/libs/guest/xg_core.h index f07584aaa6..aaca9e0a8b 100644 --- a/tools/libs/guest/xg_core.h +++ b/tools/libs/guest/xg_core.h @@ -21,7 +21,7 @@ #define XC_CORE_H #include "xen/version.h" -#include "xc_private.h" +#include "xg_private.h" #include "xen/libelf/elfstructs.h" /* section names */ diff --git a/tools/libs/guest/xg_core_x86.c b/tools/libs/guest/xg_core_x86.c index 09f5d696ce..61106b98b8 100644 --- a/tools/libs/guest/xg_core_x86.c +++ b/tools/libs/guest/xg_core_x86.c @@ -18,10 +18,24 @@ */ #include -#include "xc_private.h" +#include "xg_private.h" #include "xg_core.h" #include +/* Number of xen_pfn_t in a page */ +#define FPP (PAGE_SIZE/(dinfo->guest_width)) + +/* Number of entries in the pfn_to_mfn_frame_list_list */ +#define P2M_FLL_ENTRIES (((dinfo->p2m_size)+(FPP*FPP)-1)/(FPP*FPP)) + +/* Number of entries in the pfn_to_mfn_frame_list */ +#define P2M_FL_ENTRIES (((dinfo->p2m_size)+FPP-1)/FPP) + +/* Size in bytes of the pfn_to_mfn_frame_list */ +#define P2M_GUEST_FL_SIZE ((P2M_FL_ENTRIES) * (dinfo->guest_width)) +#define P2M_TOOLS_FL_SIZE ((P2M_FL_ENTRIES) * \ + max_t(size_t, sizeof(xen_pfn_t), dinfo->guest_width)) + int xc_core_arch_gpfn_may_present(struct xc_core_arch_context *arch_ctxt, unsigned long pfn) diff --git a/tools/libs/guest/xg_private.h b/tools/libs/guest/xg_private.h index 25e46d7ce1..03d765da21 100644 --- a/tools/libs/guest/xg_private.h +++ b/tools/libs/guest/xg_private.h @@ -42,6 +42,21 @@ #endif #endif +#define GET_FIELD(_p, _f, _w) (((_w) == 8) ? ((_p)->x64._f) : ((_p)->x32._f)) + +#define SET_FIELD(_p, _f, _v, _w) do { \ + if ((_w) == 8) \ + (_p)->x64._f = (_v); \ + else \ + (_p)->x32._f = (_v); \ +} while (0) + +struct domain_info_context { + unsigned int guest_width; + unsigned int p2m_frames; + unsigned long p2m_size; +}; + struct xc_dom_loader { const char *name; /* Sadly the error returns from these functions are not consistent: */ @@ -139,7 +154,6 @@ static inline xen_pfn_t xc_pfn_to_mfn(xen_pfn_t pfn, xen_pfn_t *p2m, /* Masks for PTE<->PFN conversions */ #define MADDR_BITS_X86 ((dinfo->guest_width == 8) ? 52 : 44) #define MFN_MASK_X86 ((1ULL << (MADDR_BITS_X86 - PAGE_SHIFT_X86)) - 1) -#define MADDR_MASK_X86 (MFN_MASK_X86 << PAGE_SHIFT_X86) int pin_table(xc_interface *xch, unsigned int type, unsigned long mfn, uint32_t dom); diff --git a/tools/libs/guest/xg_resume.c b/tools/libs/guest/xg_resume.c index d201c1488d..77e2451a3c 100644 --- a/tools/libs/guest/xg_resume.c +++ b/tools/libs/guest/xg_resume.c @@ -13,7 +13,7 @@ * License along with this library; If not, see . */ -#include "xc_private.h" +#include "xg_private.h" #include "xenguest.h" #if defined(__i386__) || defined(__x86_64__) -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:23:15 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:23:15 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146581.269713 (Exim 4.92) (envelope-from ) id 1lwMVu-0004Su-Vv; Thu, 24 Jun 2021 10:23:14 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146581.269713; Thu, 24 Jun 2021 10:23:14 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMVu-0004Sm-SS; Thu, 24 Jun 2021 10:23:14 +0000 Received: by outflank-mailman (input) for mailman id 146581; Thu, 24 Jun 2021 10:23:13 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMVt-0004S0-DZ for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:23:13 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMVt-0005yY-Ck for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:23:13 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwMVt-0006Gf-Bl for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:23:13 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=clISXLjxbjnX/E4Ak1sD0Nig1/AhQc+uV3vT6A+ST9c=; b=RPWSSMaWK0zDeWQLHHWlJ+j2xB qsZU9y6H1TGTusN0RH6BgEmPmmrc0vU80B6fuQyzZHYnh2L19TErCsBGTLIibY7upNDW3MlW/OaC+ Ub/oFgaJfOnQyh1GPAEMQE9cHUrDtFdOMcdQjYMAFrlcxj3oc3VWV+/MHktNIIk8FJf8=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] credit2: make sure we pick a runnable unit from the runq if there is one Message-Id: Date: Thu, 24 Jun 2021 10:23:13 +0000 commit 07b0eb5d0ef0be154606aa46b5b4c5c59b158505 Author: Dario Faggioli AuthorDate: Fri May 28 17:12:48 2021 +0200 Commit: George Dunlap CommitDate: Mon Jun 7 13:16:36 2021 +0100 credit2: make sure we pick a runnable unit from the runq if there is one A !runnable unit (temporarily) present in the runq may cause us to stop scanning the runq itself too early. Of course, we don't run any non-runnable vCPUs, but we end the scan and we fallback to picking the idle unit. In other word, this prevent us to find there and pick the actual unit that we're meant to start running (which might be further ahead in the runq). Depending on the vCPU pinning configuration, this may lead to such unit to be stuck in the runq for long time, causing malfunctioning inside the guest. Fix this by checking runnable/non-runnable status up-front, in the runq scanning function. Reported-by: Michał Leszczyński Reported-by: Dion Kant Signed-off-by: Dario Faggioli Reviewed-by: George Dunlap --- xen/common/sched/credit2.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) diff --git a/xen/common/sched/credit2.c b/xen/common/sched/credit2.c index eb5e5a78c5..f5c1e5b944 100644 --- a/xen/common/sched/credit2.c +++ b/xen/common/sched/credit2.c @@ -3463,6 +3463,10 @@ runq_candidate(struct csched2_runqueue_data *rqd, (unsigned char *)&d); } + /* Skip non runnable units that we (temporarily) have in the runq */ + if ( unlikely(!unit_runnable_state(svc->unit)) ) + continue; + /* Only consider vcpus that are allowed to run on this processor. */ if ( !cpumask_test_cpu(cpu, svc->unit->cpu_hard_affinity) ) continue; @@ -3496,8 +3500,7 @@ runq_candidate(struct csched2_runqueue_data *rqd, * some budget, then choose it. */ if ( (yield || svc->credit > snext->credit) && - (!has_cap(svc) || unit_grab_budget(svc)) && - unit_runnable_state(svc->unit) ) + (!has_cap(svc) || unit_grab_budget(svc)) ) snext = svc; /* In any case, if we got this far, break. */ -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:23:25 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:23:25 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146582.269716 (Exim 4.92) (envelope-from ) id 1lwMW5-0004W5-1F; Thu, 24 Jun 2021 10:23:25 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146582.269716; Thu, 24 Jun 2021 10:23:25 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMW4-0004Vw-U2; Thu, 24 Jun 2021 10:23:24 +0000 Received: by outflank-mailman (input) for mailman id 146582; Thu, 24 Jun 2021 10:23:23 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMW3-0004Vj-H6 for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:23:23 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMW3-0005yc-GJ for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:23:23 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwMW3-0006HT-FI for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:23:23 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=gfBjcYA0Hz/R4r3bHvh/hHKRbwEl4Uy0BJPMbkngRKw=; b=kr5DAYnAClRot7WuQAZvYZKSzD eXKvuJzGdkGU/mKLopINguMnbxvD6reBNPHuUiB0iWtp86LXshkzS0BrAYfzIWvt9UlFW6sm83rUp d+UCRHcZ+tf58qxXIDLlTcGtn/kRtc5Iu9+VxeJhXSRPu+zoDipg18XWQUhIEfHIHD/c=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] xen: credit2: fix per-entity load tracking when continuing running Message-Id: Date: Thu, 24 Jun 2021 10:23:23 +0000 commit 89052b9fa24bf976924e40918fc9fa3b1b940e17 Author: Dario Faggioli AuthorDate: Fri Mar 19 12:14:17 2021 +0000 Commit: George Dunlap CommitDate: Mon Jun 7 13:17:06 2021 +0100 xen: credit2: fix per-entity load tracking when continuing running If we schedule, and the current vCPU continues to run, its statistical load is not properly updated, resulting in something like this, even if all the 8 vCPUs are 100% busy: (XEN) Runqueue 0: (XEN) [...] (XEN) aveload = 2097152 (~800%) (XEN) [...] (XEN) Domain: 0 w 256 c 0 v 8 (XEN) 1: [0.0] flags=2 cpu=4 credit=9996885 [w=256] load=35 (~0%) (XEN) 2: [0.1] flags=2 cpu=2 credit=9993725 [w=256] load=796 (~0%) (XEN) 3: [0.2] flags=2 cpu=1 credit=9995885 [w=256] load=883 (~0%) (XEN) 4: [0.3] flags=2 cpu=5 credit=9998833 [w=256] load=487 (~0%) (XEN) 5: [0.4] flags=2 cpu=6 credit=9998942 [w=256] load=1595 (~0%) (XEN) 6: [0.5] flags=2 cpu=0 credit=9994669 [w=256] load=22 (~0%) (XEN) 7: [0.6] flags=2 cpu=7 credit=9997706 [w=256] load=0 (~0%) (XEN) 8: [0.7] flags=2 cpu=3 credit=9992440 [w=256] load=0 (~0%) As we can see, the average load of the runqueue as a whole is, instead, computed properly. This issue would, in theory, potentially affect Credit2 load balancing logic. In practice, however, the problem only manifests (at least with these characteristics) when there is only 1 runqueue active in the cpupool, which also means there is no need to do any load-balancing. Hence its real impact is pretty much limited to wrong per-vCPU load percentages, when looking at the output of the 'r' debug-key. With this patch, the load is updated and displayed correctly: (XEN) Runqueue 0: (XEN) [...] (XEN) aveload = 2097152 (~800%) (XEN) [...] (XEN) Domain info: (XEN) Domain: 0 w 256 c 0 v 8 (XEN) 1: [0.0] flags=2 cpu=4 credit=9995584 [w=256] load=262144 (~100%) (XEN) 2: [0.1] flags=2 cpu=6 credit=9992992 [w=256] load=262144 (~100%) (XEN) 3: [0.2] flags=2 cpu=3 credit=9998918 [w=256] load=262118 (~99%) (XEN) 4: [0.3] flags=2 cpu=5 credit=9996867 [w=256] load=262144 (~100%) (XEN) 5: [0.4] flags=2 cpu=1 credit=9998912 [w=256] load=262144 (~100%) (XEN) 6: [0.5] flags=2 cpu=2 credit=9997842 [w=256] load=262144 (~100%) (XEN) 7: [0.6] flags=2 cpu=7 credit=9994623 [w=256] load=262144 (~100%) (XEN) 8: [0.7] flags=2 cpu=0 credit=9991815 [w=256] load=262144 (~100%) Signed-off-by: Dario Faggioli Reviewed-by: George Dunlap --- xen/common/sched/credit2.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/xen/common/sched/credit2.c b/xen/common/sched/credit2.c index f5c1e5b944..ebb09ea43a 100644 --- a/xen/common/sched/credit2.c +++ b/xen/common/sched/credit2.c @@ -3649,6 +3649,8 @@ static void csched2_schedule( runq_remove(snext); __set_bit(__CSFLAG_scheduled, &snext->flags); } + else + update_load(ops, rqd, snext, 0, now); /* Clear the idle mask if necessary */ if ( cpumask_test_cpu(sched_cpu, &rqd->idle) ) -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:23:35 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:23:35 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146583.269720 (Exim 4.92) (envelope-from ) id 1lwMWF-0004Z8-2X; Thu, 24 Jun 2021 10:23:35 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146583.269720; Thu, 24 Jun 2021 10:23:35 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMWE-0004Z0-Ve; Thu, 24 Jun 2021 10:23:34 +0000 Received: by outflank-mailman (input) for mailman id 146583; Thu, 24 Jun 2021 10:23:33 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMWD-0004Yh-KC for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:23:33 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMWD-0005yg-JP for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:23:33 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwMWD-0006IC-Ic for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:23:33 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=/RF4Iy7WrP2+aq5ctIkDPe6GIr6bo3niqi5L8T96LLg=; b=6PdtqSc53436gWC58ErPEMhGo8 JjVckTJtA+cWLVrd6BGow+Yyc15K3Ofwoaw++Y7ehLTP6JaBgz/GH2+Jf3M2QBJGtJx/GzmOD/bq1 oaBvSquOJEnkHeeuKcvF74+Tv0X797nFj2pu5p5kjhKD5DUgnlEQ8gZG6kGjV41hfhSc=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] docs: release-technician-checklist: update to leaf tree version pinning Message-Id: Date: Thu, 24 Jun 2021 10:23:33 +0000 commit c4beefcada0a406681dcfb6e89f6cbe4aa368c2d Author: Jan Beulich AuthorDate: Mon Jun 7 15:40:55 2021 +0200 Commit: Jan Beulich CommitDate: Mon Jun 7 15:40:55 2021 +0200 docs: release-technician-checklist: update to leaf tree version pinning Our releases look to flip-flop between keeping or discarding the date and title of the referenced qemu-trad commit. I think with the hash replaced by a tag, the commit's date and title would better also be purged. Signed-off-by: Jan Beulich Acked-by: Ian Jackson --- docs/process/release-technician-checklist.txt | 1 + 1 file changed, 1 insertion(+) diff --git a/docs/process/release-technician-checklist.txt b/docs/process/release-technician-checklist.txt index 880630c1ed..7515da202c 100644 --- a/docs/process/release-technician-checklist.txt +++ b/docs/process/release-technician-checklist.txt @@ -53,6 +53,7 @@ t=RELEASE-$r # QEMU_UPSTREAM_REVISION, # QEMU_TRADITIONAL_REVISION # MINIOS_UPSTREAM_REVISION +# (drop any references to the specific commits, e.g. date or title) * change SUPPORT.md heading version number; -unstable or -rc tag * (empty in stable branches after .0 release). * SUPPORT.md: insert correct version number in release-notes link -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:23:45 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:23:45 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146584.269723 (Exim 4.92) (envelope-from ) id 1lwMWP-0004cE-4C; Thu, 24 Jun 2021 10:23:45 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146584.269723; Thu, 24 Jun 2021 10:23:45 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMWP-0004c6-0y; Thu, 24 Jun 2021 10:23:45 +0000 Received: by outflank-mailman (input) for mailman id 146584; Thu, 24 Jun 2021 10:23:43 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMWN-0004br-N3 for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:23:43 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMWN-0005ys-ML for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:23:43 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwMWN-0006JA-Le for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:23:43 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=a8FAIKGXEl5bft4FfwyUwztjdBVDce2jsp4ePdI16fg=; b=BoUSuFBevteNUU/qZ7AVS+eVOj Y6bUIenq1hqJHhL/m2kxqp7In1lnqLAcQ9ERxTzVXeV9t89t5XljnaFqhcncB1qLLKEGxZato/REf O16epqb+7mG+yG5a1gLvFJMG+nE6V48teWJY1+JRtykevRZCjpdpebM498K8kmWmG19U=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] x86/cpuid: Fix HLE and RTM handling (again) Message-Id: Date: Thu, 24 Jun 2021 10:23:43 +0000 commit 60fa12dbf1d4d2c4ffe1ef34b495b24aa7e41aa0 Author: Andrew Cooper AuthorDate: Mon Jun 7 13:25:09 2021 +0100 Commit: Andrew Cooper CommitDate: Mon Jun 7 15:43:35 2021 +0100 x86/cpuid: Fix HLE and RTM handling (again) For reasons which are my fault, but I don't recall why, the FDP_EXCP_ONLY/NO_FPU_SEL adjustment uses the whole special_features[] array element, not the two relevant bits. HLE and RTM were recently added to the list of special features, causing them to be always set in guest view, irrespective of the toolstacks choice on the matter. Rewrite the logic to refer to the features specifically, rather than relying on the contents of the special_features[] array. Fixes: 8fe24090d9 ("x86/cpuid: Rework HLE and RTM handling") Reported-by: Edwin Török Signed-off-by: Andrew Cooper Reviewed-by: Jan Beulich --- xen/arch/x86/cpuid.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/xen/arch/x86/cpuid.c b/xen/arch/x86/cpuid.c index f3c8950aa3..958caf35da 100644 --- a/xen/arch/x86/cpuid.c +++ b/xen/arch/x86/cpuid.c @@ -672,9 +672,11 @@ void recalculate_cpuid_policy(struct domain *d) sanitise_featureset(fs); /* Fold host's FDP_EXCP_ONLY and NO_FPU_SEL into guest's view. */ - fs[FEATURESET_7b0] &= ~special_features[FEATURESET_7b0]; + fs[FEATURESET_7b0] &= ~(cpufeat_mask(X86_FEATURE_FDP_EXCP_ONLY) | + cpufeat_mask(X86_FEATURE_NO_FPU_SEL)); fs[FEATURESET_7b0] |= (host_cpuid_policy.feat._7b0 & - special_features[FEATURESET_7b0]); + (cpufeat_mask(X86_FEATURE_FDP_EXCP_ONLY) | + cpufeat_mask(X86_FEATURE_NO_FPU_SEL))); cpuid_featureset_to_policy(fs, p); -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:23:55 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:23:55 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146585.269728 (Exim 4.92) (envelope-from ) id 1lwMWZ-0004fW-5P; Thu, 24 Jun 2021 10:23:55 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146585.269728; Thu, 24 Jun 2021 10:23:55 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMWZ-0004fO-2Y; Thu, 24 Jun 2021 10:23:55 +0000 Received: by outflank-mailman (input) for mailman id 146585; Thu, 24 Jun 2021 10:23:53 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMWX-0004f9-Q5 for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:23:53 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMWX-0005z3-PL for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:23:53 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwMWX-0006Jz-Of for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:23:53 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=QjwxkvSvRQBX9c7HlhHa/NQTlsklo3s0vhPIDkh/LtM=; b=f7bxGdckKFKZdP81k9eheW6Wdz OsfAyWq1eVepHNpDExMSYY9VnVGoGKxXOkkiuw9dc585vadd90qw5l5gq4f3ui6KIH+ElZKUoVTm5 Msn9A3BJ7ATN0t9uNnqlbGtv8EV8GkbHc9NdrdkF57ML6UJTFwdXz/h/Aa2vPlmyHaPw=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] x86/cpuid: Drop special_features[] Message-Id: Date: Thu, 24 Jun 2021 10:23:53 +0000 commit 69e1472d21cf7e5cf0795ef38b99d00de78a910e Author: Andrew Cooper AuthorDate: Mon Jun 7 13:38:53 2021 +0100 Commit: Andrew Cooper CommitDate: Mon Jun 7 15:43:35 2021 +0100 x86/cpuid: Drop special_features[] While the ! annotation is useful to indicate that something special is happening, an array of bits is not. Drop it, to prevent mistakes. Signed-off-by: Andrew Cooper Reviewed-by: Jan Beulich --- xen/arch/x86/cpuid.c | 2 -- xen/include/asm-x86/cpuid.h | 1 - xen/tools/gen-cpuid.py | 3 --- 3 files changed, 6 deletions(-) diff --git a/xen/arch/x86/cpuid.c b/xen/arch/x86/cpuid.c index 958caf35da..2079a30ae4 100644 --- a/xen/arch/x86/cpuid.c +++ b/xen/arch/x86/cpuid.c @@ -14,7 +14,6 @@ #include const uint32_t known_features[] = INIT_KNOWN_FEATURES; -const uint32_t special_features[] = INIT_SPECIAL_FEATURES; static const uint32_t __initconst pv_max_featuremask[] = INIT_PV_MAX_FEATURES; static const uint32_t hvm_shadow_max_featuremask[] = INIT_HVM_SHADOW_MAX_FEATURES; @@ -1132,7 +1131,6 @@ void guest_cpuid(const struct vcpu *v, uint32_t leaf, static void __init __maybe_unused build_assertions(void) { BUILD_BUG_ON(ARRAY_SIZE(known_features) != FSCAPINTS); - BUILD_BUG_ON(ARRAY_SIZE(special_features) != FSCAPINTS); BUILD_BUG_ON(ARRAY_SIZE(pv_max_featuremask) != FSCAPINTS); BUILD_BUG_ON(ARRAY_SIZE(hvm_shadow_max_featuremask) != FSCAPINTS); BUILD_BUG_ON(ARRAY_SIZE(hvm_hap_max_featuremask) != FSCAPINTS); diff --git a/xen/include/asm-x86/cpuid.h b/xen/include/asm-x86/cpuid.h index 7baf6c9628..46904061d0 100644 --- a/xen/include/asm-x86/cpuid.h +++ b/xen/include/asm-x86/cpuid.h @@ -14,7 +14,6 @@ #include extern const uint32_t known_features[FSCAPINTS]; -extern const uint32_t special_features[FSCAPINTS]; void init_guest_cpuid(void); diff --git a/xen/tools/gen-cpuid.py b/xen/tools/gen-cpuid.py index b953648b65..c6b5056a8d 100755 --- a/xen/tools/gen-cpuid.py +++ b/xen/tools/gen-cpuid.py @@ -362,8 +362,6 @@ def write_results(state): #define INIT_KNOWN_FEATURES { \\\n%s\n} -#define INIT_SPECIAL_FEATURES { \\\n%s\n} - #define INIT_PV_DEF_FEATURES { \\\n%s\n} #define INIT_PV_MAX_FEATURES { \\\n%s\n} @@ -384,7 +382,6 @@ def write_results(state): """ % (state.nr_entries, next(featureset_to_uint32s(state.common_1d, 1)), format_uint32s(state, state.names.keys(), 4), - format_uint32s(state, state.raw['!'], 4), format_uint32s(state, state.pv_def, 4), format_uint32s(state, state.pv_max, 4), format_uint32s(state, state.hvm_shadow_def, 4), -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:24:05 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:24:05 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146586.269731 (Exim 4.92) (envelope-from ) id 1lwMWj-0004iV-7X; Thu, 24 Jun 2021 10:24:05 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146586.269731; Thu, 24 Jun 2021 10:24:05 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMWj-0004iN-44; Thu, 24 Jun 2021 10:24:05 +0000 Received: by outflank-mailman (input) for mailman id 146586; Thu, 24 Jun 2021 10:24:03 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMWh-0004i9-TH for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:24:03 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMWh-0005zS-SO for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:24:03 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwMWh-0006LI-Rd for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:24:03 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=17edD/ne7xbUbY/GrpRlR02aDKexm/sQ8HOOM7O9JGo=; b=aaqA3n/jeo54AZ1PyLJ1XBrx9d Rk97NVHi47UFNbTKsM5K/InVKmaQnB/hPS166+avFPNSS2unSrQOAJ9z8YStcMOZksDC5Dux/R8Nk oxhljGZA9d/7t+VKh826MJvSMT1OIkz+Re+ONKDEh4LWVanKFkyawfO2OcdD/Km7ucvY=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] tools/libs/guest: fix save and restore of pv domains after 32-bit de-support Message-Id: Date: Thu, 24 Jun 2021 10:24:03 +0000 commit d21121685fac829c988e432407fb0e4ef9b19331 Author: Juergen Gross AuthorDate: Mon Jun 7 15:00:05 2021 +0200 Commit: Andrew Cooper CommitDate: Mon Jun 7 15:43:35 2021 +0100 tools/libs/guest: fix save and restore of pv domains after 32-bit de-support After 32-bit PV-guests have been security de-supported when not running under PV-shim, the hypervisor will no longer be configured to support those domains per default when not being built as PV-shim. Unfortunately libxenguest will fail saving or restoring a PV domain due to this restriction, as it is trying to get the compat MFN list even for 64 bit guests. Fix that by obtaining the compat MFN list only for 32-bit PV guests. Fixes: 1a0f2fe2297d122a08fe ("SUPPORT.md: Un-shimmed 32-bit PV guests are no longer supported") Signed-off-by: Juergen Gross Reviewed-by: Jan Beulich Reviewed-by: Andrew Cooper --- tools/libs/guest/xg_sr_common.h | 2 +- tools/libs/guest/xg_sr_common_x86_pv.c | 13 ++++++++----- 2 files changed, 9 insertions(+), 6 deletions(-) diff --git a/tools/libs/guest/xg_sr_common.h b/tools/libs/guest/xg_sr_common.h index cc3ad1c394..e2994e18ac 100644 --- a/tools/libs/guest/xg_sr_common.h +++ b/tools/libs/guest/xg_sr_common.h @@ -325,7 +325,7 @@ struct xc_sr_context xen_pfn_t max_mfn; /* Read-only machine to phys map */ xen_pfn_t *m2p; - /* first mfn of the compat m2p (Only needed for 32bit PV guests) */ + /* first mfn of the compat m2p (Only set for 32bit PV guests) */ xen_pfn_t compat_m2p_mfn0; /* Number of m2p frames mapped */ unsigned long nr_m2p_frames; diff --git a/tools/libs/guest/xg_sr_common_x86_pv.c b/tools/libs/guest/xg_sr_common_x86_pv.c index cd33406aab..c0acf00f90 100644 --- a/tools/libs/guest/xg_sr_common_x86_pv.c +++ b/tools/libs/guest/xg_sr_common_x86_pv.c @@ -149,12 +149,13 @@ int x86_pv_map_m2p(struct xc_sr_context *ctx) ctx->x86.pv.nr_m2p_frames = (M2P_CHUNK_SIZE >> PAGE_SHIFT) * m2p_chunks; + if ( ctx->x86.pv.levels == 3 ) + { #ifdef __i386__ - /* 32 bit toolstacks automatically get the compat m2p */ - ctx->x86.pv.compat_m2p_mfn0 = entries[0].mfn; + /* 32 bit toolstacks automatically get the compat m2p */ + ctx->x86.pv.compat_m2p_mfn0 = entries[0].mfn; #else - /* 64 bit toolstacks need to ask Xen specially for it */ - { + /* 64 bit toolstacks need to ask Xen specially for it */ struct xen_machphys_mfn_list xmml = { .max_extents = 1, .extent_start = { &ctx->x86.pv.compat_m2p_mfn0 }, @@ -168,8 +169,10 @@ int x86_pv_map_m2p(struct xc_sr_context *ctx) rc = -1; goto err; } - } #endif + } + else + ctx->x86.pv.compat_m2p_mfn0 = INVALID_MFN; /* All Done */ rc = 0; -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:24:15 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:24:15 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146587.269736 (Exim 4.92) (envelope-from ) id 1lwMWt-0004mE-A8; Thu, 24 Jun 2021 10:24:15 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146587.269736; Thu, 24 Jun 2021 10:24:15 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMWt-0004m6-78; Thu, 24 Jun 2021 10:24:15 +0000 Received: by outflank-mailman (input) for mailman id 146587; Thu, 24 Jun 2021 10:24:14 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMWs-0004lT-01 for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:24:14 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMWr-0005zw-VU for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:24:13 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwMWr-0006MH-Ui for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:24:13 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=Bw2wlaXMnhShGTZNVzqtU3FjdfQn5iApXKrdGKlcou4=; b=a3ZiM+z1/Yib3/sIHayjVKKGkt nTQWReUe0WhO9oKKi23pKXMXPGcN+O3w0C9H8uQdoMpMvdKbGzRR8Z0OF89H7M1Ypesn2IwudKMlr YQvOe8g4mbvtXuDmUPmAqN854jRds3GGpe10X6TTRIhtdfGXxRUPA9S6hmDRPgIGtivc=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] tools/firmware/ovmf: Use OvmfXen platform file is exist Message-Id: Date: Thu, 24 Jun 2021 10:24:13 +0000 commit aad7b5c11d51d57659978e04702ac970906894e8 Author: Anthony PERARD AuthorDate: Tue Jun 1 11:28:03 2021 +0100 Commit: Ian Jackson CommitDate: Tue Jun 8 11:31:10 2021 +0100 tools/firmware/ovmf: Use OvmfXen platform file is exist A platform introduced in EDK II named OvmfXen is now the one to use for Xen instead of OvmfX64. It comes with PVH support. Also, the Xen support in OvmfX64 is deprecated, "deprecation notice: *dynamic* multi-VMM (QEMU vs. Xen) support in OvmfPkg" https://edk2.groups.io/g/devel/message/75498 Signed-off-by: Anthony PERARD Acked-by: Ian Jackson --- tools/firmware/ovmf-makefile | 10 ++++++++-- 1 file changed, 8 insertions(+), 2 deletions(-) diff --git a/tools/firmware/ovmf-makefile b/tools/firmware/ovmf-makefile index 55f9992145..637ee509c3 100644 --- a/tools/firmware/ovmf-makefile +++ b/tools/firmware/ovmf-makefile @@ -17,8 +17,14 @@ all: build .PHONY: build build: if test -e .git ; then $(GIT) submodule update --init --recursive ; fi - OvmfPkg/build.sh -a X64 -b $(TARGET) -n 4 - cp Build/OvmfX64/$(TARGET)_GCC*/FV/OVMF.fd ovmf.bin + set -ex; \ + if test -e OvmfPkg/OvmfXen.dsc; then \ + OvmfPkg/build.sh -a X64 -b $(TARGET) -n 4 -p OvmfPkg/OvmfXen.dsc; \ + cp Build/OvmfXen/$(TARGET)_GCC*/FV/OVMF.fd ovmf.bin; \ + else \ + OvmfPkg/build.sh -a X64 -b $(TARGET) -n 4; \ + cp Build/OvmfX64/$(TARGET)_GCC*/FV/OVMF.fd ovmf.bin; \ + fi .PHONY: clean clean: -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:24:25 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:24:25 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146588.269741 (Exim 4.92) (envelope-from ) id 1lwMX3-0004p1-CA; Thu, 24 Jun 2021 10:24:25 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146588.269741; Thu, 24 Jun 2021 10:24:25 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMX3-0004op-8l; Thu, 24 Jun 2021 10:24:25 +0000 Received: by outflank-mailman (input) for mailman id 146588; Thu, 24 Jun 2021 10:24:24 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMX2-0004ob-3G for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:24:24 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMX2-000608-2R for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:24:24 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwMX2-0006N4-1a for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:24:24 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=1JBBoh6niGfGYLLNkM/25FrTcY5knucIII/d42EYico=; b=Jqrr65CUZ4Qk8xuXf+erWcR1BQ jBXpfWCCjg7HVEilAkh3XBJX57KFnLdRHm4U+/CnF/zPhhTo6H3NiAWysSgUjYclFkezWv/CbnEyf 3EPZN2s1vijzRM9Did3g79LptslpibXb7MqR4AHprV7LCImz3hl9IO1sP5E+SehBmrEg=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] evtchn: slightly defer lock acquire where possible Message-Id: Date: Thu, 24 Jun 2021 10:24:24 +0000 commit 411076f4d635f6cba5e74e8173d0ac4c576024e8 Author: Jan Beulich AuthorDate: Tue Jun 8 14:46:06 2021 +0200 Commit: Jan Beulich CommitDate: Tue Jun 8 14:46:06 2021 +0200 evtchn: slightly defer lock acquire where possible port_is_valid() and evtchn_from_port() are fine to use without holding any locks. Accordingly acquire the per-domain lock slightly later in evtchn_close() and evtchn_bind_vcpu(). Especially for the use by the former (but there are pre-existing uses) add a comment about port_is_valid()'s guarantees. Signed-off-by: Jan Beulich Acked-by: Roger Pau Monné Reviewed-by: Julien Grall --- xen/common/event_channel.c | 20 +++++++------------- xen/include/xen/event.h | 10 ++++++++++ 2 files changed, 17 insertions(+), 13 deletions(-) diff --git a/xen/common/event_channel.c b/xen/common/event_channel.c index 5479315aae..8a5f8bb9d4 100644 --- a/xen/common/event_channel.c +++ b/xen/common/event_channel.c @@ -606,17 +606,14 @@ int evtchn_close(struct domain *d1, int port1, bool guest) int port2; long rc = 0; - again: - spin_lock(&d1->event_lock); - if ( !port_is_valid(d1, port1) ) - { - rc = -EINVAL; - goto out; - } + return -EINVAL; chn1 = evtchn_from_port(d1, port1); + again: + spin_lock(&d1->event_lock); + /* Guest cannot close a Xen-attached event channel. */ if ( unlikely(consumer_is_xen(chn1)) && guest ) { @@ -1041,16 +1038,13 @@ long evtchn_bind_vcpu(unsigned int port, unsigned int vcpu_id) if ( (v = domain_vcpu(d, vcpu_id)) == NULL ) return -ENOENT; - spin_lock(&d->event_lock); - if ( !port_is_valid(d, port) ) - { - rc = -EINVAL; - goto out; - } + return -EINVAL; chn = evtchn_from_port(d, port); + spin_lock(&d->event_lock); + /* Guest cannot re-bind a Xen-attached event channel. */ if ( unlikely(consumer_is_xen(chn)) ) { diff --git a/xen/include/xen/event.h b/xen/include/xen/event.h index a0a85cdda8..a185a5b875 100644 --- a/xen/include/xen/event.h +++ b/xen/include/xen/event.h @@ -120,6 +120,16 @@ static inline void evtchn_read_unlock(struct evtchn *evtchn) read_unlock(&evtchn->lock); } +/* + * While calling the function is okay without holding a suitable lock yet + * (see the comment ahead of struct evtchn_port_ops for which ones those + * are), for a dying domain it may start returning false at any point - see + * evtchn_destroy(). This is not a fundamental problem though, as the + * struct evtchn instance won't disappear (and will continue to hold valid + * data) until final cleanup of the domain, at which point the domain itself + * cannot be looked up anymore and hence calls here can't occur any longer + * in the first place. + */ static inline bool_t port_is_valid(struct domain *d, unsigned int p) { if ( p >= read_atomic(&d->valid_evtchns) ) -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:24:35 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:24:35 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146589.269744 (Exim 4.92) (envelope-from ) id 1lwMXD-0004sF-DI; Thu, 24 Jun 2021 10:24:35 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146589.269744; Thu, 24 Jun 2021 10:24:35 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMXD-0004s7-AL; Thu, 24 Jun 2021 10:24:35 +0000 Received: by outflank-mailman (input) for mailman id 146589; Thu, 24 Jun 2021 10:24:34 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMXC-0004rw-6H for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:24:34 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMXC-00060Q-5V for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:24:34 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwMXC-0006Nr-4j for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:24:34 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=XmxKc2M74PIbtNNHeNnQbU10GLHrDokrUhQfBayM/MY=; b=l79KufX26LuSW4mlBXQXr5CSDs p1V2aNgn00MK/AugavQALHx8SJ0Qp8mxFuD19ByXkcN6b2uPIgPGL/0k5ejZmwmNau5PNWdakf2wg IYkGCHJQzIGKc9tbt7jEWTaO+iJ7NomiouAxA5eUsD1sp7VK9buxEIgb5BAHofzDPuAM=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] evtchn: add helper for port_is_valid() + evtchn_from_port() Message-Id: Date: Thu, 24 Jun 2021 10:24:34 +0000 commit 3270a2d45fbbcdf9b3dd44b4b4776c1fc4682717 Author: Jan Beulich AuthorDate: Tue Jun 8 14:47:14 2021 +0200 Commit: Jan Beulich CommitDate: Tue Jun 8 14:47:14 2021 +0200 evtchn: add helper for port_is_valid() + evtchn_from_port() The combination is pretty common, so adding a simple local helper seems worthwhile. Make it const- and type-correct, in turn requiring the two called function to also be const-correct (and at this occasion also make them type-correct). Signed-off-by: Jan Beulich Acked-by: Julien Grall --- xen/common/event_channel.c | 73 +++++++++++++++++++++------------------------- xen/include/xen/event.h | 5 ++-- 2 files changed, 36 insertions(+), 42 deletions(-) diff --git a/xen/common/event_channel.c b/xen/common/event_channel.c index 8a5f8bb9d4..886c2726d6 100644 --- a/xen/common/event_channel.c +++ b/xen/common/event_channel.c @@ -147,6 +147,12 @@ static bool virq_is_global(unsigned int virq) return true; } +static struct evtchn *_evtchn_from_port(const struct domain *d, + evtchn_port_t port) +{ + return port_is_valid(d, port) ? evtchn_from_port(d, port) : NULL; +} + static void free_evtchn_bucket(struct domain *d, struct evtchn *bucket) { if ( !bucket ) @@ -319,7 +325,6 @@ static long evtchn_alloc_unbound(evtchn_alloc_unbound_t *alloc) return rc; } - static void double_evtchn_lock(struct evtchn *lchn, struct evtchn *rchn) { ASSERT(lchn != rchn); @@ -365,9 +370,9 @@ static long evtchn_bind_interdomain(evtchn_bind_interdomain_t *bind) ERROR_EXIT(lport); lchn = evtchn_from_port(ld, lport); - if ( !port_is_valid(rd, rport) ) + rchn = _evtchn_from_port(rd, rport); + if ( !rchn ) ERROR_EXIT_DOM(-EINVAL, rd); - rchn = evtchn_from_port(rd, rport); if ( (rchn->state != ECS_UNBOUND) || (rchn->u.unbound.remote_domid != ld->domain_id) ) ERROR_EXIT_DOM(-EINVAL, rd); @@ -602,15 +607,12 @@ static long evtchn_bind_pirq(evtchn_bind_pirq_t *bind) int evtchn_close(struct domain *d1, int port1, bool guest) { struct domain *d2 = NULL; - struct evtchn *chn1, *chn2; - int port2; + struct evtchn *chn1 = _evtchn_from_port(d1, port1), *chn2; long rc = 0; - if ( !port_is_valid(d1, port1) ) + if ( !chn1 ) return -EINVAL; - chn1 = evtchn_from_port(d1, port1); - again: spin_lock(&d1->event_lock); @@ -698,10 +700,8 @@ int evtchn_close(struct domain *d1, int port1, bool guest) goto out; } - port2 = chn1->u.interdomain.remote_port; - BUG_ON(!port_is_valid(d2, port2)); - - chn2 = evtchn_from_port(d2, port2); + chn2 = _evtchn_from_port(d2, chn1->u.interdomain.remote_port); + BUG_ON(!chn2); BUG_ON(chn2->state != ECS_INTERDOMAIN); BUG_ON(chn2->u.interdomain.remote_dom != d1); @@ -739,15 +739,13 @@ int evtchn_close(struct domain *d1, int port1, bool guest) int evtchn_send(struct domain *ld, unsigned int lport) { - struct evtchn *lchn, *rchn; + struct evtchn *lchn = _evtchn_from_port(ld, lport), *rchn; struct domain *rd; int rport, ret = 0; - if ( !port_is_valid(ld, lport) ) + if ( !lchn ) return -EINVAL; - lchn = evtchn_from_port(ld, lport); - evtchn_read_lock(lchn); /* Guest cannot send via a Xen-attached event channel. */ @@ -967,15 +965,15 @@ int evtchn_status(evtchn_status_t *status) if ( d == NULL ) return -ESRCH; - spin_lock(&d->event_lock); - - if ( !port_is_valid(d, port) ) + chn = _evtchn_from_port(d, port); + if ( !chn ) { - rc = -EINVAL; - goto out; + rcu_unlock_domain(d); + return -EINVAL; } - chn = evtchn_from_port(d, port); + spin_lock(&d->event_lock); + if ( consumer_is_xen(chn) ) { rc = -EACCES; @@ -1038,11 +1036,10 @@ long evtchn_bind_vcpu(unsigned int port, unsigned int vcpu_id) if ( (v = domain_vcpu(d, vcpu_id)) == NULL ) return -ENOENT; - if ( !port_is_valid(d, port) ) + chn = _evtchn_from_port(d, port); + if ( !chn ) return -EINVAL; - chn = evtchn_from_port(d, port); - spin_lock(&d->event_lock); /* Guest cannot re-bind a Xen-attached event channel. */ @@ -1088,13 +1085,11 @@ long evtchn_bind_vcpu(unsigned int port, unsigned int vcpu_id) int evtchn_unmask(unsigned int port) { struct domain *d = current->domain; - struct evtchn *evtchn; + struct evtchn *evtchn = _evtchn_from_port(d, port); - if ( unlikely(!port_is_valid(d, port)) ) + if ( unlikely(!evtchn) ) return -EINVAL; - evtchn = evtchn_from_port(d, port); - evtchn_read_lock(evtchn); evtchn_port_unmask(d, evtchn); @@ -1177,14 +1172,12 @@ static long evtchn_set_priority(const struct evtchn_set_priority *set_priority) { struct domain *d = current->domain; unsigned int port = set_priority->port; - struct evtchn *chn; + struct evtchn *chn = _evtchn_from_port(d, port); long ret; - if ( !port_is_valid(d, port) ) + if ( !chn ) return -EINVAL; - chn = evtchn_from_port(d, port); - evtchn_read_lock(chn); ret = evtchn_port_set_priority(d, chn, set_priority->priority); @@ -1410,10 +1403,10 @@ void free_xen_event_channel(struct domain *d, int port) void notify_via_xen_event_channel(struct domain *ld, int lport) { - struct evtchn *lchn, *rchn; + struct evtchn *lchn = _evtchn_from_port(ld, lport), *rchn; struct domain *rd; - if ( !port_is_valid(ld, lport) ) + if ( !lchn ) { /* * Make sure ->is_dying is read /after/ ->valid_evtchns, pairing @@ -1424,8 +1417,6 @@ void notify_via_xen_event_channel(struct domain *ld, int lport) return; } - lchn = evtchn_from_port(ld, lport); - if ( !evtchn_read_trylock(lchn) ) return; @@ -1580,12 +1571,14 @@ static void domain_dump_evtchn_info(struct domain *d) spin_lock(&d->event_lock); - for ( port = 1; port_is_valid(d, port); ++port ) + for ( port = 1; ; ++port ) { - const struct evtchn *chn; + const struct evtchn *chn = _evtchn_from_port(d, port); char *ssid; - chn = evtchn_from_port(d, port); + if ( !chn ) + break; + if ( chn->state == ECS_FREE ) continue; diff --git a/xen/include/xen/event.h b/xen/include/xen/event.h index a185a5b875..4f4e97e9c7 100644 --- a/xen/include/xen/event.h +++ b/xen/include/xen/event.h @@ -130,7 +130,7 @@ static inline void evtchn_read_unlock(struct evtchn *evtchn) * cannot be looked up anymore and hence calls here can't occur any longer * in the first place. */ -static inline bool_t port_is_valid(struct domain *d, unsigned int p) +static inline bool port_is_valid(const struct domain *d, evtchn_port_t p) { if ( p >= read_atomic(&d->valid_evtchns) ) return false; @@ -145,7 +145,8 @@ static inline bool_t port_is_valid(struct domain *d, unsigned int p) return true; } -static inline struct evtchn *evtchn_from_port(struct domain *d, unsigned int p) +static inline struct evtchn *evtchn_from_port(const struct domain *d, + evtchn_port_t p) { if ( p < EVTCHNS_PER_BUCKET ) return &d->evtchn[array_index_nospec(p, EVTCHNS_PER_BUCKET)]; -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:24:45 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:24:45 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146590.269748 (Exim 4.92) (envelope-from ) id 1lwMXN-0004vG-Ep; Thu, 24 Jun 2021 10:24:45 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146590.269748; Thu, 24 Jun 2021 10:24:45 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMXN-0004v9-Bx; Thu, 24 Jun 2021 10:24:45 +0000 Received: by outflank-mailman (input) for mailman id 146590; Thu, 24 Jun 2021 10:24:44 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMXM-0004uz-9B for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:24:44 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMXM-00060a-8U for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:24:44 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwMXM-0006Ol-7q for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:24:44 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=kYekrM6bG/qeZMcKfdx6ekIaKdLzxOvAMANoJZzzKJE=; b=qzQdzHWupHp+0AttQPWTpaSlSg 3KR507mbzOFGnVP2HT7NjUptyx5SABHVqswHUXumJIb3mSiygZkiPEMN03xNEvSFkncW3DyJc+dCc VLHScZJr0ivRyv5kv0PsNJ9G/PtTg85gwxb3ClICMagOSYMu9+xJIVtbBPCUH9h/AR6w=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] evtchn: type adjustments Message-Id: Date: Thu, 24 Jun 2021 10:24:44 +0000 commit dd77e859686f458a5313786deae0a63683d3cba3 Author: Jan Beulich AuthorDate: Tue Jun 8 14:47:47 2021 +0200 Commit: Jan Beulich CommitDate: Tue Jun 8 14:47:47 2021 +0200 evtchn: type adjustments First of all avoid "long" when "int" suffices, i.e. in particular when merely conveying error codes. 32-bit values are slightly cheaper to deal with on x86, and their processing is at least no more expensive on Arm. Where possible use evtchn_port_t for port numbers and unsigned int for other unsigned quantities in adjacent code. In evtchn_set_priority() eliminate a local variable altogether instead of changing its type. Signed-off-by: Jan Beulich Acked-by: Roger Pau Monné --- xen/common/event_channel.c | 42 ++++++++++++++++++++---------------------- xen/include/xen/event.h | 2 +- 2 files changed, 21 insertions(+), 23 deletions(-) diff --git a/xen/common/event_channel.c b/xen/common/event_channel.c index 886c2726d6..da88ad141a 100644 --- a/xen/common/event_channel.c +++ b/xen/common/event_channel.c @@ -284,13 +284,12 @@ void evtchn_free(struct domain *d, struct evtchn *chn) xsm_evtchn_close_post(chn); } -static long evtchn_alloc_unbound(evtchn_alloc_unbound_t *alloc) +static int evtchn_alloc_unbound(evtchn_alloc_unbound_t *alloc) { struct evtchn *chn; struct domain *d; - int port; + int port, rc; domid_t dom = alloc->dom; - long rc; d = rcu_lock_domain_by_any_id(dom); if ( d == NULL ) @@ -342,13 +341,13 @@ static void double_evtchn_unlock(struct evtchn *lchn, struct evtchn *rchn) evtchn_write_unlock(rchn); } -static long evtchn_bind_interdomain(evtchn_bind_interdomain_t *bind) +static int evtchn_bind_interdomain(evtchn_bind_interdomain_t *bind) { struct evtchn *lchn, *rchn; struct domain *ld = current->domain, *rd; - int lport, rport = bind->remote_port; + int lport, rc; + evtchn_port_t rport = bind->remote_port; domid_t rdom = bind->remote_dom; - long rc; if ( (rd = rcu_lock_domain_by_any_id(rdom)) == NULL ) return -ESRCH; @@ -484,12 +483,12 @@ int evtchn_bind_virq(evtchn_bind_virq_t *bind, evtchn_port_t port) } -static long evtchn_bind_ipi(evtchn_bind_ipi_t *bind) +static int evtchn_bind_ipi(evtchn_bind_ipi_t *bind) { struct evtchn *chn; struct domain *d = current->domain; - int port, vcpu = bind->vcpu; - long rc = 0; + int port, rc = 0; + unsigned int vcpu = bind->vcpu; if ( domain_vcpu(d, vcpu) == NULL ) return -ENOENT; @@ -543,16 +542,16 @@ static void unlink_pirq_port(struct evtchn *chn, struct vcpu *v) } -static long evtchn_bind_pirq(evtchn_bind_pirq_t *bind) +static int evtchn_bind_pirq(evtchn_bind_pirq_t *bind) { struct evtchn *chn; struct domain *d = current->domain; struct vcpu *v = d->vcpu[0]; struct pirq *info; - int port = 0, pirq = bind->pirq; - long rc; + int port = 0, rc; + unsigned int pirq = bind->pirq; - if ( (pirq < 0) || (pirq >= d->nr_pirqs) ) + if ( pirq >= d->nr_pirqs ) return -EINVAL; if ( !is_hvm_domain(d) && !pirq_access_permitted(d, pirq) ) @@ -608,7 +607,7 @@ int evtchn_close(struct domain *d1, int port1, bool guest) { struct domain *d2 = NULL; struct evtchn *chn1 = _evtchn_from_port(d1, port1), *chn2; - long rc = 0; + int rc = 0; if ( !chn1 ) return -EINVAL; @@ -959,7 +958,7 @@ int evtchn_status(evtchn_status_t *status) domid_t dom = status->dom; int port = status->port; struct evtchn *chn; - long rc = 0; + int rc = 0; d = rcu_lock_domain_by_any_id(dom); if ( d == NULL ) @@ -1025,11 +1024,11 @@ int evtchn_status(evtchn_status_t *status) } -long evtchn_bind_vcpu(unsigned int port, unsigned int vcpu_id) +int evtchn_bind_vcpu(evtchn_port_t port, unsigned int vcpu_id) { struct domain *d = current->domain; struct evtchn *chn; - long rc = 0; + int rc = 0; struct vcpu *v; /* Use the vcpu info to prevent speculative out-of-bound accesses */ @@ -1168,12 +1167,11 @@ int evtchn_reset(struct domain *d, bool resuming) return rc; } -static long evtchn_set_priority(const struct evtchn_set_priority *set_priority) +static int evtchn_set_priority(const struct evtchn_set_priority *set_priority) { struct domain *d = current->domain; - unsigned int port = set_priority->port; - struct evtchn *chn = _evtchn_from_port(d, port); - long ret; + struct evtchn *chn = _evtchn_from_port(d, set_priority->port); + int ret; if ( !chn ) return -EINVAL; @@ -1189,7 +1187,7 @@ static long evtchn_set_priority(const struct evtchn_set_priority *set_priority) long do_event_channel_op(int cmd, XEN_GUEST_HANDLE_PARAM(void) arg) { - long rc; + int rc; switch ( cmd ) { diff --git a/xen/include/xen/event.h b/xen/include/xen/event.h index 4f4e97e9c7..21c95e14fd 100644 --- a/xen/include/xen/event.h +++ b/xen/include/xen/event.h @@ -54,7 +54,7 @@ void send_guest_pirq(struct domain *, const struct pirq *); int evtchn_send(struct domain *d, unsigned int lport); /* Bind a local event-channel port to the specified VCPU. */ -long evtchn_bind_vcpu(unsigned int port, unsigned int vcpu_id); +int evtchn_bind_vcpu(evtchn_port_t port, unsigned int vcpu_id); /* Bind a VIRQ. */ int evtchn_bind_virq(evtchn_bind_virq_t *bind, evtchn_port_t port); -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:24:55 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:24:55 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146591.269752 (Exim 4.92) (envelope-from ) id 1lwMXX-0004xp-GM; Thu, 24 Jun 2021 10:24:55 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146591.269752; Thu, 24 Jun 2021 10:24:55 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMXX-0004xh-DX; Thu, 24 Jun 2021 10:24:55 +0000 Received: by outflank-mailman (input) for mailman id 146591; Thu, 24 Jun 2021 10:24:54 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMXW-0004xX-CA for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:24:54 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMXW-00060m-BQ for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:24:54 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwMXW-0006Pg-Ak for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:24:54 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=qllNFDKu2YKFAhZpbBkgHcXK8r8r9VbJSOjp6ArDR0E=; b=yKHIrDuXs5Oxb2MrewwSotx948 xjKpcL9/0DA6dn5v95y9J4oteEfhrb1XG8a/LPlzpUzLQYsejSyCWKPY21uvox+xeA692xIjRfk1j tBiPJpWfkTw6g4XTFM3cBHEFmOhzmvLGYiq1XdAguWyH255DCtyNAf6IBOed0iv8upcs=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] x86/cpuid: Half revert "x86/cpuid: Drop special_features[]" Message-Id: Date: Thu, 24 Jun 2021 10:24:54 +0000 commit 0ba0663b1b32d9351890dfd02bdebb3d238897bd Author: Andrew Cooper AuthorDate: Tue Jun 8 17:13:59 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 17:27:06 2021 +0100 x86/cpuid: Half revert "x86/cpuid: Drop special_features[]" xen-cpuid does print out the list of special features, and this is helpful to keep. Fixes: ba6950fb070 ("x86/cpuid: Drop special_features[]") Reported-by: Jan Beulich Reported-by: Sander Eikelenboom Signed-off-by: Andrew Cooper Acked-by: Jan Beulich --- xen/tools/gen-cpuid.py | 3 +++ 1 file changed, 3 insertions(+) diff --git a/xen/tools/gen-cpuid.py b/xen/tools/gen-cpuid.py index c6b5056a8d..b953648b65 100755 --- a/xen/tools/gen-cpuid.py +++ b/xen/tools/gen-cpuid.py @@ -362,6 +362,8 @@ def write_results(state): #define INIT_KNOWN_FEATURES { \\\n%s\n} +#define INIT_SPECIAL_FEATURES { \\\n%s\n} + #define INIT_PV_DEF_FEATURES { \\\n%s\n} #define INIT_PV_MAX_FEATURES { \\\n%s\n} @@ -382,6 +384,7 @@ def write_results(state): """ % (state.nr_entries, next(featureset_to_uint32s(state.common_1d, 1)), format_uint32s(state, state.names.keys(), 4), + format_uint32s(state, state.raw['!'], 4), format_uint32s(state, state.pv_def, 4), format_uint32s(state, state.pv_max, 4), format_uint32s(state, state.hvm_shadow_def, 4), -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:25:05 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:25:05 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146592.269756 (Exim 4.92) (envelope-from ) id 1lwMXh-00050y-Ih; Thu, 24 Jun 2021 10:25:05 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146592.269756; Thu, 24 Jun 2021 10:25:05 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMXh-00050q-Ex; Thu, 24 Jun 2021 10:25:05 +0000 Received: by outflank-mailman (input) for mailman id 146592; Thu, 24 Jun 2021 10:25:04 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMXg-00050h-Fa for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:25:04 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMXg-000619-Er for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:25:04 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwMXg-0006Qw-Dw for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:25:04 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=Dv2uBXs1sFU5ZfbhnkwUfnp5I5ILtQKtQFcDoLPs1K0=; b=cLFaMSoeWRMMewbCFMvyAD/Oer pP/9NtR3y9DC4MwUOT6nPVzFI4oWC5YwAiqjbRVHE6TfCacClNyqgZY5oJkZs+hx2Ho9JRhm8AYZX hadqochdqjfeifuWtbtBRfRWVMx6bBS7Opl1QaSBIG4eJ+6FbO2bG2IcBIwuFlzjoZYs=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] xen/arm: Create dom0less domUs earlier Message-Id: Date: Thu, 24 Jun 2021 10:25:04 +0000 commit 371347c5b64da699d9f5a0edda5dc496fd2b7a5c Author: Julien Grall AuthorDate: Mon May 17 17:47:13 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 17:43:06 2021 +0100 xen/arm: Create dom0less domUs earlier In a follow-up patch we will need to unallocate the boot modules before heap_init_late() is called. The modules will contain the domUs kernel and initramfs. Therefore Xen will need to create extra domUs (used by dom0less) before heap_init_late(). This has two consequences on dom0less: 1) Domains will not be unpaused as soon as they are created but once all have been created. However, Xen doesn't guarantee an order to unpause, so this is not something one could rely on. 2) The memory allocated for a domU will not be scrubbed anymore when an admin select bootscrub=on. This is not something we advertised, but if this is a concern we can introduce either force scrub for all domUs or a per-domain flag in the DT. The behavior for bootscrub=off and bootscrub=idle (default) has not changed. This is part of XSA-372 / CVE-2021-28693. Signed-off-by: Julien Grall Reviewed-by: Jan Beulich Reviewed-by: Stefano Stabellini Tested-by: Stefano Stabellini --- xen/arch/arm/domain_build.c | 6 +----- xen/arch/arm/setup.c | 14 +++++++------- xen/include/asm-arm/setup.h | 2 +- 3 files changed, 9 insertions(+), 13 deletions(-) diff --git a/xen/arch/arm/domain_build.c b/xen/arch/arm/domain_build.c index 282416e74d..6c86d52781 100644 --- a/xen/arch/arm/domain_build.c +++ b/xen/arch/arm/domain_build.c @@ -2521,8 +2521,6 @@ void __init create_domUs(void) if ( construct_domU(d, node) != 0 ) panic("Could not set up domain %s\n", dt_node_name(node)); - - domain_unpause_by_systemcontroller(d); } } @@ -2584,7 +2582,7 @@ static int __init construct_dom0(struct domain *d) return construct_domain(d, &kinfo); } -struct domain* __init create_dom0(void) +void __init create_dom0(void) { struct domain *dom0; struct xen_domctl_createdomain dom0_cfg = { @@ -2615,8 +2613,6 @@ struct domain* __init create_dom0(void) if ( construct_dom0(dom0) != 0) panic("Could not set up DOM0 guest OS\n"); - - return dom0; } /* diff --git a/xen/arch/arm/setup.c b/xen/arch/arm/setup.c index 00aad1c194..e17532c132 100644 --- a/xen/arch/arm/setup.c +++ b/xen/arch/arm/setup.c @@ -836,7 +836,7 @@ void __init start_xen(unsigned long boot_phys_offset, int cpus, i; const char *cmdline; struct bootmodule *xen_bootmodule; - struct domain *dom0 = NULL; + struct domain *d; int rc; dcache_line_bytes = read_dcache_line_bytes(); @@ -992,10 +992,13 @@ void __init start_xen(unsigned long boot_phys_offset, /* Create initial domain 0. */ if ( !is_dom0less_mode() ) - dom0 = create_dom0(); + create_dom0(); else printk(XENLOG_INFO "Xen dom0less mode detected\n"); + if ( acpi_disabled ) + create_domUs(); + heap_init_late(); init_trace_bufs(); @@ -1009,11 +1012,8 @@ void __init start_xen(unsigned long boot_phys_offset, system_state = SYS_STATE_active; - if ( acpi_disabled ) - create_domUs(); - - if ( dom0 ) - domain_unpause_by_systemcontroller(dom0); + for_each_domain( d ) + domain_unpause_by_systemcontroller(d); /* Switch on to the dynamically allocated stack for the idle vcpu * since the static one we're running on is about to be freed. */ diff --git a/xen/include/asm-arm/setup.h b/xen/include/asm-arm/setup.h index 5283244015..c4b6af6029 100644 --- a/xen/include/asm-arm/setup.h +++ b/xen/include/asm-arm/setup.h @@ -94,7 +94,7 @@ void acpi_create_efi_mmap_table(struct domain *d, int acpi_make_efi_nodes(void *fdt, struct membank tbl_add[]); void create_domUs(void); -struct domain* create_dom0(void); +void create_dom0(void); void discard_initial_modules(void); void fw_unreserved_regions(paddr_t s, paddr_t e, -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:25:15 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:25:15 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146593.269759 (Exim 4.92) (envelope-from ) id 1lwMXr-00054k-Lx; Thu, 24 Jun 2021 10:25:15 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146593.269759; Thu, 24 Jun 2021 10:25:15 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMXr-00054c-J1; Thu, 24 Jun 2021 10:25:15 +0000 Received: by outflank-mailman (input) for mailman id 146593; Thu, 24 Jun 2021 10:25:14 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMXq-000549-IY for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:25:14 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMXq-00061Z-Hk for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:25:14 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwMXq-0006Rn-H6 for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:25:14 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=7wQboiyaB3L+E8RJcuFHguWkRI1x9eHfNUX7VJg1CCI=; b=YdmfFToj09YXbWRugvU3JACw92 cQqDFlIvMKhUBgzi+CC+nl0OJbBOqwtukL0KQ9NOCBaNCyEdCo9b98MhsoqYnsiN1RaJWXHi7SpZ6 sa7TFK8LyuGVpDa4KE8GNiSAIqEGCwCU7VWmiKpK9QPpMvTK7ZAD953vKZXXNNpJczHA=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] xen/arm: Boot modules should always be scrubbed if bootscrub={on, idle} Message-Id: Date: Thu, 24 Jun 2021 10:25:14 +0000 commit fd5dc41ceaed9cfcfa011cdfd50f264c89277a90 Author: Julien Grall AuthorDate: Sat Apr 17 17:38:28 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 17:43:06 2021 +0100 xen/arm: Boot modules should always be scrubbed if bootscrub={on, idle} The function to initialize the pages (see init_heap_pages()) will request scrub when the admin request idle bootscrub (default) and state == SYS_STATE_active. When bootscrub=on, Xen will scrub any free pages in heap_init_late(). Currently, the boot modules (e.g. kernels, initramfs) will be discarded/ freed after heap_init_late() is called and system_state switched to SYS_STATE_active. This means the pages associated with the boot modules will not get scrubbed before getting re-purposed. If the memory is assigned to an untrusted domU, it may be able to retrieve secrets from the modules. This is part of XSA-372 / CVE-2021-28693. Fixes: 1774e9b1df27 ("xen/arm: introduce create_domUs") Signed-off-by: Julien Grall Reviewed-by: Jan Beulich Reviewed-by: Stefano Stabellini Tested-by: Stefano Stabellini --- xen/arch/arm/setup.c | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-) diff --git a/xen/arch/arm/setup.c b/xen/arch/arm/setup.c index e17532c132..63a908e325 100644 --- a/xen/arch/arm/setup.c +++ b/xen/arch/arm/setup.c @@ -71,8 +71,6 @@ domid_t __read_mostly max_init_domid; static __used void init_done(void) { - discard_initial_modules(); - /* Must be done past setting system_state. */ unregister_init_virtual_region(); @@ -999,6 +997,12 @@ void __init start_xen(unsigned long boot_phys_offset, if ( acpi_disabled ) create_domUs(); + /* + * This needs to be called **before** heap_init_late() so modules + * will be scrubbed (unless suppressed). + */ + discard_initial_modules(); + heap_init_late(); init_trace_bufs(); -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:25:26 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:25:26 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146594.269764 (Exim 4.92) (envelope-from ) id 1lwMY2-00057j-Na; Thu, 24 Jun 2021 10:25:26 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146594.269764; Thu, 24 Jun 2021 10:25:26 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMY2-00057b-Ka; Thu, 24 Jun 2021 10:25:26 +0000 Received: by outflank-mailman (input) for mailman id 146594; Thu, 24 Jun 2021 10:25:24 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMY0-00057J-Lm for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:25:24 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMY0-00061n-L6 for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:25:24 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwMY0-0006Sj-K2 for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:25:24 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=G3J8w/0GU3AGUzSUwpnSit7EztUZ2G4aqP/VVW4nPU8=; b=fjNaJCL/v50u3cokuIqHk7qksH 21gr8w8HtzTU6z9dA3ZFSdSyPkiDimBzIEV27jmWajfsMeAkdoB8ZAKytBjZeaJQIdqiYNf2dp5jz m5turIvV708qfU0ShMmrxHvzo9nuiJrI73D0wTTYk+np9sCfrXOURU3faBNyIBmogw6k=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] VT-d: size qinval queue dynamically Message-Id: Date: Thu, 24 Jun 2021 10:25:24 +0000 commit cbfa62bb140c8f10a0ae96db3ce06e22b3b9d302 Author: Jan Beulich AuthorDate: Tue Jun 8 17:38:55 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 17:43:06 2021 +0100 VT-d: size qinval queue dynamically With the present synchronous model, we need two slots for every operation (the operation itself and a wait descriptor). There can be one such pair of requests pending per CPU. To ensure that under all normal circumstances a slot is always available when one is requested, size the queue ring according to the number of present CPUs. This is part of XSA-373 / CVE-2021-28692. Signed-off-by: Jan Beulich Reviewed-by: Paul Durrant --- xen/drivers/passthrough/vtd/iommu.h | 10 +------- xen/drivers/passthrough/vtd/qinval.c | 46 +++++++++++++++++++++++++++--------- 2 files changed, 36 insertions(+), 20 deletions(-) diff --git a/xen/drivers/passthrough/vtd/iommu.h b/xen/drivers/passthrough/vtd/iommu.h index 68b3c76ea9..cd516ac5f8 100644 --- a/xen/drivers/passthrough/vtd/iommu.h +++ b/xen/drivers/passthrough/vtd/iommu.h @@ -452,17 +452,9 @@ struct qinval_entry { }q; }; -/* Order of queue invalidation pages(max is 8) */ -#define QINVAL_PAGE_ORDER 2 - -#define QINVAL_ARCH_PAGE_ORDER (QINVAL_PAGE_ORDER + PAGE_SHIFT_4K - PAGE_SHIFT) -#define QINVAL_ARCH_PAGE_NR ( QINVAL_ARCH_PAGE_ORDER < 0 ? \ - 1 : \ - 1 << QINVAL_ARCH_PAGE_ORDER ) - /* Each entry is 16 bytes, so 2^8 entries per page */ #define QINVAL_ENTRY_ORDER ( PAGE_SHIFT - 4 ) -#define QINVAL_ENTRY_NR (1 << (QINVAL_PAGE_ORDER + 8)) +#define QINVAL_MAX_ENTRY_NR (1u << (7 + QINVAL_ENTRY_ORDER)) /* Status data flag */ #define QINVAL_STAT_INIT 0 diff --git a/xen/drivers/passthrough/vtd/qinval.c b/xen/drivers/passthrough/vtd/qinval.c index 0976def43d..69fef844bb 100644 --- a/xen/drivers/passthrough/vtd/qinval.c +++ b/xen/drivers/passthrough/vtd/qinval.c @@ -31,6 +31,9 @@ #define VTD_QI_TIMEOUT 1 +static unsigned int __read_mostly qi_pg_order; +static unsigned int __read_mostly qi_entry_nr; + static int __must_check invalidate_sync(struct vtd_iommu *iommu); static void print_qi_regs(const struct vtd_iommu *iommu) @@ -47,7 +50,7 @@ static unsigned int qinval_next_index(struct vtd_iommu *iommu) tail >>= QINVAL_INDEX_SHIFT; /* (tail+1 == head) indicates a full queue, wait for HW */ - while ( (tail + 1) % QINVAL_ENTRY_NR == + while ( ((tail + 1) & (qi_entry_nr - 1)) == (dmar_readl(iommu->reg, DMAR_IQH_REG) >> QINVAL_INDEX_SHIFT) ) cpu_relax(); @@ -60,7 +63,7 @@ static void qinval_update_qtail(struct vtd_iommu *iommu, unsigned int index) /* Need hold register lock when update tail */ ASSERT( spin_is_locked(&iommu->register_lock) ); - val = (index + 1) % QINVAL_ENTRY_NR; + val = (index + 1) & (qi_entry_nr - 1); dmar_writel(iommu->reg, DMAR_IQT_REG, val << QINVAL_INDEX_SHIFT); } @@ -395,8 +398,28 @@ int enable_qinval(struct vtd_iommu *iommu) if ( iommu->qinval_maddr == 0 ) { - iommu->qinval_maddr = alloc_pgtable_maddr(QINVAL_ARCH_PAGE_NR, - iommu->node); + if ( !qi_entry_nr ) + { + /* + * With the present synchronous model, we need two slots for every + * operation (the operation itself and a wait descriptor). There + * can be one such pair of requests pending per CPU. One extra + * entry is needed as the ring is considered full when there's + * only one entry left. + */ + BUILD_BUG_ON(CONFIG_NR_CPUS * 2 >= QINVAL_MAX_ENTRY_NR); + qi_pg_order = get_order_from_bytes((num_present_cpus() * 2 + 1) << + (PAGE_SHIFT - + QINVAL_ENTRY_ORDER)); + qi_entry_nr = 1u << (qi_pg_order + QINVAL_ENTRY_ORDER); + + dprintk(XENLOG_INFO VTDPREFIX, + "QI: using %u-entry ring(s)\n", qi_entry_nr); + } + + iommu->qinval_maddr = + alloc_pgtable_maddr(qi_entry_nr >> QINVAL_ENTRY_ORDER, + iommu->node); if ( iommu->qinval_maddr == 0 ) { dprintk(XENLOG_WARNING VTDPREFIX, @@ -410,15 +433,16 @@ int enable_qinval(struct vtd_iommu *iommu) spin_lock_irqsave(&iommu->register_lock, flags); - /* Setup Invalidation Queue Address(IQA) register with the - * address of the page we just allocated. QS field at - * bits[2:0] to indicate size of queue is one 4KB page. - * That's 256 entries. Queued Head (IQH) and Queue Tail (IQT) - * registers are automatically reset to 0 with write - * to IQA register. + /* + * Setup Invalidation Queue Address (IQA) register with the address of the + * pages we just allocated. The QS field at bits[2:0] indicates the size + * (page order) of the queue. + * + * Queued Head (IQH) and Queue Tail (IQT) registers are automatically + * reset to 0 with write to IQA register. */ dmar_writeq(iommu->reg, DMAR_IQA_REG, - iommu->qinval_maddr | QINVAL_PAGE_ORDER); + iommu->qinval_maddr | qi_pg_order); dmar_writeq(iommu->reg, DMAR_IQT_REG, 0); -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:25:37 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:25:37 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146595.269768 (Exim 4.92) (envelope-from ) id 1lwMYC-0005AO-P5; Thu, 24 Jun 2021 10:25:36 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146595.269768; Thu, 24 Jun 2021 10:25:36 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMYC-0005AE-M8; Thu, 24 Jun 2021 10:25:36 +0000 Received: by outflank-mailman (input) for mailman id 146595; Thu, 24 Jun 2021 10:25:34 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMYA-0005A2-Pv for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:25:34 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMYA-000623-OF for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:25:34 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwMYA-0006Ta-Nd for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:25:34 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=dVDQx/uGNnznxHQ33BoA8nUXm5RciPdUWCrT00U9JYs=; b=P3VbAWSlnM52g/PbIckPG3ko36 c0WNTG4VycSCM+YdgWrGK45usA5kFN0wujIWDZZZ2YPfSgE3dyvwzvuV4dCYALFzBdFc3R8HZzlXF 6HF71pOlqM4evSSW7YfCZMKHY5nb7AsOAocYFU/6LYTj7xHMnMNlELHYJ8DeVPTSAgQk=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] AMD/IOMMU: size command buffer dynamically Message-Id: Date: Thu, 24 Jun 2021 10:25:34 +0000 commit df242851ddc93ac0b0a3a20ecab34acc29e3b36b Author: Jan Beulich AuthorDate: Tue Jun 8 17:38:55 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 17:43:06 2021 +0100 AMD/IOMMU: size command buffer dynamically With the present synchronous model, we need two slots for every operation (the operation itself and a wait command). There can be one such pair of commands pending per CPU. To ensure that under all normal circumstances a slot is always available when one is requested, size the command ring according to the number of present CPUs. This is part of XSA-373 / CVE-2021-28692. Signed-off-by: Jan Beulich Reviewed-by: Paul Durrant --- xen/drivers/passthrough/amd/iommu-defs.h | 7 ++----- xen/drivers/passthrough/amd/iommu_cmd.c | 4 ++-- xen/drivers/passthrough/amd/iommu_init.c | 28 +++++++++++++++++++++++++--- 3 files changed, 29 insertions(+), 10 deletions(-) diff --git a/xen/drivers/passthrough/amd/iommu-defs.h b/xen/drivers/passthrough/amd/iommu-defs.h index 56f385f270..0c97db2a32 100644 --- a/xen/drivers/passthrough/amd/iommu-defs.h +++ b/xen/drivers/passthrough/amd/iommu-defs.h @@ -20,9 +20,6 @@ #ifndef AMD_IOMMU_DEFS_H #define AMD_IOMMU_DEFS_H -/* IOMMU Command Buffer entries: in power of 2 increments, minimum of 256 */ -#define IOMMU_CMD_BUFFER_DEFAULT_ENTRIES 512 - /* IOMMU Event Log entries: in power of 2 increments, minimum of 256 */ #define IOMMU_EVENT_LOG_DEFAULT_ENTRIES 512 @@ -164,8 +161,8 @@ struct amd_iommu_dte { #define IOMMU_CMD_BUFFER_LENGTH_MASK 0x0F000000 #define IOMMU_CMD_BUFFER_LENGTH_SHIFT 24 -#define IOMMU_CMD_BUFFER_ENTRY_SIZE 16 -#define IOMMU_CMD_BUFFER_POWER_OF2_ENTRIES_PER_PAGE 8 +#define IOMMU_CMD_BUFFER_ENTRY_ORDER 4 +#define IOMMU_CMD_BUFFER_MAX_ENTRIES (1u << 15) #define IOMMU_CMD_OPCODE_MASK 0xF0000000 #define IOMMU_CMD_OPCODE_SHIFT 28 diff --git a/xen/drivers/passthrough/amd/iommu_cmd.c b/xen/drivers/passthrough/amd/iommu_cmd.c index 6c0647c524..54d7528f96 100644 --- a/xen/drivers/passthrough/amd/iommu_cmd.c +++ b/xen/drivers/passthrough/amd/iommu_cmd.c @@ -24,7 +24,7 @@ static int queue_iommu_command(struct amd_iommu *iommu, u32 cmd[]) { uint32_t tail, head; - tail = iommu->cmd_buffer.tail + IOMMU_CMD_BUFFER_ENTRY_SIZE; + tail = iommu->cmd_buffer.tail + sizeof(cmd_entry_t); if ( tail == iommu->cmd_buffer.size ) tail = 0; @@ -33,7 +33,7 @@ static int queue_iommu_command(struct amd_iommu *iommu, u32 cmd[]) if ( head != tail ) { memcpy(iommu->cmd_buffer.buffer + iommu->cmd_buffer.tail, - cmd, IOMMU_CMD_BUFFER_ENTRY_SIZE); + cmd, sizeof(cmd_entry_t)); iommu->cmd_buffer.tail = tail; return 1; diff --git a/xen/drivers/passthrough/amd/iommu_init.c b/xen/drivers/passthrough/amd/iommu_init.c index 24d1dfec40..bb52c181f8 100644 --- a/xen/drivers/passthrough/amd/iommu_init.c +++ b/xen/drivers/passthrough/amd/iommu_init.c @@ -118,7 +118,7 @@ static void register_iommu_cmd_buffer_in_mmio_space(struct amd_iommu *iommu) writel(entry, iommu->mmio_base + IOMMU_CMD_BUFFER_BASE_LOW_OFFSET); power_of2_entries = get_order_from_bytes(iommu->cmd_buffer.size) + - IOMMU_CMD_BUFFER_POWER_OF2_ENTRIES_PER_PAGE; + PAGE_SHIFT - IOMMU_CMD_BUFFER_ENTRY_ORDER; entry = 0; iommu_set_addr_hi_to_reg(&entry, addr_hi); @@ -1018,9 +1018,31 @@ static void *__init allocate_ring_buffer(struct ring_buffer *ring_buf, static void * __init allocate_cmd_buffer(struct amd_iommu *iommu) { /* allocate 'command buffer' in power of 2 increments of 4K */ + static unsigned int __read_mostly nr_ents; + + if ( !nr_ents ) + { + unsigned int order; + + /* + * With the present synchronous model, we need two slots for every + * operation (the operation itself and a wait command). There can be + * one such pair of requests pending per CPU. One extra entry is + * needed as the ring is considered full when there's only one entry + * left. + */ + BUILD_BUG_ON(CONFIG_NR_CPUS * 2 >= IOMMU_CMD_BUFFER_MAX_ENTRIES); + order = get_order_from_bytes((num_present_cpus() * 2 + 1) << + IOMMU_CMD_BUFFER_ENTRY_ORDER); + nr_ents = 1u << (order + PAGE_SHIFT - IOMMU_CMD_BUFFER_ENTRY_ORDER); + + AMD_IOMMU_DEBUG("using %u-entry cmd ring(s)\n", nr_ents); + } + + BUILD_BUG_ON(sizeof(cmd_entry_t) != (1u << IOMMU_CMD_BUFFER_ENTRY_ORDER)); + return allocate_ring_buffer(&iommu->cmd_buffer, sizeof(cmd_entry_t), - IOMMU_CMD_BUFFER_DEFAULT_ENTRIES, - "Command Buffer", false); + nr_ents, "Command Buffer", false); } static void * __init allocate_event_log(struct amd_iommu *iommu) -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:25:47 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:25:47 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146596.269772 (Exim 4.92) (envelope-from ) id 1lwMYM-0005DU-Qs; Thu, 24 Jun 2021 10:25:46 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146596.269772; Thu, 24 Jun 2021 10:25:46 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMYM-0005DM-Nr; Thu, 24 Jun 2021 10:25:46 +0000 Received: by outflank-mailman (input) for mailman id 146596; Thu, 24 Jun 2021 10:25:44 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMYK-0005D3-Rz for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:25:44 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMYK-00062B-RG for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:25:44 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwMYK-0006UM-QW for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:25:44 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=9L7AbrOnDlGCWMGrL+ZJbsSxNgz+/tnKimN/f5dfhCA=; b=SSXC2Cw0QwFTBWhDbiwPQ1Mc3O zPz+QlGkB6lSau9zVcr4pVGRus5A7P6uzmYlKEkA6OwCsRqCiYeP06t5s0GW06QeikqqwyM2n2Bip i+eAS4cfa+kVUNm9Isxx4Q82ivDjk8FHkXTtnWyZxMrs/weU8fzOH4hw0CuqO2r0p2rY=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] VT-d: eliminate flush related timeouts Message-Id: Date: Thu, 24 Jun 2021 10:25:44 +0000 commit 7c893abc4ee7e9af62297ba6fd5f27c89d0f33f5 Author: Jan Beulich AuthorDate: Tue Jun 8 17:38:55 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 17:43:06 2021 +0100 VT-d: eliminate flush related timeouts Leaving an in-progress operation pending when it appears to take too long is problematic: If e.g. a QI command completed later, the write to the "poll slot" may instead be understood to signal a subsequently started command's completion. Also our accounting of the timeout period was actually wrong: We included the time it took for the command to actually make it to the front of the queue, which could be heavily affected by guests other than the one for which the flush is being performed. Do away with all timeout detection on all flush related code paths. Log excessively long processing times (with a progressive threshold) to have some indication of problems in this area. Additionally log (once) if qinval_next_index() didn't immediately find an available slot. Together with the earlier change sizing the queue(s) dynamically, we should now have a guarantee that with our fully synchronous model any demand for slots can actually be satisfied. This is part of XSA-373 / CVE-2021-28692. Signed-off-by: Jan Beulich Reviewed-by: Paul Durrant --- xen/drivers/passthrough/vtd/dmar.h | 28 ++++++++++++++++++++++++++++ xen/drivers/passthrough/vtd/iommu.c | 12 ++++++------ xen/drivers/passthrough/vtd/qinval.c | 33 ++++++++++++++++++++++----------- 3 files changed, 56 insertions(+), 17 deletions(-) diff --git a/xen/drivers/passthrough/vtd/dmar.h b/xen/drivers/passthrough/vtd/dmar.h index 1a9c965e59..a44820a5e7 100644 --- a/xen/drivers/passthrough/vtd/dmar.h +++ b/xen/drivers/passthrough/vtd/dmar.h @@ -127,6 +127,34 @@ do { \ } \ } while (0) +#define IOMMU_FLUSH_WAIT(what, iommu, offset, op, cond, sts) \ +do { \ + static unsigned int __read_mostly threshold = 1; \ + s_time_t start = NOW(); \ + s_time_t timeout = start + DMAR_OPERATION_TIMEOUT * threshold; \ + \ + for ( ; ; ) \ + { \ + sts = op(iommu->reg, offset); \ + if ( cond ) \ + break; \ + if ( timeout && NOW() > timeout ) \ + { \ + threshold |= threshold << 1; \ + printk(XENLOG_WARNING VTDPREFIX \ + " IOMMU#%u: %s flush taking too long\n", \ + iommu->index, what); \ + timeout = 0; \ + } \ + cpu_relax(); \ + } \ + \ + if ( !timeout ) \ + printk(XENLOG_WARNING VTDPREFIX \ + " IOMMU#%u: %s flush took %lums\n", \ + iommu->index, what, (NOW() - start) / 10000000); \ +} while ( false ) + int vtd_hw_check(void); void disable_pmr(struct vtd_iommu *iommu); int is_igd_drhd(struct acpi_drhd_unit *drhd); diff --git a/xen/drivers/passthrough/vtd/iommu.c b/xen/drivers/passthrough/vtd/iommu.c index 82ae2f5e97..b7aa15e5ef 100644 --- a/xen/drivers/passthrough/vtd/iommu.c +++ b/xen/drivers/passthrough/vtd/iommu.c @@ -373,8 +373,8 @@ static void iommu_flush_write_buffer(struct vtd_iommu *iommu) dmar_writel(iommu->reg, DMAR_GCMD_REG, val | DMA_GCMD_WBF); /* Make sure hardware complete it */ - IOMMU_WAIT_OP(iommu, DMAR_GSTS_REG, dmar_readl, - !(val & DMA_GSTS_WBFS), val); + IOMMU_FLUSH_WAIT("write buffer", iommu, DMAR_GSTS_REG, dmar_readl, + !(val & DMA_GSTS_WBFS), val); spin_unlock_irqrestore(&iommu->register_lock, flags); } @@ -423,8 +423,8 @@ int vtd_flush_context_reg(struct vtd_iommu *iommu, uint16_t did, dmar_writeq(iommu->reg, DMAR_CCMD_REG, val); /* Make sure hardware complete it */ - IOMMU_WAIT_OP(iommu, DMAR_CCMD_REG, dmar_readq, - !(val & DMA_CCMD_ICC), val); + IOMMU_FLUSH_WAIT("context", iommu, DMAR_CCMD_REG, dmar_readq, + !(val & DMA_CCMD_ICC), val); spin_unlock_irqrestore(&iommu->register_lock, flags); /* flush context entry will implicitly flush write buffer */ @@ -501,8 +501,8 @@ int vtd_flush_iotlb_reg(struct vtd_iommu *iommu, uint16_t did, uint64_t addr, dmar_writeq(iommu->reg, tlb_offset + 8, val); /* Make sure hardware complete it */ - IOMMU_WAIT_OP(iommu, (tlb_offset + 8), dmar_readq, - !(val & DMA_TLB_IVT), val); + IOMMU_FLUSH_WAIT("iotlb", iommu, (tlb_offset + 8), dmar_readq, + !(val & DMA_TLB_IVT), val); spin_unlock_irqrestore(&iommu->register_lock, flags); /* check IOTLB invalidation granularity */ diff --git a/xen/drivers/passthrough/vtd/qinval.c b/xen/drivers/passthrough/vtd/qinval.c index 69fef844bb..311bf107ed 100644 --- a/xen/drivers/passthrough/vtd/qinval.c +++ b/xen/drivers/passthrough/vtd/qinval.c @@ -29,8 +29,6 @@ #include "extern.h" #include "../ats.h" -#define VTD_QI_TIMEOUT 1 - static unsigned int __read_mostly qi_pg_order; static unsigned int __read_mostly qi_entry_nr; @@ -52,7 +50,11 @@ static unsigned int qinval_next_index(struct vtd_iommu *iommu) /* (tail+1 == head) indicates a full queue, wait for HW */ while ( ((tail + 1) & (qi_entry_nr - 1)) == (dmar_readl(iommu->reg, DMAR_IQH_REG) >> QINVAL_INDEX_SHIFT) ) + { + printk_once(XENLOG_ERR VTDPREFIX " IOMMU#%u: no QI slot available\n", + iommu->index); cpu_relax(); + } return tail; } @@ -172,23 +174,32 @@ static int __must_check queue_invalidate_wait(struct vtd_iommu *iommu, /* Now we don't support interrupt method */ if ( sw ) { - s_time_t timeout; - - /* In case all wait descriptor writes to same addr with same data */ - timeout = NOW() + MILLISECS(flush_dev_iotlb ? - iommu_dev_iotlb_timeout : VTD_QI_TIMEOUT); + static unsigned int __read_mostly threshold = 1; + s_time_t start = NOW(); + s_time_t timeout = start + (flush_dev_iotlb + ? iommu_dev_iotlb_timeout + : 100) * MILLISECS(threshold); while ( ACCESS_ONCE(*this_poll_slot) != QINVAL_STAT_DONE ) { - if ( NOW() > timeout ) + if ( timeout && NOW() > timeout ) { - print_qi_regs(iommu); + threshold |= threshold << 1; printk(XENLOG_WARNING VTDPREFIX - " Queue invalidate wait descriptor timed out\n"); - return -ETIMEDOUT; + " IOMMU#%u: QI%s wait descriptor taking too long\n", + iommu->index, flush_dev_iotlb ? " dev" : ""); + print_qi_regs(iommu); + timeout = 0; } cpu_relax(); } + + if ( !timeout ) + printk(XENLOG_WARNING VTDPREFIX + " IOMMU#%u: QI%s wait descriptor took %lums\n", + iommu->index, flush_dev_iotlb ? " dev" : "", + (NOW() - start) / 10000000); + return 0; } -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:25:56 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:25:56 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146597.269776 (Exim 4.92) (envelope-from ) id 1lwMYW-0005G1-Sx; Thu, 24 Jun 2021 10:25:56 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146597.269776; Thu, 24 Jun 2021 10:25:56 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMYW-0005Fs-PO; Thu, 24 Jun 2021 10:25:56 +0000 Received: by outflank-mailman (input) for mailman id 146597; Thu, 24 Jun 2021 10:25:55 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMYU-0005Fi-VO for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:25:54 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMYU-00062N-Ue for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:25:54 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwMYU-0006VJ-U2 for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:25:54 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=ztjX3U+kaE7KyFafJSCaJ6JuIZZ3+VWr02LY/L7ohHs=; b=JmEj2sjMwvX4V8nvIG21CecNv+ ea5JNHgtCf2jgpL8jz3dJMmHuF8/D2NpRoaqrlVWl/tzXdV5NwU+OdUUVoOy0P+1S8zlQgkAoU1DH QMHSHcoujtEk/HY+Q4p7fjn6484MLjTVp8CkSWbr0BmQnzFFvnDhUDbq1uy9scjdb6VY=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] x86/spec-ctrl: Protect against Speculative Code Store Bypass Message-Id: Date: Thu, 24 Jun 2021 10:25:54 +0000 commit 45f59ed8865318bb0356954bad067f329677ce9e Author: Andrew Cooper AuthorDate: Thu Mar 11 14:39:11 2021 +0000 Commit: Andrew Cooper CommitDate: Tue Jun 8 17:43:06 2021 +0100 x86/spec-ctrl: Protect against Speculative Code Store Bypass Modern x86 processors have far-better-than-architecturally-guaranteed self modifying code detection. Typically, when a write hits an instruction in flight, a Machine Clear occurs to flush stale content in the frontend and backend. For self modifying code, before a write which hits an instruction in flight retires, the frontend can speculatively decode and execute the old instruction stream. Speculation of this form can suffer from type confusion in registers, and potentially leak data. Furthermore, updates are typically byte-wise, rather than atomic. Depending on timing, speculation can race ahead multiple times between individual writes, and execute the transiently-malformed instruction stream. Xen has stubs which are used in certain cases for emulation purposes. Inhibit speculation between updating the stub and executing it. This is XSA-375 / CVE-2021-0089. Signed-off-by: Andrew Cooper Reviewed-by: Jan Beulich --- xen/arch/x86/pv/emul-priv-op.c | 2 ++ xen/arch/x86/x86_emulate/x86_emulate.c | 1 + 2 files changed, 3 insertions(+) diff --git a/xen/arch/x86/pv/emul-priv-op.c b/xen/arch/x86/pv/emul-priv-op.c index 8889509d2a..11467a1e3a 100644 --- a/xen/arch/x86/pv/emul-priv-op.c +++ b/xen/arch/x86/pv/emul-priv-op.c @@ -138,6 +138,8 @@ static io_emul_stub_t *io_emul_stub_setup(struct priv_op_ctxt *ctxt, u8 opcode, /* Runtime confirmation that we haven't clobbered an adjacent stub. */ BUG_ON(STUB_BUF_SIZE / 2 < (p - ctxt->io_emul_stub)); + block_speculation(); /* SCSB */ + /* Handy function-typed pointer to the stub. */ return (void *)stub_va; diff --git a/xen/arch/x86/x86_emulate/x86_emulate.c b/xen/arch/x86/x86_emulate/x86_emulate.c index 31fdec030c..45828b726b 100644 --- a/xen/arch/x86/x86_emulate/x86_emulate.c +++ b/xen/arch/x86/x86_emulate/x86_emulate.c @@ -1257,6 +1257,7 @@ static inline int mkec(uint8_t e, int32_t ec, ...) # define invoke_stub(pre, post, constraints...) do { \ stub_exn.info = (union stub_exception_token) { .raw = ~0 }; \ stub_exn.line = __LINE__; /* Utility outweighs livepatching cost */ \ + block_speculation(); /* SCSB */ \ asm volatile ( pre "\n\tINDIRECT_CALL %[stub]\n\t" post "\n" \ ".Lret%=:\n\t" \ ".pushsection .fixup,\"ax\"\n" \ -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:26:07 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:26:07 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146598.269780 (Exim 4.92) (envelope-from ) id 1lwMYg-0005JX-Vr; Thu, 24 Jun 2021 10:26:06 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146598.269780; Thu, 24 Jun 2021 10:26:06 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMYg-0005JP-Ss; Thu, 24 Jun 2021 10:26:06 +0000 Received: by outflank-mailman (input) for mailman id 146598; Thu, 24 Jun 2021 10:26:05 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMYf-0005J0-24 for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:26:05 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMYf-00062p-1O for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:26:05 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwMYf-0006WT-0l for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:26:05 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=kBGxv057G6WyonQ/jlZIKfnj84+yF/OfY9luaswW53U=; b=Ciu802X3n608GlTk1SdcPUf1Aq X+nGIUzgsG3kJWPDVK6iqxTeDvtQBbp7za/kgMz5WdH/o68gdnJxXltZZoCPRDs9pgX/wXFx8PcAJ XqKiSmFmfFiEZEBSWcogeKGlGB7maimjPixv40chjqi6gPlJvDmiO8+RCFbMTfjYlQLc=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] x86/spec-ctrl: Mitigate TAA after S3 resume Message-Id: Date: Thu, 24 Jun 2021 10:26:05 +0000 commit 8cf276cb2e0b99b96333865873f56b0b31555ff1 Author: Andrew Cooper AuthorDate: Thu May 20 01:21:39 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 17:43:06 2021 +0100 x86/spec-ctrl: Mitigate TAA after S3 resume The user chosen setting for MSR_TSX_CTRL needs restoring after S3. All APs get the correct setting via start_secondary(), but the BSP was missed out. This is XSA-377 / CVE-2021-28690. Fixes: 8c4330818f6 ("x86/spec-ctrl: Mitigate the TSX Asynchronous Abort sidechannel") Signed-off-by: Andrew Cooper Reviewed-by: Jan Beulich --- xen/arch/x86/acpi/power.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/xen/arch/x86/acpi/power.c b/xen/arch/x86/acpi/power.c index 91a8c4d0bd..31a56f02d0 100644 --- a/xen/arch/x86/acpi/power.c +++ b/xen/arch/x86/acpi/power.c @@ -288,6 +288,8 @@ static int enter_state(u32 state) microcode_update_one(); + tsx_init(); /* Needs microcode. May change HLE/RTM feature bits. */ + if ( !recheck_cpu_features(0) ) panic("Missing previously available feature(s)\n"); -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:26:17 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:26:17 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146599.269784 (Exim 4.92) (envelope-from ) id 1lwMYr-0005N7-13; Thu, 24 Jun 2021 10:26:17 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146599.269784; Thu, 24 Jun 2021 10:26:17 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMYq-0005Mq-UO; Thu, 24 Jun 2021 10:26:16 +0000 Received: by outflank-mailman (input) for mailman id 146599; Thu, 24 Jun 2021 10:26:15 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMYp-0005MY-5G for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:26:15 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMYp-00063K-4Q for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:26:15 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwMYp-0006XN-3l for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:26:15 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=JdsC/GMbXl9XUf5d9LsuFjjq8NuxqqzXJX8c4r0Bjew=; b=BsuEY/cguN949O3m+C5j9i6ZV0 zLAn7NFqzD5rf3ZLFaCad6YmdPNCtpnrxIxAeMWLR47iErSMRUhtEXQ87LqVdUCTou+TmjBVtNfm1 buBm9jKsesIFEdh/ox3cz23tO2cedqQFKkEX0Yq+DGUwPgbEt2mggmVmmoHrSJuGQbVM=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] AMD/IOMMU: wait for command slot to be available Message-Id: Date: Thu, 24 Jun 2021 10:26:15 +0000 commit c4e44343579e2c3304d676404d81b2b9bd893c5b Author: Jan Beulich AuthorDate: Tue Jun 8 18:29:40 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 18:29:40 2021 +0100 AMD/IOMMU: wait for command slot to be available No caller cared about send_iommu_command() indicating unavailability of a slot. Hence if a sufficient number prior commands timed out, we did blindly assume that the requested command was submitted to the IOMMU when really it wasn't. This could mean both a hanging system (waiting for a command to complete that was never seen by the IOMMU) or blindly propagating success back to callers, making them believe they're fine to e.g. free previously unmapped pages. Fold the three involved functions into one, add spin waiting for an available slot along the lines of VT-d's qinval_next_index(), and as a consequence drop all error indicator return types/values. This is part of XSA-373 / CVE-2021-28692. Signed-off-by: Jan Beulich Reviewed-by: Paul Durrant --- xen/drivers/passthrough/amd/iommu_cmd.c | 39 +++++++++++---------------------- 1 file changed, 13 insertions(+), 26 deletions(-) diff --git a/xen/drivers/passthrough/amd/iommu_cmd.c b/xen/drivers/passthrough/amd/iommu_cmd.c index 54d7528f96..135883aaf8 100644 --- a/xen/drivers/passthrough/amd/iommu_cmd.c +++ b/xen/drivers/passthrough/amd/iommu_cmd.c @@ -20,43 +20,30 @@ #include "iommu.h" #include "../ats.h" -static int queue_iommu_command(struct amd_iommu *iommu, u32 cmd[]) +static void send_iommu_command(struct amd_iommu *iommu, + const uint32_t cmd[4]) { - uint32_t tail, head; + uint32_t tail; tail = iommu->cmd_buffer.tail + sizeof(cmd_entry_t); if ( tail == iommu->cmd_buffer.size ) tail = 0; - head = readl(iommu->mmio_base + - IOMMU_CMD_BUFFER_HEAD_OFFSET) & IOMMU_RING_BUFFER_PTR_MASK; - if ( head != tail ) + while ( tail == (readl(iommu->mmio_base + + IOMMU_CMD_BUFFER_HEAD_OFFSET) & + IOMMU_RING_BUFFER_PTR_MASK) ) { - memcpy(iommu->cmd_buffer.buffer + iommu->cmd_buffer.tail, - cmd, sizeof(cmd_entry_t)); - - iommu->cmd_buffer.tail = tail; - return 1; + printk_once(XENLOG_ERR "AMD IOMMU %pp: no cmd slot available\n", + &PCI_SBDF2(iommu->seg, iommu->bdf)); + cpu_relax(); } - return 0; -} - -static void commit_iommu_command_buffer(struct amd_iommu *iommu) -{ - writel(iommu->cmd_buffer.tail, - iommu->mmio_base + IOMMU_CMD_BUFFER_TAIL_OFFSET); -} + memcpy(iommu->cmd_buffer.buffer + iommu->cmd_buffer.tail, + cmd, sizeof(cmd_entry_t)); -static int send_iommu_command(struct amd_iommu *iommu, u32 cmd[]) -{ - if ( queue_iommu_command(iommu, cmd) ) - { - commit_iommu_command_buffer(iommu); - return 1; - } + iommu->cmd_buffer.tail = tail; - return 0; + writel(tail, iommu->mmio_base + IOMMU_CMD_BUFFER_TAIL_OFFSET); } static void flush_command_buffer(struct amd_iommu *iommu) -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:26:26 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:26:26 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146600.269787 (Exim 4.92) (envelope-from ) id 1lwMZ0-0005QA-2Z; Thu, 24 Jun 2021 10:26:26 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146600.269787; Thu, 24 Jun 2021 10:26:26 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMYz-0005Q2-Vv; Thu, 24 Jun 2021 10:26:25 +0000 Received: by outflank-mailman (input) for mailman id 146600; Thu, 24 Jun 2021 10:26:25 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMYz-0005Pt-86 for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:26:25 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMYz-00063V-7O for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:26:25 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwMYz-0006Y6-6g for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:26:25 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=F6codjMCMt5tnAMo8sXQ+6B2sACgg87vmbfe/++KO78=; b=VcTMW2E7lf61AMqICOWipWlXu+ eSf/5YhmBsA3sm/Zb+DcqpRl3LercHNsiZf0eZj1vBCYvXQd3Rjb6DuV0gwaAazrhFmpcSW0Lpucd DLTvL3aO6ycOnoL6QQY/L6/3NYORch1LU6YB4tOndORueEqW2jnfV9ok+vbw6Ckh3eis=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] AMD/IOMMU: drop command completion timeout Message-Id: Date: Thu, 24 Jun 2021 10:26:25 +0000 commit e4fee66043120c954fc309bbb37813604c1c0eb7 Author: Jan Beulich AuthorDate: Tue Jun 8 18:29:42 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 8 18:29:42 2021 +0100 AMD/IOMMU: drop command completion timeout First and foremost - such timeouts were not signaled to callers, making them believe they're fine to e.g. free previously unmapped pages. Mirror VT-d's behavior: A fixed number of loop iterations is not a suitable way to detect timeouts in an environment (CPU and bus speeds) independent manner anyway. Furthermore, leaving an in-progress operation pending when it appears to take too long is problematic: If a command completed later, the signaling of its completion may instead be understood to signal a subsequently started command's completion. Log excessively long processing times (with a progressive threshold) to have some indication of problems in this area. Allow callers to specify a non-default timeout bias for this logging, using the same values as VT-d does, which in particular means a (by default) much larger value for device IO TLB invalidation. This is part of XSA-373 / CVE-2021-28692. Signed-off-by: Jan Beulich Reviewed-by: Paul Durrant --- xen/drivers/passthrough/amd/iommu_cmd.c | 56 +++++++++++++++++++-------------- 1 file changed, 33 insertions(+), 23 deletions(-) diff --git a/xen/drivers/passthrough/amd/iommu_cmd.c b/xen/drivers/passthrough/amd/iommu_cmd.c index 135883aaf8..92a688dc9d 100644 --- a/xen/drivers/passthrough/amd/iommu_cmd.c +++ b/xen/drivers/passthrough/amd/iommu_cmd.c @@ -46,10 +46,12 @@ static void send_iommu_command(struct amd_iommu *iommu, writel(tail, iommu->mmio_base + IOMMU_CMD_BUFFER_TAIL_OFFSET); } -static void flush_command_buffer(struct amd_iommu *iommu) +static void flush_command_buffer(struct amd_iommu *iommu, + unsigned int timeout_base) { - unsigned int cmd[4], status, loop_count; - bool comp_wait; + uint32_t cmd[4]; + s_time_t start, timeout; + static unsigned int __read_mostly threshold = 1; /* RW1C 'ComWaitInt' in status register */ writel(IOMMU_STATUS_COMP_WAIT_INT, @@ -65,22 +67,29 @@ static void flush_command_buffer(struct amd_iommu *iommu) IOMMU_COMP_WAIT_I_FLAG_SHIFT, &cmd[0]); send_iommu_command(iommu, cmd); - /* Make loop_count long enough for polling completion wait bit */ - loop_count = 1000; - do { - status = readl(iommu->mmio_base + IOMMU_STATUS_MMIO_OFFSET); - comp_wait = status & IOMMU_STATUS_COMP_WAIT_INT; - --loop_count; - } while ( !comp_wait && loop_count ); - - if ( comp_wait ) + start = NOW(); + timeout = start + (timeout_base ?: 100) * MILLISECS(threshold); + while ( !(readl(iommu->mmio_base + IOMMU_STATUS_MMIO_OFFSET) & + IOMMU_STATUS_COMP_WAIT_INT) ) { - /* RW1C 'ComWaitInt' in status register */ - writel(IOMMU_STATUS_COMP_WAIT_INT, - iommu->mmio_base + IOMMU_STATUS_MMIO_OFFSET); - return; + if ( timeout && NOW() > timeout ) + { + threshold |= threshold << 1; + printk(XENLOG_WARNING + "AMD IOMMU %pp: %scompletion wait taking too long\n", + &PCI_SBDF2(iommu->seg, iommu->bdf), + timeout_base ? "iotlb " : ""); + timeout = 0; + } + cpu_relax(); } - AMD_IOMMU_DEBUG("Warning: ComWaitInt bit did not assert!\n"); + + if ( !timeout ) + printk(XENLOG_WARNING + "AMD IOMMU %pp: %scompletion wait took %lums\n", + &PCI_SBDF2(iommu->seg, iommu->bdf), + timeout_base ? "iotlb " : "", + (NOW() - start) / 10000000); } /* Build low level iommu command messages */ @@ -291,7 +300,7 @@ void amd_iommu_flush_iotlb(u8 devfn, const struct pci_dev *pdev, /* send INVALIDATE_IOTLB_PAGES command */ spin_lock_irqsave(&iommu->lock, flags); invalidate_iotlb_pages(iommu, maxpend, 0, queueid, daddr, req_id, order); - flush_command_buffer(iommu); + flush_command_buffer(iommu, iommu_dev_iotlb_timeout); spin_unlock_irqrestore(&iommu->lock, flags); } @@ -328,7 +337,7 @@ static void _amd_iommu_flush_pages(struct domain *d, { spin_lock_irqsave(&iommu->lock, flags); invalidate_iommu_pages(iommu, daddr, dom_id, order); - flush_command_buffer(iommu); + flush_command_buffer(iommu, 0); spin_unlock_irqrestore(&iommu->lock, flags); } @@ -352,7 +361,7 @@ void amd_iommu_flush_device(struct amd_iommu *iommu, uint16_t bdf) ASSERT( spin_is_locked(&iommu->lock) ); invalidate_dev_table_entry(iommu, bdf); - flush_command_buffer(iommu); + flush_command_buffer(iommu, 0); } void amd_iommu_flush_intremap(struct amd_iommu *iommu, uint16_t bdf) @@ -360,7 +369,7 @@ void amd_iommu_flush_intremap(struct amd_iommu *iommu, uint16_t bdf) ASSERT( spin_is_locked(&iommu->lock) ); invalidate_interrupt_table(iommu, bdf); - flush_command_buffer(iommu); + flush_command_buffer(iommu, 0); } void amd_iommu_flush_all_caches(struct amd_iommu *iommu) @@ -368,7 +377,7 @@ void amd_iommu_flush_all_caches(struct amd_iommu *iommu) ASSERT( spin_is_locked(&iommu->lock) ); invalidate_iommu_all(iommu); - flush_command_buffer(iommu); + flush_command_buffer(iommu, 0); } void amd_iommu_send_guest_cmd(struct amd_iommu *iommu, u32 cmd[]) @@ -378,7 +387,8 @@ void amd_iommu_send_guest_cmd(struct amd_iommu *iommu, u32 cmd[]) spin_lock_irqsave(&iommu->lock, flags); send_iommu_command(iommu, cmd); - flush_command_buffer(iommu); + /* TBD: Timeout selection may require peeking into cmd[]. */ + flush_command_buffer(iommu, 0); spin_unlock_irqrestore(&iommu->lock, flags); } -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:26:36 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:26:36 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146601.269792 (Exim 4.92) (envelope-from ) id 1lwMZA-0005Sr-4O; Thu, 24 Jun 2021 10:26:36 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146601.269792; Thu, 24 Jun 2021 10:26:36 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMZA-0005Sj-18; Thu, 24 Jun 2021 10:26:36 +0000 Received: by outflank-mailman (input) for mailman id 146601; Thu, 24 Jun 2021 10:26:35 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMZ9-0005SX-B7 for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:26:35 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMZ9-00063j-AP for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:26:35 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwMZ9-0006Yx-9f for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:26:35 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=LD2wFnsJkPpacI6WKIFdGOG3V5EiXz9kiDT6EIlac34=; b=JxIaOd3HrdwC2ruEzSrdbHGcAF CB4MdXE/wGuGpbPl34m0jwuNdVa8w+9Le1m4l+UknI7WsAGaj0nvl1JZpdsadmCdnh65nCd/aQvtc yyti1ZpXFqc/kc+163xWX46bOkwx0uPVC8N5n2O6FIFNnDExiIXqGVBRZ7eoF8f8OOMU=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] MAINTAINERS: adjust x86/mm/shadow maintainers Message-Id: Date: Thu, 24 Jun 2021 10:26:35 +0000 commit 87d49efd3af7ea70e1b8250d619411e57da9e520 Author: Tim Deegan AuthorDate: Wed Jun 9 12:43:25 2021 +0200 Commit: Jan Beulich CommitDate: Wed Jun 9 12:43:25 2021 +0200 MAINTAINERS: adjust x86/mm/shadow maintainers Better reflect reality: Andrew and Jan are active maintainers and I review patches. Keep myself as a reviewer so I can help with historical context &c. Signed-off-by: Tim Deegan Acked-by: Jan Beulich Acked-by: Andrew Cooper --- MAINTAINERS | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/MAINTAINERS b/MAINTAINERS index d46b08a0d2..09a92534bf 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -591,7 +591,9 @@ F: xen/arch/x86/mm/mem_sharing.c F: tools/tests/mem-sharing/ X86 SHADOW PAGETABLES -M: Tim Deegan +M: Jan Beulich +M: Andrew Cooper +R: Tim Deegan S: Maintained F: xen/arch/x86/mm/shadow/ -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:26:46 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:26:46 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146602.269796 (Exim 4.92) (envelope-from ) id 1lwMZK-0005Vj-5m; Thu, 24 Jun 2021 10:26:46 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146602.269796; Thu, 24 Jun 2021 10:26:46 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMZK-0005Vb-2c; Thu, 24 Jun 2021 10:26:46 +0000 Received: by outflank-mailman (input) for mailman id 146602; Thu, 24 Jun 2021 10:26:45 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMZJ-0005VS-E9 for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:26:45 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMZJ-00063u-DP for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:26:45 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwMZJ-0006Zj-CZ for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:26:45 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=07Q8WUminXV90BybN5Y8+iNy/+MmPGmVS8xUOb49hPs=; b=qHnYdznsl4FS8rhwLjvfnIAU/m l9JJjsoV7muAhOix5eI+BVLPAtN1/VtDx/iTnij7LLOswZYCBCaSgCkvwdkhUvUMvtxcj6s012Eci tERq/hHGqueNC1C1CdheTvZ24Mkj+BrrBSo6hxHQJhr8DYIv8yK5RukgnK7S1A2skZmA=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] xen/char: default HAS_NS16550 to y only for X86 and ARM Message-Id: Date: Thu, 24 Jun 2021 10:26:45 +0000 commit 5151ce8ba1f155ecc6403f3a51e618a47a382262 Author: Connor Davis AuthorDate: Wed Jun 9 12:43:45 2021 +0200 Commit: Jan Beulich CommitDate: Wed Jun 9 12:43:45 2021 +0200 xen/char: default HAS_NS16550 to y only for X86 and ARM Defaulting to yes only for X86 and ARM reduces the requirements for a minimal build when porting new architectures. Signed-off-by: Connor Davis Acked-by: Jan Beulich Reviewed-by: Alistair Francis --- xen/drivers/char/Kconfig | 1 + 1 file changed, 1 insertion(+) diff --git a/xen/drivers/char/Kconfig b/xen/drivers/char/Kconfig index b572305657..2ff5b288e2 100644 --- a/xen/drivers/char/Kconfig +++ b/xen/drivers/char/Kconfig @@ -1,5 +1,6 @@ config HAS_NS16550 bool "NS16550 UART driver" if ARM + default n if RISCV default y help This selects the 16550-series UART support. For most systems, say Y. -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:26:56 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:26:56 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146603.269799 (Exim 4.92) (envelope-from ) id 1lwMZU-0005Ys-7H; Thu, 24 Jun 2021 10:26:56 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146603.269799; Thu, 24 Jun 2021 10:26:56 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMZU-0005Yk-4G; Thu, 24 Jun 2021 10:26:56 +0000 Received: by outflank-mailman (input) for mailman id 146603; Thu, 24 Jun 2021 10:26:55 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMZT-0005YZ-HB for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:26:55 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMZT-000647-GU for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:26:55 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwMZT-0006aZ-Fl for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:26:55 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=8zPytgZDvzr5L+9/lijPDT3ecSFWjuVd0Mqec4OdNHY=; b=C4GUlKc0S2QTtSdbibnuCoUuW2 440imjw1GbOt87sO4Si7ZcMsNYfa5f4Kf2nALjoAB1iV2CyADDp/F/tielKT/zZcav1dJU5QnZISq HmVwLcHyiKq7EufoFUuJ2u7B9ztLcL2C9PH9sGVg7RS6b68Q/xs65SkkfJhKQ83vB6CE=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] xen: add files needed for minimal riscv build Message-Id: Date: Thu, 24 Jun 2021 10:26:55 +0000 commit f5035d480f7a7033f15765b67c19df86a8ef2c69 Author: Connor Davis AuthorDate: Wed Jun 9 12:45:38 2021 +0200 Commit: Jan Beulich CommitDate: Wed Jun 9 12:45:38 2021 +0200 xen: add files needed for minimal riscv build Add arch-specific makefiles and configs needed to build for riscv. Also add a minimal head.S that is a simple infinite loop. head.o can be built with $ make XEN_TARGET_ARCH=riscv64 SUBSYSTEMS=xen -C xen tiny64_defconfig $ make XEN_TARGET_ARCH=riscv64 SUBSYSTEMS=xen -C xen TARGET=riscv64/head.o No other TARGET is supported at the moment. Signed-off-by: Connor Davis Reviewed-by: Alistair Francis Acked-by: Bobby Eshleman --- MAINTAINERS | 9 +++++++ config/riscv64.mk | 5 ++++ xen/Makefile | 8 ++++-- xen/arch/riscv/Kconfig | 48 +++++++++++++++++++++++++++++++++ xen/arch/riscv/Kconfig.debug | 0 xen/arch/riscv/Makefile | 2 ++ xen/arch/riscv/Rules.mk | 0 xen/arch/riscv/arch.mk | 14 ++++++++++ xen/arch/riscv/configs/tiny64_defconfig | 13 +++++++++ xen/arch/riscv/riscv64/asm-offsets.c | 0 xen/arch/riscv/riscv64/head.S | 6 +++++ xen/include/asm-riscv/config.h | 47 ++++++++++++++++++++++++++++++++ 12 files changed, 150 insertions(+), 2 deletions(-) diff --git a/MAINTAINERS b/MAINTAINERS index 09a92534bf..7778c64fc7 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -456,6 +456,15 @@ F: tools/libs/light/libxl_nonetbuffer.c F: tools/hotplug/Linux/remus-netbuf-setup F: tools/hotplug/Linux/block-drbd-probe +RISCV +M: Bob Eshleman +M: Alistair Francis +R: Connor Davis +S: Supported +F: config/riscv64.mk +F: xen/arch/riscv/ +F: xen/include/asm-riscv/ + RTDS SCHEDULER M: Dario Faggioli M: Meng Xu diff --git a/config/riscv64.mk b/config/riscv64.mk new file mode 100644 index 0000000000..a5a21e5fa2 --- /dev/null +++ b/config/riscv64.mk @@ -0,0 +1,5 @@ +CONFIG_RISCV := y +CONFIG_RISCV_64 := y +CONFIG_RISCV_$(XEN_OS) := y + +CONFIG_XEN_INSTALL_SUFFIX := diff --git a/xen/Makefile b/xen/Makefile index 7ce7692354..89879fad4c 100644 --- a/xen/Makefile +++ b/xen/Makefile @@ -26,7 +26,9 @@ MAKEFLAGS += -rR EFI_MOUNTPOINT ?= $(BOOT_DIR)/efi ARCH=$(XEN_TARGET_ARCH) -SRCARCH=$(shell echo $(ARCH) | sed -e 's/x86.*/x86/' -e s'/arm\(32\|64\)/arm/g') +SRCARCH=$(shell echo $(ARCH) | \ + sed -e 's/x86.*/x86/' -e s'/arm\(32\|64\)/arm/g' \ + -e s'/riscv.*/riscv/g') # Don't break if the build process wasn't called from the top level # we need XEN_TARGET_ARCH to generate the proper config @@ -35,7 +37,8 @@ include $(XEN_ROOT)/Config.mk # Set ARCH/SUBARCH appropriately. export TARGET_SUBARCH := $(XEN_TARGET_ARCH) export TARGET_ARCH := $(shell echo $(XEN_TARGET_ARCH) | \ - sed -e 's/x86.*/x86/' -e s'/arm\(32\|64\)/arm/g') + sed -e 's/x86.*/x86/' -e s'/arm\(32\|64\)/arm/g' \ + -e s'/riscv.*/riscv/g') # Allow someone to change their config file export KCONFIG_CONFIG ?= .config @@ -335,6 +338,7 @@ _clean: delete-unfresh-files $(MAKE) $(clean) xsm $(MAKE) $(clean) crypto $(MAKE) $(clean) arch/arm + $(MAKE) $(clean) arch/riscv $(MAKE) $(clean) arch/x86 $(MAKE) $(clean) test $(MAKE) -f $(BASEDIR)/tools/kconfig/Makefile.kconfig ARCH=$(ARCH) SRCARCH=$(SRCARCH) clean diff --git a/xen/arch/riscv/Kconfig b/xen/arch/riscv/Kconfig new file mode 100644 index 0000000000..468e250c86 --- /dev/null +++ b/xen/arch/riscv/Kconfig @@ -0,0 +1,48 @@ +config RISCV + def_bool y + +config RISCV_64 + def_bool y + select 64BIT + +config ARCH_DEFCONFIG + string + default "arch/riscv/configs/tiny64_defconfig" + +menu "Architecture Features" + +source "arch/Kconfig" + +endmenu + +menu "ISA Selection" + +choice + prompt "Base ISA" + default RISCV_ISA_RV64IMA if RISCV_64 + help + This selects the base ISA extensions that Xen will target. + +config RISCV_ISA_RV64IMA + bool "RV64IMA" + help + Use the RV64I base ISA, plus the "M" and "A" extensions + for integer multiply/divide and atomic instructions, respectively. + +endchoice + +config RISCV_ISA_C + bool "Compressed extension" + default y + help + Add "C" to the ISA subsets that the toolchain is allowed to + emit when building Xen, which results in compressed instructions + in the Xen binary. + + If unsure, say Y. + +endmenu + +source "common/Kconfig" + +source "drivers/Kconfig" diff --git a/xen/arch/riscv/Kconfig.debug b/xen/arch/riscv/Kconfig.debug new file mode 100644 index 0000000000..e69de29bb2 diff --git a/xen/arch/riscv/Makefile b/xen/arch/riscv/Makefile new file mode 100644 index 0000000000..942e4ffbc1 --- /dev/null +++ b/xen/arch/riscv/Makefile @@ -0,0 +1,2 @@ +.PHONY: include +include: diff --git a/xen/arch/riscv/Rules.mk b/xen/arch/riscv/Rules.mk new file mode 100644 index 0000000000..e69de29bb2 diff --git a/xen/arch/riscv/arch.mk b/xen/arch/riscv/arch.mk new file mode 100644 index 0000000000..53dadb8975 --- /dev/null +++ b/xen/arch/riscv/arch.mk @@ -0,0 +1,14 @@ +######################################## +# RISCV-specific definitions + +CFLAGS-$(CONFIG_RISCV_64) += -mabi=lp64 + +riscv-march-$(CONFIG_RISCV_ISA_RV64IMA) := rv64ima +riscv-march-$(CONFIG_RISCV_ISA_C) := $(riscv-march-y)c + +# Note that -mcmodel=medany is used so that Xen can be mapped +# into the upper half _or_ the lower half of the address space. +# -mcmodel=medlow would force Xen into the lower half. + +CFLAGS += -march=$(riscv-march-y) -mstrict-align -mcmodel=medany +CFLAGS += -I$(BASEDIR)/include diff --git a/xen/arch/riscv/configs/tiny64_defconfig b/xen/arch/riscv/configs/tiny64_defconfig new file mode 100644 index 0000000000..3c9a2ff941 --- /dev/null +++ b/xen/arch/riscv/configs/tiny64_defconfig @@ -0,0 +1,13 @@ +# CONFIG_SCHED_CREDIT is not set +# CONFIG_SCHED_RTDS is not set +# CONFIG_SCHED_NULL is not set +# CONFIG_SCHED_ARINC653 is not set +# CONFIG_TRACEBUFFER is not set +# CONFIG_HYPFS is not set +# CONFIG_GRANT_TABLE is not set +# CONFIG_SPECULATIVE_HARDEN_ARRAY is not set + +CONFIG_RISCV_64=y +CONFIG_DEBUG=y +CONFIG_DEBUG_INFO=y +CONFIG_EXPERT=y diff --git a/xen/arch/riscv/riscv64/asm-offsets.c b/xen/arch/riscv/riscv64/asm-offsets.c new file mode 100644 index 0000000000..e69de29bb2 diff --git a/xen/arch/riscv/riscv64/head.S b/xen/arch/riscv/riscv64/head.S new file mode 100644 index 0000000000..0dbc27ba75 --- /dev/null +++ b/xen/arch/riscv/riscv64/head.S @@ -0,0 +1,6 @@ +#include + + .text + +ENTRY(start) + j start diff --git a/xen/include/asm-riscv/config.h b/xen/include/asm-riscv/config.h new file mode 100644 index 0000000000..e2ae21de61 --- /dev/null +++ b/xen/include/asm-riscv/config.h @@ -0,0 +1,47 @@ +#ifndef __RISCV_CONFIG_H__ +#define __RISCV_CONFIG_H__ + +#if defined(CONFIG_RISCV_64) +# define LONG_BYTEORDER 3 +# define ELFSIZE 64 +# define MAX_VIRT_CPUS 128u +#else +# error "Unsupported RISCV variant" +#endif + +#define BYTES_PER_LONG (1 << LONG_BYTEORDER) +#define BITS_PER_LONG (BYTES_PER_LONG << 3) +#define POINTER_ALIGN BYTES_PER_LONG + +#define BITS_PER_LLONG 64 + +/* xen_ulong_t is always 64 bits */ +#define BITS_PER_XEN_ULONG 64 + +#define CONFIG_RISCV_L1_CACHE_SHIFT 6 +#define CONFIG_PAGEALLOC_MAX_ORDER 18 +#define CONFIG_DOMU_MAX_ORDER 9 +#define CONFIG_HWDOM_MAX_ORDER 10 + +#define OPT_CONSOLE_STR "dtuart" +#define INVALID_VCPU_ID MAX_VIRT_CPUS + +/* Linkage for RISCV */ +#ifdef __ASSEMBLY__ +#define ALIGN .align 2 + +#define ENTRY(name) \ + .globl name; \ + ALIGN; \ + name: +#endif + +#endif /* __RISCV_CONFIG_H__ */ +/* + * Local variables: + * mode: C + * c-file-style: "BSD" + * c-basic-offset: 4 + * indent-tabs-mode: nil + * End: + */ -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:27:08 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:27:08 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146604.269805 (Exim 4.92) (envelope-from ) id 1lwMZe-0005c5-BE; Thu, 24 Jun 2021 10:27:06 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146604.269805; Thu, 24 Jun 2021 10:27:06 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMZe-0005bt-7V; Thu, 24 Jun 2021 10:27:06 +0000 Received: by outflank-mailman (input) for mailman id 146604; Thu, 24 Jun 2021 10:27:05 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMZd-0005bj-Kk for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:27:05 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMZd-00064Y-K1 for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:27:05 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwMZd-0006bV-JI for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:27:05 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=qAhjyTWehomI3VRceXgH/0RJIi2TYFREcJYuMLFAFpI=; b=CUZKHXDUkqsItNdzKk5+FTPULW /iTRdJOUjDs9TyzQ6qgQv/ZBhGaIx2k+bNTBiiF/7+wwRJfC91XE+VA1NbG81U4GcIuQj9K6Baf9f f1iXgHtZQU048c18gdHYqohEm9pO88GT7/Of0n9lLJ9A5xEbvvoxyhXblN+YTLT6C+fA=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] x86/tsx: Cope with TSX deprecation on SKL/KBL/CFL/WHL Message-Id: Date: Thu, 24 Jun 2021 10:27:05 +0000 commit 3e09045991cde360432bc7437103f8f8a6699359 Author: Andrew Cooper AuthorDate: Wed Sep 16 16:15:52 2020 +0100 Commit: Andrew Cooper CommitDate: Wed Jun 9 14:17:58 2021 +0100 x86/tsx: Cope with TSX deprecation on SKL/KBL/CFL/WHL The June 2021 microcode is formally de-featuring TSX on the older Skylake client CPUs. The workaround from the March 2019 microcode is being dropped, and replaced with additions to MSR_TSX_FORCE_ABORT to hide the HLE/RTM CPUID bits. With this microcode in place, TSX is disabled by default on these CPUs. Backwards compatibility is provided in the same way as for TAA - RTM force aborts, rather than suffering #UD, and the CPUID bits can be hidden to recover performance. Signed-off-by: Andrew Cooper Reviewed-by: Jan Beulich --- docs/misc/xen-command-line.pandoc | 13 +++++ tools/misc/xen-cpuid.c | 2 +- xen/arch/x86/tsx.c | 76 +++++++++++++++++++++++++++-- xen/include/asm-x86/cpufeature.h | 1 + xen/include/asm-x86/msr-index.h | 2 + xen/include/public/arch-x86/cpufeatureset.h | 1 + 6 files changed, 89 insertions(+), 6 deletions(-) diff --git a/docs/misc/xen-command-line.pandoc b/docs/misc/xen-command-line.pandoc index 1fae872626..3ece83a427 100644 --- a/docs/misc/xen-command-line.pandoc +++ b/docs/misc/xen-command-line.pandoc @@ -2309,6 +2309,12 @@ Several microcode updates are relevant: and formally retiring HLE from the architecture. The user can disable TSX to mitigate TAA, and elect to hide the HLE/RTM CPUID bits. + * June 2021, removing the workaround for March 2019 on client CPUs and + formally de-featured TSX on SKL/KBL/WHL/CFL (Note: SKX still retains the + March 2019 fix). Introduced the ability to hide the HLE/RTM CPUID bits. + PCR3 works fine, and TSX is disabled by default, but the user can re-enable + TSX at their own risk, accepting that the memory order erratum is unfixed. + On systems with the ability to configure TSX, this boolean offers system wide control of whether TSX is enabled or disabled. @@ -2326,6 +2332,13 @@ control of whether TSX is enabled or disabled. ordering errata default to `true` to enable working TSX. Alternatively, selecting `tsx=0` will disable TSX and restore PCR3 to a working state. + SKX and SKL/KBL/WHL/CFL on pre-June 2021 microcode default to `true`. + Alternatively, selecting `tsx=0` will disable TSX and restore PCR3 to a + working state. + + SKL/KBL/WHL/CFL on the June 2021 microcode or later default to `false`. + Alternatively, selecting `tsx=1` will re-enable TSX at the users own risk. + ### ucode > `= List of [ | scan=, nmi=, allow-same= ]` diff --git a/tools/misc/xen-cpuid.c b/tools/misc/xen-cpuid.c index d4bc83d8c9..735bcf8f0e 100644 --- a/tools/misc/xen-cpuid.c +++ b/tools/misc/xen-cpuid.c @@ -162,7 +162,7 @@ static const char *const str_7d0[32] = [ 4] = "fsrm", [ 8] = "avx512-vp2intersect", [ 9] = "srbds-ctrl", - [10] = "md-clear", + [10] = "md-clear", [11] = "rtm-always-abort", /* 12 */ [13] = "tsx-force-abort", [14] = "serialize", diff --git a/xen/arch/x86/tsx.c b/xen/arch/x86/tsx.c index 338191df7f..88adf08c49 100644 --- a/xen/arch/x86/tsx.c +++ b/xen/arch/x86/tsx.c @@ -59,6 +59,38 @@ void tsx_init(void) * ordering erratum, with at least the March 2019 microcode. */ + /* + * Probe for the June 2021 microcode which de-features TSX on + * client parts. (Note - this is a subset of parts impacted by + * the memory ordering errata.) + * + * RTM_ALWAYS_ABORT enumerates the new functionality, but is also + * read as zero if TSX_FORCE_ABORT.ENABLE_RTM has been set before + * we run. + * + * Undo this behaviour in Xen's view of the world. + */ + bool has_rtm_always_abort = cpu_has_rtm_always_abort; + + if ( !has_rtm_always_abort ) + { + uint64_t val; + + rdmsrl(MSR_TSX_FORCE_ABORT, val); + + if ( val & TSX_ENABLE_RTM ) + has_rtm_always_abort = true; + } + + /* + * Always force RTM_ALWAYS_ABORT, even if it currently visible. + * If the user explicitly opts to enable TSX, we'll set + * TSX_FORCE_ABORT.ENABLE_RTM and cause RTM_ALWAYS_ABORT to be + * hidden from the general CPUID scan later. + */ + if ( has_rtm_always_abort ) + setup_force_cpu_cap(X86_FEATURE_RTM_ALWAYS_ABORT); + /* * If no explicit tsx= option is provided, pick a default. * @@ -67,9 +99,16 @@ void tsx_init(void) * - parse_spec_ctrl() ran before any CPU details where know. * - We now know we're running on a CPU not affected by TAA (as * TSX_FORCE_ABORT is enumerated). + * - When RTM_ALWAYS_ABORT is enumerated, TSX malfunctions, so we + * only ever want it enabled by explicit user choice. + * + * Without RTM_ALWAYS_ABORT, leave TSX active. In particular, + * this includes SKX where TSX is still supported. + * + * With RTM_ALWAYS_ABORT, disable TSX. */ if ( opt_tsx < 0 ) - opt_tsx = 1; + opt_tsx = !cpu_has_rtm_always_abort; } /* @@ -90,7 +129,7 @@ void tsx_init(void) * Force the features to be visible in Xen's view if we see any of the * infrastructure capable of hiding them. */ - if ( cpu_has_tsx_ctrl ) + if ( cpu_has_tsx_ctrl || cpu_has_tsx_force_abort ) { setup_force_cpu_cap(X86_FEATURE_HLE); setup_force_cpu_cap(X86_FEATURE_RTM); @@ -131,9 +170,36 @@ void tsx_init(void) /* Check bottom bit only. Higher bits are various sentinels. */ rtm_disabled = !(opt_tsx & 1); - lo &= ~TSX_FORCE_ABORT_RTM; - if ( rtm_disabled ) - lo |= TSX_FORCE_ABORT_RTM; + lo &= ~(TSX_FORCE_ABORT_RTM | TSX_CPUID_CLEAR | TSX_ENABLE_RTM); + + if ( cpu_has_rtm_always_abort ) + { + /* + * June 2021 microcode, on a client part with TSX de-featured: + * - There are no mitigations for the TSX memory ordering errata. + * - Performance counter 3 works. (I.e. it isn't being used by + * microcode to work around the memory ordering errata.) + * - TSX_FORCE_ABORT.FORCE_ABORT_RTM is fixed read1/write-discard. + * - TSX_FORCE_ABORT.TSX_CPUID_CLEAR can be used to hide the + * HLE/RTM CPUID bits. + * - TSX_FORCE_ABORT.ENABLE_RTM may be used to opt in to + * re-enabling RTM, at the users own risk. + */ + lo |= rtm_disabled ? TSX_CPUID_CLEAR : TSX_ENABLE_RTM; + } + else + { + /* + * Either a server part where TSX isn't de-featured, or pre-June + * 2021 microcode: + * - By default, the TSX memory ordering errata is worked around + * in microcode at the cost of Performance Counter 3. + * - "Working TSX" vs "Working PCR3" can be selected by way of + * setting TSX_FORCE_ABORT.FORCE_ABORT_RTM. + */ + if ( rtm_disabled ) + lo |= TSX_FORCE_ABORT_RTM; + } wrmsr(MSR_TSX_FORCE_ABORT, lo, hi); } diff --git a/xen/include/asm-x86/cpufeature.h b/xen/include/asm-x86/cpufeature.h index 9f5ae3aa0d..a539a4bacd 100644 --- a/xen/include/asm-x86/cpufeature.h +++ b/xen/include/asm-x86/cpufeature.h @@ -131,6 +131,7 @@ #define cpu_has_avx512_4vnniw boot_cpu_has(X86_FEATURE_AVX512_4VNNIW) #define cpu_has_avx512_4fmaps boot_cpu_has(X86_FEATURE_AVX512_4FMAPS) #define cpu_has_avx512_vp2intersect boot_cpu_has(X86_FEATURE_AVX512_VP2INTERSECT) +#define cpu_has_rtm_always_abort boot_cpu_has(X86_FEATURE_RTM_ALWAYS_ABORT) #define cpu_has_tsx_force_abort boot_cpu_has(X86_FEATURE_TSX_FORCE_ABORT) #define cpu_has_serialize boot_cpu_has(X86_FEATURE_SERIALIZE) #define cpu_has_arch_caps boot_cpu_has(X86_FEATURE_ARCH_CAPS) diff --git a/xen/include/asm-x86/msr-index.h b/xen/include/asm-x86/msr-index.h index bd3a3a1e7f..9a772c12b8 100644 --- a/xen/include/asm-x86/msr-index.h +++ b/xen/include/asm-x86/msr-index.h @@ -61,6 +61,8 @@ #define MSR_TSX_FORCE_ABORT 0x0000010f #define TSX_FORCE_ABORT_RTM (_AC(1, ULL) << 0) +#define TSX_CPUID_CLEAR (_AC(1, ULL) << 1) +#define TSX_ENABLE_RTM (_AC(1, ULL) << 2) #define MSR_TSX_CTRL 0x00000122 #define TSX_CTRL_RTM_DISABLE (_AC(1, ULL) << 0) diff --git a/xen/include/public/arch-x86/cpufeatureset.h b/xen/include/public/arch-x86/cpufeatureset.h index b65af42436..380b51b1b3 100644 --- a/xen/include/public/arch-x86/cpufeatureset.h +++ b/xen/include/public/arch-x86/cpufeatureset.h @@ -265,6 +265,7 @@ XEN_CPUFEATURE(FSRM, 9*32+ 4) /*A Fast Short REP MOVS */ XEN_CPUFEATURE(AVX512_VP2INTERSECT, 9*32+8) /*a VP2INTERSECT{D,Q} insns */ XEN_CPUFEATURE(SRBDS_CTRL, 9*32+ 9) /* MSR_MCU_OPT_CTRL and RNGDS_MITG_DIS. */ XEN_CPUFEATURE(MD_CLEAR, 9*32+10) /*A VERW clears microarchitectural buffers */ +XEN_CPUFEATURE(RTM_ALWAYS_ABORT, 9*32+11) /*! June 2021 TSX defeaturing in microcode. */ XEN_CPUFEATURE(TSX_FORCE_ABORT, 9*32+13) /* MSR_TSX_FORCE_ABORT.RTM_ABORT */ XEN_CPUFEATURE(SERIALIZE, 9*32+14) /*a SERIALIZE insn */ XEN_CPUFEATURE(CET_IBT, 9*32+20) /* CET - Indirect Branch Tracking */ -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:27:16 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:27:16 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146605.269808 (Exim 4.92) (envelope-from ) id 1lwMZo-0005f8-Br; Thu, 24 Jun 2021 10:27:16 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146605.269808; Thu, 24 Jun 2021 10:27:16 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMZo-0005f0-93; Thu, 24 Jun 2021 10:27:16 +0000 Received: by outflank-mailman (input) for mailman id 146605; Thu, 24 Jun 2021 10:27:15 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMZn-0005eu-O5 for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:27:15 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMZn-00064z-NJ for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:27:15 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwMZn-0006dq-MX for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:27:15 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=lhsihKQMeHdAB5wG74IXL63qCsX2pks6EVqWUmMaaWw=; b=4DvmRcvflbhP9GURqNS81YcgPu cD/y3ubCmHfeqhpjYzUZ15jkuJcREKPV5HlBVo2paqJ/D4BC/8o3AkblsR4y5L8D1+xHI+Npxh4ir o/uh5K6ie+NJOsx8QGzQ/lyY+XUqagj9XCrRbvE1aiw3dumVAq6RI1nYNtkjIoGpj1KE=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] xen/arm32: SPSR_hyp/SPSR Message-Id: Date: Thu, 24 Jun 2021 10:27:15 +0000 commit dfcffb128be46a3e413eaa941744536fe53c94b6 Author: Stefano Stabellini AuthorDate: Wed Jun 9 10:37:59 2021 -0700 Commit: Stefano Stabellini CommitDate: Wed Jun 9 16:52:29 2021 -0700 xen/arm32: SPSR_hyp/SPSR SPSR_hyp is not meant to be accessed from Hyp mode (EL2); accesses trigger UNPREDICTABLE behaviour. Xen should read/write SPSR instead. See: ARM DDI 0487D.b page G8-5993. This fixes booting Xen/arm32 on QEMU. Signed-off-by: Stefano Stabellini Reviewed-by: Julien Grall Reviewed-by: Edgar E. Iglesias Tested-by: Edgar E. Iglesias --- xen/arch/arm/arm32/entry.S | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/xen/arch/arm/arm32/entry.S b/xen/arch/arm/arm32/entry.S index f2f1bc7a31..4e421109db 100644 --- a/xen/arch/arm/arm32/entry.S +++ b/xen/arch/arm/arm32/entry.S @@ -170,7 +170,7 @@ ENDPROC(prepare_context_from_guest) mrc CP32(r11, HSR) /* Save exception syndrome */ str r11, [sp, #UREGS_hsr] - mrs r11, SPSR_hyp + mrs r11, SPSR str r11, [sp, #UREGS_cpsr] /* @@ -395,7 +395,7 @@ return_to_hypervisor: ldr r11, [sp, #UREGS_pc] msr ELR_hyp, r11 ldr r11, [sp, #UREGS_cpsr] - msr SPSR_hyp, r11 + msr SPSR, r11 #ifdef CONFIG_ARM32_HARDEN_BRANCH_PREDICTOR /* * Hardening branch predictor may require to setup a different -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:27:26 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:27:26 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146606.269812 (Exim 4.92) (envelope-from ) id 1lwMZy-0005jq-DK; Thu, 24 Jun 2021 10:27:26 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146606.269812; Thu, 24 Jun 2021 10:27:26 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMZy-0005ji-AW; Thu, 24 Jun 2021 10:27:26 +0000 Received: by outflank-mailman (input) for mailman id 146606; Thu, 24 Jun 2021 10:27:25 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMZx-0005jc-Qy for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:27:25 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMZx-00065G-QE for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:27:25 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwMZx-0006ey-PZ for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:27:25 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=fUzaNRyD0ntZ/JJRr/3PIQllVucfJaF2J1HKt3FLHqE=; b=zzl2Ta4zGGLIkhU9MLELe8UzzY vsZ1zCnVwKHul0J52sih/V2msHsTxtTbZhsI2F1amXu/d+1DdPlYUyIeXDaZ1R9GWGBaYI+QfpP8N S2n6LkjtkT03b812+1BeKO6dN1wR1uCsz8MF1e1VXCgg6Wk76kSOg2M3Sk0m1JL+mP8Y=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] x86: please Clang in arch_set_info_guest() Message-Id: Date: Thu, 24 Jun 2021 10:27:25 +0000 commit 2bb17a45b1814b0b6aa4646eff58e16f876281fd Author: Jan Beulich AuthorDate: Thu Jun 10 16:56:24 2021 +0200 Commit: Jan Beulich CommitDate: Thu Jun 10 16:56:24 2021 +0200 x86: please Clang in arch_set_info_guest() Clang 10 reports domain.c:1328:10: error: variable 'cr3_mfn' is used uninitialized whenever 'if' condition is false [-Werror,-Wsometimes-uninitialized] if ( !compat ) ^~~~~~~ domain.c:1334:34: note: uninitialized use occurs here cr3_page = get_page_from_mfn(cr3_mfn, d); ^~~~~~~ domain.c:1328:5: note: remove the 'if' if its condition is always true if ( !compat ) ^~~~~~~~~~~~~~ domain.c:1042:18: note: initialize the variable 'cr3_mfn' to silence this warning mfn_t cr3_mfn; ^ = 0 domain.c:1189:14: error: variable 'fail' is used uninitialized whenever 'if' condition is false [-Werror,-Wsometimes-uninitialized] if ( !compat ) ^~~~~~~ domain.c:1211:9: note: uninitialized use occurs here fail |= v->arch.pv.gdt_ents != c(gdt_ents); ^~~~ domain.c:1189:9: note: remove the 'if' if its condition is always true if ( !compat ) ^~~~~~~~~~~~~~ domain.c:1187:18: note: initialize the variable 'fail' to silence this warning bool fail; ^ = false despite this being a build with -O2 in effect, and despite "compat" being constant "false" when CONFIG_COMPAT (and hence CONFIG_PV32) is not defined, as it gets set at the top of the function from the result of is_pv_32bit_domain(). Re-arrange the two "offending" if()s such that when COMPAT=n the respective variables will be seen as unconditionally initialized. The original aim was to have the !compat cases first, though. Reported-by: Andrew Cooper Signed-off-by: Jan Beulich Acked-by: Andrew Cooper --- xen/arch/x86/domain.c | 28 ++++++++++++++-------------- 1 file changed, 14 insertions(+), 14 deletions(-) diff --git a/xen/arch/x86/domain.c b/xen/arch/x86/domain.c index 50a27197b5..ef1812dc14 100644 --- a/xen/arch/x86/domain.c +++ b/xen/arch/x86/domain.c @@ -1186,7 +1186,17 @@ int arch_set_info_guest( unsigned long pfn = pagetable_get_pfn(v->arch.guest_table); bool fail; - if ( !compat ) +#ifdef CONFIG_COMPAT + if ( compat ) + { + l4_pgentry_t *l4tab = map_domain_page(_mfn(pfn)); + + pfn = l4e_get_pfn(*l4tab); + unmap_domain_page(l4tab); + fail = compat_pfn_to_cr3(pfn) != c.cmp->ctrlreg[3]; + } + else +#endif { fail = xen_pfn_to_cr3(pfn) != c.nat->ctrlreg[3]; if ( pagetable_is_null(v->arch.guest_table_user) ) @@ -1197,16 +1207,6 @@ int arch_set_info_guest( fail |= xen_pfn_to_cr3(pfn) != c.nat->ctrlreg[1]; } } -#ifdef CONFIG_COMPAT - else - { - l4_pgentry_t *l4tab = map_domain_page(_mfn(pfn)); - - pfn = l4e_get_pfn(*l4tab); - unmap_domain_page(l4tab); - fail = compat_pfn_to_cr3(pfn) != c.cmp->ctrlreg[3]; - } -#endif fail |= v->arch.pv.gdt_ents != c(gdt_ents); for ( i = 0; !fail && i < nr_gdt_frames; ++i ) @@ -1325,12 +1325,12 @@ int arch_set_info_guest( set_bit(_VPF_in_reset, &v->pause_flags); - if ( !compat ) - cr3_mfn = _mfn(xen_cr3_to_pfn(c.nat->ctrlreg[3])); #ifdef CONFIG_COMPAT - else + if ( compat ) cr3_mfn = _mfn(compat_cr3_to_pfn(c.cmp->ctrlreg[3])); + else #endif + cr3_mfn = _mfn(xen_cr3_to_pfn(c.nat->ctrlreg[3])); cr3_page = get_page_from_mfn(cr3_mfn, d); if ( !cr3_page ) -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:27:36 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:27:36 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146607.269818 (Exim 4.92) (envelope-from ) id 1lwMa8-0005mw-Gk; Thu, 24 Jun 2021 10:27:36 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146607.269818; Thu, 24 Jun 2021 10:27:36 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMa8-0005mk-C1; Thu, 24 Jun 2021 10:27:36 +0000 Received: by outflank-mailman (input) for mailman id 146607; Thu, 24 Jun 2021 10:27:36 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMa7-0005md-UT for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:27:35 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMa7-00065R-Th for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:27:35 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwMa7-0006fw-Su for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:27:35 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=GIJtZm6ksBnsbd2QBMKLHGfvfMrTJiD050YaL2Z6+b8=; b=fJrBFkrSeBCIXuI5LVSkQCH8SN uVMmK4jilE60l/tXqosy5fvH74FR1eAjE0SdhOsvX8XxHFm2kiOc+/66UEVN421BBkqYZgLswuyFg nZEx+yGF/kJ8s9yKO7S/vUJ9iSI5/2v5cILE16XynU8MoC3qAmtrNyOXMsXWNk7weuhI=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] tools/libs/store: cleanup libxenstore interface Message-Id: Date: Thu, 24 Jun 2021 10:27:35 +0000 commit d2cad41defe4e0e9987549fbc8ebdf9ae138f90f Author: Juergen Gross AuthorDate: Wed May 12 16:48:32 2021 +0200 Commit: Ian Jackson CommitDate: Fri Jun 11 10:56:47 2021 +0100 tools/libs/store: cleanup libxenstore interface There are some internals in the libxenstore interface which should be removed. Move those functions into xs_lib.c and the related definitions into xs_lib.h. Remove the functions from the mapfile. Add xs_lib.o to xenstore_client as some of the internal functions are needed there. Bump the libxenstore version to 4.0 as the change is incompatible. Note that the removed functions should not result in any problem as they ought to be used by xenstored or xenstore_client only. Avoid an enum as part of a structure as the size of an enum is compiler implementation dependent. Signed-off-by: Juergen Gross Acked-by: Ian Jackson --- tools/include/xenstore_lib.h | 54 +++--------------- tools/libs/store/Makefile | 4 +- tools/libs/store/libxenstore.map | 10 +--- tools/libs/store/xs.c | 112 +----------------------------------- tools/xenstore/Makefile | 4 +- tools/xenstore/utils.h | 11 ++++ tools/xenstore/xenstore_client.c | 2 + tools/xenstore/xenstored_control.c | 1 + tools/xenstore/xenstored_core.c | 19 +++---- tools/xenstore/xenstored_core.h | 6 +- tools/xenstore/xenstored_watch.c | 2 +- tools/xenstore/xs_lib.c | 114 ++++++++++++++++++++++++++++++++++++- tools/xenstore/xs_lib.h | 50 ++++++++++++++++ tools/xenstore/xs_tdb_dump.c | 2 +- 14 files changed, 204 insertions(+), 187 deletions(-) diff --git a/tools/include/xenstore_lib.h b/tools/include/xenstore_lib.h index 4c9b6d1685..2266009ec1 100644 --- a/tools/include/xenstore_lib.h +++ b/tools/include/xenstore_lib.h @@ -26,42 +26,26 @@ #include #include -/* Bitmask of permissions. */ -enum xs_perm_type { - XS_PERM_NONE = 0, - XS_PERM_READ = 1, - XS_PERM_WRITE = 2, - /* Internal use. */ - XS_PERM_ENOENT_OK = 4, - XS_PERM_OWNER = 8, - XS_PERM_IGNORE = 16, -}; - struct xs_permissions { unsigned int id; - enum xs_perm_type perms; -}; - -/* Header of the node record in tdb. */ -struct xs_tdb_record_hdr { - uint64_t generation; - uint32_t num_perms; - uint32_t datalen; - uint32_t childlen; - struct xs_permissions perms[0]; + unsigned int perms; /* Bitmask of permissions. */ +#define XS_PERM_NONE 0x00 +#define XS_PERM_READ 0x01 +#define XS_PERM_WRITE 0x02 + /* Internal use. */ +#define XS_PERM_ENOENT_OK 0x04 +#define XS_PERM_OWNER 0x08 +#define XS_PERM_IGNORE 0x10 }; /* Each 10 bits takes ~ 3 digits, plus one, plus one for nul terminator. */ #define MAX_STRLEN(x) ((sizeof(x) * CHAR_BIT + CHAR_BIT-1) / 10 * 3 + 2) /* Path for various daemon things: env vars can override. */ -const char *xs_daemon_rootdir(void); const char *xs_daemon_rundir(void); const char *xs_daemon_socket(void); const char *xs_daemon_socket_ro(void); -const char *xs_domain_dev(void); -const char *xs_daemon_tdb(void); /* Simple write function: loops for you. */ bool xs_write_all(int fd, const void *data, unsigned int len); @@ -70,26 +54,4 @@ bool xs_write_all(int fd, const void *data, unsigned int len); bool xs_strings_to_perms(struct xs_permissions *perms, unsigned int num, const char *strings); -/* Convert permissions to a string (up to len MAX_STRLEN(unsigned int)+1). */ -bool xs_perm_to_string(const struct xs_permissions *perm, - char *buffer, size_t buf_len); - -/* Given a string and a length, count how many strings (nul terms). */ -unsigned int xs_count_strings(const char *strings, unsigned int len); - -/* Sanitising (quoting) possibly-binary strings. */ -struct expanding_buffer { - char *buf; - int avail; -}; - -/* Ensure that given expanding buffer has at least min_avail characters. */ -char *expanding_buffer_ensure(struct expanding_buffer *, int min_avail); - -/* sanitise_value() may return NULL if malloc fails. */ -char *sanitise_value(struct expanding_buffer *, const char *val, unsigned len); - -/* *out_len_r on entry is ignored; out must be at least strlen(in)+1 bytes. */ -void unsanitise_value(char *out, unsigned *out_len_r, const char *in); - #endif /* XENSTORE_LIB_H */ diff --git a/tools/libs/store/Makefile b/tools/libs/store/Makefile index bee57b5629..43b018aa8c 100644 --- a/tools/libs/store/Makefile +++ b/tools/libs/store/Makefile @@ -1,8 +1,8 @@ XEN_ROOT=$(CURDIR)/../../.. include $(XEN_ROOT)/tools/Rules.mk -MAJOR = 3.0 -MINOR = 3 +MAJOR = 4 +MINOR = 0 ifeq ($(CONFIG_Linux),y) APPEND_LDFLAGS += -ldl diff --git a/tools/libs/store/libxenstore.map b/tools/libs/store/libxenstore.map index 9854305a2c..7e6c7bdd30 100644 --- a/tools/libs/store/libxenstore.map +++ b/tools/libs/store/libxenstore.map @@ -1,4 +1,4 @@ -VERS_3.0.3 { +VERS_4.0 { global: xs_open; xs_close; @@ -32,18 +32,10 @@ VERS_3.0.3 { xs_control_command; xs_debug_command; xs_suspend_evtchn_port; - xs_daemon_rootdir; xs_daemon_rundir; xs_daemon_socket; xs_daemon_socket_ro; - xs_domain_dev; - xs_daemon_tdb; xs_write_all; xs_strings_to_perms; - xs_perm_to_string; - xs_count_strings; - expanding_buffer_ensure; - sanitise_value; - unsanitise_value; local: *; /* Do not expose anything by default */ }; diff --git a/tools/libs/store/xs.c b/tools/libs/store/xs.c index c91377c27f..7a9a8b1656 100644 --- a/tools/libs/store/xs.c +++ b/tools/libs/store/xs.c @@ -34,6 +34,7 @@ #include #include #include "xenstore.h" +#include "xs_lib.h" #include "list.h" #include "utils.h" @@ -1358,117 +1359,6 @@ static void *read_thread(void *arg) } #endif -char *expanding_buffer_ensure(struct expanding_buffer *ebuf, int min_avail) -{ - int want; - char *got; - - if (ebuf->avail >= min_avail) - return ebuf->buf; - - if (min_avail >= INT_MAX/3) - return 0; - - want = ebuf->avail + min_avail + 10; - got = realloc(ebuf->buf, want); - if (!got) - return 0; - - ebuf->buf = got; - ebuf->avail = want; - return ebuf->buf; -} - -char *sanitise_value(struct expanding_buffer *ebuf, - const char *val, unsigned len) -{ - int used, remain, c; - unsigned char *ip; - -#define ADD(c) (ebuf->buf[used++] = (c)) -#define ADDF(f,c) (used += sprintf(ebuf->buf+used, (f), (c))) - - assert(len < INT_MAX/5); - - ip = (unsigned char *)val; - used = 0; - remain = len; - - if (!expanding_buffer_ensure(ebuf, remain + 1)) - return NULL; - - while (remain-- > 0) { - c= *ip++; - - if (c >= ' ' && c <= '~' && c != '\\') { - ADD(c); - continue; - } - - if (!expanding_buffer_ensure(ebuf, used + remain + 5)) - /* for "\\nnn\0" */ - return 0; - - ADD('\\'); - switch (c) { - case '\t': ADD('t'); break; - case '\n': ADD('n'); break; - case '\r': ADD('r'); break; - case '\\': ADD('\\'); break; - default: - if (c < 010) ADDF("%03o", c); - else ADDF("x%02x", c); - } - } - - ADD(0); - assert(used <= ebuf->avail); - return ebuf->buf; - -#undef ADD -#undef ADDF -} - -void unsanitise_value(char *out, unsigned *out_len_r, const char *in) -{ - const char *ip; - char *op; - unsigned c; - int n; - - for (ip = in, op = out; (c = *ip++); *op++ = c) { - if (c == '\\') { - c = *ip++; - -#define GETF(f) do { \ - n = 0; \ - sscanf(ip, f "%n", &c, &n); \ - ip += n; \ - } while (0) - - switch (c) { - case 't': c= '\t'; break; - case 'n': c= '\n'; break; - case 'r': c= '\r'; break; - case '\\': c= '\\'; break; - case 'x': GETF("%2x"); break; - case '0': case '4': - case '1': case '5': - case '2': case '6': - case '3': case '7': --ip; GETF("%3o"); break; - case 0: --ip; break; - default:; - } -#undef GETF - } - } - - *op = 0; - - if (out_len_r) - *out_len_r = op - out; -} - /* * Local variables: * mode: C diff --git a/tools/xenstore/Makefile b/tools/xenstore/Makefile index 8ced12c27e..292b478fa1 100644 --- a/tools/xenstore/Makefile +++ b/tools/xenstore/Makefile @@ -78,8 +78,8 @@ xenstored.a: $(XENSTORED_OBJS) $(CLIENTS): xenstore ln -f xenstore $@ -xenstore: xenstore_client.o - $(CC) $< $(LDFLAGS) $(LDLIBS_libxenstore) $(LDLIBS_libxentoolcore) $(SOCKET_LIBS) -o $@ $(APPEND_LDFLAGS) +xenstore: xenstore_client.o xs_lib.o + $(CC) $^ $(LDFLAGS) $(LDLIBS_libxenstore) $(LDLIBS_libxentoolcore) $(SOCKET_LIBS) -o $@ $(APPEND_LDFLAGS) xenstore-control: xenstore_control.o $(CC) $< $(LDFLAGS) $(LDLIBS_libxenstore) $(LDLIBS_libxenctrl) $(LDLIBS_libxenguest) $(LDLIBS_libxentoolcore) $(SOCKET_LIBS) -o $@ $(APPEND_LDFLAGS) diff --git a/tools/xenstore/utils.h b/tools/xenstore/utils.h index 87713a8e5d..9d012b97c1 100644 --- a/tools/xenstore/utils.h +++ b/tools/xenstore/utils.h @@ -7,6 +7,17 @@ #include +#include "xenstore_lib.h" + +/* Header of the node record in tdb. */ +struct xs_tdb_record_hdr { + uint64_t generation; + uint32_t num_perms; + uint32_t datalen; + uint32_t childlen; + struct xs_permissions perms[0]; +}; + /* Is A == B ? */ #define streq(a,b) (strcmp((a),(b)) == 0) diff --git a/tools/xenstore/xenstore_client.c b/tools/xenstore/xenstore_client.c index ddbafc5175..0628ba275e 100644 --- a/tools/xenstore/xenstore_client.c +++ b/tools/xenstore/xenstore_client.c @@ -22,6 +22,8 @@ #include +#include "xs_lib.h" + #define PATH_SEP '/' #define MAX_PATH_LEN 256 diff --git a/tools/xenstore/xenstored_control.c b/tools/xenstore/xenstored_control.c index 2c63ba7ead..0d57f9f940 100644 --- a/tools/xenstore/xenstored_control.c +++ b/tools/xenstore/xenstored_control.c @@ -34,6 +34,7 @@ #include "utils.h" #include "talloc.h" +#include "xs_lib.h" #include "xenstored_core.h" #include "xenstored_control.h" #include "xenstored_domain.h" diff --git a/tools/xenstore/xenstored_core.c b/tools/xenstore/xenstored_core.c index df1e54c1f8..883a1a582a 100644 --- a/tools/xenstore/xenstored_core.c +++ b/tools/xenstore/xenstored_core.c @@ -46,7 +46,7 @@ #include "utils.h" #include "list.h" #include "talloc.h" -#include "xenstore_lib.h" +#include "xs_lib.h" #include "xenstored_core.h" #include "xenstored_watch.h" #include "xenstored_transaction.h" @@ -542,11 +542,11 @@ static int write_node(struct connection *conn, struct node *node, return write_node_raw(conn, &key, node, no_quota_check); } -enum xs_perm_type perm_for_conn(struct connection *conn, - const struct node_perms *perms) +unsigned int perm_for_conn(struct connection *conn, + const struct node_perms *perms) { unsigned int i; - enum xs_perm_type mask = XS_PERM_READ|XS_PERM_WRITE|XS_PERM_OWNER; + unsigned int mask = XS_PERM_READ|XS_PERM_WRITE|XS_PERM_OWNER; /* Owners and tools get it all... */ if (!domain_is_unprivileged(conn) || perms->p[0].id == conn->id @@ -584,7 +584,7 @@ char *get_parent(const void *ctx, const char *node) * Temporary memory allocations are done with ctx. */ static int ask_parents(struct connection *conn, const void *ctx, - const char *name, enum xs_perm_type *perm) + const char *name, unsigned int *perm) { struct node *node; @@ -618,10 +618,9 @@ static int ask_parents(struct connection *conn, const void *ctx, * Temporary memory allocations are done with ctx. */ static int errno_from_parents(struct connection *conn, const void *ctx, - const char *node, int errnum, - enum xs_perm_type perm) + const char *node, int errnum, unsigned int perm) { - enum xs_perm_type parent_perm = XS_PERM_NONE; + unsigned int parent_perm = XS_PERM_NONE; /* We always tell them about memory failures. */ if (errnum == ENOMEM) @@ -641,7 +640,7 @@ static int errno_from_parents(struct connection *conn, const void *ctx, static struct node *get_node(struct connection *conn, const void *ctx, const char *name, - enum xs_perm_type perm) + unsigned int perm) { struct node *node; @@ -873,7 +872,7 @@ static struct node *get_node_canonicalized(struct connection *conn, const void *ctx, const char *name, char **canonical_name, - enum xs_perm_type perm) + unsigned int perm) { char *tmp_name; diff --git a/tools/xenstore/xenstored_core.h b/tools/xenstore/xenstored_core.h index da43ec0e02..bb36111ecc 100644 --- a/tools/xenstore/xenstored_core.h +++ b/tools/xenstore/xenstored_core.h @@ -189,8 +189,8 @@ void send_ack(struct connection *conn, enum xsd_sockmsg_type type); char *canonicalize(struct connection *conn, const void *ctx, const char *node); /* Get access permissions. */ -enum xs_perm_type perm_for_conn(struct connection *conn, - const struct node_perms *perms); +unsigned int perm_for_conn(struct connection *conn, + const struct node_perms *perms); /* Write a node to the tdb data base. */ int write_node_raw(struct connection *conn, TDB_DATA *key, struct node *node, @@ -204,8 +204,6 @@ struct connection *new_connection(const struct interface_funcs *funcs); struct connection *get_connection_by_id(unsigned int conn_id); void check_store(void); void corrupt(struct connection *conn, const char *fmt, ...); -enum xs_perm_type perm_for_conn(struct connection *conn, - const struct node_perms *perms); /* Is this a valid node name? */ bool is_valid_nodename(const char *node); diff --git a/tools/xenstore/xenstored_watch.c b/tools/xenstore/xenstored_watch.c index db89e0141f..aca0a71bad 100644 --- a/tools/xenstore/xenstored_watch.c +++ b/tools/xenstore/xenstored_watch.c @@ -124,7 +124,7 @@ static bool watch_permitted(struct connection *conn, const void *ctx, const char *name, struct node *node, struct node_perms *perms) { - enum xs_perm_type perm; + unsigned int perm; struct node *parent; char *parent_name; diff --git a/tools/xenstore/xs_lib.c b/tools/xenstore/xs_lib.c index 80c03acbea..10fa4c3ad0 100644 --- a/tools/xenstore/xs_lib.c +++ b/tools/xenstore/xs_lib.c @@ -16,12 +16,13 @@ License along with this library; If not, see . */ +#include #include #include #include #include #include -#include "xenstore_lib.h" +#include "xs_lib.h" /* Common routines for the Xen store daemon and client library. */ @@ -179,3 +180,114 @@ unsigned int xs_count_strings(const char *strings, unsigned int len) return num; } + +char *expanding_buffer_ensure(struct expanding_buffer *ebuf, int min_avail) +{ + int want; + char *got; + + if (ebuf->avail >= min_avail) + return ebuf->buf; + + if (min_avail >= INT_MAX/3) + return 0; + + want = ebuf->avail + min_avail + 10; + got = realloc(ebuf->buf, want); + if (!got) + return 0; + + ebuf->buf = got; + ebuf->avail = want; + return ebuf->buf; +} + +char *sanitise_value(struct expanding_buffer *ebuf, + const char *val, unsigned len) +{ + int used, remain, c; + unsigned char *ip; + +#define ADD(c) (ebuf->buf[used++] = (c)) +#define ADDF(f,c) (used += sprintf(ebuf->buf+used, (f), (c))) + + assert(len < INT_MAX/5); + + ip = (unsigned char *)val; + used = 0; + remain = len; + + if (!expanding_buffer_ensure(ebuf, remain + 1)) + return NULL; + + while (remain-- > 0) { + c= *ip++; + + if (c >= ' ' && c <= '~' && c != '\\') { + ADD(c); + continue; + } + + if (!expanding_buffer_ensure(ebuf, used + remain + 5)) + /* for "\\nnn\0" */ + return 0; + + ADD('\\'); + switch (c) { + case '\t': ADD('t'); break; + case '\n': ADD('n'); break; + case '\r': ADD('r'); break; + case '\\': ADD('\\'); break; + default: + if (c < 010) ADDF("%03o", c); + else ADDF("x%02x", c); + } + } + + ADD(0); + assert(used <= ebuf->avail); + return ebuf->buf; + +#undef ADD +#undef ADDF +} + +void unsanitise_value(char *out, unsigned *out_len_r, const char *in) +{ + const char *ip; + char *op; + unsigned c; + int n; + + for (ip = in, op = out; (c = *ip++); *op++ = c) { + if (c == '\\') { + c = *ip++; + +#define GETF(f) do { \ + n = 0; \ + sscanf(ip, f "%n", &c, &n); \ + ip += n; \ + } while (0) + + switch (c) { + case 't': c= '\t'; break; + case 'n': c= '\n'; break; + case 'r': c= '\r'; break; + case '\\': c= '\\'; break; + case 'x': GETF("%2x"); break; + case '0': case '4': + case '1': case '5': + case '2': case '6': + case '3': case '7': --ip; GETF("%3o"); break; + case 0: --ip; break; + default:; + } +#undef GETF + } + } + + *op = 0; + + if (out_len_r) + *out_len_r = op - out; +} diff --git a/tools/xenstore/xs_lib.h b/tools/xenstore/xs_lib.h new file mode 100644 index 0000000000..efa05997d6 --- /dev/null +++ b/tools/xenstore/xs_lib.h @@ -0,0 +1,50 @@ +/* + Common routines between Xen store user library and daemon. + Copyright (C) 2005 Rusty Russell IBM Corporation + + This library is free software; you can redistribute it and/or + modify it under the terms of the GNU Lesser General Public + License as published by the Free Software Foundation; either + version 2.1 of the License, or (at your option) any later version. + + This library is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Lesser General Public License for more details. + + You should have received a copy of the GNU Lesser General Public + License along with this library; If not, see . +*/ + +#ifndef XS_LIB_H +#define XS_LIB_H + +#include "xenstore_lib.h" + +const char *xs_daemon_rootdir(void); +const char *xs_domain_dev(void); +const char *xs_daemon_tdb(void); + +/* Convert permissions to a string (up to len MAX_STRLEN(unsigned int)+1). */ +bool xs_perm_to_string(const struct xs_permissions *perm, + char *buffer, size_t buf_len); + +/* Given a string and a length, count how many strings (nul terms). */ +unsigned int xs_count_strings(const char *strings, unsigned int len); + +/* Sanitising (quoting) possibly-binary strings. */ +struct expanding_buffer { + char *buf; + int avail; +}; + +/* Ensure that given expanding buffer has at least min_avail characters. */ +char *expanding_buffer_ensure(struct expanding_buffer *, int min_avail); + +/* sanitise_value() may return NULL if malloc fails. */ +char *sanitise_value(struct expanding_buffer *, const char *val, unsigned len); + +/* *out_len_r on entry is ignored; out must be at least strlen(in)+1 bytes. */ +void unsanitise_value(char *out, unsigned *out_len_r, const char *in); + +#endif /* XS_LIB_H */ diff --git a/tools/xenstore/xs_tdb_dump.c b/tools/xenstore/xs_tdb_dump.c index f74676cf1c..5d2db392b4 100644 --- a/tools/xenstore/xs_tdb_dump.c +++ b/tools/xenstore/xs_tdb_dump.c @@ -17,7 +17,7 @@ static uint32_t total_size(struct xs_tdb_record_hdr *hdr) + hdr->datalen + hdr->childlen; } -static char perm_to_char(enum xs_perm_type perm) +static char perm_to_char(unsigned int perm) { return perm == XS_PERM_READ ? 'r' : perm == XS_PERM_WRITE ? 'w' : -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:27:47 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:27:47 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146608.269820 (Exim 4.92) (envelope-from ) id 1lwMaJ-0005qK-J2; Thu, 24 Jun 2021 10:27:47 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146608.269820; Thu, 24 Jun 2021 10:27:47 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMaJ-0005qC-Fu; Thu, 24 Jun 2021 10:27:47 +0000 Received: by outflank-mailman (input) for mailman id 146608; Thu, 24 Jun 2021 10:27:46 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMaI-0005pz-1H for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:27:46 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMaI-00065Z-0Q for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:27:46 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwMaH-0006gc-Vv for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:27:45 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=dlPlGdhH+o/2WlhswOvg+bbhGkH5IBmkpyZ17QUBFC0=; b=VcQtgJBN+z45eZE4Tfcjk5898c 8x7VBEbzpJKjrlHYtMI5Qjjw5IBJUZzghdW6th3MIaxCCnp913ZZF8nk7cOno2A3OWco8B6l3/ILh H3DHcxPe1RNinlzPaJKT2g3FPw7wPLdbdLWuJf0iEcOpOYsoz61lpxk52puszwph81MM=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] Arm32: MSR to SPSR needs qualification Message-Id: Date: Thu, 24 Jun 2021 10:27:45 +0000 commit 93031fbe9f4c341a2e7950a088025ea550291433 Author: Jan Beulich AuthorDate: Fri Jun 11 15:04:24 2021 +0200 Commit: Stefano Stabellini CommitDate: Fri Jun 11 15:05:24 2021 -0700 Arm32: MSR to SPSR needs qualification The Arm ARM's description of MSR (ARM DDI 0406C.d section B9.3.12) doesn't even allow for plain "SPSR" here, and while gas accepts this, it takes it to mean SPSR_cf. Yet surely all of SPSR wants updating on this path, not just the lowest and highest 8 bits. Fixes: dfcffb128be4 ("xen/arm32: SPSR_hyp/SPSR") Signed-off-by: Jan Beulich Reviewed-by: Stefano Stabellini --- xen/arch/arm/arm32/entry.S | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/xen/arch/arm/arm32/entry.S b/xen/arch/arm/arm32/entry.S index 4e421109db..38826142ad 100644 --- a/xen/arch/arm/arm32/entry.S +++ b/xen/arch/arm/arm32/entry.S @@ -395,7 +395,7 @@ return_to_hypervisor: ldr r11, [sp, #UREGS_pc] msr ELR_hyp, r11 ldr r11, [sp, #UREGS_cpsr] - msr SPSR, r11 + msr SPSR_cxsf, r11 #ifdef CONFIG_ARM32_HARDEN_BRANCH_PREDICTOR /* * Hardening branch predictor may require to setup a different -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:27:57 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:27:57 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146609.269824 (Exim 4.92) (envelope-from ) id 1lwMaT-0005tJ-KX; Thu, 24 Jun 2021 10:27:57 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146609.269824; Thu, 24 Jun 2021 10:27:57 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMaT-0005tB-HX; Thu, 24 Jun 2021 10:27:57 +0000 Received: by outflank-mailman (input) for mailman id 146609; Thu, 24 Jun 2021 10:27:56 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMaS-0005sp-4O for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:27:56 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMaS-00065j-3c for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:27:56 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwMaS-0006hU-2s for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:27:56 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=oPKbcblT0mR8T9wGpfli6Nt+R71kb/7YBkzAeCedHzE=; b=z/5+yesbG3Dp2goENO1Nie5K9W n9RjcU/YF/fJmfwCzBGOLjZqMoVy3zLFxDv+aMKzFl97hozA/5ZqdGGBsUVQ+aRtR+hrFhRAZ8G6n AY2Y+oAcacjp4JcBO/vPQXqWOLUOPUc9ToOay9beUGIWWi4QNY1c8+SCAZVZBS+FJ+hY=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] xen/grant-table: Simplify the update to the per-vCPU maptrack freelist Message-Id: Date: Thu, 24 Jun 2021 10:27:56 +0000 commit 4f1858763b7b1aeb79fa7c818eca98c96943aa69 Author: Julien Grall AuthorDate: Mon Jun 14 11:08:30 2021 +0100 Commit: Julien Grall CommitDate: Mon Jun 14 11:09:20 2021 +0100 xen/grant-table: Simplify the update to the per-vCPU maptrack freelist Since XSA-228 (commit 02cbeeb62075 "gnttab: split maptrack lock to make it fulfill its purpose again"), v->maptrack_head, v->maptrack_tail and the content of the freelist are accessed with the lock v->maptrack_freelist_lock held. Therefore it is not necessary to update the fields using cmpxchg() and also read them atomically. Note that there are two cases where v->maptrack_tail is accessed without the lock. They both happen in get_maptrack_handle() when initializing the free list of the current vCPU. Therefore there is no possible race. The code is now reworked to remove any use of cmpxch() and read_atomic() when accessing the fields v->maptrack_{head, tail} as wel as the freelist. Take the opportunity to add a comment on top of the lock definition and explain what it protects. Signed-off-by: Julien Grall Reviewed-by: Jan Beulich --- xen/common/grant_table.c | 66 ++++++++++++++++++++---------------------------- xen/include/xen/sched.h | 8 +++++- 2 files changed, 34 insertions(+), 40 deletions(-) diff --git a/xen/common/grant_table.c b/xen/common/grant_table.c index ab30e2e8cf..fab77ab9cc 100644 --- a/xen/common/grant_table.c +++ b/xen/common/grant_table.c @@ -543,33 +543,27 @@ double_gt_unlock(struct grant_table *lgt, struct grant_table *rgt) static inline grant_handle_t _get_maptrack_handle(struct grant_table *t, struct vcpu *v) { - unsigned int head, next, prev_head; + unsigned int head, next; spin_lock(&v->maptrack_freelist_lock); - do { - /* No maptrack pages allocated for this VCPU yet? */ - head = read_atomic(&v->maptrack_head); - if ( unlikely(head == MAPTRACK_TAIL) ) - { - spin_unlock(&v->maptrack_freelist_lock); - return INVALID_MAPTRACK_HANDLE; - } - - /* - * Always keep one entry in the free list to make it easier to - * add free entries to the tail. - */ - next = read_atomic(&maptrack_entry(t, head).ref); - if ( unlikely(next == MAPTRACK_TAIL) ) - { - spin_unlock(&v->maptrack_freelist_lock); - return INVALID_MAPTRACK_HANDLE; - } + /* No maptrack pages allocated for this VCPU yet? */ + head = v->maptrack_head; + if ( unlikely(head == MAPTRACK_TAIL) ) + { + spin_unlock(&v->maptrack_freelist_lock); + return INVALID_MAPTRACK_HANDLE; + } - prev_head = head; - head = cmpxchg(&v->maptrack_head, prev_head, next); - } while ( head != prev_head ); + /* + * Always keep one entry in the free list to make it easier to + * add free entries to the tail. + */ + next = maptrack_entry(t, head).ref; + if ( unlikely(next == MAPTRACK_TAIL) ) + head = INVALID_MAPTRACK_HANDLE; + else + v->maptrack_head = next; spin_unlock(&v->maptrack_freelist_lock); @@ -623,7 +617,7 @@ put_maptrack_handle( { struct domain *currd = current->domain; struct vcpu *v; - unsigned int prev_tail, cur_tail; + unsigned int tail; /* 1. Set entry to be a tail. */ maptrack_entry(t, handle).ref = MAPTRACK_TAIL; @@ -633,14 +627,11 @@ put_maptrack_handle( spin_lock(&v->maptrack_freelist_lock); - cur_tail = read_atomic(&v->maptrack_tail); - do { - prev_tail = cur_tail; - cur_tail = cmpxchg(&v->maptrack_tail, prev_tail, handle); - } while ( cur_tail != prev_tail ); + tail = v->maptrack_tail; + v->maptrack_tail = handle; /* 3. Update the old tail entry to point to the new entry. */ - write_atomic(&maptrack_entry(t, prev_tail).ref, handle); + maptrack_entry(t, tail).ref = handle; spin_unlock(&v->maptrack_freelist_lock); } @@ -650,7 +641,7 @@ get_maptrack_handle( struct grant_table *lgt) { struct vcpu *curr = current; - unsigned int i, head; + unsigned int i; grant_handle_t handle; struct grant_mapping *new_mt = NULL; @@ -686,7 +677,7 @@ get_maptrack_handle( maptrack_entry(lgt, handle).ref = MAPTRACK_TAIL; curr->maptrack_tail = handle; if ( curr->maptrack_head == MAPTRACK_TAIL ) - write_atomic(&curr->maptrack_head, handle); + curr->maptrack_head = handle; spin_unlock(&curr->maptrack_freelist_lock); } return steal_maptrack_handle(lgt, curr); @@ -707,7 +698,7 @@ get_maptrack_handle( new_mt[i].vcpu = curr->vcpu_id; } - /* Set tail directly if this is the first page for this VCPU. */ + /* Set tail directly if this is the first page for the local vCPU. */ if ( curr->maptrack_tail == MAPTRACK_TAIL ) curr->maptrack_tail = handle + MAPTRACK_PER_PAGE - 1; @@ -716,13 +707,10 @@ get_maptrack_handle( lgt->maptrack_limit += MAPTRACK_PER_PAGE; spin_unlock(&lgt->maptrack_lock); - spin_lock(&curr->maptrack_freelist_lock); - - do { - new_mt[i - 1].ref = read_atomic(&curr->maptrack_head); - head = cmpxchg(&curr->maptrack_head, new_mt[i - 1].ref, handle + 1); - } while ( head != new_mt[i - 1].ref ); + spin_lock(&curr->maptrack_freelist_lock); + new_mt[i - 1].ref = curr->maptrack_head; + curr->maptrack_head = handle + 1; spin_unlock(&curr->maptrack_freelist_lock); return handle; diff --git a/xen/include/xen/sched.h b/xen/include/xen/sched.h index 3982167144..28146ee404 100644 --- a/xen/include/xen/sched.h +++ b/xen/include/xen/sched.h @@ -255,7 +255,13 @@ struct vcpu /* VCPU paused by system controller. */ int controller_pause_count; - /* Grant table map tracking. */ + /* + * Grant table map tracking. The lock maptrack_freelist_lock + * protects: + * - entries in the freelist + * - maptrack_head + * - maptrack_tail + */ spinlock_t maptrack_freelist_lock; unsigned int maptrack_head; unsigned int maptrack_tail; -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:28:07 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:28:07 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146610.269828 (Exim 4.92) (envelope-from ) id 1lwMad-0005vz-MJ; Thu, 24 Jun 2021 10:28:07 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146610.269828; Thu, 24 Jun 2021 10:28:07 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMad-0005vr-J8; Thu, 24 Jun 2021 10:28:07 +0000 Received: by outflank-mailman (input) for mailman id 146610; Thu, 24 Jun 2021 10:28:06 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMac-0005vg-7V for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:28:06 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMac-00067g-6p for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:28:06 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwMac-0006iU-60 for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:28:06 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=C1M/ZUHepKzfbC8+0mshLQbxjvIJRDFz4HahdKJXKZs=; b=HVatcT6249CvpTfm7gLL0xqo3j cPBde4bVEaVdPV0occpfNuUf1nigKAngA2/tzzTUI+rtNyCadkHq4oxqYvZrkma3V5fJBYBoUgJbg 7AM9gMEm1NsXrNNZ3euV7QmSBMAOpzgOYjkvdELz0y9/pB4c8RpGBnXad2SaStleiiwY=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] revert "tools/libs/guest: fix max_pfn setting in map_p2m()" Message-Id: Date: Thu, 24 Jun 2021 10:28:06 +0000 commit 60d82f4bb0a22eb8765d33f87207afab25bc2396 Author: Juergen Gross AuthorDate: Mon Jun 14 14:39:52 2021 +0200 Commit: Jan Beulich CommitDate: Mon Jun 14 14:39:52 2021 +0200 revert "tools/libs/guest: fix max_pfn setting in map_p2m()" The reasoning for commit 7bd8989ab77b6a ("tools/libs/guest: fix max_pfn setting in map_p2m()") was wrong. The max_pfn field in shared_info is misnamed, it has the semantics of num_pfns, which is hidden at least partially in Linux, as the kernel is (wrongly) treating it like the highest used pfn in some places. So revert above commit. Fixes: 7bd8989ab77b6a ("tools/libs/guest: fix max_pfn setting in map_p2m()") Signed-off-by: Juergen Gross --- tools/libs/guest/xg_sr_save_x86_pv.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/tools/libs/guest/xg_sr_save_x86_pv.c b/tools/libs/guest/xg_sr_save_x86_pv.c index dae7f2817f..4964f1f7b8 100644 --- a/tools/libs/guest/xg_sr_save_x86_pv.c +++ b/tools/libs/guest/xg_sr_save_x86_pv.c @@ -468,7 +468,7 @@ static int map_p2m(struct xc_sr_context *ctx) ctx->x86.pv.p2m_generation = ~0ULL; ctx->x86.pv.max_pfn = GET_FIELD(ctx->x86.pv.shinfo, arch.max_pfn, - ctx->x86.pv.width); + ctx->x86.pv.width) - 1; p2m_cr3 = GET_FIELD(ctx->x86.pv.shinfo, arch.p2m_cr3, ctx->x86.pv.width); return p2m_cr3 ? map_p2m_list(ctx, p2m_cr3) : map_p2m_tree(ctx); -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:28:17 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:28:17 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146611.269832 (Exim 4.92) (envelope-from ) id 1lwMan-0005ze-Nw; Thu, 24 Jun 2021 10:28:17 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146611.269832; Thu, 24 Jun 2021 10:28:17 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMan-0005zV-Ki; Thu, 24 Jun 2021 10:28:17 +0000 Received: by outflank-mailman (input) for mailman id 146611; Thu, 24 Jun 2021 10:28:16 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMam-0005zI-Ac for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:28:16 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMam-00068A-9v for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:28:16 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwMam-0006jE-95 for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:28:16 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=QPuGEeaeStvqfDZCq4qZGeiyyIYG0LHXN6TCp8zyALg=; b=1BajR44kvb4kIP7Mp2BSqJFmUI iya/3ZAtSGOqeBFlZsKQxtMsTMsUVBxCTwIsEf1JtEzG8inu61FWd7JZHFjLx+mf6F4XXWwrFrLnt vnm6/AuyYC2BN1rGIh+jMiOwC0nLBr7hS2K/P92GlrjI0xd+uqNZvInB0OwkWlXHy1Zk=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] MAINTAINERS: add myself as tools/libs reviewer Message-Id: Date: Thu, 24 Jun 2021 10:28:16 +0000 commit f7079d7ef69f6bf38d6ec3bda294ed5eabcf98ba Author: Juergen Gross AuthorDate: Mon Jun 14 14:40:19 2021 +0200 Commit: Jan Beulich CommitDate: Mon Jun 14 14:40:19 2021 +0200 MAINTAINERS: add myself as tools/libs reviewer I have touched most of the Xen libraries in the past, and there is a clear lack of reviewer band width in the tools area. Add myself as a tools/libs reviewer for that reason. Signed-off-by: Juergen Gross Acked-by: Ian Jackson --- MAINTAINERS | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/MAINTAINERS b/MAINTAINERS index 7778c64fc7..39750bb75d 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -374,6 +374,13 @@ F: xen/include/{kexec,kimage}.h F: xen/arch/x86/machine_kexec.c F: xen/arch/x86/x86_64/kexec_reloc.S +LIBS: +M: Ian Jackson +M: Wei Liu +R: Juergen Gross +S: Supported +F: tools/libs/ + LIBXENLIGHT M: Ian Jackson M: Wei Liu -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:28:28 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:28:28 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146613.269856 (Exim 4.92) (envelope-from ) id 1lwMay-0006T9-7l; Thu, 24 Jun 2021 10:28:28 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146613.269856; Thu, 24 Jun 2021 10:28:28 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMay-0006T1-3Z; Thu, 24 Jun 2021 10:28:28 +0000 Received: by outflank-mailman (input) for mailman id 146613; Thu, 24 Jun 2021 10:28:26 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMaw-0006F3-DX for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:28:26 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMaw-00068L-Cs for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:28:26 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwMaw-0006k2-CC for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:28:26 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=Al+bdAugvkWlolHc6FJwvtH73cMiUPfYtcIQlalqHek=; b=6NO9BSzIIBoyO9is00Sr95Rke5 HVwWrq+mKYDOGjXaEKul2A6QkVUCOD0nwIkg7CqLMRHk9qCQD3LCZRaUIetF/yvOmrk1zw5+Gu4Zh uU8tBFsD7dFBjvTRHhMogIB/8ieqBkaCDOZTvD/nSIGjxCGUMf7cPDPy90IlatFdwDzQ=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] xen/arm32: avoid .rodata to be marked as executable Message-Id: Date: Thu, 24 Jun 2021 10:28:26 +0000 commit 163f47c14737cfa9dfb3240deea356b08caf7614 Author: Jan Beulich AuthorDate: Fri Jun 11 11:19:15 2021 +0200 Commit: Julien Grall CommitDate: Mon Jun 14 14:14:11 2021 +0100 xen/arm32: avoid .rodata to be marked as executable The section .proc.info lives in .rodata as it doesn't contain any executable code. However, the section is still marked as executable as the consequence .rodata will also be marked executable. Xen doesn't use the ELF permissions to decide the page-table mapping permission. However, this will confuse disassemblers. '#execinstr' is now removed on all the pushsection dealing with .proc.info Signed-off-by: Jan Beulich [julieng: Rework the commit message] Acked-by: Julien Grall --- xen/arch/arm/arm32/proc-v7.S | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/xen/arch/arm/arm32/proc-v7.S b/xen/arch/arm/arm32/proc-v7.S index 1efde2d72d..c90a31d80f 100644 --- a/xen/arch/arm/arm32/proc-v7.S +++ b/xen/arch/arm/arm32/proc-v7.S @@ -29,7 +29,7 @@ brahma15mp_init: mcr CP32(r0, ACTLR) mov pc, lr - .section ".proc.info", #alloc, #execinstr + .section ".proc.info", #alloc .type __v7_ca15mp_proc_info, #object __v7_ca15mp_proc_info: .long 0x410FC0F0 /* Cortex-A15 */ @@ -38,7 +38,7 @@ __v7_ca15mp_proc_info: .long caxx_processor .size __v7_ca15mp_proc_info, . - __v7_ca15mp_proc_info - .section ".proc.info", #alloc, #execinstr + .section ".proc.info", #alloc .type __v7_ca7mp_proc_info, #object __v7_ca7mp_proc_info: .long 0x410FC070 /* Cortex-A7 */ @@ -47,7 +47,7 @@ __v7_ca7mp_proc_info: .long caxx_processor .size __v7_ca7mp_proc_info, . - __v7_ca7mp_proc_info - .section ".proc.info", #alloc, #execinstr + .section ".proc.info", #alloc .type __v7_brahma15mp_proc_info, #object __v7_brahma15mp_proc_info: .long 0x420F00F0 /* Broadcom Brahma-B15 */ -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:28:37 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:28:37 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146615.269868 (Exim 4.92) (envelope-from ) id 1lwMb7-0006sB-FQ; Thu, 24 Jun 2021 10:28:37 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146615.269868; Thu, 24 Jun 2021 10:28:37 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMb7-0006s4-CH; Thu, 24 Jun 2021 10:28:37 +0000 Received: by outflank-mailman (input) for mailman id 146615; Thu, 24 Jun 2021 10:28:36 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMb6-0006r3-GX for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:28:36 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMb6-00068l-Fo for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:28:36 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwMb6-0006ks-F7 for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:28:36 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=UD+cHUaQs1h9vDRSyuIicnSMbjHWk55IVF2V3d8vBWs=; b=YHt/0XOcUtdF/slJ2ruPqAq4Pi WZ7J9Rr1/xOAkkK5kHxH5ZuflaAWJNfJmBdYceqDRt7Me5Dx2kQQm4aySwVceHqv+VQDESoNqfORb ilxSVzUqYYjo47IxEEVQpIy3D5JuMAg/rzc0rc7fGyynYqvjd3IDzBM1Hu2aqgsGRrWE=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] Arm: avoid .init.data to be marked as executable Message-Id: Date: Thu, 24 Jun 2021 10:28:36 +0000 commit 8c9ed863738ff9e8b91975d6aa4464e7e8324eb7 Author: Jan Beulich AuthorDate: Mon Jun 14 15:52:36 2021 +0200 Commit: Julien Grall CommitDate: Mon Jun 14 17:53:52 2021 +0100 Arm: avoid .init.data to be marked as executable This confuses disassemblers, at the very least. Move .altinstr_replacement to .init.text. The previously redundant ALIGN() now gets converted to page alignment, such that the hypervisor mapping won't have this as executable (it'll instead get mapped r/w, which I'm told is intended to be adjusted at some point). Note that for the actual patching logic's purposes this part of .init.text _has_ to live after _einittext (or before _sinittext), or else branch_insn_requires_update() would produce wrong results. Also, to have .altinstr_replacement have consistent attributes in the object files, add "x" to the one instance where it was missing. Signed-off-by: Jan Beulich Acked-by: Julien Grall --- xen/arch/arm/xen.lds.S | 4 ++-- xen/include/asm-arm/alternative.h | 2 +- 2 files changed, 3 insertions(+), 3 deletions(-) diff --git a/xen/arch/arm/xen.lds.S b/xen/arch/arm/xen.lds.S index fc1d39a886..b773f91f1c 100644 --- a/xen/arch/arm/xen.lds.S +++ b/xen/arch/arm/xen.lds.S @@ -148,6 +148,8 @@ SECTIONS _sinittext = .; *(.init.text) _einittext = .; + . = ALIGN(PAGE_SIZE); /* Avoid mapping alt insns executable */ + *(.altinstr_replacement) } :text . = ALIGN(PAGE_SIZE); .init.data : { @@ -169,8 +171,6 @@ SECTIONS __alt_instructions = .; *(.altinstructions) __alt_instructions_end = .; - . = ALIGN(4); - *(.altinstr_replacement) #ifdef CONFIG_DEBUG_LOCK_PROFILE . = ALIGN(POINTER_ALIGN); diff --git a/xen/include/asm-arm/alternative.h b/xen/include/asm-arm/alternative.h index 92e888aff8..1eb4b60fbb 100644 --- a/xen/include/asm-arm/alternative.h +++ b/xen/include/asm-arm/alternative.h @@ -67,7 +67,7 @@ int apply_alternatives(const struct alt_instr *start, const struct alt_instr *en ALTINSTR_ENTRY(feature,cb) \ ".popsection\n" \ " .if " __stringify(cb) " == 0\n" \ - ".pushsection .altinstr_replacement, \"a\"\n" \ + ".pushsection .altinstr_replacement, \"ax\"\n" \ "663:\n\t" \ newinstr "\n" \ "664:\n\t" \ -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:28:47 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:28:47 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146620.269872 (Exim 4.92) (envelope-from ) id 1lwMbH-00079k-HZ; Thu, 24 Jun 2021 10:28:47 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146620.269872; Thu, 24 Jun 2021 10:28:47 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMbH-00078t-E8; Thu, 24 Jun 2021 10:28:47 +0000 Received: by outflank-mailman (input) for mailman id 146620; Thu, 24 Jun 2021 10:28:46 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMbG-00076h-Jg for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:28:46 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMbG-000693-Iu for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:28:46 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwMbG-0006lf-ID for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:28:46 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=EQCZcvbdllBB67Q+hCRNt6/bzBN0FtUIuK3XeWPiSnY=; b=VWkB3oRqXB7vbz3aKF6FqVxP2Q 4P7Ai4Pbxim7CAZz2PB2TqaguLWmWa1yp04ZXjXFOFNeDejoU6rAQQ1VHKqByne+6r1jSMKcvCQq6 /zeBq3XmNBSPFGUb7w+gub4rPzEOKalPOoNi2Uv9vOEgiJwEq2BjboRnS8yIVcN/kjBo=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] libxl: Replace deprecated QMP command by "query-cpus-fast" Message-Id: Date: Thu, 24 Jun 2021 10:28:46 +0000 commit d5f54009dba11d04bfe2a28eee47b994de66b84a Author: Anthony PERARD AuthorDate: Tue May 11 10:28:03 2021 +0100 Commit: Julien Grall CommitDate: Tue Jun 15 13:56:26 2021 +0100 libxl: Replace deprecated QMP command by "query-cpus-fast" We use the deprecated QMP command "query-cpus" which is removed in the QEMU 6.0 release. There's a replacement which is "query-cpus-fast", and have been available since QEMU 2.12 (April 2018). This patch try the new command first and when the command isn't available, it fall back to the deprecated one so libxl still works with older QEMU versions. Signed-off-by: Anthony PERARD Reviewed-by: Jason Andryuk --- tools/libs/light/libxl_domain.c | 103 ++++++++++++++++++++++++++++++++++++++-- 1 file changed, 98 insertions(+), 5 deletions(-) diff --git a/tools/libs/light/libxl_domain.c b/tools/libs/light/libxl_domain.c index 5d4ec90711..8c003aa7cb 100644 --- a/tools/libs/light/libxl_domain.c +++ b/tools/libs/light/libxl_domain.c @@ -1740,6 +1740,35 @@ out: return rc; } +static int qmp_parse_query_cpus_fast(libxl__gc *gc, + libxl_domid domid, + const libxl__json_object *response, + libxl_bitmap *const map) +{ + int i; + const libxl__json_object *cpu; + + libxl_bitmap_set_none(map); + /* Parse response to QMP command "query-cpus-fast": + * [ { 'cpu-index': 'int',...} ] + */ + for (i = 0; (cpu = libxl__json_array_get(response, i)); i++) { + unsigned int cpu_index; + const libxl__json_object *o; + + o = libxl__json_map_get("cpu-index", cpu, JSON_INTEGER); + if (!o) { + LOGD(ERROR, domid, "Failed to retrieve CPU index."); + return ERROR_QEMU_API; + } + + cpu_index = libxl__json_object_get_integer(o); + libxl_bitmap_set(map, cpu_index); + } + + return 0; +} + static int qmp_parse_query_cpus(libxl__gc *gc, libxl_domid domid, const libxl__json_object *response, @@ -1778,8 +1807,13 @@ typedef struct set_vcpuonline_state { int index; /* for loop on final_map */ } set_vcpuonline_state; +static void set_vcpuonline_qmp_cpus_fast_queried(libxl__egc *, + libxl__ev_qmp *, const libxl__json_object *, int rc); static void set_vcpuonline_qmp_cpus_queried(libxl__egc *, libxl__ev_qmp *, const libxl__json_object *, int rc); +static void set_vcpuonline_qmp_query_cpus_parse(libxl__egc *, + libxl__ev_qmp *qmp, const libxl__json_object *, + bool query_cpus_fast, int rc); static void set_vcpuonline_qmp_add_cpu(libxl__egc *, libxl__ev_qmp *, const libxl__json_object *response, int rc); static void set_vcpuonline_timeout(libxl__egc *egc, @@ -1840,8 +1874,8 @@ int libxl_set_vcpuonline(libxl_ctx *ctx, uint32_t domid, set_vcpuonline_timeout, LIBXL_QMP_CMD_TIMEOUT * 1000); if (rc) goto out; - qmp->callback = set_vcpuonline_qmp_cpus_queried; - rc = libxl__ev_qmp_send(egc, qmp, "query-cpus", NULL); + qmp->callback = set_vcpuonline_qmp_cpus_fast_queried; + rc = libxl__ev_qmp_send(egc, qmp, "query-cpus-fast", NULL); if (rc) goto out; return AO_INPROGRESS; default: @@ -1860,11 +1894,39 @@ out: return AO_INPROGRESS; } +static void set_vcpuonline_qmp_cpus_fast_queried(libxl__egc *egc, + libxl__ev_qmp *qmp, const libxl__json_object *response, int rc) +{ + EGC_GC; + set_vcpuonline_state *svos = CONTAINER_OF(qmp, *svos, qmp); + + if (rc == ERROR_QMP_COMMAND_NOT_FOUND) { + /* Try again, we probably talking to a QEMU older than 2.12 */ + qmp->callback = set_vcpuonline_qmp_cpus_queried; + rc = libxl__ev_qmp_send(egc, qmp, "query-cpus", NULL); + if (rc) goto out; + return; + } + +out: + set_vcpuonline_qmp_query_cpus_parse(egc, qmp, response, true, rc); +} + static void set_vcpuonline_qmp_cpus_queried(libxl__egc *egc, libxl__ev_qmp *qmp, const libxl__json_object *response, int rc) { EGC_GC; set_vcpuonline_state *svos = CONTAINER_OF(qmp, *svos, qmp); + + set_vcpuonline_qmp_query_cpus_parse(egc, qmp, response, false, rc); +} + +static void set_vcpuonline_qmp_query_cpus_parse(libxl__egc *egc, + libxl__ev_qmp *qmp, const libxl__json_object *response, + bool query_cpus_fast, int rc) +{ + EGC_GC; + set_vcpuonline_state *svos = CONTAINER_OF(qmp, *svos, qmp); int i; libxl_bitmap current_map; @@ -1876,7 +1938,11 @@ static void set_vcpuonline_qmp_cpus_queried(libxl__egc *egc, if (rc) goto out; libxl_bitmap_alloc(CTX, ¤t_map, svos->info.vcpu_max_id + 1); - rc = qmp_parse_query_cpus(gc, qmp->domid, response, ¤t_map); + if (query_cpus_fast) { + rc = qmp_parse_query_cpus_fast(gc, qmp->domid, response, ¤t_map); + } else { + rc = qmp_parse_query_cpus(gc, qmp->domid, response, ¤t_map); + } if (rc) goto out; libxl_bitmap_copy_alloc(CTX, final_map, svos->cpumap); @@ -2121,6 +2187,9 @@ typedef struct { static void retrieve_domain_configuration_lock_acquired( libxl__egc *egc, libxl__ev_slowlock *, int rc); +static void retrieve_domain_configuration_cpu_fast_queried( + libxl__egc *egc, libxl__ev_qmp *qmp, + const libxl__json_object *response, int rc); static void retrieve_domain_configuration_cpu_queried( libxl__egc *egc, libxl__ev_qmp *qmp, const libxl__json_object *response, int rc); @@ -2198,8 +2267,8 @@ static void retrieve_domain_configuration_lock_acquired( if (rc) goto out; libxl_bitmap_alloc(CTX, &rdcs->qemuu_cpus, d_config->b_info.max_vcpus); - rdcs->qmp.callback = retrieve_domain_configuration_cpu_queried; - rc = libxl__ev_qmp_send(egc, &rdcs->qmp, "query-cpus", NULL); + rdcs->qmp.callback = retrieve_domain_configuration_cpu_fast_queried; + rc = libxl__ev_qmp_send(egc, &rdcs->qmp, "query-cpus-fast", NULL); if (rc) goto out; has_callback = true; } @@ -2210,6 +2279,30 @@ out: retrieve_domain_configuration_end(egc, rdcs, rc); } +static void retrieve_domain_configuration_cpu_fast_queried( + libxl__egc *egc, libxl__ev_qmp *qmp, + const libxl__json_object *response, int rc) +{ + EGC_GC; + retrieve_domain_configuration_state *rdcs = + CONTAINER_OF(qmp, *rdcs, qmp); + + if (rc == ERROR_QMP_COMMAND_NOT_FOUND) { + /* Try again, we probably talking to a QEMU older than 2.12 */ + rdcs->qmp.callback = retrieve_domain_configuration_cpu_queried; + rc = libxl__ev_qmp_send(egc, &rdcs->qmp, "query-cpus", NULL); + if (rc) goto out; + return; + } + + if (rc) goto out; + + rc = qmp_parse_query_cpus_fast(gc, qmp->domid, response, &rdcs->qemuu_cpus); + +out: + retrieve_domain_configuration_end(egc, rdcs, rc); +} + static void retrieve_domain_configuration_cpu_queried( libxl__egc *egc, libxl__ev_qmp *qmp, const libxl__json_object *response, int rc) -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:28:57 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:28:57 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146623.269877 (Exim 4.92) (envelope-from ) id 1lwMbR-0007QR-KX; Thu, 24 Jun 2021 10:28:57 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146623.269877; Thu, 24 Jun 2021 10:28:57 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMbR-0007QI-HR; Thu, 24 Jun 2021 10:28:57 +0000 Received: by outflank-mailman (input) for mailman id 146623; Thu, 24 Jun 2021 10:28:56 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMbQ-0007Py-Mo for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:28:56 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMbQ-00069E-M4 for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:28:56 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwMbQ-0006mN-LJ for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:28:56 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=2ZredSdF8BPfIo6vu12cDyF6TWtLhY1he/jPOcDOLkU=; b=dZAR0giZOrLH1jm2JnHoow9PXu vgJRBUIYDLQIqZzKOaCMmtlfSTw9epQlibf3iAMfQgWZYgxSf87RTku+PTtwnvzzAuXkukb/qd25z gsofa2iL86z34JzB3a1Pj9McFFGy7qOacEIhvBY6LMknXVYw7stXkNtOu7Vg7xVGsknY=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] libxl: Replace QEMU's command line short-form boolean option Message-Id: Date: Thu, 24 Jun 2021 10:28:56 +0000 commit f3f778c81769075ac0eb93b98d4b2803e7936453 Author: Anthony PERARD AuthorDate: Tue May 11 10:28:04 2021 +0100 Commit: Julien Grall CommitDate: Tue Jun 15 13:56:26 2021 +0100 libxl: Replace QEMU's command line short-form boolean option Short-form boolean options are deprecated in QEMU 6.0. Upstream commit that deprecate those: ccd3b3b8112b ("qemu-option: warn for short-form boolean options"). Signed-off-by: Anthony PERARD Reviewed-by: Jason Andryuk --- tools/libs/light/libxl_dm.c | 14 +++++++------- 1 file changed, 7 insertions(+), 7 deletions(-) diff --git a/tools/libs/light/libxl_dm.c b/tools/libs/light/libxl_dm.c index 3599a82ef0..0a0c1ef7c6 100644 --- a/tools/libs/light/libxl_dm.c +++ b/tools/libs/light/libxl_dm.c @@ -977,14 +977,14 @@ static char *dm_spice_options(libxl__gc *gc, if (spice->host) opt = GCSPRINTF("%s,addr=%s", opt, spice->host); if (libxl_defbool_val(spice->disable_ticketing)) - opt = GCSPRINTF("%s,disable-ticketing", opt); + opt = GCSPRINTF("%s,disable-ticketing=on", opt); else opt = GCSPRINTF("%s,password=%s", opt, spice->passwd); opt = GCSPRINTF("%s,agent-mouse=%s", opt, libxl_defbool_val(spice->agent_mouse) ? "on" : "off"); if (!libxl_defbool_val(spice->clipboard_sharing)) - opt = GCSPRINTF("%s,disable-copy-paste", opt); + opt = GCSPRINTF("%s,disable-copy-paste=on", opt); if (spice->image_compression) opt = GCSPRINTF("%s,image-compression=%s", opt, @@ -1224,7 +1224,7 @@ static int libxl__build_device_model_args_new(libxl__gc *gc, flexarray_append(dm_args, "-chardev"); if (state->dm_monitor_fd >= 0) { flexarray_append(dm_args, - GCSPRINTF("socket,id=libxl-cmd,fd=%d,server,nowait", + GCSPRINTF("socket,id=libxl-cmd,fd=%d,server=on,wait=off", state->dm_monitor_fd)); /* @@ -1237,7 +1237,7 @@ static int libxl__build_device_model_args_new(libxl__gc *gc, } else { flexarray_append(dm_args, GCSPRINTF("socket,id=libxl-cmd," - "path=%s,server,nowait", + "path=%s,server=on,wait=off", libxl__qemu_qmp_path(gc, guest_domid))); } @@ -1247,7 +1247,7 @@ static int libxl__build_device_model_args_new(libxl__gc *gc, flexarray_append(dm_args, "-chardev"); flexarray_append(dm_args, GCSPRINTF("socket,id=libxenstat-cmd," - "path=%s/qmp-libxenstat-%d,server,nowait", + "path=%s/qmp-libxenstat-%d,server=on,wait=off", libxl__run_dir_path(), guest_domid)); flexarray_append(dm_args, "-mon"); @@ -1264,7 +1264,7 @@ static int libxl__build_device_model_args_new(libxl__gc *gc, case LIBXL_CHANNEL_CONNECTION_SOCKET: path = guest_config->channels[i].u.socket.path; chardev = GCSPRINTF("socket,id=libxl-channel%d,path=%s," - "server,nowait", devid, path); + "server=on,wait=off", devid, path); break; default: /* We've forgotten to add the clause */ @@ -1577,7 +1577,7 @@ static int libxl__build_device_model_args_new(libxl__gc *gc, nics[i].colo_##sock_port) { \ flexarray_append(dm_args, "-chardev"); \ flexarray_append(dm_args, \ - GCSPRINTF("socket,id=%s,host=%s,port=%s,server,nowait", \ + GCSPRINTF("socket,id=%s,host=%s,port=%s,server=on,wait=off", \ nics[i].colo_##sock_id, \ nics[i].colo_##sock_ip, \ nics[i].colo_##sock_port)); \ -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:29:07 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:29:07 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146626.269881 (Exim 4.92) (envelope-from ) id 1lwMbb-0007WR-M1; Thu, 24 Jun 2021 10:29:07 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146626.269881; Thu, 24 Jun 2021 10:29:07 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMbb-0007WH-Iy; Thu, 24 Jun 2021 10:29:07 +0000 Received: by outflank-mailman (input) for mailman id 146626; Thu, 24 Jun 2021 10:29:06 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMba-0007Vz-Po for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:29:06 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMba-00069c-P1 for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:29:06 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwMba-0006nN-OP for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:29:06 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=q/vV7NMTAnbqzkxaTQSVERPsCI5Zgtl51/k3yaDMrZY=; b=XEWxMtISIQAasIoo7CVDrCUmYk jvhlDlv2TJ1s805+2xG8cdeKHIBSIiSK1ec/ojlAh3n8ja946o8uDTJ7ltQOPoifO351ZGIlHrOhn mKurL0wEpr7rH+NJXIrs5/Lu9WIrN3hivVsHkRJ6Lr97cWDpLkbX7w3nzX7UjJmYGIms=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] libxl: Replace deprecated "cpu-add" QMP command by "device_add" Message-Id: Date: Thu, 24 Jun 2021 10:29:06 +0000 commit 4e217db45e83fc3173382306c8b03da86099a25d Author: Anthony PERARD AuthorDate: Tue May 11 10:28:05 2021 +0100 Commit: Julien Grall CommitDate: Tue Jun 15 13:56:26 2021 +0100 libxl: Replace deprecated "cpu-add" QMP command by "device_add" The command "cpu-add" for CPU hotplug is deprecated and has been removed from QEMU 6.0 (April 2021). We need to add cpus with the command "device_add" now. In order to find out which parameters to pass to "device_add" we first make a call to "query-hotpluggable-cpus" which list the cpus drivers and properties. The algorithm to figure out which CPU to add, and by extension if any CPU needs to be hotplugged, is in the function that adds the cpus. Because of that, the command "query-hotpluggable-cpus" is always called, even when not needed. In case we are using a version of QEMU older than 2.7 (Sept 2016) which don't have "query-hotpluggable-cpus", we fallback to using "cpu-add". Signed-off-by: Anthony PERARD Reviewed-by: Jason Andryuk --- tools/libs/light/libxl_domain.c | 89 ++++++++++++++++++++++++++++++++++++++++- 1 file changed, 87 insertions(+), 2 deletions(-) diff --git a/tools/libs/light/libxl_domain.c b/tools/libs/light/libxl_domain.c index 8c003aa7cb..c00c36c928 100644 --- a/tools/libs/light/libxl_domain.c +++ b/tools/libs/light/libxl_domain.c @@ -1805,6 +1805,7 @@ typedef struct set_vcpuonline_state { libxl_dominfo info; libxl_bitmap final_map; int index; /* for loop on final_map */ + const char *cpu_driver; } set_vcpuonline_state; static void set_vcpuonline_qmp_cpus_fast_queried(libxl__egc *, @@ -1814,6 +1815,10 @@ static void set_vcpuonline_qmp_cpus_queried(libxl__egc *, static void set_vcpuonline_qmp_query_cpus_parse(libxl__egc *, libxl__ev_qmp *qmp, const libxl__json_object *, bool query_cpus_fast, int rc); +static void set_vcpuonline_qmp_query_hotpluggable_cpus(libxl__egc *egc, + libxl__ev_qmp *qmp, const libxl__json_object *response, int rc); +static void set_vcpuonline_qmp_device_add_cpu(libxl__egc *, + libxl__ev_qmp *, const libxl__json_object *response, int rc); static void set_vcpuonline_qmp_add_cpu(libxl__egc *, libxl__ev_qmp *, const libxl__json_object *response, int rc); static void set_vcpuonline_timeout(libxl__egc *egc, @@ -1951,13 +1956,54 @@ static void set_vcpuonline_qmp_query_cpus_parse(libxl__egc *egc, libxl_bitmap_reset(final_map, i); } + qmp->callback = set_vcpuonline_qmp_query_hotpluggable_cpus; + rc = libxl__ev_qmp_send(egc, qmp, "query-hotpluggable-cpus", NULL); + out: libxl_bitmap_dispose(¤t_map); + if (rc) + set_vcpuonline_done(egc, svos, rc); /* must be last */ +} + +static void set_vcpuonline_qmp_query_hotpluggable_cpus(libxl__egc *egc, + libxl__ev_qmp *qmp, const libxl__json_object *response, int rc) +{ + set_vcpuonline_state *svos = CONTAINER_OF(qmp, *svos, qmp); + const libxl__json_object *cpu; + const libxl__json_object *cpu_driver; + + if (rc == ERROR_QMP_COMMAND_NOT_FOUND) { + /* We are probably connected to a version of QEMU older than 2.7, + * let's fallback to using "cpu-add" command. */ + svos->index = -1; + set_vcpuonline_qmp_add_cpu(egc, qmp, NULL, 0); /* must be last */ + return; + } + + if (rc) goto out; + + /* Parse response to QMP command "query-hotpluggable-cpus" + * [ { 'type': 'str', ... ] + * + * We are looking for the driver name for CPU to be hotplug. We'll + * assume that cpus property are core-id=0, thread-id=0 and + * socket-id=$cpu_index, as we start qemu with "-smp %d,maxcpus=%d", so + * we don't parse the properties listed for each hotpluggable cpus. + */ + + cpu = libxl__json_array_get(response, 0); + cpu_driver = libxl__json_map_get("type", cpu, JSON_STRING); + svos->cpu_driver = libxl__json_object_get_string(cpu_driver); + + if (!svos->cpu_driver) + rc = ERROR_QEMU_API; + +out: svos->index = -1; - set_vcpuonline_qmp_add_cpu(egc, qmp, NULL, rc); /* must be last */ + set_vcpuonline_qmp_device_add_cpu(egc, qmp, NULL, rc); /* must be last */ } -static void set_vcpuonline_qmp_add_cpu(libxl__egc *egc, +static void set_vcpuonline_qmp_device_add_cpu(libxl__egc *egc, libxl__ev_qmp *qmp, const libxl__json_object *response, int rc) { STATE_AO_GC(qmp->ao); @@ -1969,6 +2015,45 @@ static void set_vcpuonline_qmp_add_cpu(libxl__egc *egc, if (rc) goto out; + while (libxl_bitmap_cpu_valid(map, ++svos->index)) { + if (libxl_bitmap_test(map, svos->index)) { + qmp->callback = set_vcpuonline_qmp_device_add_cpu; + libxl__qmp_param_add_string(gc, &args, "id", GCSPRINTF("cpu-%d", svos->index)); + libxl__qmp_param_add_string(gc, &args, "driver", svos->cpu_driver); + /* We'll assume that we start QEMU with -smp %d,maxcpus=%d, so + * that "core-id" and "thread-id" are always 0 so that + * "socket-id" correspond the cpu index. + * Those properties are otherwise listed by + * "query-hotpluggable-cpus". */ + libxl__qmp_param_add_integer(gc, &args, "socket-id", svos->index); + libxl__qmp_param_add_integer(gc, &args, "core-id", 0); + libxl__qmp_param_add_integer(gc, &args, "thread-id", 0); + rc = libxl__ev_qmp_send(egc, qmp, "device_add", args); + if (rc) goto out; + return; + } + } + +out: + set_vcpuonline_done(egc, svos, rc); +} + +/* Fallback function for QEMU older than 2.7, when + * 'query-hotpluggable-cpus' wasn't available and vcpu object couldn't be + * added with 'device_add'. */ +static void set_vcpuonline_qmp_add_cpu(libxl__egc *egc, libxl__ev_qmp *qmp, + const libxl__json_object *response, + int rc) +{ + STATE_AO_GC(qmp->ao); + set_vcpuonline_state *svos = CONTAINER_OF(qmp, *svos, qmp); + libxl__json_object *args = NULL; + + /* Convenience aliases */ + libxl_bitmap *map = &svos->final_map; + + if (rc) goto out; + while (libxl_bitmap_cpu_valid(map, ++svos->index)) { if (libxl_bitmap_test(map, svos->index)) { qmp->callback = set_vcpuonline_qmp_add_cpu; -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:29:17 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:29:17 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146627.269884 (Exim 4.92) (envelope-from ) id 1lwMbl-0007dy-NI; Thu, 24 Jun 2021 10:29:17 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146627.269884; Thu, 24 Jun 2021 10:29:17 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMbl-0007dq-KT; Thu, 24 Jun 2021 10:29:17 +0000 Received: by outflank-mailman (input) for mailman id 146627; Thu, 24 Jun 2021 10:29:16 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMbk-0007dL-Sw for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:29:16 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMbk-0006A9-S6 for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:29:16 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwMbk-0006oI-RL for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:29:16 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=ybtuufY+5VE5NmZJ3aI173R58jneVPYEHEZUd7iAnLk=; b=IXDWNSOz9d2zRBSszP/pHFFW4K gMt34vpadBQ6+JPjSbqZQMBFlItrKchifL+0CoAimuPrLbmIqHg7nasaTtJiUwJOl6gj2jNmzDBUu 3JmZMu9Btda1FN/sTDUX9jjfYwS97u6bw7wNGtv3x0hR6DdEDYcWbtFEKp42JQ0oHeCI=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] libxl: Use -device for cd-rom drives Message-Id: Date: Thu, 24 Jun 2021 10:29:16 +0000 commit 85760c03d664400368a3f76ae0225307c25049a7 Author: Anthony PERARD AuthorDate: Tue May 11 10:28:06 2021 +0100 Commit: Julien Grall CommitDate: Tue Jun 15 13:56:26 2021 +0100 libxl: Use -device for cd-rom drives This allows to set an `id` on the device instead of only the drive. We are going to need the `id` with the "eject" and "blockdev-change-media" QMP command as using `device` parameter on those is deprecated. (`device` is the `id` of the `-drive` on the command line). We set the same `id` on both -device and -drive as QEMU doesn't complain and we can then either do "eject id=$id" or "eject device=$id". Using "-drive + -device" instead of only "-drive" has been available since at least QEMU 0.15, and seems to be the preferred way as it separates the host part (-drive which describe the disk image location and format) from the guest part (-device which describe the emulated device). More information in qemu.git/docs/qdev-device-use.txt . Changing the command line during migration for the cdrom seems fine. Also the documentation about migration in QEMU explains that the device state ID is "been formed from a bus name and device address", so second IDE bus and first device address on bus is still thus and doesn't matter if written "-drive if=ide,index=2" or "-drive ide-cd,bus=ide.1,unit=0". See qemu.git/docs/devel/migration.rst . Signed-off-by: Anthony PERARD Reviewed-by: Jason Andryuk --- tools/libs/light/libxl_dm.c | 16 +++++++++++++--- 1 file changed, 13 insertions(+), 3 deletions(-) diff --git a/tools/libs/light/libxl_dm.c b/tools/libs/light/libxl_dm.c index 0a0c1ef7c6..5b01cf2841 100644 --- a/tools/libs/light/libxl_dm.c +++ b/tools/libs/light/libxl_dm.c @@ -1913,6 +1913,7 @@ static int libxl__build_device_model_args_new(libxl__gc *gc, } if (disks[i].is_cdrom) { + const char *drive_id; if (disk > 4) { LOGD(WARN, guest_domid, "Emulated CDROM can be only one of the first 4 disks.\n" "Disk %s will be available via PV drivers but not as an " @@ -1920,13 +1921,22 @@ static int libxl__build_device_model_args_new(libxl__gc *gc, disks[i].vdev); continue; } - drive = libxl__sprintf(gc, - "if=ide,index=%d,readonly=on,media=cdrom,id=ide-%i", - disk, dev_number); + + drive_id = GCSPRINTF("ide-%i", dev_number); + drive = GCSPRINTF("if=none,readonly=on,id=%s", drive_id); if (target_path) drive = libxl__sprintf(gc, "%s,file=%s,format=%s", drive, target_path, format); + + flexarray_vappend(dm_args, + "-drive", drive, + "-device", + GCSPRINTF("ide-cd,id=%s,drive=%s,bus=ide.%u,unit=%u", + drive_id, drive_id, + disk / 2, disk % 2), + NULL); + continue; } else { /* * Explicit sd disks are passed through as is. -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:29:27 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:29:27 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146628.269889 (Exim 4.92) (envelope-from ) id 1lwMbv-0007j6-Or; Thu, 24 Jun 2021 10:29:27 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146628.269889; Thu, 24 Jun 2021 10:29:27 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMbv-0007iy-Lz; Thu, 24 Jun 2021 10:29:27 +0000 Received: by outflank-mailman (input) for mailman id 146628; Thu, 24 Jun 2021 10:29:27 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMbu-0007ik-W4 for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:29:26 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMbu-0006AR-VI for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:29:26 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwMbu-0006oz-UN for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:29:26 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=YJswDTg2oX12ujAof32t1n6cqlBUUIS1G0VKfq5zeOM=; b=Kih+3bOmBIXrwavfueq1EmR8Cd h8kGk0lK1e22caVnSFz7nq4EDMEoK4aqwEeulqJ693di0c0prIXiTHJ5C/QyKDFiJ8luKbifNayXR /m/MP8sbVrEhLCiyTqfTUG14uI6GjCB3uk8M3Xfon/tS4zy8ta8foj6qNpgFKTERW8nQ=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] libxl: Assert qmp_ev's state in qmp_ev_qemu_compare_version Message-Id: Date: Thu, 24 Jun 2021 10:29:26 +0000 commit 0ff26a3225d69ffec76fe5aca8296852fa951204 Author: Anthony PERARD AuthorDate: Tue May 11 10:28:07 2021 +0100 Commit: Julien Grall CommitDate: Tue Jun 15 13:56:26 2021 +0100 libxl: Assert qmp_ev's state in qmp_ev_qemu_compare_version We are supposed to read the version information only when qmp_ev is in state "Connected" (that correspond to state==qmp_state_connected), assert it so that the function isn't used too early. Signed-off-by: Anthony PERARD Reviewed-by: Jason Andryuk --- tools/libs/light/libxl_qmp.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/tools/libs/light/libxl_qmp.c b/tools/libs/light/libxl_qmp.c index 9b638e6f54..d0967c9f02 100644 --- a/tools/libs/light/libxl_qmp.c +++ b/tools/libs/light/libxl_qmp.c @@ -292,6 +292,8 @@ static int qmp_handle_response(libxl__gc *gc, libxl__qmp_handler *qmp, static int qmp_ev_qemu_compare_version(libxl__ev_qmp *ev, int major, int minor, int micro) { + assert(ev->state == qmp_state_connected); + #define CHECK_VERSION(level) do { \ if (ev->qemu_version.level > (level)) return +1; \ if (ev->qemu_version.level < (level)) return -1; \ -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:29:37 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:29:37 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146632.269892 (Exim 4.92) (envelope-from ) id 1lwMc5-0007ov-Qf; Thu, 24 Jun 2021 10:29:37 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146632.269892; Thu, 24 Jun 2021 10:29:37 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMc5-0007on-Na; Thu, 24 Jun 2021 10:29:37 +0000 Received: by outflank-mailman (input) for mailman id 146632; Thu, 24 Jun 2021 10:29:37 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMc5-0007oY-2x for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:29:37 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMc5-0006Ae-2E for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:29:37 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwMc5-0006pi-1a for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:29:37 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=7VJAhYu1ws+qioPI8vAxHZMdmMhrqN3wDgNxInD0Fpo=; b=TaEo0qJHcCU9YQsmTM9hLbRC4z bNp50Q6OIGyHaRieu4GGGchVShGXbhqpv11dihQwKweM9S9MdDPm7U32IMP+kRdIFC0GveMtX4dux FTtN6HNqB4VAInPjpQiXxTSd0f85b532Fh5RmlRJuV9rr9osR+dLkMYhl2uCdsm+2/vw=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] libxl: Export libxl__qmp_ev_qemu_compare_version Message-Id: Date: Thu, 24 Jun 2021 10:29:37 +0000 commit 7c313e8365eb663311a0cf39f77b4f5880244765 Author: Anthony PERARD AuthorDate: Tue May 11 10:28:08 2021 +0100 Commit: Julien Grall CommitDate: Tue Jun 15 13:56:26 2021 +0100 libxl: Export libxl__qmp_ev_qemu_compare_version We are going to want to check QEMU's version in other places where we can use libxl__ev_qmp_send. Signed-off-by: Anthony PERARD Reviewed-by: Jason Andryuk --- tools/libs/light/libxl_internal.h | 8 ++++++++ tools/libs/light/libxl_qmp.c | 4 ++-- 2 files changed, 10 insertions(+), 2 deletions(-) diff --git a/tools/libs/light/libxl_internal.h b/tools/libs/light/libxl_internal.h index 44a2f3c8fe..0b4671318c 100644 --- a/tools/libs/light/libxl_internal.h +++ b/tools/libs/light/libxl_internal.h @@ -492,6 +492,14 @@ _hidden int libxl__ev_qmp_send(libxl__egc *egc, libxl__ev_qmp *ev, const char *cmd, libxl__json_object *args); _hidden void libxl__ev_qmp_dispose(libxl__gc *gc, libxl__ev_qmp *ev); +/* return values: + * < 0 if qemu's version < asked version + * = 0 if qemu's version == asked version + * > 0 if qemu's version > asked version + */ +_hidden int libxl__qmp_ev_qemu_compare_version(libxl__ev_qmp *ev, int major, + int minor, int micro); + typedef enum { /* initial state */ qmp_state_disconnected = 1, diff --git a/tools/libs/light/libxl_qmp.c b/tools/libs/light/libxl_qmp.c index d0967c9f02..fb146a54cb 100644 --- a/tools/libs/light/libxl_qmp.c +++ b/tools/libs/light/libxl_qmp.c @@ -289,7 +289,7 @@ static int qmp_handle_response(libxl__gc *gc, libxl__qmp_handler *qmp, * = 0 if qemu's version == asked version * > 0 if qemu's version > asked version */ -static int qmp_ev_qemu_compare_version(libxl__ev_qmp *ev, int major, +int libxl__qmp_ev_qemu_compare_version(libxl__ev_qmp *ev, int major, int minor, int micro) { assert(ev->state == qmp_state_connected); @@ -1073,7 +1073,7 @@ static void dm_state_save_to_fdset(libxl__egc *egc, libxl__ev_qmp *ev, int fdset /* The `live` parameter was added to QEMU 2.11. It signals QEMU that * the save operation is for a live migration rather than for taking a * snapshot. */ - if (qmp_ev_qemu_compare_version(ev, 2, 11, 0) >= 0) + if (libxl__qmp_ev_qemu_compare_version(ev, 2, 11, 0) >= 0) libxl__qmp_param_add_bool(gc, &args, "live", dsps->live); QMP_PARAMETERS_SPRINTF(&args, "filename", "/dev/fdset/%d", fdset); rc = libxl__ev_qmp_send(egc, ev, "xen-save-devices-state", args); -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:29:47 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:29:47 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146635.269897 (Exim 4.92) (envelope-from ) id 1lwMcF-0007tN-SD; Thu, 24 Jun 2021 10:29:47 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146635.269897; Thu, 24 Jun 2021 10:29:47 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMcF-0007tG-PA; Thu, 24 Jun 2021 10:29:47 +0000 Received: by outflank-mailman (input) for mailman id 146635; Thu, 24 Jun 2021 10:29:47 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMcF-0007sz-60 for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:29:47 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMcF-0006Av-5D for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:29:47 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwMcF-0006qI-4Z for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:29:47 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=llj2KJiHLX/v8bZ6aVGsXldO7YjoATEJhNetSyVO2uI=; b=OK34Ki/Rzc6Z2Zqg3izimJresG VgqJhL2y6ceSrjzFn256kAs3lTnWRVjtDbv+1dzS3PIFWFmL+zoWD+Rs2vLWz3M52Bnk4pq91DQaY ku6RL2MxC5gc9YqQQTTavURBIAc99gJ/1e2l32IUwBxP0+0muwtSdjeitn1DtdBxyTeI=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] libxl: Use `id` with the "eject" QMP command Message-Id: Date: Thu, 24 Jun 2021 10:29:47 +0000 commit 0c0b3a7e4a2d65fd252b89b46bdcdb048bb24b6c Author: Anthony PERARD AuthorDate: Tue May 11 10:28:09 2021 +0100 Commit: Julien Grall CommitDate: Tue Jun 15 13:56:26 2021 +0100 libxl: Use `id` with the "eject" QMP command `device` parameter is deprecated since QEMU 2.8. This requires changes to the command line introduced by: "libxl: Use -device for cd-rom drives" Signed-off-by: Anthony PERARD Reviewed-by: Jason Andryuk --- tools/libs/light/libxl_disk.c | 43 ++++++++++++++++++++++++++++++++++++------- 1 file changed, 36 insertions(+), 7 deletions(-) diff --git a/tools/libs/light/libxl_disk.c b/tools/libs/light/libxl_disk.c index 411ffeaca6..faabdea7a4 100644 --- a/tools/libs/light/libxl_disk.c +++ b/tools/libs/light/libxl_disk.c @@ -656,6 +656,8 @@ typedef struct { static void cdrom_insert_lock_acquired(libxl__egc *, libxl__ev_slowlock *, int rc); +static void cdrom_insert_qmp_connected(libxl__egc *, libxl__ev_qmp *, + const libxl__json_object *, int rc); static void cdrom_insert_ejected(libxl__egc *egc, libxl__ev_qmp *, const libxl__json_object *, int rc); static void cdrom_insert_addfd_cb(libxl__egc *egc, libxl__ev_qmp *, @@ -770,13 +772,12 @@ static void cdrom_insert_lock_acquired(libxl__egc *egc, */ if (cis->dm_ver == LIBXL_DEVICE_MODEL_VERSION_QEMU_XEN) { - libxl__json_object *args = NULL; - int devid = libxl__device_disk_dev_number(cis->disk->vdev, - NULL, NULL); - - QMP_PARAMETERS_SPRINTF(&args, "device", "ide-%i", devid); - cis->qmp.callback = cdrom_insert_ejected; - rc = libxl__ev_qmp_send(egc, &cis->qmp, "eject", args); + /* Before running the "eject" command, we need to know QEMU's + * version to find out which command to issue. + * cis->qmp isn't in Connected state yet, so run a dummy command + * to have QEMU's version available. */ + cis->qmp.callback = cdrom_insert_qmp_connected; + rc = libxl__ev_qmp_send(egc, &cis->qmp, "query-version", NULL); if (rc) goto out; } else { cdrom_insert_ejected(egc, &cis->qmp, NULL, 0); /* must be last */ @@ -787,6 +788,34 @@ out: cdrom_insert_done(egc, cis, rc); /* must be last */ } +static void cdrom_insert_qmp_connected(libxl__egc *egc, libxl__ev_qmp *qmp, + const libxl__json_object *response, + int rc) +{ + libxl__cdrom_insert_state *cis = CONTAINER_OF(qmp, *cis, qmp); + STATE_AO_GC(cis->ao); + libxl__json_object *args = NULL; + int devid = libxl__device_disk_dev_number(cis->disk->vdev, + NULL, NULL); + + if (rc) goto out; + + /* Using `device` parameter is deprecated since QEMU 2.8, we should + * use `id` now. They both have different meaning but we set the + * same `id` on -drive and -device on the command line. + */ + if (libxl__qmp_ev_qemu_compare_version(qmp, 2, 8, 0) >= 0) + QMP_PARAMETERS_SPRINTF(&args, "id", "ide-%i", devid); + else + QMP_PARAMETERS_SPRINTF(&args, "device", "ide-%i", devid); + qmp->callback = cdrom_insert_ejected; + rc = libxl__ev_qmp_send(egc, qmp, "eject", args); + if (rc) goto out; + return; +out: + cdrom_insert_done(egc, cis, rc); /* must be last */ +} + static void cdrom_insert_ejected(libxl__egc *egc, libxl__ev_qmp *qmp, const libxl__json_object *response, -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:29:58 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:29:58 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146636.269901 (Exim 4.92) (envelope-from ) id 1lwMcP-0007y9-VI; Thu, 24 Jun 2021 10:29:57 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146636.269901; Thu, 24 Jun 2021 10:29:57 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMcP-0007y1-SG; Thu, 24 Jun 2021 10:29:57 +0000 Received: by outflank-mailman (input) for mailman id 146636; Thu, 24 Jun 2021 10:29:57 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMcP-0007xv-93 for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:29:57 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMcP-0006B6-8I for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:29:57 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwMcP-0006qw-7Z for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:29:57 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=72tR/WJz78Znmm/rA9Qa3r71r6oMUexgwhSBhD/3lLk=; b=pHX9YK+3wxmGEWbQTAdsSy/+Fg P/+e0C2rQxG9pA7ez3N/0McRbHuNyUUIPR+IE88uKxJRwfXH0B7QQUEUgjvcxv62qIVfLYDmqhdut r1s97t+iltGi60b8Nyn3tsnA4imo8vv/Te/RMpptrRt2aEMM667WsGioJvERPZhMcErI=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] libxl: Replace QMP command "change" by "blockdev-change-media" Message-Id: Date: Thu, 24 Jun 2021 10:29:57 +0000 commit fe6630ddc4e8a8fbf8dd28a1bc58e3881393f9c1 Author: Anthony PERARD AuthorDate: Tue May 11 10:28:10 2021 +0100 Commit: Julien Grall CommitDate: Tue Jun 15 13:56:26 2021 +0100 libxl: Replace QMP command "change" by "blockdev-change-media" "change" command as been removed in QEMU 6.0. We can use "blockdev-change-medium" instead. Using `id` with "blockdev-change-medium" requires a change to the QEMU command line, introduced by: "libxl: Use -device for cd-rom drives" Signed-off-by: Anthony PERARD Reviewed-by: Jason Andryuk --- tools/libs/light/libxl_disk.c | 24 +++++++++++++++++++----- 1 file changed, 19 insertions(+), 5 deletions(-) diff --git a/tools/libs/light/libxl_disk.c b/tools/libs/light/libxl_disk.c index faabdea7a4..93936d0dd0 100644 --- a/tools/libs/light/libxl_disk.c +++ b/tools/libs/light/libxl_disk.c @@ -962,12 +962,26 @@ static void cdrom_insert_addfd_cb(libxl__egc *egc, fdset = libxl__json_object_get_integer(o); devid = libxl__device_disk_dev_number(disk->vdev, NULL, NULL); - QMP_PARAMETERS_SPRINTF(&args, "device", "ide-%i", devid); - QMP_PARAMETERS_SPRINTF(&args, "target", "/dev/fdset/%d", fdset); - libxl__qmp_param_add_string(gc, &args, "arg", - libxl__qemu_disk_format_string(disk->format)); qmp->callback = cdrom_insert_inserted; - rc = libxl__ev_qmp_send(egc, qmp, "change", args); + + /* "change" is deprecated since QEMU 2.5 and the `device` parameter for + * for "blockdev-change-medium" is deprecated in QEMU 2.8. + * But `id` is only available in 2.8 we'll start using the new command + * with `id` with QEMU 2.8. + */ + if (libxl__qmp_ev_qemu_compare_version(qmp, 2, 8, 0) >= 0) { + QMP_PARAMETERS_SPRINTF(&args, "id", "ide-%i", devid); + QMP_PARAMETERS_SPRINTF(&args, "filename", "/dev/fdset/%d", fdset); + libxl__qmp_param_add_string(gc, &args, "format", + libxl__qemu_disk_format_string(disk->format)); + rc = libxl__ev_qmp_send(egc, qmp, "blockdev-change-medium", args); + } else { + QMP_PARAMETERS_SPRINTF(&args, "device", "ide-%i", devid); + QMP_PARAMETERS_SPRINTF(&args, "target", "/dev/fdset/%d", fdset); + libxl__qmp_param_add_string(gc, &args, "arg", + libxl__qemu_disk_format_string(disk->format)); + rc = libxl__ev_qmp_send(egc, qmp, "change", args); + } out: if (rc) cdrom_insert_done(egc, cis, rc); /* must be last */ -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:30:08 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:30:08 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146637.269905 (Exim 4.92) (envelope-from ) id 1lwMca-0000FL-1h; Thu, 24 Jun 2021 10:30:08 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146637.269905; Thu, 24 Jun 2021 10:30:08 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMcZ-0000FD-U5; Thu, 24 Jun 2021 10:30:07 +0000 Received: by outflank-mailman (input) for mailman id 146637; Thu, 24 Jun 2021 10:30:07 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMcZ-0000Eu-Bv for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:30:07 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMcZ-0006Bh-B8 for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:30:07 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwMcZ-0006rl-AP for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:30:07 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=iUTW29Y9CnDp5RV5yNmbYNKtVfJQuUtBweUQCQyU3oQ=; b=3UjHZjSMAjf9fXHBeY9gdeAcGl gcXqEABCGiSkii/gnBIUbN3EKDGlhmqhqEiGsAyiskjRzp5se/8sD2sKTw1+XMikQ5v3QYTtfbzN8 oaFTUdE0ev6avEtMezAsSP0c5IgFFq9THZ/rpB8pdfdcGGcMigbtt7Xtw75sqSCAxKZY=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] x86/vpt: fully init timers before putting onto list Message-Id: Date: Thu, 24 Jun 2021 10:30:07 +0000 commit 6d622f3a96bbd76ce8422c6e3805e6609417ec76 Author: Jan Beulich AuthorDate: Tue Jun 15 15:14:20 2021 +0200 Commit: Jan Beulich CommitDate: Tue Jun 15 15:14:20 2021 +0200 x86/vpt: fully init timers before putting onto list With pt_vcpu_lock() no longer acquiring the pt_migrate lock, parties iterating the list and acting on the timers of the list entries will no longer be kept from entering their loops by create_periodic_time()'s holding of that lock. Therefore at least init_timer() needs calling ahead of list insertion, but keep this and set_timer() together. Fixes: 8113b02f0bf8 ("x86/vpt: do not take pt_migrate rwlock in some cases") Reported-by: Igor Druzhinin Signed-off-by: Jan Beulich Reviewed-by: Roger Pau Monné Reviewed-by: Boris Ostrovsky --- xen/arch/x86/hvm/vpt.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/xen/arch/x86/hvm/vpt.c b/xen/arch/x86/hvm/vpt.c index 4cc0a0848b..6fdc3e19fe 100644 --- a/xen/arch/x86/hvm/vpt.c +++ b/xen/arch/x86/hvm/vpt.c @@ -554,14 +554,14 @@ void create_periodic_time( pt->cb = cb; pt->priv = data; + init_timer(&pt->timer, pt_timer_fn, pt, v->processor); + set_timer(&pt->timer, pt->scheduled); + pt_vcpu_lock(v); pt->on_list = 1; list_add(&pt->list, &v->arch.hvm.tm_list); pt_vcpu_unlock(v); - init_timer(&pt->timer, pt_timer_fn, pt, v->processor); - set_timer(&pt->timer, pt->scheduled); - write_unlock(&v->domain->arch.hvm.pl_time->pt_migrate); } -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:30:19 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:30:19 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146638.269909 (Exim 4.92) (envelope-from ) id 1lwMcl-0000Nl-33; Thu, 24 Jun 2021 10:30:19 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146638.269909; Thu, 24 Jun 2021 10:30:19 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMck-0000Nd-W8; Thu, 24 Jun 2021 10:30:18 +0000 Received: by outflank-mailman (input) for mailman id 146638; Thu, 24 Jun 2021 10:30:17 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMcj-0000NI-Eu for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:30:17 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMcj-0006CE-EA for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:30:17 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwMcj-0006sY-DP for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:30:17 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=bXnVVRGbOgZBNrmqVNw6OeJ97X0u0s1aIIEMJkjKtko=; b=1rY3HAiXLIFL5kqvVfSk8Rbq+s SPOOYA2jr0uRdBcwAh7zdJm3/qawhODOVfb1zPIpKrbJppIHSHBkOfZusTnuv/XP2jYhQ6SELwz15 yvYfSIsmR0PBeVNP38cWhJBobFh23sZDOHge34qwfsmqO/Fpa9KmCf+MorH41JR8vhrQ=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] x86: move .altinstr_replacement past _einittext Message-Id: Date: Thu, 24 Jun 2021 10:30:17 +0000 commit 93c5f98296fc78de79d621418a1e62fd413e73d1 Author: Jan Beulich AuthorDate: Tue Jun 15 15:15:26 2021 +0200 Commit: Jan Beulich CommitDate: Tue Jun 15 15:15:26 2021 +0200 x86: move .altinstr_replacement past _einittext This section's contents do not represent part of actual hypervisor text, so shouldn't be included in what is_kernel_inittext() or (while still booting) is_active_kernel_text() report "true" for. Keep them in .init.text though, as there's no real reason to have a separate section for this in the final binary. Signed-off-by: Jan Beulich Acked-by: Andrew Cooper --- xen/arch/x86/xen.lds.S | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/xen/arch/x86/xen.lds.S b/xen/arch/x86/xen.lds.S index 34f647a44e..9c6c1c8005 100644 --- a/xen/arch/x86/xen.lds.S +++ b/xen/arch/x86/xen.lds.S @@ -185,13 +185,13 @@ SECTIONS #endif _sinittext = .; *(.init.text) + _einittext = .; /* * Here are the replacement instructions. The linker sticks them * as binary blobs. The .altinstructions has enough data to get * the address and the length of them to patch the kernel safely. */ *(.altinstr_replacement) - _einittext = .; #ifdef EFI /* EFI wants to merge all of .init.* ELF doesn't. */ . = ALIGN(SMP_CACHE_BYTES); -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:30:29 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:30:29 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146639.269913 (Exim 4.92) (envelope-from ) id 1lwMcv-0000RX-51; Thu, 24 Jun 2021 10:30:29 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146639.269913; Thu, 24 Jun 2021 10:30:29 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMcv-0000RP-1X; Thu, 24 Jun 2021 10:30:29 +0000 Received: by outflank-mailman (input) for mailman id 146639; Thu, 24 Jun 2021 10:30:27 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMct-0000R4-Hk for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:30:27 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMct-0006CP-H1 for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:30:27 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwMct-0006tT-GL for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:30:27 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=xbAA3SFywmo8Y3tdS8fSI67CaldxkYdY/B355n1EqcA=; b=g3XWjDROpwXynw6XuUfzAq8yen /RjKliIBfvWm6/1a0riNyGvja/cJlJo2bAkyvixwx0d+g45sgmdk6iEvMzd1nx9kg26NhG8uFrx8f FyowRm7fWlXoNaTfXbS+Qd2uqvheDDO0E4HOdyYHXXMT13sVR/jSjXvImSDC2+teZcfE=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] libs/foreignmemory: Fix osdep_xenforeignmemory_map prototype Message-Id: Date: Thu, 24 Jun 2021 10:30:27 +0000 commit 5d3e4ebb5c71477d74a0c503438545a0126d3863 Author: Anthony PERARD AuthorDate: Tue Jun 1 16:41:47 2021 +0100 Commit: Julien Grall CommitDate: Tue Jun 15 18:07:58 2021 +0100 libs/foreignmemory: Fix osdep_xenforeignmemory_map prototype Commit cf8c4d3d13b8 made some preparation to have one day variable-length-array argument, but didn't declare the array in the function prototype the same way as in the function definition. And now GCC 11 complains about it. Fixes: cf8c4d3d13b8 ("tools/libs/foreignmemory: pull array length argument to map forward") Signed-off-by: Anthony PERARD Reviewed-by: Jan Beulich --- tools/libs/foreignmemory/private.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/tools/libs/foreignmemory/private.h b/tools/libs/foreignmemory/private.h index 1ee3626dd2..5bb0cefb09 100644 --- a/tools/libs/foreignmemory/private.h +++ b/tools/libs/foreignmemory/private.h @@ -32,7 +32,7 @@ int osdep_xenforeignmemory_close(xenforeignmemory_handle *fmem); void *osdep_xenforeignmemory_map(xenforeignmemory_handle *fmem, uint32_t dom, void *addr, int prot, int flags, size_t num, - const xen_pfn_t arr[num], int err[num]); + const xen_pfn_t arr[/*num*/], int err[/*num*/]); int osdep_xenforeignmemory_unmap(xenforeignmemory_handle *fmem, void *addr, size_t num); -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:30:39 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:30:39 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146640.269917 (Exim 4.92) (envelope-from ) id 1lwMd5-0000V6-6q; Thu, 24 Jun 2021 10:30:39 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146640.269917; Thu, 24 Jun 2021 10:30:39 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMd5-0000Uw-3W; Thu, 24 Jun 2021 10:30:39 +0000 Received: by outflank-mailman (input) for mailman id 146640; Thu, 24 Jun 2021 10:30:37 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMd3-0000Ud-Mj for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:30:37 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMd3-0006Cf-KK for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:30:37 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwMd3-0006uc-JT for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:30:37 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=9bH66801GxCceHTZNAGXBZrH9962tH1iH+EeC4Biv+8=; b=xgNqWgmImetS4PiH2hbigpvnLO xFac+3ojyItitDR87vESvWrzR28XZKqGV2lutWMO0UF0D2A0TnVZVZ5DSYonV4dq+/LRHUHz2V5iU uAqUdokC4OtAs7Rx1fI7hUGdneq+xyY1HrEx3f6lNH+jMU3w5uIToeXF6b1sVob56bvE=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] x86/platform: Improve MSR permission handling for XENPF_resource_op Message-Id: Date: Thu, 24 Jun 2021 10:30:37 +0000 commit 2cf3b4b92ab1a34e8cb1e859196e1492c89299de Author: Andrew Cooper AuthorDate: Thu Jun 10 11:01:06 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 15 20:50:32 2021 +0100 x86/platform: Improve MSR permission handling for XENPF_resource_op The logic to disallow writes to the TSC is out-of-place, and should be in check_resource_access() rather than in resource_access(). Split the existing allow_access_msr() into two - msr_{read,write}_allowed() - and move all permissions checks here. Furthermore, guard access to MSR_IA32_CMT_{EVTSEL,CTR} to prohibit their use on hardware which is lacking the QoS Monitoring feature. Introduce cpu_has_pqe to help with the logic. Signed-off-by: Andrew Cooper Reviewed-by: Jan Beulich --- xen/arch/x86/platform_hypercall.c | 41 ++++++++++++++++++++++++++++----------- xen/arch/x86/psr.c | 2 +- xen/include/asm-x86/cpufeature.h | 1 + 3 files changed, 32 insertions(+), 12 deletions(-) diff --git a/xen/arch/x86/platform_hypercall.c b/xen/arch/x86/platform_hypercall.c index 23fadbc782..41d8e59563 100644 --- a/xen/arch/x86/platform_hypercall.c +++ b/xen/arch/x86/platform_hypercall.c @@ -64,17 +64,33 @@ long cpu_frequency_change_helper(void *data) return cpu_frequency_change((uint64_t)data); } -static bool allow_access_msr(unsigned int msr) +static bool msr_read_allowed(unsigned int msr) { switch ( msr ) { - /* MSR for CMT, refer to chapter 17.14 of Intel SDM. */ case MSR_IA32_CMT_EVTSEL: case MSR_IA32_CMT_CTR: + return cpu_has_pqe; + case MSR_IA32_TSC: return true; } + if ( ppin_msr && msr == ppin_msr ) + return true; + + return false; +} + +static bool msr_write_allowed(unsigned int msr) +{ + switch ( msr ) + { + case MSR_IA32_CMT_EVTSEL: + case MSR_IA32_CMT_CTR: + return cpu_has_pqe; + } + return false; } @@ -96,15 +112,19 @@ void check_resource_access(struct resource_access *ra) switch ( entry->u.cmd ) { case XEN_RESOURCE_OP_MSR_READ: - if ( ppin_msr && entry->idx == ppin_msr ) - break; - /* fall through */ + if ( entry->idx >> 32 ) + ret = -EINVAL; + else if ( !msr_read_allowed(entry->idx) ) + ret = -EPERM; + break; + case XEN_RESOURCE_OP_MSR_WRITE: if ( entry->idx >> 32 ) ret = -EINVAL; - else if ( !allow_access_msr(entry->idx) ) - ret = -EACCES; + else if ( !msr_write_allowed(entry->idx) ) + ret = -EPERM; break; + default: ret = -EOPNOTSUPP; break; @@ -163,12 +183,11 @@ void resource_access(void *info) } } break; + case XEN_RESOURCE_OP_MSR_WRITE: - if ( unlikely(entry->idx == MSR_IA32_TSC) ) - ret = -EPERM; - else - ret = wrmsr_safe(entry->idx, entry->val); + ret = wrmsr_safe(entry->idx, entry->val); break; + default: BUG(); break; diff --git a/xen/arch/x86/psr.c b/xen/arch/x86/psr.c index d7f8864651..d805b85dc6 100644 --- a/xen/arch/x86/psr.c +++ b/xen/arch/x86/psr.c @@ -1558,7 +1558,7 @@ static void psr_cpu_init(void) struct cpuid_leaf regs; uint32_t feat_mask; - if ( !psr_alloc_feat_enabled() || !boot_cpu_has(X86_FEATURE_PQE) ) + if ( !psr_alloc_feat_enabled() || !cpu_has_pqe ) goto assoc_init; if ( boot_cpu_data.cpuid_level < PSR_CPUID_LEVEL_CAT ) diff --git a/xen/include/asm-x86/cpufeature.h b/xen/include/asm-x86/cpufeature.h index a539a4bacd..5f6b83f71c 100644 --- a/xen/include/asm-x86/cpufeature.h +++ b/xen/include/asm-x86/cpufeature.h @@ -94,6 +94,7 @@ #define cpu_has_bmi2 boot_cpu_has(X86_FEATURE_BMI2) #define cpu_has_invpcid boot_cpu_has(X86_FEATURE_INVPCID) #define cpu_has_rtm boot_cpu_has(X86_FEATURE_RTM) +#define cpu_has_pqe boot_cpu_has(X86_FEATURE_PQE) #define cpu_has_fpu_sel (!boot_cpu_has(X86_FEATURE_NO_FPU_SEL)) #define cpu_has_mpx boot_cpu_has(X86_FEATURE_MPX) #define cpu_has_avx512f boot_cpu_has(X86_FEATURE_AVX512F) -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:30:49 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:30:49 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146641.269921 (Exim 4.92) (envelope-from ) id 1lwMdF-0000Yf-98; Thu, 24 Jun 2021 10:30:49 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146641.269921; Thu, 24 Jun 2021 10:30:49 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMdF-0000YU-5N; Thu, 24 Jun 2021 10:30:49 +0000 Received: by outflank-mailman (input) for mailman id 146641; Thu, 24 Jun 2021 10:30:47 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMdD-0000YA-Nk for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:30:47 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMdD-0006Cq-N3 for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:30:47 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwMdD-0006w3-MU for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:30:47 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=n7OGhyW1VaGZG1hNkMdpNlOGuxGOuvofl59Wb4yF0Gs=; b=FjtGIRa1Npgmqq0DJ+lAbqbeER 10WBnIaKaim+TCPnp2SN7pWMY0lp/QYIuSi7AFFuO+qpEmRmYRRLq76cg2/X8pSpC/qWCMQbjxIOk wLpC0iKNeIRRQEjK/dHGpPXAKiNqmTz0oRTCki6+naqgtwei9ThMaxOyinp8T6Nh/x8w=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] x86/platform: Permit reading the TSX control MSRs via XENPF_resource_op Message-Id: Date: Thu, 24 Jun 2021 10:30:47 +0000 commit 3ccaa174faeb8fbd68f79c670e32c6689cfc7ffc Author: Andrew Cooper AuthorDate: Thu Jun 10 11:02:47 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 15 20:50:32 2021 +0100 x86/platform: Permit reading the TSX control MSRs via XENPF_resource_op We are going to want this to write some tests with. Signed-off-by: Andrew Cooper Reviewed-by: Jan Beulich --- xen/arch/x86/platform_hypercall.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/xen/arch/x86/platform_hypercall.c b/xen/arch/x86/platform_hypercall.c index 41d8e59563..284c2dfb9e 100644 --- a/xen/arch/x86/platform_hypercall.c +++ b/xen/arch/x86/platform_hypercall.c @@ -74,6 +74,12 @@ static bool msr_read_allowed(unsigned int msr) case MSR_IA32_TSC: return true; + + case MSR_TSX_FORCE_ABORT: + return cpu_has_tsx_force_abort; + + case MSR_TSX_CTRL: + return cpu_has_tsx_ctrl; } if ( ppin_msr && msr == ppin_msr ) -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:30:59 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:30:59 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146642.269924 (Exim 4.92) (envelope-from ) id 1lwMdP-0000cl-9n; Thu, 24 Jun 2021 10:30:59 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146642.269924; Thu, 24 Jun 2021 10:30:59 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMdP-0000cd-72; Thu, 24 Jun 2021 10:30:59 +0000 Received: by outflank-mailman (input) for mailman id 146642; Thu, 24 Jun 2021 10:30:57 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMdN-0000cK-QV for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:30:57 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMdN-0006D3-Pn for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:30:57 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwMdN-0006ww-PD for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:30:57 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=9s5D9H2/kIaEVywVyEUKPJJuBPMAIt7Lh0gbrBUlrrU=; b=5Xgk26q96oKBwSXPKEZ0gLX0Z3 ns5njWOGX5uIOraE6gGe2VdaDVR08V5dVk7lxRG6nMEqiH7EAZIcCkMLC0HvaZT18E/b9xS2tXTmS /ABFx/XMRHd/voQmdHzGIxlp/AVaQOtXRlbm2Z68fkbiKsRfh4KPLRZZHrqk7PYtH95U=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] x86/msr: Expose MSR_ARCH_CAPS in the raw and host policies Message-Id: Date: Thu, 24 Jun 2021 10:30:57 +0000 commit b672695e748869b2e2e2cb03f671f12003d2b079 Author: Andrew Cooper AuthorDate: Fri Jun 11 11:37:53 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 15 20:50:32 2021 +0100 x86/msr: Expose MSR_ARCH_CAPS in the raw and host policies MSR_ARCH_CAPS is still not supported for guests yet (other than the hardware domain), until the toolstack learns how to construct an MSR policy. However, we want access to the host ARCH_CAPS_TSX_CTRL value in particular for testing purposes. Signed-off-by: Andrew Cooper Reviewed-by: Jan Beulich --- xen/arch/x86/msr.c | 15 +++++++++++++++ 1 file changed, 15 insertions(+) diff --git a/xen/arch/x86/msr.c b/xen/arch/x86/msr.c index 374f92b2c5..72bbe46949 100644 --- a/xen/arch/x86/msr.c +++ b/xen/arch/x86/msr.c @@ -47,8 +47,13 @@ struct msr_policy __read_mostly hvm_def_msr_policy; static void __init calculate_raw_policy(void) { + struct msr_policy *mp = &raw_msr_policy; + /* 0x000000ce MSR_INTEL_PLATFORM_INFO */ /* Was already added by probe_cpuid_faulting() */ + + if ( cpu_has_arch_caps ) + rdmsrl(MSR_ARCH_CAPABILITIES, mp->arch_caps.raw); } static void __init calculate_host_policy(void) @@ -60,6 +65,12 @@ static void __init calculate_host_policy(void) /* 0x000000ce MSR_INTEL_PLATFORM_INFO */ /* probe_cpuid_faulting() sanity checks presence of MISC_FEATURES_ENABLES */ mp->platform_info.cpuid_faulting = cpu_has_cpuid_faulting; + + /* Temporary, until we have known_features[] for feature bits in MSRs. */ + mp->arch_caps.raw &= + (ARCH_CAPS_RDCL_NO | ARCH_CAPS_IBRS_ALL | ARCH_CAPS_RSBA | + ARCH_CAPS_SKIP_L1DFL | ARCH_CAPS_SSB_NO | ARCH_CAPS_MDS_NO | + ARCH_CAPS_IF_PSCHANGE_MC_NO | ARCH_CAPS_TSX_CTRL | ARCH_CAPS_TAA_NO); } static void __init calculate_pv_max_policy(void) @@ -67,6 +78,8 @@ static void __init calculate_pv_max_policy(void) struct msr_policy *mp = &pv_max_msr_policy; *mp = host_msr_policy; + + mp->arch_caps.raw = 0; /* Not supported yet. */ } static void __init calculate_pv_def_policy(void) @@ -84,6 +97,8 @@ static void __init calculate_hvm_max_policy(void) /* It's always possible to emulate CPUID faulting for HVM guests */ mp->platform_info.cpuid_faulting = true; + + mp->arch_caps.raw = 0; /* Not supported yet. */ } static void __init calculate_hvm_def_policy(void) -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:31:09 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:31:09 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146643.269929 (Exim 4.92) (envelope-from ) id 1lwMdZ-0000hR-Cv; Thu, 24 Jun 2021 10:31:09 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146643.269929; Thu, 24 Jun 2021 10:31:09 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMdZ-0000hJ-A1; Thu, 24 Jun 2021 10:31:09 +0000 Received: by outflank-mailman (input) for mailman id 146643; Thu, 24 Jun 2021 10:31:07 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMdX-0000gy-Te for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:31:07 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMdX-0006Dj-Sp for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:31:07 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwMdX-0006yP-S5 for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:31:07 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=95nry1eB1v2ueP/Kbnf+kWnqYavfi0KPcoOgJP0w7Is=; b=G2VtKENIocFZX108RSXZWKHkZJ fEStjTvsS4YWiyGUKMHO93/bgb+lHoq8ccZ4JlU6L7z4JeaPdQHyuXW8VhvoQnTN9uManl9BeiAGv kbnx7580UorsrvR6JTaick4ty5T1oEa/1oBdDXshdBAfUzqA9cLxKdc6zEfwvv4L92DM=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] libs/guest: Move struct xc_cpu_policy into xg_private.h Message-Id: Date: Thu, 24 Jun 2021 10:31:07 +0000 commit f7ad9ee7cafa679946beaa40d5eb3eab500e86b5 Author: Andrew Cooper AuthorDate: Thu Jun 10 12:57:42 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 15 20:50:32 2021 +0100 libs/guest: Move struct xc_cpu_policy into xg_private.h ... so tests can peek at the internals, without the structure being generally available to users of the library. Signed-off-by: Andrew Cooper Acked-by: Jan Beulich --- tools/libs/guest/xg_cpuid_x86.c | 11 +---------- tools/libs/guest/xg_private.h | 11 +++++++++++ 2 files changed, 12 insertions(+), 10 deletions(-) diff --git a/tools/libs/guest/xg_cpuid_x86.c b/tools/libs/guest/xg_cpuid_x86.c index ec5a47fde4..e01d657e03 100644 --- a/tools/libs/guest/xg_cpuid_x86.c +++ b/tools/libs/guest/xg_cpuid_x86.c @@ -22,7 +22,7 @@ #include #include #include -#include "xc_private.h" +#include "xg_private.h" #include "xc_bitops.h" #include #include @@ -34,18 +34,9 @@ enum { #include -#include - #define bitmaskof(idx) (1u << ((idx) & 31)) #define featureword_of(idx) ((idx) >> 5) -struct xc_cpu_policy { - struct cpuid_policy cpuid; - struct msr_policy msr; - xen_cpuid_leaf_t leaves[CPUID_MAX_SERIALISED_LEAVES]; - xen_msr_entry_t entries[MSR_MAX_SERIALISED_ENTRIES]; -}; - int xc_get_cpu_levelling_caps(xc_interface *xch, uint32_t *caps) { DECLARE_SYSCTL; diff --git a/tools/libs/guest/xg_private.h b/tools/libs/guest/xg_private.h index 03d765da21..28441ee13f 100644 --- a/tools/libs/guest/xg_private.h +++ b/tools/libs/guest/xg_private.h @@ -168,4 +168,15 @@ int pin_table(xc_interface *xch, unsigned int type, unsigned long mfn, #define M2P_SIZE(_m) ROUNDUP(((_m) * sizeof(xen_pfn_t)), M2P_SHIFT) #define M2P_CHUNKS(_m) (M2P_SIZE((_m)) >> M2P_SHIFT) +#if defined(__x86_64__) || defined(__i386__) +#include + +struct xc_cpu_policy { + struct cpuid_policy cpuid; + struct msr_policy msr; + xen_cpuid_leaf_t leaves[CPUID_MAX_SERIALISED_LEAVES]; + xen_msr_entry_t entries[MSR_MAX_SERIALISED_ENTRIES]; +}; +#endif /* x86 */ + #endif /* XG_PRIVATE_H */ -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:31:19 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:31:19 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146644.269934 (Exim 4.92) (envelope-from ) id 1lwMdj-0000lG-FL; Thu, 24 Jun 2021 10:31:19 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146644.269934; Thu, 24 Jun 2021 10:31:19 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMdj-0000l6-BX; Thu, 24 Jun 2021 10:31:19 +0000 Received: by outflank-mailman (input) for mailman id 146644; Thu, 24 Jun 2021 10:31:18 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMdi-0000kj-0J for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:31:18 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMdh-0006Du-Vp for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:31:17 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwMdh-0006zP-VB for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:31:17 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=VGssTNeMve8VIQupJzc8MPbPGZwh6FxEBHjasJQQZic=; b=rrS/4h4eDh7UFgJzcMpfCtmxF1 rJeMKAdlsvfQafJxS290DC7NA5qjdhKv7ycUbWEjHGSZ9Hfocd84MvaqtPtn5sm+WNa46VUflslOj zZxgGLrpjwnsIwvJgHMtacCWZXFQ7C0oMeXUFpafdp40ydHr55GyMh5iPMWXbZ5brRTU=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] tests: Introduce a TSX test Message-Id: Date: Thu, 24 Jun 2021 10:31:17 +0000 commit 4bcf6433eed3d9cbc00865ec62380a33ca832dac Author: Andrew Cooper AuthorDate: Thu Jun 10 12:34:45 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 15 20:50:32 2021 +0100 tests: Introduce a TSX test See the comment at the top of test-tsx.c for details. This covers various complexities encountered while trying to address the recent TSX deprecation on client parts. A sample run on KabyLake with latest microcode and default tsx= looks like this: root@host# ./test-tsx TSX tests Got 8 CPUs Testing MSR_TSX_FORCE_ABORT consistency CPU0 val 0x3 Testing MSR_TSX_CTRL consistency Testing RTM behaviour Got Abort Testing PV default/max policies Max: RTM 1, HLE 1, TSX_FORCE_ABORT 0, RTM_ALWAYS_ABORT 0, TSX_CTRL 0 Def: RTM 0, HLE 0, TSX_FORCE_ABORT 0, RTM_ALWAYS_ABORT 0, TSX_CTRL 0 Testing HVM default/max policies Max: RTM 1, HLE 1, TSX_FORCE_ABORT 0, RTM_ALWAYS_ABORT 0, TSX_CTRL 0 Def: RTM 0, HLE 0, TSX_FORCE_ABORT 0, RTM_ALWAYS_ABORT 0, TSX_CTRL 0 Testing PV guest Created d7 Cur: RTM 0, HLE 0, TSX_FORCE_ABORT 0, RTM_ALWAYS_ABORT 0, TSX_CTRL 0 Cur: RTM 1, HLE 1, TSX_FORCE_ABORT 0, RTM_ALWAYS_ABORT 0, TSX_CTRL 0 Testing HVM guest Created d8 Cur: RTM 0, HLE 0, TSX_FORCE_ABORT 0, RTM_ALWAYS_ABORT 0, TSX_CTRL 0 Cur: RTM 1, HLE 1, TSX_FORCE_ABORT 0, RTM_ALWAYS_ABORT 0, TSX_CTRL 0 Signed-off-by: Andrew Cooper Acked-by: Jan Beulich --- tools/tests/Makefile | 1 + tools/tests/tsx/.gitignore | 1 + tools/tests/tsx/Makefile | 45 ++++ tools/tests/tsx/test-tsx.c | 538 +++++++++++++++++++++++++++++++++++++++++++++ 4 files changed, 585 insertions(+) diff --git a/tools/tests/Makefile b/tools/tests/Makefile index 8746aabe6b..25531a984a 100644 --- a/tools/tests/Makefile +++ b/tools/tests/Makefile @@ -5,6 +5,7 @@ SUBDIRS-y := SUBDIRS-y += resource SUBDIRS-$(CONFIG_X86) += cpu-policy SUBDIRS-$(CONFIG_X86) += mce-test +SUBDIRS-$(CONFIG_X86) += tsx ifneq ($(clang),y) SUBDIRS-$(CONFIG_X86) += x86_emulator endif diff --git a/tools/tests/tsx/.gitignore b/tools/tests/tsx/.gitignore new file mode 100644 index 0000000000..97ec4db7ff --- /dev/null +++ b/tools/tests/tsx/.gitignore @@ -0,0 +1 @@ +test-tsx diff --git a/tools/tests/tsx/Makefile b/tools/tests/tsx/Makefile new file mode 100644 index 0000000000..d7d2a5d95e --- /dev/null +++ b/tools/tests/tsx/Makefile @@ -0,0 +1,45 @@ +XEN_ROOT = $(CURDIR)/../../.. +include $(XEN_ROOT)/tools/Rules.mk + +TARGET := test-tsx + +.PHONY: all +all: $(TARGET) + +.PHONY: clean +clean: + $(RM) -- *.o $(TARGET) $(DEPS_RM) + +.PHONY: distclean +distclean: clean + $(RM) -- *~ + +.PHONY: install +install: all + $(INSTALL_DIR) $(DESTDIR)$(LIBEXEC_BIN) + $(INSTALL_PROG) $(TARGET) $(DESTDIR)$(LIBEXEC_BIN) + +.PHONY: uninstall +uninstall: + $(RM) -- $(DESTDIR)$(LIBEXEC_BIN)/$(TARGET) + +.PHONY: uninstall +uninstall: + +CFLAGS += -Werror +CFLAGS += -I$(XEN_ROOT)/tools/libs/ctrl -I$(XEN_ROOT)/tools/libs/guest +CFLAGS += $(CFLAGS_xeninclude) +CFLAGS += $(CFLAGS_libxenctrl) +CFLAGS += $(CFLAGS_libxenguest) +CFLAGS += $(APPEND_CFLAGS) + +LDFLAGS += $(LDLIBS_libxenctrl) +LDFLAGS += $(LDLIBS_libxenguest) +LDFLAGS += $(APPEND_LDFLAGS) + +%.o: Makefile + +$(TARGET): test-tsx.o + $(CC) -o $@ $< $(LDFLAGS) + +-include $(DEPS_INCLUDE) diff --git a/tools/tests/tsx/test-tsx.c b/tools/tests/tsx/test-tsx.c new file mode 100644 index 0000000000..fab99c135e --- /dev/null +++ b/tools/tests/tsx/test-tsx.c @@ -0,0 +1,538 @@ +/* + * TSX settings and consistency tests + * + * This tests various behaviours and invariants with regards to TSX. It + * ideally wants running for several microcode versions, and all applicable + * tsx= commandline settings, on a single CPU, including after an S3 + * suspend/resume event. + * + * It tests specifically: + * - The consistency of MSR_TSX_CTRL/MSR_TSX_FORCE_ABORT values across the + * system, and their accessibility WRT data in the host CPU policy. + * - The actual behaviour of RTM on the system. + * - Cross-check the default/max policies based on the actual RTM behaviour. + * - Create some guests, check their defaults, and check that the defaults + * can be changed. + */ + +#define _GNU_SOURCE + +#include +#include +#include +#include +#include +#include +#include +#include + +#include +#include +#include + +#include "xg_private.h" + +enum { +#define XEN_CPUFEATURE(name, value) X86_FEATURE_##name = value, +#include +}; +#define bitmaskof(idx) (1u << ((idx) & 31)) + +#define MSR_ARCH_CAPABILITIES 0x0000010a +#define ARCH_CAPS_TSX_CTRL (1 << 7) +#define MSR_TSX_FORCE_ABORT 0x0000010f +#define MSR_TSX_CTRL 0x00000122 + +static unsigned int nr_failures; +#define fail(fmt, ...) \ +({ \ + nr_failures++; \ + (void)printf(fmt, ##__VA_ARGS__); \ +}) + +static xc_interface *xch; + +/* + * Policies, arranged as an array for easy collection of all of them. We + * don't care about the raw policy (index 0) so reuse that for the guest + * policy. + */ +static struct xc_cpu_policy policies[6]; +#define guest_policy policies[0] +#define host policies[XEN_SYSCTL_cpu_policy_host] +#define pv_max policies[XEN_SYSCTL_cpu_policy_pv_max] +#define hvm_max policies[XEN_SYSCTL_cpu_policy_hvm_max] +#define pv_default policies[XEN_SYSCTL_cpu_policy_pv_default] +#define hvm_default policies[XEN_SYSCTL_cpu_policy_hvm_default] + +static bool xen_has_pv = true, xen_has_hvm = true; + +static xc_physinfo_t physinfo; + +static enum rtm_behaviour { + RTM_UD, + RTM_OK, + RTM_ABORT, +} rtm_behaviour; + +/* + * Test a specific TSX MSR for consistency across the system, taking into + * account whether it ought to be accessible or not. + * + * We can't query offline CPUs, so skip those if encountered. We don't care + * particularly for the exact MSR value, but we do care that it is the same + * everywhere. + */ +static void test_tsx_msr_consistency(unsigned int msr, bool accessible) +{ + uint64_t cpu0_val = ~0; + + for ( unsigned int cpu = 0; cpu <= physinfo.max_cpu_id; ++cpu ) + { + xc_resource_entry_t ent = { + .u.cmd = XEN_RESOURCE_OP_MSR_READ, + .idx = msr, + }; + xc_resource_op_t op = { + .cpu = cpu, + .entries = &ent, + .nr_entries = 1, + }; + int rc = xc_resource_op(xch, 1, &op); + + if ( rc < 0 ) + { + /* Don't emit a message for offline CPUs */ + if ( errno != ENODEV ) + fail(" xc_resource_op() for CPU%u failed: rc %d, errno %d - %s\n", + cpu, rc, errno, strerror(errno)); + continue; + } + + if ( accessible ) + { + if ( rc != 1 ) + { + fail(" Expected 1 result, got %d\n", rc); + continue; + } + if ( ent.u.ret != 0 ) + { + fail(" Expected ok, got %d\n", ent.u.ret); + continue; + } + } + else + { + if ( rc != 0 ) + fail(" Expected 0 results, got %u\n", rc); + else if ( ent.u.ret != -EPERM ) + fail(" Expected -EPERM, got %d\n", ent.u.ret); + continue; + } + + if ( cpu == 0 ) + { + cpu0_val = ent.val; + printf(" CPU0 val %#"PRIx64"\n", cpu0_val); + } + else if ( ent.val != cpu0_val ) + fail(" CPU%u val %#"PRIx64" differs from CPU0 %#"PRIx64"\n", + cpu, ent.val, cpu0_val); + } +} + +/* + * Check all TSX MSRs, and in particular that their accessibility matches what + * is expressed in the host CPU policy. + */ +static void test_tsx_msrs(void) +{ + printf("Testing MSR_TSX_FORCE_ABORT consistency\n"); + test_tsx_msr_consistency( + MSR_TSX_FORCE_ABORT, host.cpuid.feat.tsx_force_abort); + + printf("Testing MSR_TSX_CTRL consistency\n"); + test_tsx_msr_consistency( + MSR_TSX_CTRL, host.msr.arch_caps.tsx_ctrl); +} + +/* + * Probe for how RTM behaves, deliberately not inspecting CPUID. + * Distinguishes between "no support at all" (i.e. XBEGIN suffers #UD), + * working ok, and appearing to always abort. + */ +static enum rtm_behaviour probe_rtm_behaviour(void) +{ + for ( unsigned int i = 0; i < 1000; ++i ) + { + /* + * Opencoding the RTM infrastructure from immintrin.h, because we + * still support older versions of GCC. Also so we can include #UD + * detection logic. + */ +#define XBEGIN_STARTED -1 +#define XBEGIN_UD -2 + unsigned int status = XBEGIN_STARTED; + + asm volatile ( ".Lxbegin: .byte 0xc7,0xf8,0,0,0,0" /* XBEGIN 1f; 1: */ + : "+a" (status) :: "memory" ); + if ( status == XBEGIN_STARTED ) + { + asm volatile ( ".byte 0x0f,0x01,0xd5" ::: "memory" ); /* XEND */ + return RTM_OK; + } + else if ( status == XBEGIN_UD ) + return RTM_UD; + } + + return RTM_ABORT; +} + +static struct sigaction old_sigill; + +static void sigill_handler(int signo, siginfo_t *info, void *extra) +{ + extern const char xbegin_label[] asm(".Lxbegin"); + + if ( info->si_addr == xbegin_label && + memcmp(info->si_addr, "\xc7\xf8\x00\x00\x00\x00", 6) == 0 ) + { + ucontext_t *context = extra; + + /* + * Found the XBEGIN instruction. Step over it, and update `status` to + * signal #UD. + */ +#if defined(__linux__) +# ifdef __x86_64__ + context->uc_mcontext.gregs[REG_RIP] += 6; + context->uc_mcontext.gregs[REG_RAX] = XBEGIN_UD; +# else + context->uc_mcontext.gregs[REG_EIP] += 6; + context->uc_mcontext.gregs[REG_EAX] = XBEGIN_UD; +# endif + +#elif defined(__FreeBSD__) +# ifdef __x86_64__ + context->uc_mcontext.mc_rip += 6; + context->uc_mcontext.mc_rax = XBEGIN_UD; +# else + context->uc_mcontext.mc_eip += 6; + context->uc_mcontext.mc_eax = XBEGIN_UD; +# endif + +#elif defined(__NetBSD__) +# ifdef __x86_64__ + context->uc_mcontext.__gregs[_REG_RIP] += 6; + context->uc_mcontext.__gregs[_REG_RAX] = XBEGIN_UD; +# else + context->uc_mcontext.__gregs[_REG_EIP] += 6; + context->uc_mcontext.__gregs[_REG_EAX] = XBEGIN_UD; +# endif + +#else +# error Unknown environment - please adjust +#endif + } + else + { + /* + * Not the SIGILL we're looking for... Restore the old handler and + * try again. Will likely coredump as a result. + */ + sigaction(SIGILL, &old_sigill, NULL); + } +} + +static void test_rtm_behaviour(void) +{ + struct sigaction new_sigill = { + .sa_flags = SA_SIGINFO, + .sa_sigaction = sigill_handler, + }; + const char *str; + + printf("Testing RTM behaviour\n"); + + /* + * Install a custom SIGILL handler while probing for RTM behaviour, as the + * XBEGIN instruction might suffer #UD. + */ + sigaction(SIGILL, &new_sigill, &old_sigill); + rtm_behaviour = probe_rtm_behaviour(); + sigaction(SIGILL, &old_sigill, NULL); + + switch ( rtm_behaviour ) + { + case RTM_UD: str = "#UD"; break; + case RTM_OK: str = "OK"; break; + case RTM_ABORT: str = "Abort"; break; + default: str = NULL; break; + } + + if ( str ) + printf(" Got %s\n", str); + else + return fail(" Got unexpected behaviour %d\n", rtm_behaviour); + + if ( host.cpuid.feat.rtm ) + { + if ( rtm_behaviour == RTM_UD ) + fail(" Host reports RTM, but appears unavailable\n"); + } + else + { + if ( rtm_behaviour != RTM_UD ) + fail(" Host reports no RTM, but appears available\n"); + } +} + +static void dump_tsx_details(const struct xc_cpu_policy *p, const char *pref) +{ + printf(" %s RTM %u, HLE %u, TSX_FORCE_ABORT %u, RTM_ALWAYS_ABORT %u, TSX_CTRL %u\n", + pref, + p->cpuid.feat.rtm, + p->cpuid.feat.hle, + p->cpuid.feat.tsx_force_abort, + p->cpuid.feat.rtm_always_abort, + p->msr.arch_caps.tsx_ctrl); +} + +/* Sanity test various invariants we expect in the default/max policies. */ +static void test_guest_policies(const struct xc_cpu_policy *max, + const struct xc_cpu_policy *def) +{ + const struct cpuid_policy *cm = &max->cpuid; + const struct cpuid_policy *cd = &def->cpuid; + const struct msr_policy *mm = &max->msr; + const struct msr_policy *md = &def->msr; + + dump_tsx_details(max, "Max:"); + dump_tsx_details(def, "Def:"); + + if ( ((cm->feat.raw[0].d | cd->feat.raw[0].d) & + (bitmaskof(X86_FEATURE_TSX_FORCE_ABORT) | + bitmaskof(X86_FEATURE_RTM_ALWAYS_ABORT))) || + ((mm->arch_caps.raw | md->arch_caps.raw) & ARCH_CAPS_TSX_CTRL) ) + fail(" Xen-only TSX controls offered to guest\n"); + + switch ( rtm_behaviour ) + { + case RTM_UD: + if ( (cm->feat.raw[0].b | cd->feat.raw[0].b) & + (bitmaskof(X86_FEATURE_HLE) | bitmaskof(X86_FEATURE_RTM)) ) + fail(" HLE/RTM offered to guests despite not being available\n"); + break; + + case RTM_ABORT: + if ( cd->feat.raw[0].b & + (bitmaskof(X86_FEATURE_HLE) | bitmaskof(X86_FEATURE_RTM)) ) + fail(" HLE/RTM offered to guests by default despite not being usable\n"); + break; + + case RTM_OK: + if ( !cm->feat.rtm || !cd->feat.rtm ) + fail(" RTM not offered to guests despite being available\n"); + break; + } + + if ( cd->feat.hle ) + fail(" Fail: HLE offered in default policy\n"); +} + +static void test_def_max_policies(void) +{ + if ( xen_has_pv ) + { + printf("Testing PV default/max policies\n"); + test_guest_policies(&pv_max, &pv_default); + } + + if ( xen_has_hvm ) + { + printf("Testing HVM default/max policies\n"); + test_guest_policies(&hvm_max, &hvm_default); + } +} + +static void test_guest(struct xen_domctl_createdomain *c) +{ + uint32_t domid = 0; + int rc; + + rc = xc_domain_create(xch, &domid, c); + if ( rc ) + return fail(" Domain create failure: %d - %s\n", + errno, strerror(errno)); + + printf(" Created d%u\n", domid); + + rc = xc_cpu_policy_get_domain(xch, domid, &guest_policy); + if ( rc ) + { + fail(" Failed to obtain domain policy: %d - %s\n", + errno, strerror(errno)); + goto out; + } + + dump_tsx_details(&guest_policy, "Cur:"); + + /* + * Check defaults given to the guest. + */ + if ( guest_policy.cpuid.feat.rtm != (rtm_behaviour == RTM_OK) ) + fail(" RTM %u in guest, despite rtm behaviour\n", + guest_policy.cpuid.feat.rtm); + + if ( guest_policy.cpuid.feat.hle || + guest_policy.cpuid.feat.tsx_force_abort || + guest_policy.cpuid.feat.rtm_always_abort || + guest_policy.msr.arch_caps.tsx_ctrl ) + fail(" Unexpected features advertised\n"); + + if ( host.cpuid.feat.rtm ) + { + unsigned int _7b0; + + /* + * If host RTM is available, all combinations of guest flags should be + * possible. Flip both HLE/RTM to check non-default settings. + */ + _7b0 = (guest_policy.cpuid.feat.raw[0].b ^= + (bitmaskof(X86_FEATURE_HLE) | bitmaskof(X86_FEATURE_RTM))); + + /* Set the new policy. */ + rc = xc_cpu_policy_set_domain(xch, domid, &guest_policy); + if ( rc ) + { + fail(" Failed to set domain policy: %d - %s\n", + errno, strerror(errno)); + goto out; + } + + /* Re-get the new policy. */ + rc = xc_cpu_policy_get_domain(xch, domid, &guest_policy); + if ( rc ) + { + fail(" Failed to obtain domain policy: %d - %s\n", + errno, strerror(errno)); + goto out; + } + + dump_tsx_details(&guest_policy, "Cur:"); + + if ( guest_policy.cpuid.feat.raw[0].b != _7b0 ) + { + fail(" Expected CPUID.7[1].b 0x%08x differs from actual 0x%08x\n", + _7b0, guest_policy.cpuid.feat.raw[0].b); + goto out; + } + } + + out: + rc = xc_domain_destroy(xch, domid); + if ( rc ) + fail(" Failed to destroy domain: %d - %s\n", + errno, strerror(errno)); +} + +static void test_guests(void) +{ + if ( xen_has_pv ) + { + struct xen_domctl_createdomain c = { + .max_vcpus = 1, + .max_grant_frames = 1, + }; + + printf("Testing PV guest\n"); + test_guest(&c); + } + + if ( xen_has_hvm ) + { + struct xen_domctl_createdomain c = { + .flags = XEN_DOMCTL_CDF_hvm, + .max_vcpus = 1, + .max_grant_frames = 1, + .arch = { + .emulation_flags = XEN_X86_EMU_LAPIC, + }, + }; + + if ( physinfo.capabilities & XEN_SYSCTL_PHYSCAP_hap ) + c.flags |= XEN_DOMCTL_CDF_hap; + else if ( !(physinfo.capabilities & XEN_SYSCTL_PHYSCAP_shadow) ) + return fail(" HVM available, but neither HAP nor Shadow\n"); + + printf("Testing HVM guest\n"); + test_guest(&c); + } +} + +/* Obtain some general data, then run the tests. */ +static void test_tsx(void) +{ + int rc; + + /* Read all policies except raw. */ + for ( unsigned int i = XEN_SYSCTL_cpu_policy_host; + i <= XEN_SYSCTL_cpu_policy_hvm_default; ++i ) + { + rc = xc_cpu_policy_get_system(xch, i, &policies[i]); + + if ( rc == -1 && errno == EOPNOTSUPP ) + { + /* + * Use EOPNOTSUPP to spot Xen missing CONFIG_{PV,HVM}, and adjust + * later testing accordingly. + */ + switch ( i ) + { + case XEN_SYSCTL_cpu_policy_pv_max: + case XEN_SYSCTL_cpu_policy_pv_default: + if ( xen_has_pv ) + printf(" Xen doesn't support PV\n"); + xen_has_pv = false; + continue; + + case XEN_SYSCTL_cpu_policy_hvm_max: + case XEN_SYSCTL_cpu_policy_hvm_default: + if ( xen_has_hvm ) + printf(" Xen doesn't support HVM\n"); + xen_has_hvm = false; + continue; + } + } + if ( rc ) + return fail("Failed to obtain policy[%u]: %d - %s\n", + i, errno, strerror(errno)); + } + + rc = xc_physinfo(xch, &physinfo); + if ( rc ) + return fail("Failed to obtain physinfo: %d - %s\n", + errno, strerror(errno)); + + printf(" Got %u CPUs\n", physinfo.max_cpu_id + 1); + + test_tsx_msrs(); + test_rtm_behaviour(); + test_def_max_policies(); + test_guests(); +} + +int main(int argc, char **argv) +{ + printf("TSX tests\n"); + + xch = xc_interface_open(NULL, NULL, 0); + + if ( !xch ) + err(1, "xc_interface_open"); + + test_tsx(); + + return !!nr_failures; +} -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:31:29 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:31:29 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146646.269937 (Exim 4.92) (envelope-from ) id 1lwMdt-0000ov-GJ; Thu, 24 Jun 2021 10:31:29 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146646.269937; Thu, 24 Jun 2021 10:31:29 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMdt-0000of-DC; Thu, 24 Jun 2021 10:31:29 +0000 Received: by outflank-mailman (input) for mailman id 146646; Thu, 24 Jun 2021 10:31:28 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMds-0000oQ-3e for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:31:28 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMds-0006E5-2s for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:31:28 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwMds-00070I-2E for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:31:28 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=9emrz29oIS9X5Yq6xgogZoxlsjto4W6m2gGiHQi6438=; b=imIICX/PrOADlgBSLkPjZbXR7O T1T4huvL21AZ7CHR/OO8alXpDMrLqxl0Pqv4czA+u9JmBofeXADa+GW1pwkRuGgzbjWJT71GYHB9g eee7fEMfgG34Ni1dnyqZkPArjr2ylqlnujPh8Zqdzl+y8sJa8kv2B/MCIo4iX1R25+SY=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] x86/mtrr: move epte_get_entry_emt to p2m-ept.c Message-Id: Date: Thu, 24 Jun 2021 10:31:28 +0000 commit ed464d4e8a9a49559307c96ee6aa59e97820f692 Author: Roger Pau Monné AuthorDate: Thu Jun 17 17:58:11 2021 +0200 Commit: Jan Beulich CommitDate: Thu Jun 17 17:58:11 2021 +0200 x86/mtrr: move epte_get_entry_emt to p2m-ept.c This is an EPT specific function, so it shouldn't live in the generic mtrr file. Such movement is also needed for future work that will require passing a p2m_type_t parameter to epte_get_entry_emt, and making that type visible to the mtrr users is cumbersome and unneeded. Moving epte_get_entry_emt out of mtrr.c requires making the helper to get the MTRR type of an address from the mtrr state public. While there rename the function to start with the mtrr prefix, like other mtrr related functions. While there fix some of the types of the function parameters. No functional change intended. Signed-off-by: Roger Pau Monné Reviewed-by: Jan Beulich Reviewed-by: Kevin Tian --- xen/arch/x86/hvm/mtrr.c | 107 +---------------------------------- xen/arch/x86/hvm/vmx/vmx.c | 4 +- xen/arch/x86/mm/p2m-ept.c | 114 ++++++++++++++++++++++++++++++++++++-- xen/include/asm-x86/hvm/vmx/vmx.h | 2 + xen/include/asm-x86/mtrr.h | 5 +- 5 files changed, 118 insertions(+), 114 deletions(-) diff --git a/xen/arch/x86/hvm/mtrr.c b/xen/arch/x86/hvm/mtrr.c index 82ded1635c..4a9f3177ed 100644 --- a/xen/arch/x86/hvm/mtrr.c +++ b/xen/arch/x86/hvm/mtrr.c @@ -194,8 +194,7 @@ void hvm_vcpu_cacheattr_destroy(struct vcpu *v) * May return a negative value when order > 0, indicating to the caller * that the respective mapping needs splitting. */ -static int get_mtrr_type(const struct mtrr_state *m, - paddr_t pa, unsigned int order) +int mtrr_get_type(const struct mtrr_state *m, paddr_t pa, unsigned int order) { uint8_t overlap_mtrr = 0; uint8_t overlap_mtrr_pos = 0; @@ -323,7 +322,7 @@ static uint8_t effective_mm_type(struct mtrr_state *m, * just use it */ if ( gmtrr_mtype == NO_HARDCODE_MEM_TYPE ) - mtrr_mtype = get_mtrr_type(m, gpa, 0); + mtrr_mtype = mtrr_get_type(m, gpa, 0); else mtrr_mtype = gmtrr_mtype; @@ -350,7 +349,7 @@ uint32_t get_pat_flags(struct vcpu *v, guest_eff_mm_type = effective_mm_type(g, pat, gpaddr, gl1e_flags, gmtrr_mtype); /* 2. Get the memory type of host physical address, with MTRR */ - shadow_mtrr_type = get_mtrr_type(&mtrr_state, spaddr, 0); + shadow_mtrr_type = mtrr_get_type(&mtrr_state, spaddr, 0); /* 3. Find the memory type in PAT, with host MTRR memory type * and guest effective memory type. @@ -789,106 +788,6 @@ void memory_type_changed(struct domain *d) } } -int epte_get_entry_emt(struct domain *d, unsigned long gfn, mfn_t mfn, - unsigned int order, uint8_t *ipat, bool_t direct_mmio) -{ - int gmtrr_mtype, hmtrr_mtype; - struct vcpu *v = current; - unsigned long i; - - *ipat = 0; - - if ( v->domain != d ) - v = d->vcpu ? d->vcpu[0] : NULL; - - /* Mask, not add, for order so it works with INVALID_MFN on unmapping */ - if ( rangeset_overlaps_range(mmio_ro_ranges, mfn_x(mfn), - mfn_x(mfn) | ((1UL << order) - 1)) ) - { - if ( !order || rangeset_contains_range(mmio_ro_ranges, mfn_x(mfn), - mfn_x(mfn) | ((1UL << order) - 1)) ) - { - *ipat = 1; - return MTRR_TYPE_UNCACHABLE; - } - /* Force invalid memory type so resolve_misconfig() will split it */ - return -1; - } - - if ( !mfn_valid(mfn) ) - { - *ipat = 1; - return MTRR_TYPE_UNCACHABLE; - } - - if ( !direct_mmio && !is_iommu_enabled(d) && !cache_flush_permitted(d) ) - { - *ipat = 1; - return MTRR_TYPE_WRBACK; - } - - for ( i = 0; i < (1ul << order); i++ ) - { - if ( is_special_page(mfn_to_page(mfn_add(mfn, i))) ) - { - if ( order ) - return -1; - *ipat = 1; - return MTRR_TYPE_WRBACK; - } - } - - if ( direct_mmio ) - return MTRR_TYPE_UNCACHABLE; - - gmtrr_mtype = hvm_get_mem_pinned_cacheattr(d, _gfn(gfn), order); - if ( gmtrr_mtype >= 0 ) - { - *ipat = 1; - return gmtrr_mtype != PAT_TYPE_UC_MINUS ? gmtrr_mtype - : MTRR_TYPE_UNCACHABLE; - } - if ( gmtrr_mtype == -EADDRNOTAVAIL ) - return -1; - - gmtrr_mtype = v ? get_mtrr_type(&v->arch.hvm.mtrr, gfn << PAGE_SHIFT, order) - : MTRR_TYPE_WRBACK; - hmtrr_mtype = get_mtrr_type(&mtrr_state, mfn_x(mfn) << PAGE_SHIFT, order); - if ( gmtrr_mtype < 0 || hmtrr_mtype < 0 ) - return -1; - - /* If both types match we're fine. */ - if ( likely(gmtrr_mtype == hmtrr_mtype) ) - return hmtrr_mtype; - - /* If either type is UC, we have to go with that one. */ - if ( gmtrr_mtype == MTRR_TYPE_UNCACHABLE || - hmtrr_mtype == MTRR_TYPE_UNCACHABLE ) - return MTRR_TYPE_UNCACHABLE; - - /* If either type is WB, we have to go with the other one. */ - if ( gmtrr_mtype == MTRR_TYPE_WRBACK ) - return hmtrr_mtype; - if ( hmtrr_mtype == MTRR_TYPE_WRBACK ) - return gmtrr_mtype; - - /* - * At this point we have disagreeing WC, WT, or WP types. The only - * combination that can be cleanly resolved is WT:WP. The ones involving - * WC need to be converted to UC, both due to the memory ordering - * differences and because WC disallows reads to be cached (WT and WP - * permit this), while WT and WP require writes to go straight to memory - * (WC can buffer them). - */ - if ( (gmtrr_mtype == MTRR_TYPE_WRTHROUGH && - hmtrr_mtype == MTRR_TYPE_WRPROT) || - (gmtrr_mtype == MTRR_TYPE_WRPROT && - hmtrr_mtype == MTRR_TYPE_WRTHROUGH) ) - return MTRR_TYPE_WRPROT; - - return MTRR_TYPE_UNCACHABLE; -} - /* * Local variables: * mode: C diff --git a/xen/arch/x86/hvm/vmx/vmx.c b/xen/arch/x86/hvm/vmx/vmx.c index 7e3e67fdc3..0d4b47681b 100644 --- a/xen/arch/x86/hvm/vmx/vmx.c +++ b/xen/arch/x86/hvm/vmx/vmx.c @@ -417,12 +417,12 @@ static int vmx_domain_initialise(struct domain *d) static void domain_creation_finished(struct domain *d) { gfn_t gfn = gaddr_to_gfn(APIC_DEFAULT_PHYS_BASE); - uint8_t ipat; + bool ipat; if ( !has_vlapic(d) || mfn_eq(apic_access_mfn, INVALID_MFN) ) return; - ASSERT(epte_get_entry_emt(d, gfn_x(gfn), apic_access_mfn, 0, &ipat, + ASSERT(epte_get_entry_emt(d, gfn, apic_access_mfn, 0, &ipat, true) == MTRR_TYPE_WRBACK); ASSERT(ipat); diff --git a/xen/arch/x86/mm/p2m-ept.c b/xen/arch/x86/mm/p2m-ept.c index a3beaf91e2..f1d1d07e92 100644 --- a/xen/arch/x86/mm/p2m-ept.c +++ b/xen/arch/x86/mm/p2m-ept.c @@ -20,6 +20,7 @@ #include #include #include +#include #include #include #include @@ -485,6 +486,108 @@ static int ept_invalidate_emt_range(struct p2m_domain *p2m, return rc; } +int epte_get_entry_emt(struct domain *d, gfn_t gfn, mfn_t mfn, + unsigned int order, bool *ipat, bool direct_mmio) +{ + int gmtrr_mtype, hmtrr_mtype; + struct vcpu *v = current; + unsigned long i; + + *ipat = false; + + if ( v->domain != d ) + v = d->vcpu ? d->vcpu[0] : NULL; + + /* Mask, not add, for order so it works with INVALID_MFN on unmapping */ + if ( rangeset_overlaps_range(mmio_ro_ranges, mfn_x(mfn), + mfn_x(mfn) | ((1UL << order) - 1)) ) + { + if ( !order || rangeset_contains_range(mmio_ro_ranges, mfn_x(mfn), + mfn_x(mfn) | ((1UL << order) - 1)) ) + { + *ipat = true; + return MTRR_TYPE_UNCACHABLE; + } + /* Force invalid memory type so resolve_misconfig() will split it */ + return -1; + } + + if ( !mfn_valid(mfn) ) + { + *ipat = true; + return MTRR_TYPE_UNCACHABLE; + } + + if ( !direct_mmio && !is_iommu_enabled(d) && !cache_flush_permitted(d) ) + { + *ipat = true; + return MTRR_TYPE_WRBACK; + } + + for ( i = 0; i < (1ul << order); i++ ) + { + if ( is_special_page(mfn_to_page(mfn_add(mfn, i))) ) + { + if ( order ) + return -1; + *ipat = true; + return MTRR_TYPE_WRBACK; + } + } + + if ( direct_mmio ) + return MTRR_TYPE_UNCACHABLE; + + gmtrr_mtype = hvm_get_mem_pinned_cacheattr(d, gfn, order); + if ( gmtrr_mtype >= 0 ) + { + *ipat = true; + return gmtrr_mtype != PAT_TYPE_UC_MINUS ? gmtrr_mtype + : MTRR_TYPE_UNCACHABLE; + } + if ( gmtrr_mtype == -EADDRNOTAVAIL ) + return -1; + + gmtrr_mtype = v ? mtrr_get_type(&v->arch.hvm.mtrr, + gfn_x(gfn) << PAGE_SHIFT, order) + : MTRR_TYPE_WRBACK; + hmtrr_mtype = mtrr_get_type(&mtrr_state, mfn_x(mfn) << PAGE_SHIFT, + order); + if ( gmtrr_mtype < 0 || hmtrr_mtype < 0 ) + return -1; + + /* If both types match we're fine. */ + if ( likely(gmtrr_mtype == hmtrr_mtype) ) + return hmtrr_mtype; + + /* If either type is UC, we have to go with that one. */ + if ( gmtrr_mtype == MTRR_TYPE_UNCACHABLE || + hmtrr_mtype == MTRR_TYPE_UNCACHABLE ) + return MTRR_TYPE_UNCACHABLE; + + /* If either type is WB, we have to go with the other one. */ + if ( gmtrr_mtype == MTRR_TYPE_WRBACK ) + return hmtrr_mtype; + if ( hmtrr_mtype == MTRR_TYPE_WRBACK ) + return gmtrr_mtype; + + /* + * At this point we have disagreeing WC, WT, or WP types. The only + * combination that can be cleanly resolved is WT:WP. The ones involving + * WC need to be converted to UC, both due to the memory ordering + * differences and because WC disallows reads to be cached (WT and WP + * permit this), while WT and WP require writes to go straight to memory + * (WC can buffer them). + */ + if ( (gmtrr_mtype == MTRR_TYPE_WRTHROUGH && + hmtrr_mtype == MTRR_TYPE_WRPROT) || + (gmtrr_mtype == MTRR_TYPE_WRPROT && + hmtrr_mtype == MTRR_TYPE_WRTHROUGH) ) + return MTRR_TYPE_WRPROT; + + return MTRR_TYPE_UNCACHABLE; +} + /* * Resolve deliberately mis-configured (EMT field set to an invalid value) * entries in the page table hierarchy for the given GFN: @@ -519,7 +622,7 @@ static int resolve_misconfig(struct p2m_domain *p2m, unsigned long gfn) if ( level == 0 || is_epte_superpage(&e) ) { - uint8_t ipat = 0; + bool ipat; if ( e.emt != MTRR_NUM_TYPES ) break; @@ -535,7 +638,7 @@ static int resolve_misconfig(struct p2m_domain *p2m, unsigned long gfn) e.emt = 0; if ( !is_epte_valid(&e) || !is_epte_present(&e) ) continue; - e.emt = epte_get_entry_emt(p2m->domain, gfn + i, + e.emt = epte_get_entry_emt(p2m->domain, _gfn(gfn + i), _mfn(e.mfn), 0, &ipat, e.sa_p2mt == p2m_mmio_direct); e.ipat = ipat; @@ -553,7 +656,8 @@ static int resolve_misconfig(struct p2m_domain *p2m, unsigned long gfn) } else { - int emt = epte_get_entry_emt(p2m->domain, gfn, _mfn(e.mfn), + int emt = epte_get_entry_emt(p2m->domain, _gfn(gfn), + _mfn(e.mfn), level * EPT_TABLE_ORDER, &ipat, e.sa_p2mt == p2m_mmio_direct); bool_t recalc = e.recalc; @@ -788,8 +892,8 @@ ept_set_entry(struct p2m_domain *p2m, gfn_t gfn_, mfn_t mfn, if ( mfn_valid(mfn) || p2m_allows_invalid_mfn(p2mt) ) { - uint8_t ipat = 0; - int emt = epte_get_entry_emt(p2m->domain, gfn, mfn, + bool ipat; + int emt = epte_get_entry_emt(p2m->domain, _gfn(gfn), mfn, i * EPT_TABLE_ORDER, &ipat, p2mt == p2m_mmio_direct); diff --git a/xen/include/asm-x86/hvm/vmx/vmx.h b/xen/include/asm-x86/hvm/vmx/vmx.h index 534e9fc221..f668ee1f09 100644 --- a/xen/include/asm-x86/hvm/vmx/vmx.h +++ b/xen/include/asm-x86/hvm/vmx/vmx.h @@ -599,6 +599,8 @@ void ept_p2m_uninit(struct p2m_domain *p2m); void ept_walk_table(struct domain *d, unsigned long gfn); bool_t ept_handle_misconfig(uint64_t gpa); +int epte_get_entry_emt(struct domain *d, gfn_t gfn, mfn_t mfn, + unsigned int order, bool *ipat, bool direct_mmio); void setup_ept_dump(void); void p2m_init_altp2m_ept(struct domain *d, unsigned int i); /* Locate an alternate p2m by its EPTP */ diff --git a/xen/include/asm-x86/mtrr.h b/xen/include/asm-x86/mtrr.h index 24e5de5c22..e0fd1005ce 100644 --- a/xen/include/asm-x86/mtrr.h +++ b/xen/include/asm-x86/mtrr.h @@ -72,12 +72,11 @@ extern int mtrr_add_page(unsigned long base, unsigned long size, unsigned int type, char increment); extern int mtrr_del(int reg, unsigned long base, unsigned long size); extern int mtrr_del_page(int reg, unsigned long base, unsigned long size); +extern int mtrr_get_type(const struct mtrr_state *m, paddr_t pa, + unsigned int order); extern void mtrr_centaur_report_mcr(int mcr, u32 lo, u32 hi); extern u32 get_pat_flags(struct vcpu *v, u32 gl1e_flags, paddr_t gpaddr, paddr_t spaddr, uint8_t gmtrr_mtype); -extern int epte_get_entry_emt(struct domain *, unsigned long gfn, mfn_t mfn, - unsigned int order, uint8_t *ipat, - bool_t direct_mmio); extern unsigned char pat_type_2_pte_flags(unsigned char pat_type); extern int hold_mtrr_updates_on_aps; extern void mtrr_aps_sync_begin(void); -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:31:39 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:31:39 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146647.269941 (Exim 4.92) (envelope-from ) id 1lwMe3-0000tM-JP; Thu, 24 Jun 2021 10:31:39 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146647.269941; Thu, 24 Jun 2021 10:31:39 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMe3-0000t9-GQ; Thu, 24 Jun 2021 10:31:39 +0000 Received: by outflank-mailman (input) for mailman id 146647; Thu, 24 Jun 2021 10:31:38 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMe2-0000sp-6t for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:31:38 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMe2-0006EI-6A for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:31:38 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwMe2-00071L-5T for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:31:38 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=5c4U288m2LO7ynxW5Vq5qt1C4rNv6NKfuADGFCsdde8=; b=kI5RKAmFEG6PbLQ0VKnCeJ4ZRk 3Xdj5CE0mTbJqAa+J8tp00R9X1rQcYX/JRah9N4puFo7r4QOLRktvBH1Jd3Fz0hxAcyIrp+oM5Yot nbr8QoTvgn2IjaVs+9eR8+8vprlOPT4LpUnAeDB5bb4qH2MGGAWL3GfA8hphrU6udNGI=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] x86/ept: force WB cache attributes for grant and foreign maps Message-Id: Date: Thu, 24 Jun 2021 10:31:38 +0000 commit 8af4b47f061edf6450f2b0a0a8134fdf1c13b3e5 Author: Roger Pau Monné AuthorDate: Thu Jun 17 18:00:57 2021 +0200 Commit: Jan Beulich CommitDate: Thu Jun 17 18:00:57 2021 +0200 x86/ept: force WB cache attributes for grant and foreign maps Force WB type for grants and foreign pages. Those are usually mapped over unpopulated physical ranges in the p2m, and those ranges would usually be UC in the MTRR state, which is unlikely to be the correct cache attribute. It's also cumbersome (or even impossible) for the guest to be setting the MTRR type for all those mappings as WB, as MTRR ranges are finite. Note that this is not an issue on AMD because WB cache attribute is already set on grants and foreign mappings in the p2m and MTRR types are ignored. Also on AMD Xen cannot force a cache attribute because of the lack of ignore PAT equivalent, so the behavior here slightly diverges between AMD and Intel (or EPT vs NPT/shadow). Signed-off-by: Roger Pau Monné Reviewed-by: Jan Beulich Reviewed-by: Kevin Tian --- xen/arch/x86/hvm/vmx/vmx.c | 2 +- xen/arch/x86/mm/p2m-ept.c | 37 +++++++++++++++++++++++++++++++------ xen/include/asm-x86/hvm/vmx/vmx.h | 2 +- 3 files changed, 33 insertions(+), 8 deletions(-) diff --git a/xen/arch/x86/hvm/vmx/vmx.c b/xen/arch/x86/hvm/vmx/vmx.c index 0d4b47681b..e09b7e3af9 100644 --- a/xen/arch/x86/hvm/vmx/vmx.c +++ b/xen/arch/x86/hvm/vmx/vmx.c @@ -423,7 +423,7 @@ static void domain_creation_finished(struct domain *d) return; ASSERT(epte_get_entry_emt(d, gfn, apic_access_mfn, 0, &ipat, - true) == MTRR_TYPE_WRBACK); + p2m_mmio_direct) == MTRR_TYPE_WRBACK); ASSERT(ipat); if ( set_mmio_p2m_entry(d, gfn, apic_access_mfn, PAGE_ORDER_4K) ) diff --git a/xen/arch/x86/mm/p2m-ept.c b/xen/arch/x86/mm/p2m-ept.c index f1d1d07e92..289001b66b 100644 --- a/xen/arch/x86/mm/p2m-ept.c +++ b/xen/arch/x86/mm/p2m-ept.c @@ -487,7 +487,7 @@ static int ept_invalidate_emt_range(struct p2m_domain *p2m, } int epte_get_entry_emt(struct domain *d, gfn_t gfn, mfn_t mfn, - unsigned int order, bool *ipat, bool direct_mmio) + unsigned int order, bool *ipat, p2m_type_t type) { int gmtrr_mtype, hmtrr_mtype; struct vcpu *v = current; @@ -518,7 +518,8 @@ int epte_get_entry_emt(struct domain *d, gfn_t gfn, mfn_t mfn, return MTRR_TYPE_UNCACHABLE; } - if ( !direct_mmio && !is_iommu_enabled(d) && !cache_flush_permitted(d) ) + if ( type != p2m_mmio_direct && !is_iommu_enabled(d) && + !cache_flush_permitted(d) ) { *ipat = true; return MTRR_TYPE_WRBACK; @@ -535,9 +536,33 @@ int epte_get_entry_emt(struct domain *d, gfn_t gfn, mfn_t mfn, } } - if ( direct_mmio ) + switch ( type ) + { + case p2m_mmio_direct: return MTRR_TYPE_UNCACHABLE; + case p2m_grant_map_ro: + case p2m_grant_map_rw: + case p2m_map_foreign: + /* + * Force WB type for grants and foreign pages. Those are usually mapped + * over unpopulated physical ranges in the p2m, and those would usually + * be UC in the MTRR state, which is unlikely to be the correct cache + * attribute. It's also cumbersome (or even impossible) for the guest + * to be setting the MTRR type for all those mappings as WB, as MTRR + * ranges are finite. + * + * Note that on AMD we cannot force a cache attribute because of the + * lack of ignore PAT equivalent, so the behavior here slightly + * diverges. See p2m_type_to_flags for the AMD attributes. + */ + *ipat = true; + return MTRR_TYPE_WRBACK; + + default: + break; + } + gmtrr_mtype = hvm_get_mem_pinned_cacheattr(d, gfn, order); if ( gmtrr_mtype >= 0 ) { @@ -640,7 +665,7 @@ static int resolve_misconfig(struct p2m_domain *p2m, unsigned long gfn) continue; e.emt = epte_get_entry_emt(p2m->domain, _gfn(gfn + i), _mfn(e.mfn), 0, &ipat, - e.sa_p2mt == p2m_mmio_direct); + e.sa_p2mt); e.ipat = ipat; nt = p2m_recalc_type(e.recalc, e.sa_p2mt, p2m, gfn + i); @@ -659,7 +684,7 @@ static int resolve_misconfig(struct p2m_domain *p2m, unsigned long gfn) int emt = epte_get_entry_emt(p2m->domain, _gfn(gfn), _mfn(e.mfn), level * EPT_TABLE_ORDER, &ipat, - e.sa_p2mt == p2m_mmio_direct); + e.sa_p2mt); bool_t recalc = e.recalc; if ( recalc && p2m_is_changeable(e.sa_p2mt) ) @@ -895,7 +920,7 @@ ept_set_entry(struct p2m_domain *p2m, gfn_t gfn_, mfn_t mfn, bool ipat; int emt = epte_get_entry_emt(p2m->domain, _gfn(gfn), mfn, i * EPT_TABLE_ORDER, &ipat, - p2mt == p2m_mmio_direct); + p2mt); if ( emt >= 0 ) new_entry.emt = emt; diff --git a/xen/include/asm-x86/hvm/vmx/vmx.h b/xen/include/asm-x86/hvm/vmx/vmx.h index f668ee1f09..0deb507490 100644 --- a/xen/include/asm-x86/hvm/vmx/vmx.h +++ b/xen/include/asm-x86/hvm/vmx/vmx.h @@ -600,7 +600,7 @@ void ept_p2m_uninit(struct p2m_domain *p2m); void ept_walk_table(struct domain *d, unsigned long gfn); bool_t ept_handle_misconfig(uint64_t gpa); int epte_get_entry_emt(struct domain *d, gfn_t gfn, mfn_t mfn, - unsigned int order, bool *ipat, bool direct_mmio); + unsigned int order, bool *ipat, p2m_type_t type); void setup_ept_dump(void); void p2m_init_altp2m_ept(struct domain *d, unsigned int i); /* Locate an alternate p2m by its EPTP */ -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:31:49 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:31:49 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146648.269945 (Exim 4.92) (envelope-from ) id 1lwMeD-0000x6-Ku; Thu, 24 Jun 2021 10:31:49 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146648.269945; Thu, 24 Jun 2021 10:31:49 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMeD-0000wy-I2; Thu, 24 Jun 2021 10:31:49 +0000 Received: by outflank-mailman (input) for mailman id 146648; Thu, 24 Jun 2021 10:31:48 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMeC-0000wk-9x for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:31:48 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMeC-0006ES-9G for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:31:48 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwMeC-00072D-8c for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:31:48 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=TEcup6W6DEsBXqrsIi/G17t8u6iF7XQDOJ32AV6zk7E=; b=LYO0GOpwXKspdmB9kQ4a7GTFBD H98d9muFt1yuewzsRYPLZJMB5J4YhhcqBddZZMaUsn4FAPEW+ESHVkpsRxX1w8Nbp+hxdn4GfONBw NiRlolXv9BpjGbQUVszvuTyglOWmpBnAmU/XY4YhMId9GZAVsMLQR9yDDuPevjj6YG2M=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] golang/xenlight: update generated code Message-Id: Date: Thu, 24 Jun 2021 10:31:48 +0000 commit b291ce703b9cebef0800267446334e867588354a Author: Nick Rosbrook AuthorDate: Mon May 24 16:36:42 2021 -0400 Commit: George Dunlap CommitDate: Mon Jun 21 16:49:08 2021 +0100 golang/xenlight: update generated code Re-generate code to reflect changes to libxl_types.idl from the following commits: 0570d7f276 x86/msr: introduce an option for compatible MSR behavior selection 7e5cffcd1e viridian: allow vCPU hotplug for Windows VMs 9835246710 viridian: remove implicit limit of 64 VPs per partition Signed-off-by: Nick Rosbrook Reviewed-by: George Dunlap --- tools/golang/xenlight/helpers.gen.go | 6 ++++++ tools/golang/xenlight/types.gen.go | 5 +++++ 2 files changed, 11 insertions(+) diff --git a/tools/golang/xenlight/helpers.gen.go b/tools/golang/xenlight/helpers.gen.go index 4c60d27a9c..b454b12d52 100644 --- a/tools/golang/xenlight/helpers.gen.go +++ b/tools/golang/xenlight/helpers.gen.go @@ -1113,6 +1113,9 @@ default: return fmt.Errorf("invalid union key '%v'", x.Type)} x.ArchArm.GicVersion = GicVersion(xc.arch_arm.gic_version) x.ArchArm.Vuart = VuartType(xc.arch_arm.vuart) +if err := x.ArchX86.MsrRelaxed.fromC(&xc.arch_x86.msr_relaxed);err != nil { +return fmt.Errorf("converting field ArchX86.MsrRelaxed: %v", err) +} x.Altp2M = Altp2MMode(xc.altp2m) x.VmtraceBufKb = int(xc.vmtrace_buf_kb) @@ -1589,6 +1592,9 @@ default: return fmt.Errorf("invalid union key '%v'", x.Type)} xc.arch_arm.gic_version = C.libxl_gic_version(x.ArchArm.GicVersion) xc.arch_arm.vuart = C.libxl_vuart_type(x.ArchArm.Vuart) +if err := x.ArchX86.MsrRelaxed.toC(&xc.arch_x86.msr_relaxed); err != nil { +return fmt.Errorf("converting field ArchX86.MsrRelaxed: %v", err) +} xc.altp2m = C.libxl_altp2m_mode(x.Altp2M) xc.vmtrace_buf_kb = C.int(x.VmtraceBufKb) diff --git a/tools/golang/xenlight/types.gen.go b/tools/golang/xenlight/types.gen.go index cb13002fdb..f2ceceb61c 100644 --- a/tools/golang/xenlight/types.gen.go +++ b/tools/golang/xenlight/types.gen.go @@ -211,6 +211,8 @@ ViridianEnlightenmentSynic ViridianEnlightenment = 7 ViridianEnlightenmentStimer ViridianEnlightenment = 8 ViridianEnlightenmentHcallIpi ViridianEnlightenment = 9 ViridianEnlightenmentExProcessorMasks ViridianEnlightenment = 10 +ViridianEnlightenmentNoVpLimit ViridianEnlightenment = 11 +ViridianEnlightenmentCpuHotplug ViridianEnlightenment = 12 ) type Hdtype int @@ -513,6 +515,9 @@ ArchArm struct { GicVersion GicVersion Vuart VuartType } +ArchX86 struct { +MsrRelaxed Defbool +} Altp2M Altp2MMode VmtraceBufKb int } -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:31:59 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:31:59 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146649.269949 (Exim 4.92) (envelope-from ) id 1lwMeN-0000zt-MR; Thu, 24 Jun 2021 10:31:59 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146649.269949; Thu, 24 Jun 2021 10:31:59 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMeN-0000zl-JS; Thu, 24 Jun 2021 10:31:59 +0000 Received: by outflank-mailman (input) for mailman id 146649; Thu, 24 Jun 2021 10:31:58 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMeM-0000za-Cs for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:31:58 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMeM-0006Eg-C9 for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:31:58 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwMeM-000731-BO for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:31:58 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=MAQOC67yqk0KZoMiAJqDE8EwLZJPLKtNHU5SnElMvfw=; b=5Wa4iquVs5XXCTg4pDcnMUT9yb VmnRVmuHnFq7B6KpPbkvxUhXGq8J0ocH8Irj9TrueNuUP2mdg8DPm6EWzKJPxNz9X9YBEef5TKx53 lR9ecQeavwPA5FGH6mnZwTdPrGq9crctnfCt0Vewo+o9/Kl6nW7KgOile1eVSFyOF1x0=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] golang/xenlight: fix StringList toC conversion Message-Id: Date: Thu, 24 Jun 2021 10:31:58 +0000 commit 1422d8db1b3dfdf7d9179944e594876e5e356a4b Author: Nick Rosbrook AuthorDate: Mon May 24 16:36:43 2021 -0400 Commit: George Dunlap CommitDate: Mon Jun 21 16:49:08 2021 +0100 golang/xenlight: fix StringList toC conversion The current implementation of StringList.toC does not correctly account for how libxl_string_list is expected to be laid out in C, which is clear when one looks at libxl_string_list_length in libxl.c. In particular, StringList.toC does not account for the extra memory that should be allocated for the "sentinel" entry. And, when using the "slice trick" to create a slice that can address C memory, the unsafe.Pointer conversion should be on a C.libxl_string_list, not *C.libxl_string_list. Fix these problems by (1) allocating an extra slot in the slice used to address the C memory, and explicity set the last entry to nil so the C memory will be zeroed out, and (2) dereferencing csl in the unsafe.Pointer conversion. Signed-off-by: Nick Rosbrook Reviewed-by: George Dunlap --- tools/golang/xenlight/xenlight.go | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/tools/golang/xenlight/xenlight.go b/tools/golang/xenlight/xenlight.go index b9189dec5c..13171d0ad1 100644 --- a/tools/golang/xenlight/xenlight.go +++ b/tools/golang/xenlight/xenlight.go @@ -491,13 +491,14 @@ func (sl *StringList) fromC(csl *C.libxl_string_list) error { func (sl StringList) toC(csl *C.libxl_string_list) error { var char *C.char - size := len(sl) + size := len(sl) + 1 *csl = (C.libxl_string_list)(C.malloc(C.ulong(size) * C.ulong(unsafe.Sizeof(char)))) - clist := (*[1 << 30]*C.char)(unsafe.Pointer(csl))[:size:size] + clist := (*[1 << 30]*C.char)(unsafe.Pointer(*csl))[:size:size] for i, v := range sl { clist[i] = C.CString(v) } + clist[len(clist)-1] = nil return nil } -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:32:12 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:32:12 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146650.269953 (Exim 4.92) (envelope-from ) id 1lwMeX-00013F-Nr; Thu, 24 Jun 2021 10:32:09 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146650.269953; Thu, 24 Jun 2021 10:32:09 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMeX-000137-L0; Thu, 24 Jun 2021 10:32:09 +0000 Received: by outflank-mailman (input) for mailman id 146650; Thu, 24 Jun 2021 10:32:08 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMeW-00012s-Fn for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:32:08 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMeW-0006FQ-F6 for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:32:08 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwMeW-0007Ek-EN for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:32:08 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=BG3gYNwLyAmwmNz/H9lCH5+ERcm3nTivxTPhFHigzV4=; b=AkI5KQWWeM34RWfODWdU6eZ/Gq 607jlds/V1TbRlUC9N9q+LKxd1OpGxTioTXQKiqAmoALuf3iqsBwt+GZqkcFWiTD2EYyK5VLeu8Fs g8ApLDvlE2UhCv2Sr+qmgBPv2IBxNTzwr9P8ixQS1aJjl30LqZy1qRWGhklU2OamGPn4=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] golang/xenlight: export keyed union interface types Message-Id: Date: Thu, 24 Jun 2021 10:32:08 +0000 commit bc9f632e31ee66be3f1860fc7303fe91a42e56a6 Author: Nick Rosbrook AuthorDate: Mon May 24 16:36:45 2021 -0400 Commit: George Dunlap CommitDate: Mon Jun 21 16:49:08 2021 +0100 golang/xenlight: export keyed union interface types For structs that have a keyed union, e.g. DomainBuildInfo, the TypeUnion field must be exported so that package users can get/set the fields within. This means that users are aware of the existence of the interface type used in those fields (see [1]), so it is awkward that the interface itself is not exported. However, the single method within the interface must remain unexported so that users cannot mistakenly "implement" those interfaces. Since there seems to be no reason to do otherwise, export the keyed union interface types. [1] https://pkg.go.dev/xenbits.xenproject.org/git-http/xen.git/tools/golang/xenlight?tab=doc#DeviceUsbdev Signed-off-by: Nick Rosbrook Reviewed-by: George Dunlap --- tools/golang/xenlight/gengotypes.py | 6 ++-- tools/golang/xenlight/types.gen.go | 58 ++++++++++++++++++------------------- 2 files changed, 32 insertions(+), 32 deletions(-) diff --git a/tools/golang/xenlight/gengotypes.py b/tools/golang/xenlight/gengotypes.py index 3e40c3d5dc..ef3279a910 100644 --- a/tools/golang/xenlight/gengotypes.py +++ b/tools/golang/xenlight/gengotypes.py @@ -159,7 +159,7 @@ def xenlight_golang_define_union(ty = None, struct_name = '', union_name = ''): extras = [] interface_name = '{0}_{1}_union'.format(struct_name, ty.keyvar.name) - interface_name = xenlight_golang_fmt_name(interface_name, exported=False) + interface_name = xenlight_golang_fmt_name(interface_name) s += 'type {0} interface {{\n'.format(interface_name) s += 'is{0}()\n'.format(interface_name) @@ -341,7 +341,7 @@ def xenlight_golang_union_from_C(ty = None, union_name = '', struct_name = ''): field_name = xenlight_golang_fmt_name('{0}_union'.format(keyname)) interface_name = '{0}_{1}_union'.format(struct_name, keyname) - interface_name = xenlight_golang_fmt_name(interface_name, exported=False) + interface_name = xenlight_golang_fmt_name(interface_name) cgo_keyname = keyname if cgo_keyname in go_keywords: @@ -545,7 +545,7 @@ def xenlight_golang_union_to_C(ty = None, union_name = '', gokeytype = xenlight_golang_fmt_name(keytype) interface_name = '{0}_{1}_union'.format(struct_name, keyname) - interface_name = xenlight_golang_fmt_name(interface_name, exported=False) + interface_name = xenlight_golang_fmt_name(interface_name) cgo_keyname = keyname if cgo_keyname in go_keywords: diff --git a/tools/golang/xenlight/types.gen.go b/tools/golang/xenlight/types.gen.go index f2ceceb61c..a214dd9df6 100644 --- a/tools/golang/xenlight/types.gen.go +++ b/tools/golang/xenlight/types.gen.go @@ -337,18 +337,18 @@ State int Evtch int Rref int Connection ChannelConnection -ConnectionUnion channelinfoConnectionUnion +ConnectionUnion ChannelinfoConnectionUnion } -type channelinfoConnectionUnion interface { -ischannelinfoConnectionUnion() +type ChannelinfoConnectionUnion interface { +isChannelinfoConnectionUnion() } type ChannelinfoConnectionUnionPty struct { Path string } -func (x ChannelinfoConnectionUnionPty) ischannelinfoConnectionUnion(){} +func (x ChannelinfoConnectionUnionPty) isChannelinfoConnectionUnion(){} type Vminfo struct { Uuid Uuid @@ -510,7 +510,7 @@ Apic Defbool DmRestrict Defbool Tee TeeType Type DomainType -TypeUnion domainBuildInfoTypeUnion +TypeUnion DomainBuildInfoTypeUnion ArchArm struct { GicVersion GicVersion Vuart VuartType @@ -522,8 +522,8 @@ Altp2M Altp2MMode VmtraceBufKb int } -type domainBuildInfoTypeUnion interface { -isdomainBuildInfoTypeUnion() +type DomainBuildInfoTypeUnion interface { +isDomainBuildInfoTypeUnion() } type DomainBuildInfoTypeUnionHvm struct { @@ -575,7 +575,7 @@ RdmMemBoundaryMemkb uint64 McaCaps uint64 } -func (x DomainBuildInfoTypeUnionHvm) isdomainBuildInfoTypeUnion(){} +func (x DomainBuildInfoTypeUnionHvm) isDomainBuildInfoTypeUnion(){} type DomainBuildInfoTypeUnionPv struct { Kernel string @@ -588,7 +588,7 @@ Features string E820Host Defbool } -func (x DomainBuildInfoTypeUnionPv) isdomainBuildInfoTypeUnion(){} +func (x DomainBuildInfoTypeUnionPv) isDomainBuildInfoTypeUnion(){} type DomainBuildInfoTypeUnionPvh struct { Pvshim Defbool @@ -597,7 +597,7 @@ PvshimCmdline string PvshimExtra string } -func (x DomainBuildInfoTypeUnionPvh) isdomainBuildInfoTypeUnion(){} +func (x DomainBuildInfoTypeUnionPvh) isDomainBuildInfoTypeUnion(){} type DeviceVfb struct { BackendDomid Domid @@ -761,11 +761,11 @@ type DeviceUsbdev struct { Ctrl Devid Port int Type UsbdevType -TypeUnion deviceUsbdevTypeUnion +TypeUnion DeviceUsbdevTypeUnion } -type deviceUsbdevTypeUnion interface { -isdeviceUsbdevTypeUnion() +type DeviceUsbdevTypeUnion interface { +isDeviceUsbdevTypeUnion() } type DeviceUsbdevTypeUnionHostdev struct { @@ -773,7 +773,7 @@ Hostbus byte Hostaddr byte } -func (x DeviceUsbdevTypeUnionHostdev) isdeviceUsbdevTypeUnion(){} +func (x DeviceUsbdevTypeUnionHostdev) isDeviceUsbdevTypeUnion(){} type DeviceDtdev struct { Path string @@ -807,18 +807,18 @@ BackendDomname string Devid Devid Name string Connection ChannelConnection -ConnectionUnion deviceChannelConnectionUnion +ConnectionUnion DeviceChannelConnectionUnion } -type deviceChannelConnectionUnion interface { -isdeviceChannelConnectionUnion() +type DeviceChannelConnectionUnion interface { +isDeviceChannelConnectionUnion() } type DeviceChannelConnectionUnionSocket struct { Path string } -func (x DeviceChannelConnectionUnionSocket) isdeviceChannelConnectionUnion(){} +func (x DeviceChannelConnectionUnionSocket) isDeviceChannelConnectionUnion(){} type ConnectorParam struct { UniqueId string @@ -1116,31 +1116,31 @@ Domid Domid Domuuid Uuid ForUser uint64 Type EventType -TypeUnion eventTypeUnion +TypeUnion EventTypeUnion } -type eventTypeUnion interface { -iseventTypeUnion() +type EventTypeUnion interface { +isEventTypeUnion() } type EventTypeUnionDomainShutdown struct { ShutdownReason byte } -func (x EventTypeUnionDomainShutdown) iseventTypeUnion(){} +func (x EventTypeUnionDomainShutdown) isEventTypeUnion(){} type EventTypeUnionDiskEject struct { Vdev string Disk DeviceDisk } -func (x EventTypeUnionDiskEject) iseventTypeUnion(){} +func (x EventTypeUnionDiskEject) isEventTypeUnion(){} type EventTypeUnionOperationComplete struct { Rc int } -func (x EventTypeUnionOperationComplete) iseventTypeUnion(){} +func (x EventTypeUnionOperationComplete) isEventTypeUnion(){} type PsrCmtType int const( @@ -1175,11 +1175,11 @@ PsrFeatTypeMba PsrFeatType = 2 type PsrHwInfo struct { Id uint32 Type PsrFeatType -TypeUnion psrHwInfoTypeUnion +TypeUnion PsrHwInfoTypeUnion } -type psrHwInfoTypeUnion interface { -ispsrHwInfoTypeUnion() +type PsrHwInfoTypeUnion interface { +isPsrHwInfoTypeUnion() } type PsrHwInfoTypeUnionCat struct { @@ -1188,7 +1188,7 @@ CbmLen uint32 CdpEnabled bool } -func (x PsrHwInfoTypeUnionCat) ispsrHwInfoTypeUnion(){} +func (x PsrHwInfoTypeUnionCat) isPsrHwInfoTypeUnion(){} type PsrHwInfoTypeUnionMba struct { CosMax uint32 @@ -1196,5 +1196,5 @@ ThrtlMax uint32 Linear bool } -func (x PsrHwInfoTypeUnionMba) ispsrHwInfoTypeUnion(){} +func (x PsrHwInfoTypeUnionMba) isPsrHwInfoTypeUnion(){} -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:32:19 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:32:19 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146652.269958 (Exim 4.92) (envelope-from ) id 1lwMeh-00016b-R5; Thu, 24 Jun 2021 10:32:19 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146652.269958; Thu, 24 Jun 2021 10:32:19 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMeh-00016T-MZ; Thu, 24 Jun 2021 10:32:19 +0000 Received: by outflank-mailman (input) for mailman id 146652; Thu, 24 Jun 2021 10:32:18 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMeg-00016E-Iz for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:32:18 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMeg-0006Fh-ID for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:32:18 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwMeg-0007FT-HT for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:32:18 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=uKdaHJyInxG6kodDLRmuTcdvOtQugOECHW7lWlwN6+g=; b=0tVlsKPzQGqGakBiiSa380wPzl D34PRdDIGzTi+HHbZEhXcjWqnj2rBAfNWlqgo8gSHkeNNfGo8IE6f3qMEsWqibqznou7rkyxq6jR2 4Hw5zg1smKmeCr0KlncN54Q+UURBgFU40RnR5kiLOI0rafmrj6xSDM9ddC7sH5IU+q+Y=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] golang/xenlight: use struct pointers in keyed union fields Message-Id: Date: Thu, 24 Jun 2021 10:32:18 +0000 commit 1997940ad25e3566d1ab38496b8c7b07a086695a Author: Nick Rosbrook AuthorDate: Mon May 24 16:36:46 2021 -0400 Commit: George Dunlap CommitDate: Mon Jun 21 16:49:08 2021 +0100 golang/xenlight: use struct pointers in keyed union fields Currently, when marshalig Go types with keyed union fields, we assign the value of the struct (e.g. DomainBuildInfoTypeUnionHvm) which implements the interface of the keyed union field (e.g. DomainBuildInfoTypeUnion). As-is, this means that if a populated DomainBuildInfo is marshaled to e.g. JSON, unmarshaling back to DomainBuildInfo will fail. When the encoding/json is unmarshaling data into a Go type, and encounters a JSON object, it basically can either marshal the data into an empty interface, a map, or a struct. It cannot, however, marshal data into an interface with at least one method defined on it (e.g. DomainBuildInfoTypeUnion). Before this check is done, however, the decoder will check if the Go type is a pointer, and dereference it if so. It will then use the type of this value as the "target" type. This means that if the TypeUnion field is populated with a DomainBuildInfoTypeUnion, the decoder will see a non-empty interface and fail. If the TypeUnion field is populated with a *DomainBuildInfoTypeUnionHvm, it dereferences the pointer and sees a struct instead, allowing decoding to continue normally. Since there does not appear to be a strict need for NOT using pointers in these fields, update code generation to set keyed union fields to pointers of their implementing structs. Signed-off-by: Nick Rosbrook Reviewed-by: George Dunlap --- tools/golang/xenlight/gengotypes.py | 4 ++-- tools/golang/xenlight/helpers.gen.go | 44 ++++++++++++++++++------------------ 2 files changed, 24 insertions(+), 24 deletions(-) diff --git a/tools/golang/xenlight/gengotypes.py b/tools/golang/xenlight/gengotypes.py index ef3279a910..ac1cf060dd 100644 --- a/tools/golang/xenlight/gengotypes.py +++ b/tools/golang/xenlight/gengotypes.py @@ -404,7 +404,7 @@ def xenlight_golang_union_from_C(ty = None, union_name = '', struct_name = ''): s += 'if err := {0}.fromC(xc);'.format(goname) s += 'err != nil {{\n return fmt.Errorf("converting field {0}: %v", err)\n}}\n'.format(goname) - s += 'x.{0} = {1}\n'.format(field_name, goname) + s += 'x.{0} = &{1}\n'.format(field_name, goname) # End switch statement s += 'default:\n' @@ -570,7 +570,7 @@ def xenlight_golang_union_to_C(ty = None, union_name = '', gotype = xenlight_golang_fmt_name(cgotype) field_name = xenlight_golang_fmt_name('{0}_union'.format(keyname)) - s += 'tmp, ok := x.{0}.({1})\n'.format(field_name,gotype) + s += 'tmp, ok := x.{0}.(*{1})\n'.format(field_name,gotype) s += 'if !ok {\n' s += 'return errors.New("wrong type for union key {0}")\n'.format(keyname) s += '}\n' diff --git a/tools/golang/xenlight/helpers.gen.go b/tools/golang/xenlight/helpers.gen.go index b454b12d52..db82537b42 100644 --- a/tools/golang/xenlight/helpers.gen.go +++ b/tools/golang/xenlight/helpers.gen.go @@ -433,7 +433,7 @@ var connectionPty ChannelinfoConnectionUnionPty if err := connectionPty.fromC(xc);err != nil { return fmt.Errorf("converting field connectionPty: %v", err) } -x.ConnectionUnion = connectionPty +x.ConnectionUnion = &connectionPty case ChannelConnectionSocket: x.ConnectionUnion = nil case ChannelConnectionUnknown: @@ -473,7 +473,7 @@ switch x.Connection{ case ChannelConnectionUnknown: break case ChannelConnectionPty: -tmp, ok := x.ConnectionUnion.(ChannelinfoConnectionUnionPty) +tmp, ok := x.ConnectionUnion.(*ChannelinfoConnectionUnionPty) if !ok { return errors.New("wrong type for union key connection") } @@ -1094,7 +1094,7 @@ var typeHvm DomainBuildInfoTypeUnionHvm if err := typeHvm.fromC(xc);err != nil { return fmt.Errorf("converting field typeHvm: %v", err) } -x.TypeUnion = typeHvm +x.TypeUnion = &typeHvm case DomainTypeInvalid: x.TypeUnion = nil case DomainTypePv: @@ -1102,13 +1102,13 @@ var typePv DomainBuildInfoTypeUnionPv if err := typePv.fromC(xc);err != nil { return fmt.Errorf("converting field typePv: %v", err) } -x.TypeUnion = typePv +x.TypeUnion = &typePv case DomainTypePvh: var typePvh DomainBuildInfoTypeUnionPvh if err := typePvh.fromC(xc);err != nil { return fmt.Errorf("converting field typePvh: %v", err) } -x.TypeUnion = typePvh +x.TypeUnion = &typePvh default: return fmt.Errorf("invalid union key '%v'", x.Type)} x.ArchArm.GicVersion = GicVersion(xc.arch_arm.gic_version) @@ -1427,7 +1427,7 @@ xc.tee = C.libxl_tee_type(x.Tee) xc._type = C.libxl_domain_type(x.Type) switch x.Type{ case DomainTypeHvm: -tmp, ok := x.TypeUnion.(DomainBuildInfoTypeUnionHvm) +tmp, ok := x.TypeUnion.(*DomainBuildInfoTypeUnionHvm) if !ok { return errors.New("wrong type for union key type") } @@ -1545,7 +1545,7 @@ hvm.mca_caps = C.uint64_t(tmp.McaCaps) hvmBytes := C.GoBytes(unsafe.Pointer(&hvm),C.sizeof_libxl_domain_build_info_type_union_hvm) copy(xc.u[:],hvmBytes) case DomainTypePv: -tmp, ok := x.TypeUnion.(DomainBuildInfoTypeUnionPv) +tmp, ok := x.TypeUnion.(*DomainBuildInfoTypeUnionPv) if !ok { return errors.New("wrong type for union key type") } @@ -1570,7 +1570,7 @@ return fmt.Errorf("converting field E820Host: %v", err) pvBytes := C.GoBytes(unsafe.Pointer(&pv),C.sizeof_libxl_domain_build_info_type_union_pv) copy(xc.u[:],pvBytes) case DomainTypePvh: -tmp, ok := x.TypeUnion.(DomainBuildInfoTypeUnionPvh) +tmp, ok := x.TypeUnion.(*DomainBuildInfoTypeUnionPvh) if !ok { return errors.New("wrong type for union key type") } @@ -2157,7 +2157,7 @@ var typeHostdev DeviceUsbdevTypeUnionHostdev if err := typeHostdev.fromC(xc);err != nil { return fmt.Errorf("converting field typeHostdev: %v", err) } -x.TypeUnion = typeHostdev +x.TypeUnion = &typeHostdev default: return fmt.Errorf("invalid union key '%v'", x.Type)} @@ -2184,7 +2184,7 @@ xc.port = C.int(x.Port) xc._type = C.libxl_usbdev_type(x.Type) switch x.Type{ case UsbdevTypeHostdev: -tmp, ok := x.TypeUnion.(DeviceUsbdevTypeUnionHostdev) +tmp, ok := x.TypeUnion.(*DeviceUsbdevTypeUnionHostdev) if !ok { return errors.New("wrong type for union key type") } @@ -2375,7 +2375,7 @@ var connectionSocket DeviceChannelConnectionUnionSocket if err := connectionSocket.fromC(xc);err != nil { return fmt.Errorf("converting field connectionSocket: %v", err) } -x.ConnectionUnion = connectionSocket +x.ConnectionUnion = &connectionSocket case ChannelConnectionUnknown: x.ConnectionUnion = nil default: @@ -2411,7 +2411,7 @@ break case ChannelConnectionPty: break case ChannelConnectionSocket: -tmp, ok := x.ConnectionUnion.(DeviceChannelConnectionUnionSocket) +tmp, ok := x.ConnectionUnion.(*DeviceChannelConnectionUnionSocket) if !ok { return errors.New("wrong type for union key connection") } @@ -3948,7 +3948,7 @@ var typeDiskEject EventTypeUnionDiskEject if err := typeDiskEject.fromC(xc);err != nil { return fmt.Errorf("converting field typeDiskEject: %v", err) } -x.TypeUnion = typeDiskEject +x.TypeUnion = &typeDiskEject case EventTypeDomainCreateConsoleAvailable: x.TypeUnion = nil case EventTypeDomainDeath: @@ -3958,13 +3958,13 @@ var typeDomainShutdown EventTypeUnionDomainShutdown if err := typeDomainShutdown.fromC(xc);err != nil { return fmt.Errorf("converting field typeDomainShutdown: %v", err) } -x.TypeUnion = typeDomainShutdown +x.TypeUnion = &typeDomainShutdown case EventTypeOperationComplete: var typeOperationComplete EventTypeUnionOperationComplete if err := typeOperationComplete.fromC(xc);err != nil { return fmt.Errorf("converting field typeOperationComplete: %v", err) } -x.TypeUnion = typeOperationComplete +x.TypeUnion = &typeOperationComplete default: return fmt.Errorf("invalid union key '%v'", x.Type)} @@ -4019,7 +4019,7 @@ xc.for_user = C.uint64_t(x.ForUser) xc._type = C.libxl_event_type(x.Type) switch x.Type{ case EventTypeDomainShutdown: -tmp, ok := x.TypeUnion.(EventTypeUnionDomainShutdown) +tmp, ok := x.TypeUnion.(*EventTypeUnionDomainShutdown) if !ok { return errors.New("wrong type for union key type") } @@ -4030,7 +4030,7 @@ copy(xc.u[:],domain_shutdownBytes) case EventTypeDomainDeath: break case EventTypeDiskEject: -tmp, ok := x.TypeUnion.(EventTypeUnionDiskEject) +tmp, ok := x.TypeUnion.(*EventTypeUnionDiskEject) if !ok { return errors.New("wrong type for union key type") } @@ -4043,7 +4043,7 @@ return fmt.Errorf("converting field Disk: %v", err) disk_ejectBytes := C.GoBytes(unsafe.Pointer(&disk_eject),C.sizeof_libxl_event_type_union_disk_eject) copy(xc.u[:],disk_ejectBytes) case EventTypeOperationComplete: -tmp, ok := x.TypeUnion.(EventTypeUnionOperationComplete) +tmp, ok := x.TypeUnion.(*EventTypeUnionOperationComplete) if !ok { return errors.New("wrong type for union key type") } @@ -4118,13 +4118,13 @@ var typeCat PsrHwInfoTypeUnionCat if err := typeCat.fromC(xc);err != nil { return fmt.Errorf("converting field typeCat: %v", err) } -x.TypeUnion = typeCat +x.TypeUnion = &typeCat case PsrFeatTypeMba: var typeMba PsrHwInfoTypeUnionMba if err := typeMba.fromC(xc);err != nil { return fmt.Errorf("converting field typeMba: %v", err) } -x.TypeUnion = typeMba +x.TypeUnion = &typeMba default: return fmt.Errorf("invalid union key '%v'", x.Type)} @@ -4163,7 +4163,7 @@ xc.id = C.uint32_t(x.Id) xc._type = C.libxl_psr_feat_type(x.Type) switch x.Type{ case PsrFeatTypeCat: -tmp, ok := x.TypeUnion.(PsrHwInfoTypeUnionCat) +tmp, ok := x.TypeUnion.(*PsrHwInfoTypeUnionCat) if !ok { return errors.New("wrong type for union key type") } @@ -4174,7 +4174,7 @@ cat.cdp_enabled = C.bool(tmp.CdpEnabled) catBytes := C.GoBytes(unsafe.Pointer(&cat),C.sizeof_libxl_psr_hw_info_type_union_cat) copy(xc.u[:],catBytes) case PsrFeatTypeMba: -tmp, ok := x.TypeUnion.(PsrHwInfoTypeUnionMba) +tmp, ok := x.TypeUnion.(*PsrHwInfoTypeUnionMba) if !ok { return errors.New("wrong type for union key type") } -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:32:29 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:32:29 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146654.269960 (Exim 4.92) (envelope-from ) id 1lwMer-0001Bx-TG; Thu, 24 Jun 2021 10:32:29 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146654.269960; Thu, 24 Jun 2021 10:32:29 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMer-0001Bp-QN; Thu, 24 Jun 2021 10:32:29 +0000 Received: by outflank-mailman (input) for mailman id 146654; Thu, 24 Jun 2021 10:32:28 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMeq-0001Bd-M6 for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:32:28 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMeq-0006Fs-LM for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:32:28 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwMeq-0007GF-Kb for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:32:28 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=Fdz1sTFsS7hYpKQEFN41q3NqilyvtgFMzqYSQDThkec=; b=Gx1kFN7gkCDqwifiSb9ZUvIcCq eiU89b8ZA9JPYiTfr4M0TbrnGCKzDD+tdYJZhlDRbw5BIGUBDX0Z35o6TMdr2f/8AMHHugGcKZrZ9 1wakIVViNxv9gPv47M7a6q1HoNN+xXa5O8NfkTkoby8qDZD7QBN8DoD2nwGbsrtS9Xlk=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] golang/xenlight: rename Ctx receivers to ctx Message-Id: Date: Thu, 24 Jun 2021 10:32:28 +0000 commit c089de0e2fa56d846cfb658b7b5efc3426895973 Author: Nick Rosbrook AuthorDate: Mon May 24 16:36:47 2021 -0400 Commit: George Dunlap CommitDate: Mon Jun 21 16:49:08 2021 +0100 golang/xenlight: rename Ctx receivers to ctx As a matter of style, it is strange to see capitalized receiver names, due to the significance of capitalized symbols in Go (although there is in fact nothing special about a capitalized receiver name). Fix this in xenlight.go by running: gofmt -w -r 'Ctx -> ctx' xenlight.go from tools/golang/xenlight. There is no functional change. Signed-off-by: Nick Rosbrook Reviewed-by: George Dunlap --- tools/golang/xenlight/xenlight.go | 154 +++++++++++++++++++------------------- 1 file changed, 77 insertions(+), 77 deletions(-) diff --git a/tools/golang/xenlight/xenlight.go b/tools/golang/xenlight/xenlight.go index 13171d0ad1..fc3eb0bf3f 100644 --- a/tools/golang/xenlight/xenlight.go +++ b/tools/golang/xenlight/xenlight.go @@ -203,13 +203,13 @@ type Domid uint32 // NameToDomid does not guarantee that the domid associated with name at // the time NameToDomid is called is the same as the domid associated with // name at the time NameToDomid returns. -func (Ctx *Context) NameToDomid(name string) (Domid, error) { +func (ctx *Context) NameToDomid(name string) (Domid, error) { var domid C.uint32_t cname := C.CString(name) defer C.free(unsafe.Pointer(cname)) - if ret := C.libxl_name_to_domid(Ctx.ctx, cname, &domid); ret != 0 { + if ret := C.libxl_name_to_domid(ctx.ctx, cname, &domid); ret != 0 { return DomidInvalid, Error(ret) } @@ -223,8 +223,8 @@ func (Ctx *Context) NameToDomid(name string) (Domid, error) { // DomidToName does not guarantee that the name (if any) associated with domid // at the time DomidToName is called is the same as the name (if any) associated // with domid at the time DomidToName returns. -func (Ctx *Context) DomidToName(domid Domid) string { - cname := C.libxl_domid_to_name(Ctx.ctx, C.uint32_t(domid)) +func (ctx *Context) DomidToName(domid Domid) string { + cname := C.libxl_domid_to_name(ctx.ctx, C.uint32_t(domid)) defer C.free(unsafe.Pointer(cname)) return C.GoString(cname) @@ -594,10 +594,10 @@ func SchedulerFromString(name string) (s Scheduler, err error) { // libxl_cpupoolinfo * libxl_list_cpupool(libxl_ctx*, int *nb_pool_out); // void libxl_cpupoolinfo_list_free(libxl_cpupoolinfo *list, int nb_pool); -func (Ctx *Context) ListCpupool() (list []Cpupoolinfo) { +func (ctx *Context) ListCpupool() (list []Cpupoolinfo) { var nbPool C.int - c_cpupool_list := C.libxl_list_cpupool(Ctx.ctx, &nbPool) + c_cpupool_list := C.libxl_list_cpupool(ctx.ctx, &nbPool) defer C.libxl_cpupoolinfo_list_free(c_cpupool_list, nbPool) @@ -617,10 +617,10 @@ func (Ctx *Context) ListCpupool() (list []Cpupoolinfo) { } // int libxl_cpupool_info(libxl_ctx *ctx, libxl_cpupoolinfo *info, uint32_t poolid); -func (Ctx *Context) CpupoolInfo(Poolid uint32) (pool Cpupoolinfo, err error) { +func (ctx *Context) CpupoolInfo(Poolid uint32) (pool Cpupoolinfo, err error) { var c_cpupool C.libxl_cpupoolinfo - ret := C.libxl_cpupool_info(Ctx.ctx, &c_cpupool, C.uint32_t(Poolid)) + ret := C.libxl_cpupool_info(ctx.ctx, &c_cpupool, C.uint32_t(Poolid)) if ret != 0 { err = Error(-ret) return @@ -638,7 +638,7 @@ func (Ctx *Context) CpupoolInfo(Poolid uint32) (pool Cpupoolinfo, err error) { // uint32_t *poolid); // FIXME: uuid // FIXME: Setting poolid -func (Ctx *Context) CpupoolCreate(Name string, Scheduler Scheduler, Cpumap Bitmap) (err error, Poolid uint32) { +func (ctx *Context) CpupoolCreate(Name string, Scheduler Scheduler, Cpumap Bitmap) (err error, Poolid uint32) { poolid := C.uint32_t(C.LIBXL_CPUPOOL_POOLID_ANY) name := C.CString(Name) defer C.free(unsafe.Pointer(name)) @@ -653,7 +653,7 @@ func (Ctx *Context) CpupoolCreate(Name string, Scheduler Scheduler, Cpumap Bitma } defer C.libxl_bitmap_dispose(&cbm) - ret := C.libxl_cpupool_create(Ctx.ctx, name, C.libxl_scheduler(Scheduler), + ret := C.libxl_cpupool_create(ctx.ctx, name, C.libxl_scheduler(Scheduler), cbm, &uuid, &poolid) if ret != 0 { err = Error(-ret) @@ -666,8 +666,8 @@ func (Ctx *Context) CpupoolCreate(Name string, Scheduler Scheduler, Cpumap Bitma } // int libxl_cpupool_destroy(libxl_ctx *ctx, uint32_t poolid); -func (Ctx *Context) CpupoolDestroy(Poolid uint32) (err error) { - ret := C.libxl_cpupool_destroy(Ctx.ctx, C.uint32_t(Poolid)) +func (ctx *Context) CpupoolDestroy(Poolid uint32) (err error) { + ret := C.libxl_cpupool_destroy(ctx.ctx, C.uint32_t(Poolid)) if ret != 0 { err = Error(-ret) return @@ -677,8 +677,8 @@ func (Ctx *Context) CpupoolDestroy(Poolid uint32) (err error) { } // int libxl_cpupool_cpuadd(libxl_ctx *ctx, uint32_t poolid, int cpu); -func (Ctx *Context) CpupoolCpuadd(Poolid uint32, Cpu int) (err error) { - ret := C.libxl_cpupool_cpuadd(Ctx.ctx, C.uint32_t(Poolid), C.int(Cpu)) +func (ctx *Context) CpupoolCpuadd(Poolid uint32, Cpu int) (err error) { + ret := C.libxl_cpupool_cpuadd(ctx.ctx, C.uint32_t(Poolid), C.int(Cpu)) if ret != 0 { err = Error(-ret) return @@ -689,14 +689,14 @@ func (Ctx *Context) CpupoolCpuadd(Poolid uint32, Cpu int) (err error) { // int libxl_cpupool_cpuadd_cpumap(libxl_ctx *ctx, uint32_t poolid, // const libxl_bitmap *cpumap); -func (Ctx *Context) CpupoolCpuaddCpumap(Poolid uint32, Cpumap Bitmap) (err error) { +func (ctx *Context) CpupoolCpuaddCpumap(Poolid uint32, Cpumap Bitmap) (err error) { var cbm C.libxl_bitmap if err = Cpumap.toC(&cbm); err != nil { return } defer C.libxl_bitmap_dispose(&cbm) - ret := C.libxl_cpupool_cpuadd_cpumap(Ctx.ctx, C.uint32_t(Poolid), &cbm) + ret := C.libxl_cpupool_cpuadd_cpumap(ctx.ctx, C.uint32_t(Poolid), &cbm) if ret != 0 { err = Error(-ret) return @@ -706,8 +706,8 @@ func (Ctx *Context) CpupoolCpuaddCpumap(Poolid uint32, Cpumap Bitmap) (err error } // int libxl_cpupool_cpuremove(libxl_ctx *ctx, uint32_t poolid, int cpu); -func (Ctx *Context) CpupoolCpuremove(Poolid uint32, Cpu int) (err error) { - ret := C.libxl_cpupool_cpuremove(Ctx.ctx, C.uint32_t(Poolid), C.int(Cpu)) +func (ctx *Context) CpupoolCpuremove(Poolid uint32, Cpu int) (err error) { + ret := C.libxl_cpupool_cpuremove(ctx.ctx, C.uint32_t(Poolid), C.int(Cpu)) if ret != 0 { err = Error(-ret) return @@ -718,14 +718,14 @@ func (Ctx *Context) CpupoolCpuremove(Poolid uint32, Cpu int) (err error) { // int libxl_cpupool_cpuremove_cpumap(libxl_ctx *ctx, uint32_t poolid, // const libxl_bitmap *cpumap); -func (Ctx *Context) CpupoolCpuremoveCpumap(Poolid uint32, Cpumap Bitmap) (err error) { +func (ctx *Context) CpupoolCpuremoveCpumap(Poolid uint32, Cpumap Bitmap) (err error) { var cbm C.libxl_bitmap if err = Cpumap.toC(&cbm); err != nil { return } defer C.libxl_bitmap_dispose(&cbm) - ret := C.libxl_cpupool_cpuremove_cpumap(Ctx.ctx, C.uint32_t(Poolid), &cbm) + ret := C.libxl_cpupool_cpuremove_cpumap(ctx.ctx, C.uint32_t(Poolid), &cbm) if ret != 0 { err = Error(-ret) return @@ -735,11 +735,11 @@ func (Ctx *Context) CpupoolCpuremoveCpumap(Poolid uint32, Cpumap Bitmap) (err er } // int libxl_cpupool_rename(libxl_ctx *ctx, const char *name, uint32_t poolid); -func (Ctx *Context) CpupoolRename(Name string, Poolid uint32) (err error) { +func (ctx *Context) CpupoolRename(Name string, Poolid uint32) (err error) { name := C.CString(Name) defer C.free(unsafe.Pointer(name)) - ret := C.libxl_cpupool_rename(Ctx.ctx, name, C.uint32_t(Poolid)) + ret := C.libxl_cpupool_rename(ctx.ctx, name, C.uint32_t(Poolid)) if ret != 0 { err = Error(-ret) return @@ -749,10 +749,10 @@ func (Ctx *Context) CpupoolRename(Name string, Poolid uint32) (err error) { } // int libxl_cpupool_cpuadd_node(libxl_ctx *ctx, uint32_t poolid, int node, int *cpus); -func (Ctx *Context) CpupoolCpuaddNode(Poolid uint32, Node int) (Cpus int, err error) { +func (ctx *Context) CpupoolCpuaddNode(Poolid uint32, Node int) (Cpus int, err error) { ccpus := C.int(0) - ret := C.libxl_cpupool_cpuadd_node(Ctx.ctx, C.uint32_t(Poolid), C.int(Node), &ccpus) + ret := C.libxl_cpupool_cpuadd_node(ctx.ctx, C.uint32_t(Poolid), C.int(Node), &ccpus) if ret != 0 { err = Error(-ret) return @@ -764,10 +764,10 @@ func (Ctx *Context) CpupoolCpuaddNode(Poolid uint32, Node int) (Cpus int, err er } // int libxl_cpupool_cpuremove_node(libxl_ctx *ctx, uint32_t poolid, int node, int *cpus); -func (Ctx *Context) CpupoolCpuremoveNode(Poolid uint32, Node int) (Cpus int, err error) { +func (ctx *Context) CpupoolCpuremoveNode(Poolid uint32, Node int) (Cpus int, err error) { ccpus := C.int(0) - ret := C.libxl_cpupool_cpuremove_node(Ctx.ctx, C.uint32_t(Poolid), C.int(Node), &ccpus) + ret := C.libxl_cpupool_cpuremove_node(ctx.ctx, C.uint32_t(Poolid), C.int(Node), &ccpus) if ret != 0 { err = Error(-ret) return @@ -779,8 +779,8 @@ func (Ctx *Context) CpupoolCpuremoveNode(Poolid uint32, Node int) (Cpus int, err } // int libxl_cpupool_movedomain(libxl_ctx *ctx, uint32_t poolid, uint32_t domid); -func (Ctx *Context) CpupoolMovedomain(Poolid uint32, Id Domid) (err error) { - ret := C.libxl_cpupool_movedomain(Ctx.ctx, C.uint32_t(Poolid), C.uint32_t(Id)) +func (ctx *Context) CpupoolMovedomain(Poolid uint32, Id Domid) (err error) { + ret := C.libxl_cpupool_movedomain(ctx.ctx, C.uint32_t(Poolid), C.uint32_t(Id)) if ret != 0 { err = Error(-ret) return @@ -792,8 +792,8 @@ func (Ctx *Context) CpupoolMovedomain(Poolid uint32, Id Domid) (err error) { // // Utility functions // -func (Ctx *Context) CpupoolFindByName(name string) (info Cpupoolinfo, found bool) { - plist := Ctx.ListCpupool() +func (ctx *Context) CpupoolFindByName(name string) (info Cpupoolinfo, found bool) { + plist := ctx.ListCpupool() for i := range plist { if plist[i].PoolName == name { @@ -805,14 +805,14 @@ func (Ctx *Context) CpupoolFindByName(name string) (info Cpupoolinfo, found bool return } -func (Ctx *Context) CpupoolMakeFree(Cpumap Bitmap) (err error) { - plist := Ctx.ListCpupool() +func (ctx *Context) CpupoolMakeFree(Cpumap Bitmap) (err error) { + plist := ctx.ListCpupool() for i := range plist { var Intersection Bitmap Intersection = Cpumap.And(plist[i].Cpumap) if !Intersection.IsEmpty() { - err = Ctx.CpupoolCpuremoveCpumap(plist[i].Poolid, Intersection) + err = ctx.CpupoolCpuremoveCpumap(plist[i].Poolid, Intersection) if err != nil { return } @@ -940,8 +940,8 @@ func (bm Bitmap) String() (s string) { } //int libxl_get_max_cpus(libxl_ctx *ctx); -func (Ctx *Context) GetMaxCpus() (maxCpus int, err error) { - ret := C.libxl_get_max_cpus(Ctx.ctx) +func (ctx *Context) GetMaxCpus() (maxCpus int, err error) { + ret := C.libxl_get_max_cpus(ctx.ctx) if ret < 0 { err = Error(-ret) return @@ -951,8 +951,8 @@ func (Ctx *Context) GetMaxCpus() (maxCpus int, err error) { } //int libxl_get_online_cpus(libxl_ctx *ctx); -func (Ctx *Context) GetOnlineCpus() (onCpus int, err error) { - ret := C.libxl_get_online_cpus(Ctx.ctx) +func (ctx *Context) GetOnlineCpus() (onCpus int, err error) { + ret := C.libxl_get_online_cpus(ctx.ctx) if ret < 0 { err = Error(-ret) return @@ -962,8 +962,8 @@ func (Ctx *Context) GetOnlineCpus() (onCpus int, err error) { } //int libxl_get_max_nodes(libxl_ctx *ctx); -func (Ctx *Context) GetMaxNodes() (maxNodes int, err error) { - ret := C.libxl_get_max_nodes(Ctx.ctx) +func (ctx *Context) GetMaxNodes() (maxNodes int, err error) { + ret := C.libxl_get_max_nodes(ctx.ctx) if ret < 0 { err = Error(-ret) return @@ -973,9 +973,9 @@ func (Ctx *Context) GetMaxNodes() (maxNodes int, err error) { } //int libxl_get_free_memory(libxl_ctx *ctx, uint64_t *memkb); -func (Ctx *Context) GetFreeMemory() (memkb uint64, err error) { +func (ctx *Context) GetFreeMemory() (memkb uint64, err error) { var cmem C.uint64_t - ret := C.libxl_get_free_memory(Ctx.ctx, &cmem) + ret := C.libxl_get_free_memory(ctx.ctx, &cmem) if ret < 0 { err = Error(-ret) @@ -988,12 +988,12 @@ func (Ctx *Context) GetFreeMemory() (memkb uint64, err error) { } //int libxl_get_physinfo(libxl_ctx *ctx, libxl_physinfo *physinfo) -func (Ctx *Context) GetPhysinfo() (physinfo *Physinfo, err error) { +func (ctx *Context) GetPhysinfo() (physinfo *Physinfo, err error) { var cphys C.libxl_physinfo C.libxl_physinfo_init(&cphys) defer C.libxl_physinfo_dispose(&cphys) - ret := C.libxl_get_physinfo(Ctx.ctx, &cphys) + ret := C.libxl_get_physinfo(ctx.ctx, &cphys) if ret < 0 { err = Error(ret) @@ -1005,22 +1005,22 @@ func (Ctx *Context) GetPhysinfo() (physinfo *Physinfo, err error) { } //const libxl_version_info* libxl_get_version_info(libxl_ctx *ctx); -func (Ctx *Context) GetVersionInfo() (info *VersionInfo, err error) { +func (ctx *Context) GetVersionInfo() (info *VersionInfo, err error) { var cinfo *C.libxl_version_info - cinfo = C.libxl_get_version_info(Ctx.ctx) + cinfo = C.libxl_get_version_info(ctx.ctx) err = info.fromC(cinfo) return } -func (Ctx *Context) DomainInfo(Id Domid) (di *Dominfo, err error) { +func (ctx *Context) DomainInfo(Id Domid) (di *Dominfo, err error) { var cdi C.libxl_dominfo C.libxl_dominfo_init(&cdi) defer C.libxl_dominfo_dispose(&cdi) - ret := C.libxl_domain_info(Ctx.ctx, &cdi, C.uint32_t(Id)) + ret := C.libxl_domain_info(ctx.ctx, &cdi, C.uint32_t(Id)) if ret != 0 { err = Error(-ret) @@ -1032,8 +1032,8 @@ func (Ctx *Context) DomainInfo(Id Domid) (di *Dominfo, err error) { return } -func (Ctx *Context) DomainUnpause(Id Domid) (err error) { - ret := C.libxl_domain_unpause(Ctx.ctx, C.uint32_t(Id), nil) +func (ctx *Context) DomainUnpause(Id Domid) (err error) { + ret := C.libxl_domain_unpause(ctx.ctx, C.uint32_t(Id), nil) if ret != 0 { err = Error(-ret) @@ -1042,8 +1042,8 @@ func (Ctx *Context) DomainUnpause(Id Domid) (err error) { } //int libxl_domain_pause(libxl_ctx *ctx, uint32_t domain); -func (Ctx *Context) DomainPause(id Domid) (err error) { - ret := C.libxl_domain_pause(Ctx.ctx, C.uint32_t(id), nil) +func (ctx *Context) DomainPause(id Domid) (err error) { + ret := C.libxl_domain_pause(ctx.ctx, C.uint32_t(id), nil) if ret != 0 { err = Error(-ret) @@ -1052,8 +1052,8 @@ func (Ctx *Context) DomainPause(id Domid) (err error) { } //int libxl_domain_shutdown(libxl_ctx *ctx, uint32_t domid); -func (Ctx *Context) DomainShutdown(id Domid) (err error) { - ret := C.libxl_domain_shutdown(Ctx.ctx, C.uint32_t(id), nil) +func (ctx *Context) DomainShutdown(id Domid) (err error) { + ret := C.libxl_domain_shutdown(ctx.ctx, C.uint32_t(id), nil) if ret != 0 { err = Error(-ret) @@ -1062,8 +1062,8 @@ func (Ctx *Context) DomainShutdown(id Domid) (err error) { } //int libxl_domain_reboot(libxl_ctx *ctx, uint32_t domid); -func (Ctx *Context) DomainReboot(id Domid) (err error) { - ret := C.libxl_domain_reboot(Ctx.ctx, C.uint32_t(id), nil) +func (ctx *Context) DomainReboot(id Domid) (err error) { + ret := C.libxl_domain_reboot(ctx.ctx, C.uint32_t(id), nil) if ret != 0 { err = Error(-ret) @@ -1073,9 +1073,9 @@ func (Ctx *Context) DomainReboot(id Domid) (err error) { //libxl_dominfo * libxl_list_domain(libxl_ctx*, int *nb_domain_out); //void libxl_dominfo_list_free(libxl_dominfo *list, int nb_domain); -func (Ctx *Context) ListDomain() (glist []Dominfo) { +func (ctx *Context) ListDomain() (glist []Dominfo) { var nbDomain C.int - clist := C.libxl_list_domain(Ctx.ctx, &nbDomain) + clist := C.libxl_list_domain(ctx.ctx, &nbDomain) defer C.libxl_dominfo_list_free(clist, nbDomain) if int(nbDomain) == 0 { @@ -1095,11 +1095,11 @@ func (Ctx *Context) ListDomain() (glist []Dominfo) { //libxl_vcpuinfo *libxl_list_vcpu(libxl_ctx *ctx, uint32_t domid, // int *nb_vcpu, int *nr_cpus_out); //void libxl_vcpuinfo_list_free(libxl_vcpuinfo *, int nr_vcpus); -func (Ctx *Context) ListVcpu(id Domid) (glist []Vcpuinfo) { +func (ctx *Context) ListVcpu(id Domid) (glist []Vcpuinfo) { var nbVcpu C.int var nrCpu C.int - clist := C.libxl_list_vcpu(Ctx.ctx, C.uint32_t(id), &nbVcpu, &nrCpu) + clist := C.libxl_list_vcpu(ctx.ctx, C.uint32_t(id), &nbVcpu, &nrCpu) defer C.libxl_vcpuinfo_list_free(clist, nbVcpu) if int(nbVcpu) == 0 { @@ -1125,9 +1125,9 @@ func (ct ConsoleType) String() (str string) { //int libxl_console_get_tty(libxl_ctx *ctx, uint32_t domid, int cons_num, //libxl_console_type type, char **path); -func (Ctx *Context) ConsoleGetTty(id Domid, consNum int, conType ConsoleType) (path string, err error) { +func (ctx *Context) ConsoleGetTty(id Domid, consNum int, conType ConsoleType) (path string, err error) { var cpath *C.char - ret := C.libxl_console_get_tty(Ctx.ctx, C.uint32_t(id), C.int(consNum), C.libxl_console_type(conType), &cpath) + ret := C.libxl_console_get_tty(ctx.ctx, C.uint32_t(id), C.int(consNum), C.libxl_console_type(conType), &cpath) if ret != 0 { err = Error(-ret) return @@ -1140,9 +1140,9 @@ func (Ctx *Context) ConsoleGetTty(id Domid, consNum int, conType ConsoleType) (p //int libxl_primary_console_get_tty(libxl_ctx *ctx, uint32_t domid_vm, // char **path); -func (Ctx *Context) PrimaryConsoleGetTty(domid uint32) (path string, err error) { +func (ctx *Context) PrimaryConsoleGetTty(domid uint32) (path string, err error) { var cpath *C.char - ret := C.libxl_primary_console_get_tty(Ctx.ctx, C.uint32_t(domid), &cpath) + ret := C.libxl_primary_console_get_tty(ctx.ctx, C.uint32_t(domid), &cpath) if ret != 0 { err = Error(-ret) return @@ -1154,7 +1154,7 @@ func (Ctx *Context) PrimaryConsoleGetTty(domid uint32) (path string, err error) } // DeviceNicAdd adds a nic to a domain. -func (Ctx *Context) DeviceNicAdd(domid Domid, nic *DeviceNic) error { +func (ctx *Context) DeviceNicAdd(domid Domid, nic *DeviceNic) error { var cnic C.libxl_device_nic if err := nic.toC(&cnic); err != nil { @@ -1162,7 +1162,7 @@ func (Ctx *Context) DeviceNicAdd(domid Domid, nic *DeviceNic) error { } defer C.libxl_device_nic_dispose(&cnic) - ret := C.libxl_device_nic_add(Ctx.ctx, C.uint32_t(domid), &cnic, nil) + ret := C.libxl_device_nic_add(ctx.ctx, C.uint32_t(domid), &cnic, nil) if ret != 0 { return Error(ret) } @@ -1171,7 +1171,7 @@ func (Ctx *Context) DeviceNicAdd(domid Domid, nic *DeviceNic) error { } // DeviceNicRemove removes a nic from a domain. -func (Ctx *Context) DeviceNicRemove(domid Domid, nic *DeviceNic) error { +func (ctx *Context) DeviceNicRemove(domid Domid, nic *DeviceNic) error { var cnic C.libxl_device_nic if err := nic.toC(&cnic); err != nil { @@ -1179,7 +1179,7 @@ func (Ctx *Context) DeviceNicRemove(domid Domid, nic *DeviceNic) error { } defer C.libxl_device_nic_dispose(&cnic) - ret := C.libxl_device_nic_remove(Ctx.ctx, C.uint32_t(domid), &cnic, nil) + ret := C.libxl_device_nic_remove(ctx.ctx, C.uint32_t(domid), &cnic, nil) if ret != 0 { return Error(ret) } @@ -1188,7 +1188,7 @@ func (Ctx *Context) DeviceNicRemove(domid Domid, nic *DeviceNic) error { } // DevicePciAdd is used to passthrough a PCI device to a domain. -func (Ctx *Context) DevicePciAdd(domid Domid, pci *DevicePci) error { +func (ctx *Context) DevicePciAdd(domid Domid, pci *DevicePci) error { var cpci C.libxl_device_pci if err := pci.toC(&cpci); err != nil { @@ -1196,7 +1196,7 @@ func (Ctx *Context) DevicePciAdd(domid Domid, pci *DevicePci) error { } defer C.libxl_device_pci_dispose(&cpci) - ret := C.libxl_device_pci_add(Ctx.ctx, C.uint32_t(domid), &cpci, nil) + ret := C.libxl_device_pci_add(ctx.ctx, C.uint32_t(domid), &cpci, nil) if ret != 0 { return Error(ret) } @@ -1205,7 +1205,7 @@ func (Ctx *Context) DevicePciAdd(domid Domid, pci *DevicePci) error { } // DevicePciRemove removes a PCI device from a domain. -func (Ctx *Context) DevicePciRemove(domid Domid, pci *DevicePci) error { +func (ctx *Context) DevicePciRemove(domid Domid, pci *DevicePci) error { var cpci C.libxl_device_pci if err := pci.toC(&cpci); err != nil { @@ -1213,7 +1213,7 @@ func (Ctx *Context) DevicePciRemove(domid Domid, pci *DevicePci) error { } defer C.libxl_device_pci_dispose(&cpci) - ret := C.libxl_device_pci_remove(Ctx.ctx, C.uint32_t(domid), &cpci, nil) + ret := C.libxl_device_pci_remove(ctx.ctx, C.uint32_t(domid), &cpci, nil) if ret != 0 { return Error(ret) } @@ -1222,7 +1222,7 @@ func (Ctx *Context) DevicePciRemove(domid Domid, pci *DevicePci) error { } // DeviceUsbdevAdd adds a USB device to a domain. -func (Ctx *Context) DeviceUsbdevAdd(domid Domid, usbdev *DeviceUsbdev) error { +func (ctx *Context) DeviceUsbdevAdd(domid Domid, usbdev *DeviceUsbdev) error { var cusbdev C.libxl_device_usbdev if err := usbdev.toC(&cusbdev); err != nil { @@ -1230,7 +1230,7 @@ func (Ctx *Context) DeviceUsbdevAdd(domid Domid, usbdev *DeviceUsbdev) error { } defer C.libxl_device_usbdev_dispose(&cusbdev) - ret := C.libxl_device_usbdev_add(Ctx.ctx, C.uint32_t(domid), &cusbdev, nil) + ret := C.libxl_device_usbdev_add(ctx.ctx, C.uint32_t(domid), &cusbdev, nil) if ret != 0 { return Error(ret) } @@ -1239,7 +1239,7 @@ func (Ctx *Context) DeviceUsbdevAdd(domid Domid, usbdev *DeviceUsbdev) error { } // DeviceUsbdevRemove removes a USB device from a domain. -func (Ctx *Context) DeviceUsbdevRemove(domid Domid, usbdev *DeviceUsbdev) error { +func (ctx *Context) DeviceUsbdevRemove(domid Domid, usbdev *DeviceUsbdev) error { var cusbdev C.libxl_device_usbdev if err := usbdev.toC(&cusbdev); err != nil { @@ -1247,7 +1247,7 @@ func (Ctx *Context) DeviceUsbdevRemove(domid Domid, usbdev *DeviceUsbdev) error } defer C.libxl_device_usbdev_dispose(&cusbdev) - ret := C.libxl_device_usbdev_remove(Ctx.ctx, C.uint32_t(domid), &cusbdev, nil) + ret := C.libxl_device_usbdev_remove(ctx.ctx, C.uint32_t(domid), &cusbdev, nil) if ret != 0 { return Error(ret) } @@ -1256,7 +1256,7 @@ func (Ctx *Context) DeviceUsbdevRemove(domid Domid, usbdev *DeviceUsbdev) error } // DomainCreateNew creates a new domain. -func (Ctx *Context) DomainCreateNew(config *DomainConfig) (Domid, error) { +func (ctx *Context) DomainCreateNew(config *DomainConfig) (Domid, error) { var cdomid C.uint32_t var cconfig C.libxl_domain_config err := config.toC(&cconfig) @@ -1265,7 +1265,7 @@ func (Ctx *Context) DomainCreateNew(config *DomainConfig) (Domid, error) { } defer C.libxl_domain_config_dispose(&cconfig) - ret := C.libxl_domain_create_new(Ctx.ctx, &cconfig, &cdomid, nil, nil) + ret := C.libxl_domain_create_new(ctx.ctx, &cconfig, &cdomid, nil, nil) if ret != 0 { return Domid(0), Error(ret) } -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:32:40 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:32:40 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146655.269965 (Exim 4.92) (envelope-from ) id 1lwMf2-0001Ey-Uv; Thu, 24 Jun 2021 10:32:40 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146655.269965; Thu, 24 Jun 2021 10:32:40 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMf2-0001Eq-Ry; Thu, 24 Jun 2021 10:32:40 +0000 Received: by outflank-mailman (input) for mailman id 146655; Thu, 24 Jun 2021 10:32:38 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMf0-0001Ed-Ot for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:32:38 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMf0-0006G3-OF for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:32:38 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwMf0-0007H9-NZ for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:32:38 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=2fo9+dmBJYdkVMvLrsCBnJSR0j+1k/uI3o4MBAjOLwU=; b=pQh4aIgdRViwT7A0JJFzLwkYdi 1D17fBzxin/h382+SoAANU+H8Y7IqNqXMSvQWHi96NFTg9UMf2s9Qdu1pEbB4knXnJHlM2ZMQ8Kw9 UNGOvfTeFTF75yQDlXtMm+vnCd/TN4n3kU9AECSFaN0ExHg+vGl9qHh8/doRMnUzYnkM=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] golang/xenlight: add DomainDestroy wrapper Message-Id: Date: Thu, 24 Jun 2021 10:32:38 +0000 commit 9b6d865e0af56e376740ba03b1ccdf316362a71e Author: Nick Rosbrook AuthorDate: Mon May 24 16:36:50 2021 -0400 Commit: George Dunlap CommitDate: Mon Jun 21 16:49:08 2021 +0100 golang/xenlight: add DomainDestroy wrapper Add a wrapper around libxl_domain_destroy. Signed-off-by: Nick Rosbrook Reviewed-by: George Dunlap --- tools/golang/xenlight/xenlight.go | 10 ++++++++++ 1 file changed, 10 insertions(+) diff --git a/tools/golang/xenlight/xenlight.go b/tools/golang/xenlight/xenlight.go index fc3eb0bf3f..f605f17613 100644 --- a/tools/golang/xenlight/xenlight.go +++ b/tools/golang/xenlight/xenlight.go @@ -1272,3 +1272,13 @@ func (ctx *Context) DomainCreateNew(config *DomainConfig) (Domid, error) { return Domid(cdomid), nil } + +// DomainDestroy destroys a domain given a domid. +func (ctx *Context) DomainDestroy(domid Domid) error { + ret := C.libxl_domain_destroy(ctx.ctx, C.uint32_t(domid), nil) + if ret != 0 { + return Error(ret) + } + + return nil +} -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:32:51 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:32:51 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146656.269968 (Exim 4.92) (envelope-from ) id 1lwMfD-0001Hs-08; Thu, 24 Jun 2021 10:32:51 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146656.269968; Thu, 24 Jun 2021 10:32:50 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMfC-0001Hk-TU; Thu, 24 Jun 2021 10:32:50 +0000 Received: by outflank-mailman (input) for mailman id 146656; Thu, 24 Jun 2021 10:32:48 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMfA-0001HU-Ro for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:32:48 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMfA-0006GE-R9 for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:32:48 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwMfA-0007Hu-QR for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:32:48 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=jol5VRDhw/lJy7s39DXQbDgmeiigR5ab6rG/7U2NbhA=; b=NqSXf4Jhs+s/rfHceWA973Qp8+ DXzYlshzH0lxefkl8ayktqc94phXe+JIHs0omcQA7rvAkfrd3cmz0u/7ijHKrU18dvgVdN1MbO5S+ gerJT2HK7Ry8Rpnm+d7AvRt0xK0gGB/XRPyrZ71eoyblFxfSp+EAYR8Bi0zY4CJgu3D4=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] golang/xenlight: add SendTrigger wrapper Message-Id: Date: Thu, 24 Jun 2021 10:32:48 +0000 commit 1d95fd75df18bf25cb445feb47caf62da25c00e8 Author: Nick Rosbrook AuthorDate: Mon May 24 16:36:51 2021 -0400 Commit: George Dunlap CommitDate: Mon Jun 21 16:49:08 2021 +0100 golang/xenlight: add SendTrigger wrapper Add a warpper around libxl_send_trigger. Signed-off-by: Nick Rosbrook Reviewed-by: George Dunlap --- tools/golang/xenlight/xenlight.go | 11 +++++++++++ 1 file changed, 11 insertions(+) diff --git a/tools/golang/xenlight/xenlight.go b/tools/golang/xenlight/xenlight.go index f605f17613..a651554d0c 100644 --- a/tools/golang/xenlight/xenlight.go +++ b/tools/golang/xenlight/xenlight.go @@ -1282,3 +1282,14 @@ func (ctx *Context) DomainDestroy(domid Domid) error { return nil } + +// SendTrigger sends a Trigger to the domain specified by domid. +func (ctx *Context) SendTrigger(domid Domid, trigger Trigger, vcpuid int) error { + ret := C.libxl_send_trigger(ctx.ctx, C.uint32_t(domid), + C.libxl_trigger(trigger), C.uint32_t(vcpuid), nil) + if ret != 0 { + return Error(ret) + } + + return nil +} -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:33:00 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:33:00 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146658.269973 (Exim 4.92) (envelope-from ) id 1lwMfM-0001LR-4f; Thu, 24 Jun 2021 10:33:00 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146658.269973; Thu, 24 Jun 2021 10:33:00 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMfM-0001LG-1K; Thu, 24 Jun 2021 10:33:00 +0000 Received: by outflank-mailman (input) for mailman id 146658; Thu, 24 Jun 2021 10:32:59 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMfK-0001L9-Ug for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:32:58 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMfK-0006Ht-U1 for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:32:58 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwMfK-0007Ib-TL for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:32:58 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=VU65F4X/JAmNIu57TNz5uY4opIqVIQtzEj4qyY74Jzo=; b=2x6Aq5MbJYp0442xxFG3v9/1Tx 8uPpEAs/I+ZB4Sy1a7CXbxpgn1e1BAn1SGhxNjJe2xHjkTubazfBkA85gmgCjIs31tsZE+hfnb/si hsKaTkUbEaac4rdQfNE/Pvtr1+EfuS8Q963APVvGkZN6XSNaRUG2IkUAw05RJ8xydG3I=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] golang/xenlight: do not negate ret when converting to Error Message-Id: Date: Thu, 24 Jun 2021 10:32:58 +0000 commit c9b59f9032d41be8bade8a25d9148cf6ed203704 Author: Nick Rosbrook AuthorDate: Mon May 24 16:36:52 2021 -0400 Commit: George Dunlap CommitDate: Mon Jun 21 16:49:08 2021 +0100 golang/xenlight: do not negate ret when converting to Error Commit 871e51d2d4 changed the sign on the xenlight error types (making the values negative, same as the C-generated constants), but failed to remove the code changing the sign before casting to Error(). This results in error strings like "libxl error: ", rather than the correct message. Fix all occurrances of this by running: gofmt -w -r 'Error(-ret) -> Error(ret)' xenlight.go from tools/golang/xenlight. Signed-off-by: Nick Rosbrook Acked-by: George Dunlap --- tools/golang/xenlight/xenlight.go | 46 +++++++++++++++++++-------------------- 1 file changed, 23 insertions(+), 23 deletions(-) diff --git a/tools/golang/xenlight/xenlight.go b/tools/golang/xenlight/xenlight.go index a651554d0c..a45c636952 100644 --- a/tools/golang/xenlight/xenlight.go +++ b/tools/golang/xenlight/xenlight.go @@ -583,7 +583,7 @@ func SchedulerFromString(name string) (s Scheduler, err error) { ret := C.libxl_scheduler_from_string(cname, &cs) if ret != 0 { - err = Error(-ret) + err = Error(ret) return } @@ -622,7 +622,7 @@ func (ctx *Context) CpupoolInfo(Poolid uint32) (pool Cpupoolinfo, err error) { ret := C.libxl_cpupool_info(ctx.ctx, &c_cpupool, C.uint32_t(Poolid)) if ret != 0 { - err = Error(-ret) + err = Error(ret) return } defer C.libxl_cpupoolinfo_dispose(&c_cpupool) @@ -656,7 +656,7 @@ func (ctx *Context) CpupoolCreate(Name string, Scheduler Scheduler, Cpumap Bitma ret := C.libxl_cpupool_create(ctx.ctx, name, C.libxl_scheduler(Scheduler), cbm, &uuid, &poolid) if ret != 0 { - err = Error(-ret) + err = Error(ret) return } @@ -669,7 +669,7 @@ func (ctx *Context) CpupoolCreate(Name string, Scheduler Scheduler, Cpumap Bitma func (ctx *Context) CpupoolDestroy(Poolid uint32) (err error) { ret := C.libxl_cpupool_destroy(ctx.ctx, C.uint32_t(Poolid)) if ret != 0 { - err = Error(-ret) + err = Error(ret) return } @@ -680,7 +680,7 @@ func (ctx *Context) CpupoolDestroy(Poolid uint32) (err error) { func (ctx *Context) CpupoolCpuadd(Poolid uint32, Cpu int) (err error) { ret := C.libxl_cpupool_cpuadd(ctx.ctx, C.uint32_t(Poolid), C.int(Cpu)) if ret != 0 { - err = Error(-ret) + err = Error(ret) return } @@ -698,7 +698,7 @@ func (ctx *Context) CpupoolCpuaddCpumap(Poolid uint32, Cpumap Bitmap) (err error ret := C.libxl_cpupool_cpuadd_cpumap(ctx.ctx, C.uint32_t(Poolid), &cbm) if ret != 0 { - err = Error(-ret) + err = Error(ret) return } @@ -709,7 +709,7 @@ func (ctx *Context) CpupoolCpuaddCpumap(Poolid uint32, Cpumap Bitmap) (err error func (ctx *Context) CpupoolCpuremove(Poolid uint32, Cpu int) (err error) { ret := C.libxl_cpupool_cpuremove(ctx.ctx, C.uint32_t(Poolid), C.int(Cpu)) if ret != 0 { - err = Error(-ret) + err = Error(ret) return } @@ -727,7 +727,7 @@ func (ctx *Context) CpupoolCpuremoveCpumap(Poolid uint32, Cpumap Bitmap) (err er ret := C.libxl_cpupool_cpuremove_cpumap(ctx.ctx, C.uint32_t(Poolid), &cbm) if ret != 0 { - err = Error(-ret) + err = Error(ret) return } @@ -741,7 +741,7 @@ func (ctx *Context) CpupoolRename(Name string, Poolid uint32) (err error) { ret := C.libxl_cpupool_rename(ctx.ctx, name, C.uint32_t(Poolid)) if ret != 0 { - err = Error(-ret) + err = Error(ret) return } @@ -754,7 +754,7 @@ func (ctx *Context) CpupoolCpuaddNode(Poolid uint32, Node int) (Cpus int, err er ret := C.libxl_cpupool_cpuadd_node(ctx.ctx, C.uint32_t(Poolid), C.int(Node), &ccpus) if ret != 0 { - err = Error(-ret) + err = Error(ret) return } @@ -769,7 +769,7 @@ func (ctx *Context) CpupoolCpuremoveNode(Poolid uint32, Node int) (Cpus int, err ret := C.libxl_cpupool_cpuremove_node(ctx.ctx, C.uint32_t(Poolid), C.int(Node), &ccpus) if ret != 0 { - err = Error(-ret) + err = Error(ret) return } @@ -782,7 +782,7 @@ func (ctx *Context) CpupoolCpuremoveNode(Poolid uint32, Node int) (Cpus int, err func (ctx *Context) CpupoolMovedomain(Poolid uint32, Id Domid) (err error) { ret := C.libxl_cpupool_movedomain(ctx.ctx, C.uint32_t(Poolid), C.uint32_t(Id)) if ret != 0 { - err = Error(-ret) + err = Error(ret) return } @@ -943,7 +943,7 @@ func (bm Bitmap) String() (s string) { func (ctx *Context) GetMaxCpus() (maxCpus int, err error) { ret := C.libxl_get_max_cpus(ctx.ctx) if ret < 0 { - err = Error(-ret) + err = Error(ret) return } maxCpus = int(ret) @@ -954,7 +954,7 @@ func (ctx *Context) GetMaxCpus() (maxCpus int, err error) { func (ctx *Context) GetOnlineCpus() (onCpus int, err error) { ret := C.libxl_get_online_cpus(ctx.ctx) if ret < 0 { - err = Error(-ret) + err = Error(ret) return } onCpus = int(ret) @@ -965,7 +965,7 @@ func (ctx *Context) GetOnlineCpus() (onCpus int, err error) { func (ctx *Context) GetMaxNodes() (maxNodes int, err error) { ret := C.libxl_get_max_nodes(ctx.ctx) if ret < 0 { - err = Error(-ret) + err = Error(ret) return } maxNodes = int(ret) @@ -978,7 +978,7 @@ func (ctx *Context) GetFreeMemory() (memkb uint64, err error) { ret := C.libxl_get_free_memory(ctx.ctx, &cmem) if ret < 0 { - err = Error(-ret) + err = Error(ret) return } @@ -1023,7 +1023,7 @@ func (ctx *Context) DomainInfo(Id Domid) (di *Dominfo, err error) { ret := C.libxl_domain_info(ctx.ctx, &cdi, C.uint32_t(Id)) if ret != 0 { - err = Error(-ret) + err = Error(ret) return } @@ -1036,7 +1036,7 @@ func (ctx *Context) DomainUnpause(Id Domid) (err error) { ret := C.libxl_domain_unpause(ctx.ctx, C.uint32_t(Id), nil) if ret != 0 { - err = Error(-ret) + err = Error(ret) } return } @@ -1046,7 +1046,7 @@ func (ctx *Context) DomainPause(id Domid) (err error) { ret := C.libxl_domain_pause(ctx.ctx, C.uint32_t(id), nil) if ret != 0 { - err = Error(-ret) + err = Error(ret) } return } @@ -1056,7 +1056,7 @@ func (ctx *Context) DomainShutdown(id Domid) (err error) { ret := C.libxl_domain_shutdown(ctx.ctx, C.uint32_t(id), nil) if ret != 0 { - err = Error(-ret) + err = Error(ret) } return } @@ -1066,7 +1066,7 @@ func (ctx *Context) DomainReboot(id Domid) (err error) { ret := C.libxl_domain_reboot(ctx.ctx, C.uint32_t(id), nil) if ret != 0 { - err = Error(-ret) + err = Error(ret) } return } @@ -1129,7 +1129,7 @@ func (ctx *Context) ConsoleGetTty(id Domid, consNum int, conType ConsoleType) (p var cpath *C.char ret := C.libxl_console_get_tty(ctx.ctx, C.uint32_t(id), C.int(consNum), C.libxl_console_type(conType), &cpath) if ret != 0 { - err = Error(-ret) + err = Error(ret) return } defer C.free(unsafe.Pointer(cpath)) @@ -1144,7 +1144,7 @@ func (ctx *Context) PrimaryConsoleGetTty(domid uint32) (path string, err error) var cpath *C.char ret := C.libxl_primary_console_get_tty(ctx.ctx, C.uint32_t(domid), &cpath) if ret != 0 { - err = Error(-ret) + err = Error(ret) return } defer C.free(unsafe.Pointer(cpath)) -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:33:10 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:33:10 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146659.269977 (Exim 4.92) (envelope-from ) id 1lwMfW-0001OJ-62; Thu, 24 Jun 2021 10:33:10 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146659.269977; Thu, 24 Jun 2021 10:33:10 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMfW-0001O9-2r; Thu, 24 Jun 2021 10:33:10 +0000 Received: by outflank-mailman (input) for mailman id 146659; Thu, 24 Jun 2021 10:33:09 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMfV-0001Nw-1M for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:33:09 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMfV-0006Ic-0d for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:33:09 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwMfU-0007Jj-W5 for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:33:08 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=N2CcNVFatsN2lj1LOQT0dTAZajtwxwL1icr4ph2LcOQ=; b=06dNjbdfIHvoOWgR+D9PXgdMMX bOE3luWJ7I1XvQc07p4aHkE+FSS3+ydFWDtNn9mXI5P1tg9eg3EoeW54aT0MO4gD22zVrbVHM74GM vR5iMw++ehgQiACvqyap5tNVF1jTy97TiN4WCMDovdOtjKbAoMzSmFGWIFcEMHb7A4Ec=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] Revert "tools/firmware/ovmf: Use OvmfXen platform file is exist" Message-Id: Date: Thu, 24 Jun 2021 10:33:08 +0000 commit c7691f5e340f3b579d40c77548f63133cdf5aac7 Author: Andrew Cooper AuthorDate: Tue Jun 22 16:14:53 2021 +0100 Commit: Andrew Cooper CommitDate: Tue Jun 22 18:00:08 2021 +0100 Revert "tools/firmware/ovmf: Use OvmfXen platform file is exist" This reverts commit aad7b5c11d51d57659978e04702ac970906894e8. The change from OvmfX64 to OvmfXen causes a change in behaviour, whereby OvmfXen maps its shared info page at the top of address space. When trying to migrate such a domain, XENMEM_maximum_gpfn returns a very large value. This has uncovered multiple issues: 1) The userspace hypercall wrappers truncate all return values to int on Linux and Solaris, even on 64bit. This needs fixing in libxenctrl. 2) 32bit toolstacks can't migrate any domain with RAM above the 2^40 mark, because of virtual address constraints. This needs fixing in OVMF. Fixes for both of these aren't completely trivial. Revert the change to unblock staging in the meantime. Signed-off-by: Andrew Cooper Acked-by: Anthony PERARD --- tools/firmware/ovmf-makefile | 10 ++-------- 1 file changed, 2 insertions(+), 8 deletions(-) diff --git a/tools/firmware/ovmf-makefile b/tools/firmware/ovmf-makefile index 637ee509c3..55f9992145 100644 --- a/tools/firmware/ovmf-makefile +++ b/tools/firmware/ovmf-makefile @@ -17,14 +17,8 @@ all: build .PHONY: build build: if test -e .git ; then $(GIT) submodule update --init --recursive ; fi - set -ex; \ - if test -e OvmfPkg/OvmfXen.dsc; then \ - OvmfPkg/build.sh -a X64 -b $(TARGET) -n 4 -p OvmfPkg/OvmfXen.dsc; \ - cp Build/OvmfXen/$(TARGET)_GCC*/FV/OVMF.fd ovmf.bin; \ - else \ - OvmfPkg/build.sh -a X64 -b $(TARGET) -n 4; \ - cp Build/OvmfX64/$(TARGET)_GCC*/FV/OVMF.fd ovmf.bin; \ - fi + OvmfPkg/build.sh -a X64 -b $(TARGET) -n 4 + cp Build/OvmfX64/$(TARGET)_GCC*/FV/OVMF.fd ovmf.bin .PHONY: clean clean: -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:44:09 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:44:09 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146667.269992 (Exim 4.92) (envelope-from ) id 1lwMq5-0002tk-DS; Thu, 24 Jun 2021 10:44:05 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146667.269992; Thu, 24 Jun 2021 10:44:05 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMq5-0002tc-AE; Thu, 24 Jun 2021 10:44:05 +0000 Received: by outflank-mailman (input) for mailman id 146667; Thu, 24 Jun 2021 10:44:04 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMq4-0002tU-Et for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:44:04 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMq4-0006TB-E0 for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:44:04 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwMq4-0000GJ-DN for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:44:04 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=V0uepzjH8611hYu0WDA+k58io1nhGfB/C63NnsJQNtc=; b=oF28iuKLY1WVjpp4nyBo3HwPnG pQC9VeJmoQMB+FYkuozw9/8pdDLafcw+yIte0wiZBwJGqur6d0W+2L5OyYD1Ol6Q9twT3lOvBmg42 xI1k6ye0gqCcHRlfERkvLFjyUvB2eHHoVjycvUz5KoJI1EBDIttIHOLgq8rONMl7j5gE=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] MAINTAINERS: Add myself as reviewers for tools/xenstore Message-Id: Date: Thu, 24 Jun 2021 10:44:04 +0000 commit 82b1b19469cb5595c589898c26b5f33f37f90963 Author: Julien Grall AuthorDate: Thu Jun 24 09:02:56 2021 +0100 Commit: Julien Grall CommitDate: Thu Jun 24 09:02:56 2021 +0100 MAINTAINERS: Add myself as reviewers for tools/xenstore I would like to help reviewing Xenstored patches. It is more convenient to find them if I am CCed. Signed-off-by: Julien Grall Acked-by: Juergen Gross --- MAINTAINERS | 1 + 1 file changed, 1 insertion(+) diff --git a/MAINTAINERS b/MAINTAINERS index 39750bb75d..dd8c011456 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -628,6 +628,7 @@ XENSTORE M: Ian Jackson M: Wei Liu M: Juergen Gross +R: Julien Grall S: Supported F: tools/xenstore/ -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:44:15 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:44:15 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146668.269995 (Exim 4.92) (envelope-from ) id 1lwMqF-0002wr-Eg; Thu, 24 Jun 2021 10:44:15 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146668.269995; Thu, 24 Jun 2021 10:44:15 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMqF-0002wj-Bl; Thu, 24 Jun 2021 10:44:15 +0000 Received: by outflank-mailman (input) for mailman id 146668; Thu, 24 Jun 2021 10:44:14 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMqE-0002wC-Hu for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:44:14 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMqE-0006Tg-HA for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:44:14 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwMqE-0000HC-GN for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:44:14 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=iS6a7Y0gTDktHUARgP9v0C6Lc3pzoxl2mtP644UIfmk=; b=JSJTEmxu8Y9jJ0bEacR0CQ7yda iF1BoIy6ZqTOfoAuiRGyTbM94s9FML/TRCCdYQpQNsOTgA/N7N833qLs2WM10ofI9Y1Y7k4rh3Gc9 8clBxQohSNQcUoCxm/o7y5RMwnNyUxzwGy0nklHFkaJEFj1YTzjo3r0shvNZ3UPrzRro=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] tools/xenstored: Introduce lu_get_connection() and use it Message-Id: Date: Thu, 24 Jun 2021 10:44:14 +0000 commit 3ce271ef535cfe3c0839fdea56f33abf5fdbf23c Author: Julien Grall AuthorDate: Thu Jun 24 09:02:59 2021 +0100 Commit: Julien Grall CommitDate: Thu Jun 24 09:02:59 2021 +0100 tools/xenstored: Introduce lu_get_connection() and use it At the moment, dump_state_buffered_data() is taking two connections in parameters (one is the connection to dump, the other is the connection used to request LU). The naming doesn't help to distinguish (c vs conn) them and this already lead to several mistake while modifying the function. To remove the confusion, introduce an help lu_get_connection() that will return the connection used to request LU and use it in place of the existing parameter. Signed-off-by: Julien Grall Reviewed-by: Luca Fancellu Reviewed-by: Juergen Gross --- tools/xenstore/xenstored_control.c | 13 ++++++++++++- tools/xenstore/xenstored_control.h | 2 ++ tools/xenstore/xenstored_core.c | 7 +++---- tools/xenstore/xenstored_core.h | 1 - tools/xenstore/xenstored_domain.c | 6 +++--- tools/xenstore/xenstored_domain.h | 2 +- 6 files changed, 21 insertions(+), 10 deletions(-) diff --git a/tools/xenstore/xenstored_control.c b/tools/xenstore/xenstored_control.c index 0d57f9f940..d08a2b9614 100644 --- a/tools/xenstore/xenstored_control.c +++ b/tools/xenstore/xenstored_control.c @@ -104,6 +104,17 @@ static const char *lu_begin(struct connection *conn) return NULL; } + +struct connection *lu_get_connection(void) +{ + return lu_status ? lu_status->conn : NULL; +} + +#else +struct connection *lu_get_connection(void) +{ + return NULL; +} #endif struct cmd_s { @@ -516,7 +527,7 @@ static const char *lu_dump_state(const void *ctx, struct connection *conn) ret = dump_state_global(fp); if (ret) goto out; - ret = dump_state_connections(fp, conn); + ret = dump_state_connections(fp); if (ret) goto out; ret = dump_state_special_nodes(fp); diff --git a/tools/xenstore/xenstored_control.h b/tools/xenstore/xenstored_control.h index aac61f0590..6842b8d887 100644 --- a/tools/xenstore/xenstored_control.h +++ b/tools/xenstore/xenstored_control.h @@ -18,3 +18,5 @@ int do_control(struct connection *conn, struct buffered_data *in); void lu_read_state(void); + +struct connection *lu_get_connection(void); diff --git a/tools/xenstore/xenstored_core.c b/tools/xenstore/xenstored_core.c index 883a1a582a..607187361d 100644 --- a/tools/xenstore/xenstored_core.c +++ b/tools/xenstore/xenstored_core.c @@ -2369,14 +2369,13 @@ const char *dump_state_global(FILE *fp) /* Called twice: first with fp == NULL to get length, then for writing data. */ const char *dump_state_buffered_data(FILE *fp, const struct connection *c, - const struct connection *conn, struct xs_state_connection *sc) { unsigned int len = 0, used; struct buffered_data *out, *in = c->in; bool partial = true; - if (in && c != conn) { + if (in && c != lu_get_connection()) { len = in->inhdr ? in->used : sizeof(in->hdr); if (fp && fwrite(&in->hdr, len, 1, fp) != 1) return "Dump read data error"; @@ -2416,8 +2415,8 @@ const char *dump_state_buffered_data(FILE *fp, const struct connection *c, } /* Add "OK" for live-update command. */ - if (c == conn) { - struct xsd_sockmsg msg = conn->in->hdr.msg; + if (c == lu_get_connection()) { + struct xsd_sockmsg msg = c->in->hdr.msg; msg.len = sizeof("OK"); if (fp && fwrite(&msg, sizeof(msg), 1, fp) != 1) diff --git a/tools/xenstore/xenstored_core.h b/tools/xenstore/xenstored_core.h index bb36111ecc..89ce155e75 100644 --- a/tools/xenstore/xenstored_core.h +++ b/tools/xenstore/xenstored_core.h @@ -269,7 +269,6 @@ void set_tdb_key(const char *name, TDB_DATA *key); const char *dump_state_global(FILE *fp); const char *dump_state_buffered_data(FILE *fp, const struct connection *c, - const struct connection *conn, struct xs_state_connection *sc); const char *dump_state_nodes(FILE *fp, const void *ctx); const char *dump_state_node_perms(FILE *fp, const struct xs_permissions *perms, diff --git a/tools/xenstore/xenstored_domain.c b/tools/xenstore/xenstored_domain.c index 322b0dbca4..6d8d29cbe4 100644 --- a/tools/xenstore/xenstored_domain.c +++ b/tools/xenstore/xenstored_domain.c @@ -1183,7 +1183,7 @@ void wrl_apply_debit_trans_commit(struct connection *conn) wrl_apply_debit_actual(conn->domain); } -const char *dump_state_connections(FILE *fp, struct connection *conn) +const char *dump_state_connections(FILE *fp) { const char *ret = NULL; unsigned int conn_id = 1; @@ -1209,7 +1209,7 @@ const char *dump_state_connections(FILE *fp, struct connection *conn) sc.spec.socket_fd = c->fd; } - ret = dump_state_buffered_data(NULL, c, conn, &sc); + ret = dump_state_buffered_data(NULL, c, &sc); if (ret) return ret; head.length += sc.data_in_len + sc.data_out_len; @@ -1219,7 +1219,7 @@ const char *dump_state_connections(FILE *fp, struct connection *conn) if (fwrite(&sc, offsetof(struct xs_state_connection, data), 1, fp) != 1) return "Dump connection state error"; - ret = dump_state_buffered_data(fp, c, conn, NULL); + ret = dump_state_buffered_data(fp, c, NULL); if (ret) return ret; ret = dump_state_align(fp); diff --git a/tools/xenstore/xenstored_domain.h b/tools/xenstore/xenstored_domain.h index cc5147d7e7..62ee471ea6 100644 --- a/tools/xenstore/xenstored_domain.h +++ b/tools/xenstore/xenstored_domain.h @@ -101,7 +101,7 @@ void wrl_log_periodic(struct wrl_timestampt now); void wrl_apply_debit_direct(struct connection *conn); void wrl_apply_debit_trans_commit(struct connection *conn); -const char *dump_state_connections(FILE *fp, struct connection *conn); +const char *dump_state_connections(FILE *fp); const char *dump_state_special_nodes(FILE *fp); void read_state_connection(const void *ctx, const void *state); -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:44:25 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:44:25 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146669.270000 (Exim 4.92) (envelope-from ) id 1lwMqP-00030u-Gj; Thu, 24 Jun 2021 10:44:25 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146669.270000; Thu, 24 Jun 2021 10:44:25 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMqP-00030m-DK; Thu, 24 Jun 2021 10:44:25 +0000 Received: by outflank-mailman (input) for mailman id 146669; Thu, 24 Jun 2021 10:44:24 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMqO-00030a-LM for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:44:24 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMqO-0006Tq-KY for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:44:24 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwMqO-0000IE-Ja for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:44:24 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=CAsx6+KHKdIsYsor55n+b+9JCU0CbSjRKHB0WNQ+a2s=; b=2L6NjTZLXJ4U5pVZPSd/53lhcN BIwEbBynxDllacUfvp55M0ueQnEster9xquPhN4IZYRTjudYctffceKnXEpzpTrp06jBDxDk5keSE er1be08HTx1pJsVpLyV7bu568uTsjE9ls5Cg7ZAJlcFUiebyk7BUPXXmdzUcV+TUQt4U=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] tools/xenstore: Don't assume conn->in points to the LU request Message-Id: Date: Thu, 24 Jun 2021 10:44:24 +0000 commit 65f19ed62aa1efa75c0dfccbf5d0514daa85be18 Author: Julien Grall AuthorDate: Thu Jun 24 09:06:58 2021 +0100 Commit: Julien Grall CommitDate: Thu Jun 24 09:06:58 2021 +0100 tools/xenstore: Don't assume conn->in points to the LU request call_delayed() is currently assuming that conn->in is NULL when handling delayed request. However, the connection is not paused. Therefore new request can be processed and conn->in may be non-NULL if we have only received a partial request. Furthermore, as we overwrite conn->in, the current partial request will not be transferred. This will result to corrupt the connection. Rather than updating conn->in, stash the LU request in lu_status and let each callback for delayed request to update conn->in when necessary. To keep a sane interface, the code to write the "OK" response the LU request is moved in xenstored_core.c. Fixes: c5ca1404b4 ("tools/xenstore: add support for delaying execution of a xenstore request") Fixes: ed6eebf17d ("tools/xenstore: dump the xenstore state for live update") Signed-off-by: Julien Grall Reviewed-by: Juergen Gross --- tools/xenstore/xenstored_control.c | 41 ++++++++++++++++++++++++++++++++++++-- tools/xenstore/xenstored_control.h | 3 +++ tools/xenstore/xenstored_core.c | 21 ++++++------------- 3 files changed, 48 insertions(+), 17 deletions(-) diff --git a/tools/xenstore/xenstored_control.c b/tools/xenstore/xenstored_control.c index d08a2b9614..7acc2d134f 100644 --- a/tools/xenstore/xenstored_control.c +++ b/tools/xenstore/xenstored_control.c @@ -50,6 +50,9 @@ struct live_update { /* For verification the correct connection is acting. */ struct connection *conn; + /* Pointer to the command used to request LU */ + struct buffered_data *in; + #ifdef __MINIOS__ void *kernel; unsigned int kernel_size; @@ -100,6 +103,7 @@ static const char *lu_begin(struct connection *conn) if (!lu_status) return "Allocation failure."; lu_status->conn = conn; + lu_status->in = conn->in; talloc_set_destructor(lu_status, lu_destroy); return NULL; @@ -110,11 +114,34 @@ struct connection *lu_get_connection(void) return lu_status ? lu_status->conn : NULL; } +unsigned int lu_write_response(FILE *fp) +{ + struct xsd_sockmsg msg; + + assert(lu_status); + + msg = lu_status->in->hdr.msg; + + msg.len = sizeof("OK"); + if (fp && fwrite(&msg, sizeof(msg), 1, fp) != 1) + return 0; + if (fp && fwrite("OK", msg.len, 1, fp) != 1) + return 0; + + return sizeof(msg) + msg.len; +} + #else struct connection *lu_get_connection(void) { return NULL; } + +unsigned int lu_write_response(FILE *fp) +{ + /* Unsupported */ + return 0; +} #endif struct cmd_s { @@ -658,6 +685,8 @@ static bool do_lu_start(struct delayed_request *req) { time_t now = time(NULL); const char *ret; + struct buffered_data *saved_in; + struct connection *conn = lu_status->conn; if (!lu_check_lu_allowed()) { if (now < lu_status->started_at + lu_status->timeout) @@ -668,8 +697,9 @@ static bool do_lu_start(struct delayed_request *req) } } + assert(req->in == lu_status->in); /* Dump out internal state, including "OK" for live update. */ - ret = lu_dump_state(req->in, lu_status->conn); + ret = lu_dump_state(req->in, conn); if (!ret) { /* Perform the activation of new binary. */ ret = lu_activate_binary(req->in); @@ -677,7 +707,14 @@ static bool do_lu_start(struct delayed_request *req) /* We will reach this point only in case of failure. */ out: - send_reply(lu_status->conn, XS_CONTROL, ret, strlen(ret) + 1); + /* + * send_reply() will send the response for conn->in. Save the current + * conn->in and restore it afterwards. + */ + saved_in = conn->in; + conn->in = req->in; + send_reply(conn, XS_CONTROL, ret, strlen(ret) + 1); + conn->in = saved_in; talloc_free(lu_status); return true; diff --git a/tools/xenstore/xenstored_control.h b/tools/xenstore/xenstored_control.h index 6842b8d887..27d7f19e4b 100644 --- a/tools/xenstore/xenstored_control.h +++ b/tools/xenstore/xenstored_control.h @@ -20,3 +20,6 @@ int do_control(struct connection *conn, struct buffered_data *in); void lu_read_state(void); struct connection *lu_get_connection(void); + +/* Write the "OK" response for the live-update command */ +unsigned int lu_write_response(FILE *fp); diff --git a/tools/xenstore/xenstored_core.c b/tools/xenstore/xenstored_core.c index 607187361d..4b6509b90d 100644 --- a/tools/xenstore/xenstored_core.c +++ b/tools/xenstore/xenstored_core.c @@ -270,17 +270,12 @@ static int undelay_request(void *_req) return 0; } -static void call_delayed(struct connection *conn, struct delayed_request *req) +static void call_delayed(struct delayed_request *req) { - assert(conn->in == NULL); - conn->in = req->in; - if (req->func(req)) { undelay_request(req); talloc_set_destructor(req, NULL); } - - conn->in = NULL; } int delay_request(struct connection *conn, struct buffered_data *in, @@ -2342,7 +2337,7 @@ int main(int argc, char *argv[]) list_for_each_entry_safe(req, tmp, &conn->delayed, list) - call_delayed(conn, req); + call_delayed(req); } } @@ -2375,7 +2370,7 @@ const char *dump_state_buffered_data(FILE *fp, const struct connection *c, struct buffered_data *out, *in = c->in; bool partial = true; - if (in && c != lu_get_connection()) { + if (in) { len = in->inhdr ? in->used : sizeof(in->hdr); if (fp && fwrite(&in->hdr, len, 1, fp) != 1) return "Dump read data error"; @@ -2416,16 +2411,12 @@ const char *dump_state_buffered_data(FILE *fp, const struct connection *c, /* Add "OK" for live-update command. */ if (c == lu_get_connection()) { - struct xsd_sockmsg msg = c->in->hdr.msg; + unsigned int rc = lu_write_response(fp); - msg.len = sizeof("OK"); - if (fp && fwrite(&msg, sizeof(msg), 1, fp) != 1) + if (!rc) return "Dump buffered data error"; - len += sizeof(msg); - if (fp && fwrite("OK", msg.len, 1, fp) != 1) - return "Dump buffered data error"; - len += msg.len; + len += rc; } if (sc) -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:44:35 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:44:35 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146671.270015 (Exim 4.92) (envelope-from ) id 1lwMqZ-0003MC-RG; Thu, 24 Jun 2021 10:44:35 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146671.270015; Thu, 24 Jun 2021 10:44:35 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMqZ-0003M3-No; Thu, 24 Jun 2021 10:44:35 +0000 Received: by outflank-mailman (input) for mailman id 146671; Thu, 24 Jun 2021 10:44:34 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMqY-0003LQ-OJ for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:44:34 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMqY-0006UK-NX for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:44:34 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwMqY-0000JA-Mr for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:44:34 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=RinfPo+kik4wT7mKuBfI+mvISzBRsfB8b7prIJSNktA=; b=dueSXqpRAa9+WnPTETHZIm6aXq wF4biIHdlRAwRQjutM1iKydYqnaNW2yKjYnibRS2z1cQ1Hxsbe2ZEMJlImKv17i9DsOb4Oe2rGbkQ 8BkBU7+rQfsEhf4Q1o6Ml7jDKO3DHGtxr6v77nDcD5qBOej0XIol/5GeDxbvG0sDhwq8=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] tools/xenstored: Limit the number of requests a connection can delay Message-Id: Date: Thu, 24 Jun 2021 10:44:34 +0000 commit 2aa9e00bd3d0431fc99d4baaee2cddbf40bbee61 Author: Julien Grall AuthorDate: Thu Jun 24 09:07:30 2021 +0100 Commit: Julien Grall CommitDate: Thu Jun 24 09:07:30 2021 +0100 tools/xenstored: Limit the number of requests a connection can delay Currently, only liveupdate request can be delayed. The request can only be performed by a privileged connection (e.g. dom0). So it is fine to have no limits. In a follow-up patch we will want to delay request for unprivileged connection as well. So it is best to apply a limit. For now and for simplicity, only a single request can be delayed for a given unprivileged connection. Take the opportunity to tweak the prototype and provide a way to bypass the quota check. This would be useful when the function is called from the restore code. Signed-off-by: Julien Grall Reviewed-by: Luca Fancellu Reviewed-by: Juergen Gross --- tools/xenstore/xenstored_control.c | 2 +- tools/xenstore/xenstored_core.c | 11 ++++++++++- tools/xenstore/xenstored_core.h | 3 ++- 3 files changed, 13 insertions(+), 3 deletions(-) diff --git a/tools/xenstore/xenstored_control.c b/tools/xenstore/xenstored_control.c index 7acc2d134f..1c24d4869e 100644 --- a/tools/xenstore/xenstored_control.c +++ b/tools/xenstore/xenstored_control.c @@ -737,7 +737,7 @@ static const char *lu_start(const void *ctx, struct connection *conn, lu_status->timeout = to; lu_status->started_at = time(NULL); - errno = delay_request(conn, conn->in, do_lu_start, NULL); + errno = delay_request(conn, conn->in, do_lu_start, NULL, false); return NULL; } diff --git a/tools/xenstore/xenstored_core.c b/tools/xenstore/xenstored_core.c index 4b6509b90d..268b0187e6 100644 --- a/tools/xenstore/xenstored_core.c +++ b/tools/xenstore/xenstored_core.c @@ -279,10 +279,19 @@ static void call_delayed(struct delayed_request *req) } int delay_request(struct connection *conn, struct buffered_data *in, - bool (*func)(struct delayed_request *), void *data) + bool (*func)(struct delayed_request *), void *data, + bool no_quota_check) { struct delayed_request *req; + /* + * Only allow one request can be delayed for an unprivileged + * connection. + */ + if (!no_quota_check && domain_is_unprivileged(conn) && + !list_empty(&conn->delayed)) + return ENOSPC; + req = talloc(in, struct delayed_request); if (!req) return ENOMEM; diff --git a/tools/xenstore/xenstored_core.h b/tools/xenstore/xenstored_core.h index 89ce155e75..34839b34f6 100644 --- a/tools/xenstore/xenstored_core.h +++ b/tools/xenstore/xenstored_core.h @@ -213,7 +213,8 @@ char *get_parent(const void *ctx, const char *node); /* Delay a request. */ int delay_request(struct connection *conn, struct buffered_data *in, - bool (*func)(struct delayed_request *), void *data); + bool (*func)(struct delayed_request *), void *data, + bool no_quota_check); /* Tracing infrastructure. */ void trace_create(const void *data, const char *type); -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:44:46 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:44:46 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146672.270019 (Exim 4.92) (envelope-from ) id 1lwMqk-0003Wu-Si; Thu, 24 Jun 2021 10:44:46 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146672.270019; Thu, 24 Jun 2021 10:44:46 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMqk-0003Wm-PS; Thu, 24 Jun 2021 10:44:46 +0000 Received: by outflank-mailman (input) for mailman id 146672; Thu, 24 Jun 2021 10:44:44 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMqi-0003UF-RA for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:44:44 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMqi-0006UY-QJ for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:44:44 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwMqi-0000K3-Pa for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:44:44 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=RfJc75BSP/xuI4MQhInXNIrK0cbmUvlnkf93Q7gss0o=; b=N5THhfcmKgIzdSEvAepC/vyqVg SS3u4Ej0uln02P+sTsfWqMOpp5LU8sqRpj8j/aycsUaoIJcg+q6cOmtV+MbN6PJLf9Lkefs7Sombc H2pyrzzb+ZIlTMSUi/bnijj7xGPq4zMTAO+bbE1ljOelEVaZhC8woo99uaBj24hcmcvo=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] tools/xenstored: xenstored_core.h should include fcntl.h Message-Id: Date: Thu, 24 Jun 2021 10:44:44 +0000 commit a8002238b49c3423c7081dfbac3580651f5242ef Author: Julien Grall AuthorDate: Thu Jun 24 09:07:52 2021 +0100 Commit: Julien Grall CommitDate: Thu Jun 24 09:07:52 2021 +0100 tools/xenstored: xenstored_core.h should include fcntl.h xenstored_core.h will consider live-udpate is not supported if O_CLOEXEC doesn't exist. However, the header doesn't include the one defining O_CLOEXEC (i.e. fcntl.h). This means that depending on the header included, some source file will think Live-Update is not supported. I am not aware of any issue with the existing. Therefore this is just a latent bug so far. Prevent any potential issue by including fcntl.h in xenstored_core.h Fixes: cd831ee438 ("tools/xenstore: handle CLOEXEC flag for local files and pipes") Signed-off-by: Julien Grall Reviewed-by: Luca Fancellu Reviewed-by: Juergen Gross --- tools/xenstore/xenstored_core.h | 1 + 1 file changed, 1 insertion(+) diff --git a/tools/xenstore/xenstored_core.h b/tools/xenstore/xenstored_core.h index 34839b34f6..dac5171569 100644 --- a/tools/xenstore/xenstored_core.h +++ b/tools/xenstore/xenstored_core.h @@ -24,6 +24,7 @@ #include #include +#include #include #include #include -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:44:56 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:44:56 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146675.270023 (Exim 4.92) (envelope-from ) id 1lwMqu-0003iM-Tx; Thu, 24 Jun 2021 10:44:56 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146675.270023; Thu, 24 Jun 2021 10:44:56 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMqu-0003iE-Qz; Thu, 24 Jun 2021 10:44:56 +0000 Received: by outflank-mailman (input) for mailman id 146675; Thu, 24 Jun 2021 10:44:54 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMqs-0003hf-UD for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:44:54 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMqs-0006Us-TI for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:44:54 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwMqs-0000Kk-SY for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:44:54 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=jZpm4K8y/5oGK64BtH/roAGU1Drle8XcH2hlQB2GhNs=; b=0PzO1bdT7kNcZ7FTKVwrncdpXU abG8/R+KCEKN5HvtHPSRwqQt4Ip3kEXM/CeXSlid8dYMbbbOaKKrz/5mH1nFYHCy6FHsGu7a2Vpqk bxmp5r2fzvuDL2AGF3Af/6mwKCbctnIcECpCpDy4DwBhZ9baIMBHUsdqaczL8gRevP+g=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] tools/xenstored: Introduce a wrapper for conn->funcs->can_{read, write} Message-Id: Date: Thu, 24 Jun 2021 10:44:54 +0000 commit 3adfb50315d9b5643bec7a8a5378625f15986200 Author: Julien Grall AuthorDate: Thu Jun 24 09:08:42 2021 +0100 Commit: Julien Grall CommitDate: Thu Jun 24 09:08:42 2021 +0100 tools/xenstored: Introduce a wrapper for conn->funcs->can_{read, write} Currently, the callbacks can_read and can_write are called directly. This doesn't allow us to add generic check and therefore requires duplication. At the moment, one check that could benefit to be common is whether the connection should ignored. The position is slightly different between domain and socket because for the latter we want to check the state of the file descriptor first. In follow-up patches, there will be more potential generic checks. This patch provides wrappers to read/write a connection and move the check ->is_ignored after the callback for everyone. This also requires to replace the direct call to domain_can_read() and domain_can_write() with the new wrapper. At the same time, both functions can now be static. Note that the implementations need to be moved earlier in the file xenstored_domain.c to avoid declaring the prototype. Signed-off-by: Julien Grall Reviewed-by: Luca Fancellu Reviewed-by: Juergen Gross --- tools/xenstore/xenstored_core.c | 18 ++++++++++++++---- tools/xenstore/xenstored_domain.c | 40 +++++++++++++++++---------------------- tools/xenstore/xenstored_domain.h | 4 ---- 3 files changed, 31 insertions(+), 31 deletions(-) diff --git a/tools/xenstore/xenstored_core.c b/tools/xenstore/xenstored_core.c index 268b0187e6..4c0d6e1874 100644 --- a/tools/xenstore/xenstored_core.c +++ b/tools/xenstore/xenstored_core.c @@ -334,6 +334,16 @@ static int destroy_conn(void *_conn) return 0; } +static bool conn_can_read(struct connection *conn) +{ + return conn->funcs->can_read(conn) && !conn->is_ignored; +} + +static bool conn_can_write(struct connection *conn) +{ + return conn->funcs->can_write(conn) && !conn->is_ignored; +} + /* This function returns index inside the array if succeed, -1 if fail */ static int set_fd(int fd, short events) { @@ -396,8 +406,8 @@ static void initialize_fds(int *p_sock_pollfd_idx, int *ptimeout) list_for_each_entry(conn, &connections, list) { if (conn->domain) { wrl_check_timeout(conn->domain, now, ptimeout); - if (domain_can_read(conn) || - (domain_can_write(conn) && + if (conn_can_read(conn) || + (conn_can_write(conn) && !list_empty(&conn->out_list))) *ptimeout = 0; } else { @@ -2325,14 +2335,14 @@ int main(int argc, char *argv[]) if (&next->list != &connections) talloc_increase_ref_count(next); - if (conn->funcs->can_read(conn)) + if (conn_can_read(conn)) handle_input(conn); if (talloc_free(conn) == 0) continue; talloc_increase_ref_count(conn); - if (conn->funcs->can_write(conn)) + if (conn_can_write(conn)) handle_output(conn); if (talloc_free(conn) == 0) continue; diff --git a/tools/xenstore/xenstored_domain.c b/tools/xenstore/xenstored_domain.c index 6d8d29cbe4..47e9107c14 100644 --- a/tools/xenstore/xenstored_domain.c +++ b/tools/xenstore/xenstored_domain.c @@ -172,6 +172,23 @@ static int readchn(struct connection *conn, void *data, unsigned int len) return len; } +static bool domain_can_write(struct connection *conn) +{ + struct xenstore_domain_interface *intf = conn->domain->interface; + + return ((intf->rsp_prod - intf->rsp_cons) != XENSTORE_RING_SIZE); +} + +static bool domain_can_read(struct connection *conn) +{ + struct xenstore_domain_interface *intf = conn->domain->interface; + + if (domain_is_unprivileged(conn) && conn->domain->wrl_credit < 0) + return false; + + return (intf->req_cons != intf->req_prod); +} + static const struct interface_funcs domain_funcs = { .write = writechn, .read = readchn, @@ -290,19 +307,6 @@ void handle_event(void) barf_perror("Failed to write to event fd"); } -bool domain_can_read(struct connection *conn) -{ - struct xenstore_domain_interface *intf = conn->domain->interface; - - if (domain_is_unprivileged(conn) && conn->domain->wrl_credit < 0) - return false; - - if (conn->is_ignored) - return false; - - return (intf->req_cons != intf->req_prod); -} - static bool domid_is_unprivileged(unsigned int domid) { return domid != 0 && domid != priv_domid; @@ -314,16 +318,6 @@ bool domain_is_unprivileged(struct connection *conn) domid_is_unprivileged(conn->domain->domid); } -bool domain_can_write(struct connection *conn) -{ - struct xenstore_domain_interface *intf = conn->domain->interface; - - if (conn->is_ignored) - return false; - - return ((intf->rsp_prod - intf->rsp_cons) != XENSTORE_RING_SIZE); -} - static char *talloc_domain_path(void *context, unsigned int domid) { return talloc_asprintf(context, "/local/domain/%u", domid); diff --git a/tools/xenstore/xenstored_domain.h b/tools/xenstore/xenstored_domain.h index 62ee471ea6..1e929b8f8c 100644 --- a/tools/xenstore/xenstored_domain.h +++ b/tools/xenstore/xenstored_domain.h @@ -51,10 +51,6 @@ void domain_deinit(void); /* Returns the implicit path of a connection (only domains have this) */ const char *get_implicit_path(const struct connection *conn); -/* Can connection attached to domain read/write. */ -bool domain_can_read(struct connection *conn); -bool domain_can_write(struct connection *conn); - bool domain_is_unprivileged(struct connection *conn); /* Remove node permissions for no longer existing domains. */ -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:45:07 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:45:07 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146678.270027 (Exim 4.92) (envelope-from ) id 1lwMr4-0003oM-VL; Thu, 24 Jun 2021 10:45:06 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146678.270027; Thu, 24 Jun 2021 10:45:06 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMr4-0003oE-SV; Thu, 24 Jun 2021 10:45:06 +0000 Received: by outflank-mailman (input) for mailman id 146678; Thu, 24 Jun 2021 10:45:05 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMr3-0003nW-0t for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:45:05 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMr3-0006VP-06 for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:45:05 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwMr2-0000Lo-Va for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:45:04 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=JDKwaHb08ols1PJ6kUSJCA0uYpei/EoW0Oedea4/yD0=; b=Go28/b0zedUsk6OUTM9+DPLo6n suSFoN6nih8ix8XoYAxAPq0l0wNZp9DHhplRGVt3QXkQQIekYvbQSPQDsvC9xciu3THU0wwxaeWvO IKqQ6sIlqQOfYiXGxdkG223vv+W22wpw3P7vEHMAsj3aBndLMzy0EV+RXiVJUPmNdovg=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] tools/xenstored: delay_request: don't assume conn->in == in Message-Id: Date: Thu, 24 Jun 2021 10:45:04 +0000 commit 361f1deb42ec32ea97df7e79f8df12bfc2d0e049 Author: Julien Grall AuthorDate: Thu Jun 24 09:08:56 2021 +0100 Commit: Julien Grall CommitDate: Thu Jun 24 09:08:56 2021 +0100 tools/xenstored: delay_request: don't assume conn->in == in delay_request() is currently assuming that the request delayed is always conn->in. This is currently correct, but it is a call for a latent bug as the function allows the caller to specify any request. To prevent any future surprise, check if the request delayed is the current one. Fixes: c5ca1404b4 ("tools/xenstore: add support for delaying execution of a xenstore request") Signed-off-by: Julien Grall Reviewed-by: Luca Fancellu Reviewed-by: Juergen Gross --- tools/xenstore/xenstored_core.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/tools/xenstore/xenstored_core.c b/tools/xenstore/xenstored_core.c index 4c0d6e1874..3286eb55d3 100644 --- a/tools/xenstore/xenstored_core.c +++ b/tools/xenstore/xenstored_core.c @@ -306,7 +306,9 @@ int delay_request(struct connection *conn, struct buffered_data *in, delayed_requests++; list_add(&req->list, &conn->delayed); - conn->in = NULL; + /* Unlink the request from conn if this is the current one */ + if (conn->in == in) + conn->in = NULL; return 0; } -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:45:17 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:45:17 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146679.270031 (Exim 4.92) (envelope-from ) id 1lwMrF-0003sU-13; Thu, 24 Jun 2021 10:45:17 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146679.270031; Thu, 24 Jun 2021 10:45:17 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMrE-0003sM-UP; Thu, 24 Jun 2021 10:45:16 +0000 Received: by outflank-mailman (input) for mailman id 146679; Thu, 24 Jun 2021 10:45:15 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMrD-0003s0-3p for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:45:15 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwMrD-0006Vt-34 for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:45:15 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwMrD-0000Mg-2N for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:45:15 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=TfzNeu+AJQ7utFD5VWj/clot9O2NUkqrq5VAXttkW2M=; b=fv7QvHqoQW1QAh+sZrf/CurUXI SQZ9yzbr/7/CWszeQFB/QksUduT8dYwzJihs4dmI/fJjww683SBSfW2kflJXAZgICSoclyG1CD8WW vCoPaVBC7x4ALG7a2nR1v86PQLJf5+tQGRJ8OLGtpfyNd8aHuLZ/NWRsIH/NGHFAKRc8=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] tools/xenstored: Extend restore code to handle multiple input buffer Message-Id: Date: Thu, 24 Jun 2021 10:45:15 +0000 commit c0fe360f42245da6869362584b1f115a71d36ad6 Author: Julien Grall AuthorDate: Thu Jun 24 11:41:00 2021 +0100 Commit: Julien Grall CommitDate: Thu Jun 24 11:41:00 2021 +0100 tools/xenstored: Extend restore code to handle multiple input buffer Currently, the restore code is considering the stream will contain at most one in-flight request per connection. In a follow-up changes, we will want to transfer multiple in-flight requests. The function read_state_buffered() is now extended to restore multiple in-flight request. Complete requests will be queued as delayed requests, if there a partial request (only the last one can) then it will used as the current in-flight request. Note that we want to bypass the quota check for delayed requests as the new Xenstore may have a lower limit. Lastly, there is no need to change the specification as there was no restriction on the number of in-flight requests preserved. Signed-off-by: Julien Grall Reviewed-by: Luca Fancellu Reviewed-by: Juergen Gross --- tools/xenstore/xenstored_core.c | 56 +++++++++++++++++++++++++++++++++++------ 1 file changed, 48 insertions(+), 8 deletions(-) diff --git a/tools/xenstore/xenstored_core.c b/tools/xenstore/xenstored_core.c index 3286eb55d3..a83593b566 100644 --- a/tools/xenstore/xenstored_core.c +++ b/tools/xenstore/xenstored_core.c @@ -1486,6 +1486,10 @@ static void process_message(struct connection *conn, struct buffered_data *in) enum xsd_sockmsg_type type = in->hdr.msg.type; int ret; + /* At least send_error() and send_reply() expects conn->in == in */ + assert(conn->in == in); + trace_io(conn, in, 0); + if ((unsigned int)type >= XS_TYPE_COUNT || !wire_funcs[type].func) { eprintf("Client unknown operation %i", type); send_error(conn, ENOSYS); @@ -1515,6 +1519,23 @@ static void process_message(struct connection *conn, struct buffered_data *in) conn->transaction = NULL; } +static bool process_delayed_message(struct delayed_request *req) +{ + struct connection *conn = req->data; + struct buffered_data *saved_in = conn->in; + + /* + * Part of process_message() expects conn->in to contains the + * processed response. So save the current conn->in and restore it + * afterwards. + */ + conn->in = req->in; + process_message(req->data, req->in); + conn->in = saved_in; + + return true; +} + static void consider_message(struct connection *conn) { if (verbose) @@ -1582,7 +1603,6 @@ static void handle_input(struct connection *conn) if (in->used != in->hdr.msg.len) return; - trace_io(conn, in, 0); consider_message(conn); return; @@ -2611,14 +2631,20 @@ void read_state_buffered_data(const void *ctx, struct connection *conn, unsigned int len; bool partial = sc->data_resp_len; - if (sc->data_in_len) { + for (data = sc->data; data < sc->data + sc->data_in_len; data += len) { bdata = new_buffer(conn); if (!bdata) barf("error restoring read data"); - if (sc->data_in_len < sizeof(bdata->hdr)) { + + /* + * We don't know yet if there is more than one message + * to process. So the len is the size of the leftover data. + */ + len = sc->data_in_len - (data - sc->data); + if (len < sizeof(bdata->hdr)) { bdata->inhdr = true; - memcpy(&bdata->hdr, sc->data, sc->data_in_len); - bdata->used = sc->data_in_len; + memcpy(&bdata->hdr, sc->data, len); + bdata->used = len; } else { bdata->inhdr = false; memcpy(&bdata->hdr, sc->data, sizeof(bdata->hdr)); @@ -2629,12 +2655,26 @@ void read_state_buffered_data(const void *ctx, struct connection *conn, bdata->hdr.msg.len); if (!bdata->buffer) barf("Error allocating in buffer"); - bdata->used = sc->data_in_len - sizeof(bdata->hdr); - memcpy(bdata->buffer, sc->data + sizeof(bdata->hdr), + bdata->used = min_t(unsigned int, + len - sizeof(bdata->hdr), + bdata->hdr.msg.len); + memcpy(bdata->buffer, data + sizeof(bdata->hdr), bdata->used); + /* Update len to match the size of the message. */ + len = bdata->used + sizeof(bdata->hdr); } - conn->in = bdata; + /* + * If the message is not complete, then it means this was + * the current processed message. All the other messages + * will be queued to be handled after restoring. + */ + if (bdata->inhdr || bdata->used != bdata->hdr.msg.len) { + assert(conn->in == NULL); + conn->in = bdata; + } else if (delay_request(conn, bdata, process_delayed_message, + conn, true)) + barf("Unable to delay the request"); } for (data = sc->data + sc->data_in_len; -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:55:10 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:55:10 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146696.270068 (Exim 4.92) (envelope-from ) id 1lwN0k-0006dY-Hi; Thu, 24 Jun 2021 10:55:06 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146696.270068; Thu, 24 Jun 2021 10:55:06 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwN0k-0006dP-E0; Thu, 24 Jun 2021 10:55:06 +0000 Received: by outflank-mailman (input) for mailman id 146696; Thu, 24 Jun 2021 10:55:05 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwN0j-0006dJ-HK for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:55:05 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwN0j-0006hw-EI for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:55:05 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwN0j-00015u-D7 for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:55:05 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=40egODoUnFKuUNKIK/sN46pDbje8PjOK2G/yuQa0OmQ=; b=yfrFXJL4JGOt9DvWBIcU69lpmd zvwopqFtD5xWN5EyjwIvy74FpbsaIB8L/1Xfjw4Swhy9wN0BSS8MZXWvoYlEFgGc0F6kRz8Ko9tBi a+pRCmGa/L18jyEbrFcSi5upEgnjmnIDBiEE/FWi2wU1MMbvGj2WfMWs3KlJjcrZR4mk=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] iommu/arm: ipmmu-vmsa: Add compatible for Renesas R-Car M3-W+ SoC Message-Id: Date: Thu, 24 Jun 2021 10:55:05 +0000 commit ae8d47ede97707af07e772a164a475aa5f447708 Author: Oleksandr Tyshchenko AuthorDate: Mon Jun 14 22:18:12 2021 +0300 Commit: Julien Grall CommitDate: Thu Jun 24 11:43:53 2021 +0100 iommu/arm: ipmmu-vmsa: Add compatible for Renesas R-Car M3-W+ SoC The "renesas,r8a77961" string identifies M3-W+ (aka M3-W ES3.0) instead of "renesas,r8a7796" since Linux commit: "9c9f7891093b02eb64ca4e1c7ab776a4296c058f soc: renesas: Identify R-Car M3-W+". Add new compatible to the Xen driver. Signed-off-by: Oleksandr Tyshchenko Acked-by: Julien Grall --- xen/drivers/passthrough/arm/ipmmu-vmsa.c | 1 + 1 file changed, 1 insertion(+) diff --git a/xen/drivers/passthrough/arm/ipmmu-vmsa.c b/xen/drivers/passthrough/arm/ipmmu-vmsa.c index 8b8e3a00ba..1255b0d4fe 100644 --- a/xen/drivers/passthrough/arm/ipmmu-vmsa.c +++ b/xen/drivers/passthrough/arm/ipmmu-vmsa.c @@ -1316,6 +1316,7 @@ static const struct dt_device_match ipmmu_dt_match[] __initconst = DT_MATCH_COMPATIBLE("renesas,ipmmu-r8a7795"), DT_MATCH_COMPATIBLE("renesas,ipmmu-r8a77965"), DT_MATCH_COMPATIBLE("renesas,ipmmu-r8a7796"), + DT_MATCH_COMPATIBLE("renesas,ipmmu-r8a77961"), { /* sentinel */ }, }; -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 10:55:16 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 10:55:16 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146697.270072 (Exim 4.92) (envelope-from ) id 1lwN0u-0006gV-Jy; Thu, 24 Jun 2021 10:55:16 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146697.270072; Thu, 24 Jun 2021 10:55:16 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwN0u-0006gN-Gz; Thu, 24 Jun 2021 10:55:16 +0000 Received: by outflank-mailman (input) for mailman id 146697; Thu, 24 Jun 2021 10:55:15 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwN0t-0006g7-IH for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:55:15 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwN0t-0006iK-HL for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:55:15 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwN0t-00016k-GU for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 10:55:15 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=kVEUavvjZ213fqksZUcELvE2Cei6TB2IlMwrdaUeYF0=; b=ook8QdSuumc5Ib2WJ6ga/nnkuQ MWnTpZ8U7cGsXuetaBCxHefasSnnO7s3NFR4/PO0AwVUXLIAKb7MvGoEIQCBxtTS0tEGyYXWZc5SY jeM2yTYmRoDXOl6bQX4xQkMnYIgRbYCb+ZjlP6rGpXNk55Lvu6R+mGjyUbq9fe52Abus=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] maintainers: adding new reviewer for xsm Message-Id: Date: Thu, 24 Jun 2021 10:55:15 +0000 commit 76a0aa9c4d7a9fc6fee1158fd9df82ae9b8b605d Author: Daniel P. Smith AuthorDate: Thu Jun 17 19:49:55 2021 -0400 Commit: Julien Grall CommitDate: Thu Jun 24 11:45:32 2021 +0100 maintainers: adding new reviewer for xsm Would like to add myself as a reviewer for XSM. Signed-off-by: Daniel P. Smith Acked-by: Stefano Stabellini Acked-by: Andrew Cooper --- MAINTAINERS | 1 + 1 file changed, 1 insertion(+) diff --git a/MAINTAINERS b/MAINTAINERS index dd8c011456..8a52a03969 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -641,6 +641,7 @@ F: xen/include/xen/trace.h XSM/FLASK M: Daniel De Graaf +R: Daniel P. Smith S: Supported F: tools/flask/ F: xen/include/xsm/ -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 11:22:09 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 11:22:09 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146739.270149 (Exim 4.92) (envelope-from ) id 1lwNQs-0004uO-G6; Thu, 24 Jun 2021 11:22:06 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146739.270149; Thu, 24 Jun 2021 11:22:06 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwNQs-0004uG-Cx; Thu, 24 Jun 2021 11:22:06 +0000 Received: by outflank-mailman (input) for mailman id 146739; Thu, 24 Jun 2021 11:22:05 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwNQr-0004uA-3L for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 11:22:05 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwNQr-0007Bs-0e for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 11:22:05 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwNQq-0003m7-W3 for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 11:22:04 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=vzEIa6TQt7v2xU9miHcTAog4rFbteMOZqyTL8d+FUsw=; b=a3YmI3jjFdGfdqtdYFhqZ92JIG tNooOfIr3EuPMyGpFXXzyVOgo62qM18jdbqaDByFLC54PZpvWCovd7A7aV9lYh9lFWUBmdUeRRT0x S4QZn8S2nBk3HG1IUKB6hFjnAYdhgArxZ4M5BoHWS7WOZkfVOrQf6A52ypjkiRtkS0Ek=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] tools/xenstored: Dump delayed requests Message-Id: Date: Thu, 24 Jun 2021 11:22:04 +0000 commit 54f73e6fc6ef3fbd287cd030fb70d43f8f92244d Author: Julien Grall AuthorDate: Thu Jun 24 12:12:54 2021 +0100 Commit: Julien Grall CommitDate: Thu Jun 24 12:12:54 2021 +0100 tools/xenstored: Dump delayed requests Currently, only Live-Update request can be delayed. In a follow-up, we will want to delay more requests (e.g. transaction start). Therefore we want to preserve delayed requests across Live-Update. Delayed requests are just complete "in" buffer. So the code is refactored to allow sharing the code to dump "in" buffer. Signed-off-by: Julien Grall Reviewed-by: Luca Fancellu Reviewed-by: Juergen Gross --- tools/xenstore/xenstored_core.c | 47 ++++++++++++++++++++++++++++++++--------- 1 file changed, 37 insertions(+), 10 deletions(-) diff --git a/tools/xenstore/xenstored_core.c b/tools/xenstore/xenstored_core.c index a83593b566..0cb46d8c64 100644 --- a/tools/xenstore/xenstored_core.c +++ b/tools/xenstore/xenstored_core.c @@ -2403,25 +2403,52 @@ const char *dump_state_global(FILE *fp) return NULL; } +static const char *dump_input_buffered_data(FILE *fp, + const struct buffered_data *in, + unsigned int *total_len) +{ + unsigned int hlen = in->inhdr ? in->used : sizeof(in->hdr); + + *total_len += hlen; + if (fp && fwrite(&in->hdr, hlen, 1, fp) != 1) + return "Dump read data error"; + if (!in->inhdr && in->used) { + *total_len += in->used; + if (fp && fwrite(in->buffer, in->used, 1, fp) != 1) + return "Dump read data error"; + } + + return NULL; +} + /* Called twice: first with fp == NULL to get length, then for writing data. */ const char *dump_state_buffered_data(FILE *fp, const struct connection *c, struct xs_state_connection *sc) { unsigned int len = 0, used; - struct buffered_data *out, *in = c->in; + struct buffered_data *out; bool partial = true; + struct delayed_request *req; + const char *ret; - if (in) { - len = in->inhdr ? in->used : sizeof(in->hdr); - if (fp && fwrite(&in->hdr, len, 1, fp) != 1) - return "Dump read data error"; - if (!in->inhdr && in->used) { - len += in->used; - if (fp && fwrite(in->buffer, in->used, 1, fp) != 1) - return "Dump read data error"; - } + /* Dump any command that was delayed */ + list_for_each_entry(req, &c->delayed, list) { + /* + * We only want to preserve commands that weren't processed at + * all. All the other delayed requests (such as do_lu_start()) + * must be processed before Live-Update. + */ + if (req->func != process_delayed_message) + continue; + + assert(!req->in->inhdr); + if ((ret = dump_input_buffered_data(fp, req->in, &len))) + return ret; } + if (c->in && (ret = dump_input_buffered_data(fp, c->in, &len))) + return ret; + if (sc) { sc->data_in_len = len; sc->data_resp_len = 0; -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 11:22:16 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 11:22:16 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146740.270153 (Exim 4.92) (envelope-from ) id 1lwNR2-0004wx-Hi; Thu, 24 Jun 2021 11:22:16 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146740.270153; Thu, 24 Jun 2021 11:22:16 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwNR2-0004wp-EV; Thu, 24 Jun 2021 11:22:16 +0000 Received: by outflank-mailman (input) for mailman id 146740; Thu, 24 Jun 2021 11:22:15 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwNR1-0004wI-BX for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 11:22:15 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwNR1-0007CJ-94 for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 11:22:15 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwNR1-0003p8-67 for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 11:22:15 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=Cs0Rqu9nHK8w4NO2RGuyPoPq8wYlfbVm3mhQGqODTBk=; b=zjA7KsPdIL+g+ugMvjcY17qGNX Gvpy2UXETfmTEFW6FwBY8renAw5cS0qeXnCxf/5VPn/eq+FswsfNcWXKn+wlMz8xnZJUCvnLuNOoT peK8WgGtf/hA/1I8/CzbMLYDXHQh7gX1u4BJxPaT4TQbEUgfWvwkxURW+Zn9uva+hr5A=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] tools/xenstored: Delay new transaction while Live-Update is pending Message-Id: Date: Thu, 24 Jun 2021 11:22:15 +0000 commit c8be420025ada0f90e98ae0408226e29017757df Author: Julien Grall AuthorDate: Thu Jun 24 12:14:56 2021 +0100 Commit: Julien Grall CommitDate: Thu Jun 24 12:14:56 2021 +0100 tools/xenstored: Delay new transaction while Live-Update is pending At the moment, Live-Update will, by default, not proceed if there are in-flight transactions. It is possible force it by passing -F but this will break any connection with in-flight transactions. There are PV drivers out that may never terminate some transaction. On host running such guest, we would need to use -F. Unfortunately, this also risks to break well-behaving guests (and even dom0) because Live-Update will happen as soon as the timeout is hit. Ideally, we would want to preserve transactions but this requires some work and a lot of testing to be able to use it in production. As a stop gap, we want to limit the damage of -F. This patch will delay any transactions that are started after Live-Update has been requested. If the request cannot be delayed, the connection will be stalled to avoid loosing requests. If the connection has already a pending transaction before Live-Update, then new transaction will not be delayed. This is to avoid the connection to stall. With this stop gap in place, domains with long running transactions will still break when using -F, but other domains which starts a transaction in the middle of Live-Update will continue to work. Signed-off-by: Julien Grall Reviewed-by: Luca Fancellu Reviewed-by: Juergen Gross --- tools/xenstore/xenstored_control.c | 10 ++++++++ tools/xenstore/xenstored_control.h | 2 ++ tools/xenstore/xenstored_core.c | 49 +++++++++++++++++++++++++++++++++++++- tools/xenstore/xenstored_core.h | 3 +++ 4 files changed, 63 insertions(+), 1 deletion(-) diff --git a/tools/xenstore/xenstored_control.c b/tools/xenstore/xenstored_control.c index 1c24d4869e..a045f102a4 100644 --- a/tools/xenstore/xenstored_control.c +++ b/tools/xenstore/xenstored_control.c @@ -131,6 +131,11 @@ unsigned int lu_write_response(FILE *fp) return sizeof(msg) + msg.len; } +bool lu_is_pending(void) +{ + return lu_status != NULL; +} + #else struct connection *lu_get_connection(void) { @@ -142,6 +147,11 @@ unsigned int lu_write_response(FILE *fp) /* Unsupported */ return 0; } + +bool lu_is_pending(void) +{ + return false; +} #endif struct cmd_s { diff --git a/tools/xenstore/xenstored_control.h b/tools/xenstore/xenstored_control.h index 27d7f19e4b..98b6fbcea2 100644 --- a/tools/xenstore/xenstored_control.h +++ b/tools/xenstore/xenstored_control.h @@ -23,3 +23,5 @@ struct connection *lu_get_connection(void); /* Write the "OK" response for the live-update command */ unsigned int lu_write_response(FILE *fp); + +bool lu_is_pending(void); diff --git a/tools/xenstore/xenstored_core.c b/tools/xenstore/xenstored_core.c index 0cb46d8c64..9ffd2ac66d 100644 --- a/tools/xenstore/xenstored_core.c +++ b/tools/xenstore/xenstored_core.c @@ -338,7 +338,20 @@ static int destroy_conn(void *_conn) static bool conn_can_read(struct connection *conn) { - return conn->funcs->can_read(conn) && !conn->is_ignored; + if (!conn->funcs->can_read(conn)) + return false; + + if (conn->is_ignored) + return false; + + /* + * For stalled connection, we want to process the pending + * command as soon as live-update has aborted. + */ + if (conn->is_stalled) + return !lu_is_pending(); + + return true; } static bool conn_can_write(struct connection *conn) @@ -417,6 +430,12 @@ static void initialize_fds(int *p_sock_pollfd_idx, int *ptimeout) if (!list_empty(&conn->out_list)) events |= POLLOUT; conn->pollfd_idx = set_fd(conn->fd, events); + /* + * For stalled connection, we want to process the + * pending command as soon as live-update has aborted. + */ + if (conn->is_stalled && !lu_is_pending()) + *ptimeout = 0; } } } @@ -1524,6 +1543,9 @@ static bool process_delayed_message(struct delayed_request *req) struct connection *conn = req->data; struct buffered_data *saved_in = conn->in; + if (lu_is_pending()) + return false; + /* * Part of process_message() expects conn->in to contains the * processed response. So save the current conn->in and restore it @@ -1543,6 +1565,30 @@ static void consider_message(struct connection *conn) sockmsg_string(conn->in->hdr.msg.type), conn->in->hdr.msg.len, conn); + conn->is_stalled = false; + /* + * Currently, Live-Update is not supported if there is active + * transactions. In order to reduce the number of retry, delay + * any new request to start a transaction if Live-Update is pending + * and there are no transactions in-flight. + * + * If we can't delay the request, then mark the connection as + * stalled. This will ignore new requests until Live-Update happened + * or it was aborted. + */ + if (lu_is_pending() && conn->transaction_started == 0 && + conn->in->hdr.msg.type == XS_TRANSACTION_START) { + trace("Delaying transaction start for connection %p req_id %u\n", + conn, conn->in->hdr.msg.req_id); + + if (delay_request(conn, conn->in, process_delayed_message, + conn, false) != 0) { + trace("Stalling connection %p\n", conn); + conn->is_stalled = true; + } + return; + } + process_message(conn, conn->in); assert(conn->in == NULL); @@ -1629,6 +1675,7 @@ struct connection *new_connection(const struct interface_funcs *funcs) new->pollfd_idx = -1; new->funcs = funcs; new->is_ignored = false; + new->is_stalled = false; new->transaction_started = 0; INIT_LIST_HEAD(&new->out_list); INIT_LIST_HEAD(&new->watches); diff --git a/tools/xenstore/xenstored_core.h b/tools/xenstore/xenstored_core.h index dac5171569..258f6ff382 100644 --- a/tools/xenstore/xenstored_core.h +++ b/tools/xenstore/xenstored_core.h @@ -110,6 +110,9 @@ struct connection /* Is this connection ignored? */ bool is_ignored; + /* Is the connection stalled? */ + bool is_stalled; + /* Buffered incoming data. */ struct buffered_data *in; -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 11:22:26 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 11:22:26 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146741.270157 (Exim 4.92) (envelope-from ) id 1lwNRC-00051m-J5; Thu, 24 Jun 2021 11:22:26 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146741.270157; Thu, 24 Jun 2021 11:22:26 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwNRC-00051f-G6; Thu, 24 Jun 2021 11:22:26 +0000 Received: by outflank-mailman (input) for mailman id 146741; Thu, 24 Jun 2021 11:22:25 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwNRB-00051R-Fb for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 11:22:25 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwNRB-0007Cd-Eo for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 11:22:25 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwNRB-0003qm-Dj for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 11:22:25 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=Uzh6FG/8L/6VMUEs2rRZajqJm2N95p86UogIkOJJoGk=; b=RrErYAf+fvjuzAfKXga8Hhk5tB 8IMmz2dJqqahht4OkUp3VYsJ+9IEorZbBp/t1Fg2KEnoa1JigqNlAaQX7OZTk9s751bc0Osdi4c9W F8Evp4BD4LhulGf8RqTV7kHo2RbjhvcsKWpm7YXlTVPKeAWBt3dwtIoziazbCcPoBZRo=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] tools/xenstored: Don't crash xenstored when Live-Update is cancelled Message-Id: Date: Thu, 24 Jun 2021 11:22:25 +0000 commit 90bafdba8ebf41c9af31b5c725a938da2a75d292 Author: Julien GralL AuthorDate: Thu Jun 24 12:15:49 2021 +0100 Commit: Julien Grall CommitDate: Thu Jun 24 12:15:49 2021 +0100 tools/xenstored: Don't crash xenstored when Live-Update is cancelled As Live-Update is asynchronous, it is possible to receive a request to cancel it (either on the same connection or from a different one). Currently, this will crash xenstored because do_lu_start() assumes lu_status will be valid. This is not the case when Live-Update has been cancelled. This will result to dereference a NULL pointer and crash Xenstored. Rework do_lu_start() to check if lu_status is NULL and return an error in this case. Fixes: af216a99fb ("tools/xenstore: add the basic framework for doing the live update") Signed-off-by: Julien Grall Reviewed-by: Luca Fancellu Reviewed-by: Juergen Gross --- tools/xenstore/xenstored_control.c | 15 +++++++++++++-- 1 file changed, 13 insertions(+), 2 deletions(-) diff --git a/tools/xenstore/xenstored_control.c b/tools/xenstore/xenstored_control.c index a045f102a4..a1b1bd5a71 100644 --- a/tools/xenstore/xenstored_control.c +++ b/tools/xenstore/xenstored_control.c @@ -696,7 +696,18 @@ static bool do_lu_start(struct delayed_request *req) time_t now = time(NULL); const char *ret; struct buffered_data *saved_in; - struct connection *conn = lu_status->conn; + struct connection *conn = req->data; + + /* + * Cancellation may have been requested asynchronously. In this + * case, lu_status will be NULL. + */ + if (!lu_status) { + ret = "Cancellation was requested"; + goto out; + } + + assert(lu_status->conn == conn); if (!lu_check_lu_allowed()) { if (now < lu_status->started_at + lu_status->timeout) @@ -747,7 +758,7 @@ static const char *lu_start(const void *ctx, struct connection *conn, lu_status->timeout = to; lu_status->started_at = time(NULL); - errno = delay_request(conn, conn->in, do_lu_start, NULL, false); + errno = delay_request(conn, conn->in, do_lu_start, conn, false); return NULL; } -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 14:33:10 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 14:33:10 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146833.270308 (Exim 4.92) (envelope-from ) id 1lwQPi-0003rA-W4; Thu, 24 Jun 2021 14:33:06 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146833.270308; Thu, 24 Jun 2021 14:33:06 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwQPi-0003r2-T9; Thu, 24 Jun 2021 14:33:06 +0000 Received: by outflank-mailman (input) for mailman id 146833; Thu, 24 Jun 2021 14:33:05 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwQPh-0003qw-NP for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 14:33:05 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwQPh-0001zq-L2 for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 14:33:05 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwQPh-0002hJ-K1 for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 14:33:05 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=wFWyZxNZkYng5g6HZoC/4zq7Tjy8I/W9+cg+14nfD9A=; b=vqVCgZ75dHcrg6EgFUIRo83lN9 dXqovN2EIarZKiJkO82Q+JwfHoHr+JkLr3kmvCaVNEd68rOOdFT4nx3rc0jazRBlX7jYMp4LVeNnJ QzLHyMJqJFpnu5M1j5vDdZhN8Pz94G1O8gSKamfAzi7a5Avs2shzsxNAjHsETwnjI8E0=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] VT-d: undo device mappings upon error Message-Id: Date: Thu, 24 Jun 2021 14:33:05 +0000 commit f3401d65d9f0dce508c3d7da55de4a093d748ae1 Author: Jan Beulich AuthorDate: Thu Jun 24 16:28:25 2021 +0200 Commit: Jan Beulich CommitDate: Thu Jun 24 16:28:25 2021 +0200 VT-d: undo device mappings upon error When - flushes (supposedly not possible anymore after XSA-373), - secondary mappings for legacy PCI devices behind bridges, - secondary mappings for chipset quirks, or - find_upstream_bridge() invocations fail, the successfully established device mappings should not be left around. Further, when (parts of) unmapping fail, simply returning an error is typically not enough. Crash the domain instead in such cases, arranging for domain cleanup to continue in a best effort manner despite such failures. Finally make domain_context_unmap()'s error behavior consistent in the legacy PCI device case: Don't bail from the function in one special case, but always just exit the switch statement. Signed-off-by: Jan Beulich Reviewed-by: Paul Durrant Reviewed-by: Kevin Tian --- xen/drivers/passthrough/vtd/iommu.c | 57 +++++++++++++++++++++++++++++++------ 1 file changed, 48 insertions(+), 9 deletions(-) diff --git a/xen/drivers/passthrough/vtd/iommu.c b/xen/drivers/passthrough/vtd/iommu.c index b7aa15e5ef..d9062422d5 100644 --- a/xen/drivers/passthrough/vtd/iommu.c +++ b/xen/drivers/passthrough/vtd/iommu.c @@ -1442,9 +1442,15 @@ int domain_context_mapping_one( if ( !seg && !rc ) rc = me_wifi_quirk(domain, bus, devfn, MAP_ME_PHANTOM_FUNC); + if ( rc ) + domain_context_unmap_one(domain, iommu, bus, devfn); + return rc; } +static int domain_context_unmap(struct domain *d, uint8_t devfn, + struct pci_dev *pdev); + static int domain_context_mapping(struct domain *domain, u8 devfn, struct pci_dev *pdev) { @@ -1505,16 +1511,21 @@ static int domain_context_mapping(struct domain *domain, u8 devfn, if ( ret ) break; - if ( find_upstream_bridge(seg, &bus, &devfn, &secbus) < 1 ) - break; + if ( (ret = find_upstream_bridge(seg, &bus, &devfn, &secbus)) < 1 ) + { + if ( !ret ) + break; + ret = -ENXIO; + } /* * Mapping a bridge should, if anything, pass the struct pci_dev of * that bridge. Since bridges don't normally get assigned to guests, * their owner would be the wrong one. Pass NULL instead. */ - ret = domain_context_mapping_one(domain, drhd->iommu, bus, devfn, - NULL); + if ( ret >= 0 ) + ret = domain_context_mapping_one(domain, drhd->iommu, bus, devfn, + NULL); /* * Devices behind PCIe-to-PCI/PCIx bridge may generate different @@ -1531,6 +1542,9 @@ static int domain_context_mapping(struct domain *domain, u8 devfn, ret = domain_context_mapping_one(domain, drhd->iommu, secbus, 0, NULL); + if ( ret ) + domain_context_unmap(domain, devfn, pdev); + break; default: @@ -1609,6 +1623,19 @@ int domain_context_unmap_one( if ( !iommu->drhd->segment && !rc ) rc = me_wifi_quirk(domain, bus, devfn, UNMAP_ME_PHANTOM_FUNC); + if ( rc && !is_hardware_domain(domain) && domain != dom_io ) + { + if ( domain->is_dying ) + { + printk(XENLOG_ERR "%pd: error %d unmapping %04x:%02x:%02x.%u\n", + domain, rc, iommu->drhd->segment, bus, + PCI_SLOT(devfn), PCI_FUNC(devfn)); + rc = 0; /* Make upper layers continue in a best effort manner. */ + } + else + domain_crash(domain); + } + return rc; } @@ -1661,17 +1688,29 @@ static int domain_context_unmap(struct domain *domain, u8 devfn, tmp_bus = bus; tmp_devfn = devfn; - if ( find_upstream_bridge(seg, &tmp_bus, &tmp_devfn, &secbus) < 1 ) + if ( (ret = find_upstream_bridge(seg, &tmp_bus, &tmp_devfn, + &secbus)) < 1 ) + { + if ( ret ) + { + ret = -ENXIO; + if ( !domain->is_dying && + !is_hardware_domain(domain) && domain != dom_io ) + { + domain_crash(domain); + /* Make upper layers continue in a best effort manner. */ + ret = 0; + } + } break; + } /* PCIe to PCI/PCIx bridge */ if ( pdev_type(seg, tmp_bus, tmp_devfn) == DEV_TYPE_PCIe2PCI_BRIDGE ) { ret = domain_context_unmap_one(domain, iommu, tmp_bus, tmp_devfn); - if ( ret ) - return ret; - - ret = domain_context_unmap_one(domain, iommu, secbus, 0); + if ( !ret ) + ret = domain_context_unmap_one(domain, iommu, secbus, 0); } else /* Legacy PCI bridge */ ret = domain_context_unmap_one(domain, iommu, tmp_bus, tmp_devfn); -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 14:33:17 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 14:33:17 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146834.270311 (Exim 4.92) (envelope-from ) id 1lwQPt-0003t4-1Z; Thu, 24 Jun 2021 14:33:17 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146834.270311; Thu, 24 Jun 2021 14:33:17 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwQPs-0003sw-Uo; Thu, 24 Jun 2021 14:33:16 +0000 Received: by outflank-mailman (input) for mailman id 146834; Thu, 24 Jun 2021 14:33:15 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwQPr-0003sm-Po for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 14:33:15 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwQPr-00020E-OY for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 14:33:15 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwQPr-0002iM-Nd for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 14:33:15 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=/MUv6LJwZAROSjphcARtwBuLPVdZy4AzsOlPRKYmjZ0=; b=7Ma9/MvCi5l+PYwbKcUUJ7VM8e HFzG4c+vq/R6Y9xZu3OfrmhGTS4JAR/VILhusv42PJ7ZsD/X5WCFW+8+itykA4S9Y+CZKa2wm6koK Scbf0cTOsaH7ytAFicN/ULrGXV8pSHqqbCtgmNIPh1wCYMsl7XKXf1fiC0O8vOsVtvIw=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] VT-d: adjust domid map updating when unmapping context Message-Id: Date: Thu, 24 Jun 2021 14:33:15 +0000 commit 32655880057ce2829f962d46916ea6cec60f98d3 Author: Jan Beulich AuthorDate: Thu Jun 24 16:29:13 2021 +0200 Commit: Jan Beulich CommitDate: Thu Jun 24 16:29:13 2021 +0200 VT-d: adjust domid map updating when unmapping context When an earlier error occurred, cleaning up the domid mapping data is wrong, as references likely still exist. The only exception to this is when the actual unmapping worked, but some flush failed (supposedly impossible after XSA-373). The guest will get crashed in such a case though, so add fallback cleanup to domain destruction to cover this case. This in turn makes it desirable to silence the dprintk() in domain_iommu_domid(). Note that no error will be returned anymore when the lookup fails - in the common case lookup failure would already have caused domain_context_unmap_one() to fail, yet even from a more general perspective it doesn't look right to fail domain_context_unmap() in such a case when this was the last device, but not when any earlier unmap was otherwise successful. Signed-off-by: Jan Beulich Reviewed-by: Kevin Tian --- xen/drivers/passthrough/vtd/iommu.c | 39 +++++++++++++++++++++++-------------- 1 file changed, 24 insertions(+), 15 deletions(-) diff --git a/xen/drivers/passthrough/vtd/iommu.c b/xen/drivers/passthrough/vtd/iommu.c index d9062422d5..0342daaa3c 100644 --- a/xen/drivers/passthrough/vtd/iommu.c +++ b/xen/drivers/passthrough/vtd/iommu.c @@ -80,9 +80,11 @@ static int domain_iommu_domid(struct domain *d, i = find_next_bit(iommu->domid_bitmap, nr_dom, i+1); } - dprintk(XENLOG_ERR VTDPREFIX, - "Cannot get valid iommu domid: domid=%d iommu->index=%d\n", - d->domain_id, iommu->index); + if ( !d->is_dying ) + dprintk(XENLOG_ERR VTDPREFIX, + "Cannot get valid iommu %u domid: %pd\n", + iommu->index, d); + return -1; } @@ -147,6 +149,17 @@ static int context_get_domain_id(struct context_entry *context, return domid; } +static void cleanup_domid_map(struct domain *domain, struct vtd_iommu *iommu) +{ + int iommu_domid = domain_iommu_domid(domain, iommu); + + if ( iommu_domid >= 0 ) + { + clear_bit(iommu_domid, iommu->domid_bitmap); + iommu->domid_map[iommu_domid] = 0; + } +} + static void sync_cache(const void *addr, unsigned int size) { static unsigned long clflush_size = 0; @@ -1724,6 +1737,9 @@ static int domain_context_unmap(struct domain *domain, u8 devfn, goto out; } + if ( ret ) + goto out; + /* * if no other devices under the same iommu owned by this domain, * clear iommu in iommu_bitmap and clear domain_id in domid_bitmp @@ -1743,19 +1759,8 @@ static int domain_context_unmap(struct domain *domain, u8 devfn, if ( found == 0 ) { - int iommu_domid; - clear_bit(iommu->index, &dom_iommu(domain)->arch.vtd.iommu_bitmap); - - iommu_domid = domain_iommu_domid(domain, iommu); - if ( iommu_domid == -1 ) - { - ret = -EINVAL; - goto out; - } - - clear_bit(iommu_domid, iommu->domid_bitmap); - iommu->domid_map[iommu_domid] = 0; + cleanup_domid_map(domain, iommu); } out: @@ -1775,6 +1780,7 @@ static void iommu_domain_teardown(struct domain *d) { struct domain_iommu *hd = dom_iommu(d); struct mapped_rmrr *mrmrr, *tmp; + const struct acpi_drhd_unit *drhd; if ( list_empty(&acpi_drhd_units) ) return; @@ -1786,6 +1792,9 @@ static void iommu_domain_teardown(struct domain *d) } ASSERT(!hd->arch.vtd.pgd_maddr); + + for_each_drhd_unit ( drhd ) + cleanup_domid_map(d, drhd->iommu); } static int __must_check intel_iommu_map_page(struct domain *d, dfn_t dfn, -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 14:33:27 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 14:33:27 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146835.270315 (Exim 4.92) (envelope-from ) id 1lwQQ3-0003vv-3Q; Thu, 24 Jun 2021 14:33:27 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146835.270315; Thu, 24 Jun 2021 14:33:27 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwQQ3-0003vl-0K; Thu, 24 Jun 2021 14:33:27 +0000 Received: by outflank-mailman (input) for mailman id 146835; Thu, 24 Jun 2021 14:33:25 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwQQ1-0003va-T5 for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 14:33:25 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwQQ1-00020O-SA for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 14:33:25 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwQQ1-0002jC-R7 for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 14:33:25 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=Jq5BZD5MlfHu9t7+gnnCEPjYXpjKMWKXZvCokEYFSvw=; b=JEbh5ZOo9Z5O5NSPU/eIvmufTf vHufoU+vVDWLD7bT3VkZSkwm3Nm6TnoaRsC4PTkwFnKvYtFaI8O4TBfGttKw9Mq5dpGapIFC1VUW+ tNP+N9jYr9K4KEmDAkxZWtDDsb0jU1YIWL7hkMBhu/XNYpI36N1dwxtdXpXHMhsHudXI=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] VT-d: clear_fault_bits() should clear all fault bits Message-Id: Date: Thu, 24 Jun 2021 14:33:25 +0000 commit 80589800ae62fce43fd3921e8fbd362465fe5ba3 Author: Jan Beulich AuthorDate: Thu Jun 24 16:29:42 2021 +0200 Commit: Jan Beulich CommitDate: Thu Jun 24 16:29:42 2021 +0200 VT-d: clear_fault_bits() should clear all fault bits If there is any way for one fault to be left set in the recording registers, there's no reason there couldn't also be multiple ones. If PPF set set (being the OR or all F fields), simply loop over the entire range of fault recording registers, clearing F everywhere. Since PPF is a r/o bit, also remove it from DMA_FSTS_FAULTS (arguably the constant's name is ambiguous as well). Signed-off-by: Jan Beulich Reviewed-by: Kevin Tian --- xen/drivers/passthrough/vtd/iommu.c | 16 +++++++++++++--- xen/drivers/passthrough/vtd/iommu.h | 5 ++--- 2 files changed, 15 insertions(+), 6 deletions(-) diff --git a/xen/drivers/passthrough/vtd/iommu.c b/xen/drivers/passthrough/vtd/iommu.c index 0342daaa3c..f3524a1def 100644 --- a/xen/drivers/passthrough/vtd/iommu.c +++ b/xen/drivers/passthrough/vtd/iommu.c @@ -2094,13 +2094,23 @@ static int __hwdom_init setup_hwdom_device(u8 devfn, struct pci_dev *pdev) void clear_fault_bits(struct vtd_iommu *iommu) { - u64 val; unsigned long flags; spin_lock_irqsave(&iommu->register_lock, flags); - val = dmar_readq(iommu->reg, cap_fault_reg_offset(iommu->cap) + 8); - dmar_writeq(iommu->reg, cap_fault_reg_offset(iommu->cap) + 8, val); + + if ( dmar_readl(iommu->reg, DMAR_FSTS_REG) & DMA_FSTS_PPF ) + { + unsigned int reg = cap_fault_reg_offset(iommu->cap); + unsigned int end = reg + cap_num_fault_regs(iommu->cap); + + do { + dmar_writel(iommu->reg, reg + 12, DMA_FRCD_F); + reg += PRIMARY_FAULT_REG_LEN; + } while ( reg < end ); + } + dmar_writel(iommu->reg, DMAR_FSTS_REG, DMA_FSTS_FAULTS); + spin_unlock_irqrestore(&iommu->register_lock, flags); } diff --git a/xen/drivers/passthrough/vtd/iommu.h b/xen/drivers/passthrough/vtd/iommu.h index cd516ac5f8..02390cb84b 100644 --- a/xen/drivers/passthrough/vtd/iommu.h +++ b/xen/drivers/passthrough/vtd/iommu.h @@ -174,9 +174,8 @@ #define DMA_FSTS_IQE (1u << 4) #define DMA_FSTS_ICE (1u << 5) #define DMA_FSTS_ITE (1u << 6) -#define DMA_FSTS_FAULTS (DMA_FSTS_PFO | DMA_FSTS_PPF | DMA_FSTS_AFO | \ - DMA_FSTS_APF | DMA_FSTS_IQE | DMA_FSTS_ICE | \ - DMA_FSTS_ITE) +#define DMA_FSTS_FAULTS (DMA_FSTS_PFO | DMA_FSTS_AFO | DMA_FSTS_APF | \ + DMA_FSTS_IQE | DMA_FSTS_ICE | DMA_FSTS_ITE) #define dma_fsts_fault_record_index(s) (((s) >> 8) & 0xff) /* FRCD_REG, 32 bits access */ -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 14:33:37 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 14:33:37 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146836.270320 (Exim 4.92) (envelope-from ) id 1lwQQD-0003yS-4o; Thu, 24 Jun 2021 14:33:37 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146836.270320; Thu, 24 Jun 2021 14:33:37 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwQQD-0003yK-1q; Thu, 24 Jun 2021 14:33:37 +0000 Received: by outflank-mailman (input) for mailman id 146836; Thu, 24 Jun 2021 14:33:36 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwQQC-0003y9-0A for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 14:33:36 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwQQB-00020b-VM for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 14:33:35 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwQQB-0002k5-UW for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 14:33:35 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=fvvYio8oqIiwq3agORl7te/zJMmGaDZCk4hkQqYjVpI=; b=r9GvjzffBfmZQzg2+y2LvuILxH +VWXMIocMmpwKP5GsvpL01f94kqEnd6D3t47wBPQydpn6sjIeYtMpzuAOt8iY3uli9O9uI5gVKjah RJMZ3gTYhhHdDiJOXZGgeJCzW+tPDBJuzKEliBTwXm+MGEiQpxSIj9HEt3ImhfRO9RJA=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] VT-d: don't lose errors when flushing TLBs on multiple IOMMUs Message-Id: Date: Thu, 24 Jun 2021 14:33:35 +0000 commit e7059776f9755b989a992d229c68c3d7778412be Author: Jan Beulich AuthorDate: Thu Jun 24 16:30:06 2021 +0200 Commit: Jan Beulich CommitDate: Thu Jun 24 16:30:06 2021 +0200 VT-d: don't lose errors when flushing TLBs on multiple IOMMUs While no longer an immediate problem with flushes no longer timing out, errors (if any) get properly reported by iommu_flush_iotlb_{dsi,psi}(). Overwriting such an error with, perhaps, a success indicator received from another IOMMU will misguide callers. Record the first error, but don't bail from the loop (such that further necessary invalidation gets carried out on a best effort basis). Signed-off-by: Jan Beulich Reviewed-by: Kevin Tian --- xen/drivers/passthrough/vtd/iommu.c | 11 ++++++----- 1 file changed, 6 insertions(+), 5 deletions(-) diff --git a/xen/drivers/passthrough/vtd/iommu.c b/xen/drivers/passthrough/vtd/iommu.c index f3524a1def..0f10951ce9 100644 --- a/xen/drivers/passthrough/vtd/iommu.c +++ b/xen/drivers/passthrough/vtd/iommu.c @@ -643,7 +643,7 @@ static int __must_check iommu_flush_iotlb(struct domain *d, dfn_t dfn, struct vtd_iommu *iommu; bool_t flush_dev_iotlb; int iommu_domid; - int rc = 0; + int ret = 0; /* * No need pcideves_lock here because we have flush @@ -651,6 +651,8 @@ static int __must_check iommu_flush_iotlb(struct domain *d, dfn_t dfn, */ for_each_drhd_unit ( drhd ) { + int rc; + iommu = drhd->iommu; if ( !test_bit(iommu->index, &hd->arch.vtd.iommu_bitmap) ) @@ -673,13 +675,12 @@ static int __must_check iommu_flush_iotlb(struct domain *d, dfn_t dfn, flush_dev_iotlb); if ( rc > 0 ) - { iommu_flush_write_buffer(iommu); - rc = 0; - } + else if ( !ret ) + ret = rc; } - return rc; + return ret; } static int __must_check iommu_flush_iotlb_pages(struct domain *d, -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 14:33:47 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 14:33:47 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146837.270324 (Exim 4.92) (envelope-from ) id 1lwQQN-00041X-6m; Thu, 24 Jun 2021 14:33:47 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146837.270324; Thu, 24 Jun 2021 14:33:47 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwQQN-00041M-3K; Thu, 24 Jun 2021 14:33:47 +0000 Received: by outflank-mailman (input) for mailman id 146837; Thu, 24 Jun 2021 14:33:46 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwQQM-000418-2u for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 14:33:46 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwQQM-00020u-1t for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 14:33:46 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwQQM-0002l3-1H for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 14:33:46 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=DoYwri8M50H1HeA4yVJukP1HVRrOW+0AddG3UwNIZRo=; b=If9aDVBcSCoHLwRwqyxWSvU4Re N1APQ+iCucbdnaZilJ+kBJiNz6OeiSNou920f6/E/GM+uNS9n87DZGSasz8L7/f5SA/tJZXY4aY0Z XYETgfVBAKwq3H38rtTJNIXlCnxx6kGNdTzGHdcE59/GCZS9WxdE6anbvaImKYttaizs=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] VT-d: centralize mapping of QI entries Message-Id: Date: Thu, 24 Jun 2021 14:33:46 +0000 commit a0eb19723d9a6006bf465f0af60e229206d39333 Author: Jan Beulich AuthorDate: Thu Jun 24 16:30:32 2021 +0200 Commit: Jan Beulich CommitDate: Thu Jun 24 16:30:32 2021 +0200 VT-d: centralize mapping of QI entries Introduce a helper function to reduce redundancy. Take the opportunity to express the logic without using the somewhat odd QINVAL_ENTRY_ORDER. Also take the opportunity to uniformly unmap after updating queue tail and dropping the lock (like was done so far only by queue_invalidate_context_sync()). Signed-off-by: Jan Beulich Reviewed-by: Kevin Tian --- xen/drivers/passthrough/vtd/qinval.c | 64 ++++++++++++++++-------------------- 1 file changed, 29 insertions(+), 35 deletions(-) diff --git a/xen/drivers/passthrough/vtd/qinval.c b/xen/drivers/passthrough/vtd/qinval.c index 311bf107ed..2e4a55a869 100644 --- a/xen/drivers/passthrough/vtd/qinval.c +++ b/xen/drivers/passthrough/vtd/qinval.c @@ -69,6 +69,16 @@ static void qinval_update_qtail(struct vtd_iommu *iommu, unsigned int index) dmar_writel(iommu->reg, DMAR_IQT_REG, val << QINVAL_INDEX_SHIFT); } +static struct qinval_entry *qi_map_entry(const struct vtd_iommu *iommu, + unsigned int index) +{ + paddr_t base = iommu->qinval_maddr + + ((index * sizeof(struct qinval_entry)) & PAGE_MASK); + struct qinval_entry *entries = map_vtd_domain_page(base); + + return &entries[index % (PAGE_SIZE / sizeof(*entries))]; +} + static int __must_check queue_invalidate_context_sync(struct vtd_iommu *iommu, u16 did, u16 source_id, u8 function_mask, @@ -76,15 +86,11 @@ static int __must_check queue_invalidate_context_sync(struct vtd_iommu *iommu, { unsigned long flags; unsigned int index; - u64 entry_base; - struct qinval_entry *qinval_entry, *qinval_entries; + struct qinval_entry *qinval_entry; spin_lock_irqsave(&iommu->register_lock, flags); index = qinval_next_index(iommu); - entry_base = iommu->qinval_maddr + - ((index >> QINVAL_ENTRY_ORDER) << PAGE_SHIFT); - qinval_entries = map_vtd_domain_page(entry_base); - qinval_entry = &qinval_entries[index % (1 << QINVAL_ENTRY_ORDER)]; + qinval_entry = qi_map_entry(iommu, index); qinval_entry->q.cc_inv_dsc.lo.type = TYPE_INVAL_CONTEXT; qinval_entry->q.cc_inv_dsc.lo.granu = granu; @@ -98,7 +104,7 @@ static int __must_check queue_invalidate_context_sync(struct vtd_iommu *iommu, qinval_update_qtail(iommu, index); spin_unlock_irqrestore(&iommu->register_lock, flags); - unmap_vtd_domain_page(qinval_entries); + unmap_vtd_domain_page(qinval_entry); return invalidate_sync(iommu); } @@ -110,15 +116,11 @@ static int __must_check queue_invalidate_iotlb_sync(struct vtd_iommu *iommu, { unsigned long flags; unsigned int index; - u64 entry_base; - struct qinval_entry *qinval_entry, *qinval_entries; + struct qinval_entry *qinval_entry; spin_lock_irqsave(&iommu->register_lock, flags); index = qinval_next_index(iommu); - entry_base = iommu->qinval_maddr + - ((index >> QINVAL_ENTRY_ORDER) << PAGE_SHIFT); - qinval_entries = map_vtd_domain_page(entry_base); - qinval_entry = &qinval_entries[index % (1 << QINVAL_ENTRY_ORDER)]; + qinval_entry = qi_map_entry(iommu, index); qinval_entry->q.iotlb_inv_dsc.lo.type = TYPE_INVAL_IOTLB; qinval_entry->q.iotlb_inv_dsc.lo.granu = granu; @@ -133,10 +135,11 @@ static int __must_check queue_invalidate_iotlb_sync(struct vtd_iommu *iommu, qinval_entry->q.iotlb_inv_dsc.hi.res_1 = 0; qinval_entry->q.iotlb_inv_dsc.hi.addr = addr >> PAGE_SHIFT_4K; - unmap_vtd_domain_page(qinval_entries); qinval_update_qtail(iommu, index); spin_unlock_irqrestore(&iommu->register_lock, flags); + unmap_vtd_domain_page(qinval_entry); + return invalidate_sync(iommu); } @@ -147,17 +150,13 @@ static int __must_check queue_invalidate_wait(struct vtd_iommu *iommu, static DEFINE_PER_CPU(uint32_t, poll_slot); unsigned int index; unsigned long flags; - u64 entry_base; - struct qinval_entry *qinval_entry, *qinval_entries; + struct qinval_entry *qinval_entry; uint32_t *this_poll_slot = &this_cpu(poll_slot); spin_lock_irqsave(&iommu->register_lock, flags); ACCESS_ONCE(*this_poll_slot) = QINVAL_STAT_INIT; index = qinval_next_index(iommu); - entry_base = iommu->qinval_maddr + - ((index >> QINVAL_ENTRY_ORDER) << PAGE_SHIFT); - qinval_entries = map_vtd_domain_page(entry_base); - qinval_entry = &qinval_entries[index % (1 << QINVAL_ENTRY_ORDER)]; + qinval_entry = qi_map_entry(iommu, index); qinval_entry->q.inv_wait_dsc.lo.type = TYPE_INVAL_WAIT; qinval_entry->q.inv_wait_dsc.lo.iflag = iflag; @@ -167,10 +166,11 @@ static int __must_check queue_invalidate_wait(struct vtd_iommu *iommu, qinval_entry->q.inv_wait_dsc.lo.sdata = QINVAL_STAT_DONE; qinval_entry->q.inv_wait_dsc.hi.saddr = virt_to_maddr(this_poll_slot); - unmap_vtd_domain_page(qinval_entries); qinval_update_qtail(iommu, index); spin_unlock_irqrestore(&iommu->register_lock, flags); + unmap_vtd_domain_page(qinval_entry); + /* Now we don't support interrupt method */ if ( sw ) { @@ -246,16 +246,12 @@ int qinval_device_iotlb_sync(struct vtd_iommu *iommu, struct pci_dev *pdev, { unsigned long flags; unsigned int index; - u64 entry_base; - struct qinval_entry *qinval_entry, *qinval_entries; + struct qinval_entry *qinval_entry; ASSERT(pdev); spin_lock_irqsave(&iommu->register_lock, flags); index = qinval_next_index(iommu); - entry_base = iommu->qinval_maddr + - ((index >> QINVAL_ENTRY_ORDER) << PAGE_SHIFT); - qinval_entries = map_vtd_domain_page(entry_base); - qinval_entry = &qinval_entries[index % (1 << QINVAL_ENTRY_ORDER)]; + qinval_entry = qi_map_entry(iommu, index); qinval_entry->q.dev_iotlb_inv_dsc.lo.type = TYPE_INVAL_DEVICE_IOTLB; qinval_entry->q.dev_iotlb_inv_dsc.lo.res_1 = 0; @@ -268,10 +264,11 @@ int qinval_device_iotlb_sync(struct vtd_iommu *iommu, struct pci_dev *pdev, qinval_entry->q.dev_iotlb_inv_dsc.hi.res_1 = 0; qinval_entry->q.dev_iotlb_inv_dsc.hi.addr = addr >> PAGE_SHIFT_4K; - unmap_vtd_domain_page(qinval_entries); qinval_update_qtail(iommu, index); spin_unlock_irqrestore(&iommu->register_lock, flags); + unmap_vtd_domain_page(qinval_entry); + return dev_invalidate_sync(iommu, pdev, did); } @@ -280,16 +277,12 @@ static int __must_check queue_invalidate_iec_sync(struct vtd_iommu *iommu, { unsigned long flags; unsigned int index; - u64 entry_base; - struct qinval_entry *qinval_entry, *qinval_entries; + struct qinval_entry *qinval_entry; int ret; spin_lock_irqsave(&iommu->register_lock, flags); index = qinval_next_index(iommu); - entry_base = iommu->qinval_maddr + - ((index >> QINVAL_ENTRY_ORDER) << PAGE_SHIFT); - qinval_entries = map_vtd_domain_page(entry_base); - qinval_entry = &qinval_entries[index % (1 << QINVAL_ENTRY_ORDER)]; + qinval_entry = qi_map_entry(iommu, index); qinval_entry->q.iec_inv_dsc.lo.type = TYPE_INVAL_IEC; qinval_entry->q.iec_inv_dsc.lo.granu = granu; @@ -299,10 +292,11 @@ static int __must_check queue_invalidate_iec_sync(struct vtd_iommu *iommu, qinval_entry->q.iec_inv_dsc.lo.res_2 = 0; qinval_entry->q.iec_inv_dsc.hi.res = 0; - unmap_vtd_domain_page(qinval_entries); qinval_update_qtail(iommu, index); spin_unlock_irqrestore(&iommu->register_lock, flags); + unmap_vtd_domain_page(qinval_entry); + ret = invalidate_sync(iommu); /* -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 14:33:58 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 14:33:58 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146838.270328 (Exim 4.92) (envelope-from ) id 1lwQQY-000468-91; Thu, 24 Jun 2021 14:33:58 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146838.270328; Thu, 24 Jun 2021 14:33:58 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwQQY-000460-68; Thu, 24 Jun 2021 14:33:58 +0000 Received: by outflank-mailman (input) for mailman id 146838; Thu, 24 Jun 2021 14:33:56 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwQQW-00045a-6x for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 14:33:56 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwQQW-000215-5v for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 14:33:56 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwQQW-0002mQ-4f for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 14:33:56 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=kMW7dNuZc30iDXKjoLa6L1bgfMk8UJN3uXCaYawLYxA=; b=HuyMlpanwdwhgd5y+TVc2s+rpl bdS6IX49kcFEdcj1o6G/bJcFKNfgwwKwv50UrO/OS25x+hPOwAjHHHzcSNyRjdGidilKpwTrva/SL Tin+tLR//ZH3WzLr+14Vsw+OPb/iM8ccWIaP8knh2S6dng3BWqidQx+qAym8XCt7snCo=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] VT-d: drop/move a few QI related constants Message-Id: Date: Thu, 24 Jun 2021 14:33:56 +0000 commit f7db924309c25b18951c058f80d3e5979114b055 Author: Jan Beulich AuthorDate: Thu Jun 24 16:30:51 2021 +0200 Commit: Jan Beulich CommitDate: Thu Jun 24 16:30:51 2021 +0200 VT-d: drop/move a few QI related constants Replace uses of QINVAL_ENTRY_ORDER and QINVAL_INDEX_SHIFT, such that the constants can be dropped. Move the remaining QINVAL_* ones to the single source file using them. Signed-off-by: Jan Beulich Reviewed-by: Kevin Tian --- xen/drivers/passthrough/vtd/iommu.h | 11 ----------- xen/drivers/passthrough/vtd/qinval.c | 24 ++++++++++++++++-------- 2 files changed, 16 insertions(+), 19 deletions(-) diff --git a/xen/drivers/passthrough/vtd/iommu.h b/xen/drivers/passthrough/vtd/iommu.h index 02390cb84b..ee28ecac6d 100644 --- a/xen/drivers/passthrough/vtd/iommu.h +++ b/xen/drivers/passthrough/vtd/iommu.h @@ -451,17 +451,6 @@ struct qinval_entry { }q; }; -/* Each entry is 16 bytes, so 2^8 entries per page */ -#define QINVAL_ENTRY_ORDER ( PAGE_SHIFT - 4 ) -#define QINVAL_MAX_ENTRY_NR (1u << (7 + QINVAL_ENTRY_ORDER)) - -/* Status data flag */ -#define QINVAL_STAT_INIT 0 -#define QINVAL_STAT_DONE 1 - -/* Queue invalidation head/tail shift */ -#define QINVAL_INDEX_SHIFT 4 - #define TYPE_INVAL_CONTEXT 0x1 #define TYPE_INVAL_IOTLB 0x2 #define TYPE_INVAL_DEVICE_IOTLB 0x3 diff --git a/xen/drivers/passthrough/vtd/qinval.c b/xen/drivers/passthrough/vtd/qinval.c index 2e4a55a869..b0e3672231 100644 --- a/xen/drivers/passthrough/vtd/qinval.c +++ b/xen/drivers/passthrough/vtd/qinval.c @@ -29,6 +29,13 @@ #include "extern.h" #include "../ats.h" +/* Each entry is 16 bytes, and there can be up to 2^7 pages. */ +#define QINVAL_MAX_ENTRY_NR (1u << (7 + PAGE_SHIFT_4K - 4)) + +/* Status data flag */ +#define QINVAL_STAT_INIT 0 +#define QINVAL_STAT_DONE 1 + static unsigned int __read_mostly qi_pg_order; static unsigned int __read_mostly qi_entry_nr; @@ -45,11 +52,11 @@ static unsigned int qinval_next_index(struct vtd_iommu *iommu) { unsigned int tail = dmar_readl(iommu->reg, DMAR_IQT_REG); - tail >>= QINVAL_INDEX_SHIFT; + tail /= sizeof(struct qinval_entry); /* (tail+1 == head) indicates a full queue, wait for HW */ while ( ((tail + 1) & (qi_entry_nr - 1)) == - (dmar_readl(iommu->reg, DMAR_IQH_REG) >> QINVAL_INDEX_SHIFT) ) + (dmar_readl(iommu->reg, DMAR_IQH_REG) / sizeof(struct qinval_entry)) ) { printk_once(XENLOG_ERR VTDPREFIX " IOMMU#%u: no QI slot available\n", iommu->index); @@ -66,7 +73,7 @@ static void qinval_update_qtail(struct vtd_iommu *iommu, unsigned int index) /* Need hold register lock when update tail */ ASSERT( spin_is_locked(&iommu->register_lock) ); val = (index + 1) & (qi_entry_nr - 1); - dmar_writel(iommu->reg, DMAR_IQT_REG, val << QINVAL_INDEX_SHIFT); + dmar_writel(iommu->reg, DMAR_IQT_REG, val * sizeof(struct qinval_entry)); } static struct qinval_entry *qi_map_entry(const struct vtd_iommu *iommu, @@ -413,17 +420,18 @@ int enable_qinval(struct vtd_iommu *iommu) * only one entry left. */ BUILD_BUG_ON(CONFIG_NR_CPUS * 2 >= QINVAL_MAX_ENTRY_NR); - qi_pg_order = get_order_from_bytes((num_present_cpus() * 2 + 1) << - (PAGE_SHIFT - - QINVAL_ENTRY_ORDER)); - qi_entry_nr = 1u << (qi_pg_order + QINVAL_ENTRY_ORDER); + qi_pg_order = get_order_from_bytes((num_present_cpus() * 2 + 1) * + sizeof(struct qinval_entry)); + qi_entry_nr = (PAGE_SIZE << qi_pg_order) / + sizeof(struct qinval_entry); dprintk(XENLOG_INFO VTDPREFIX, "QI: using %u-entry ring(s)\n", qi_entry_nr); } iommu->qinval_maddr = - alloc_pgtable_maddr(qi_entry_nr >> QINVAL_ENTRY_ORDER, + alloc_pgtable_maddr(PFN_DOWN(qi_entry_nr * + sizeof(struct qinval_entry)), iommu->node); if ( iommu->qinval_maddr == 0 ) { -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 14:44:10 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 14:44:10 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146852.270350 (Exim 4.92) (envelope-from ) id 1lwQaN-00060S-EN; Thu, 24 Jun 2021 14:44:07 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146852.270350; Thu, 24 Jun 2021 14:44:07 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwQaN-00060K-BQ; Thu, 24 Jun 2021 14:44:07 +0000 Received: by outflank-mailman (input) for mailman id 146852; Thu, 24 Jun 2021 14:44:05 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwQaL-00060E-Tp for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 14:44:05 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwQaL-0002Ap-Rn for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 14:44:05 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwQaL-0003th-Qo for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 14:44:05 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=ILQ3rvwGNNnH4ULFcgPWSye/yYHORVxkUx0tqFe6igs=; b=s+o/QKcUNAzhQcFmOBPHN5GRor KNPaa3nuPMQtzQwWPPvTfmt4Clgw2Wgg98LG6LCgF/u+TSfZFzsGcQb6a08+bx1OPsEKNI1Y0gYhK wkqjlvu1cO/yBXdeB2nIFNhGwFQasLlXg8GtYJqto7dGYYtvR06rBiWYTK2x3T9WRfCg=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] x86/HVM: wire up multicalls Message-Id: Date: Thu, 24 Jun 2021 14:44:05 +0000 commit 198a2bc6f149ca41e16d90fc73de2c81affe4490 Author: Jan Beulich AuthorDate: Thu Jun 24 16:35:39 2021 +0200 Commit: Jan Beulich CommitDate: Thu Jun 24 16:35:39 2021 +0200 x86/HVM: wire up multicalls To be able to use them from, in particular, the tool stack, they need to be supported for all guest types. Note that xc_resource_op() already does, so would not work without this on PVH Dom0. Signed-off-by: Jan Beulich Begrudingly acked-by: Andrew Cooper Acked-by: Ian Jackson --- xen/arch/x86/hvm/hypercall.c | 35 +++++++++++++++++++++++++++++++++++ xen/arch/x86/hypercall.c | 14 ++++++++++---- xen/arch/x86/pv/hypercall.c | 3 ++- xen/include/asm-x86/multicall.h | 12 ++++++++++++ 4 files changed, 59 insertions(+), 5 deletions(-) diff --git a/xen/arch/x86/hvm/hypercall.c b/xen/arch/x86/hvm/hypercall.c index 2d27fcbd2c..9e1f7249f3 100644 --- a/xen/arch/x86/hvm/hypercall.c +++ b/xen/arch/x86/hvm/hypercall.c @@ -26,6 +26,7 @@ #include #include #include +#include #include #include @@ -125,6 +126,7 @@ static const struct { hypercall_fn_t *native, *compat; } hvm_hypercall_table[] = { HVM_CALL(memory_op), + COMPAT_CALL(multicall), #ifdef CONFIG_GRANT_TABLE HVM_CALL(grant_table_op), #endif @@ -334,6 +336,39 @@ int hvm_hypercall(struct cpu_user_regs *regs) return curr->hcall_preempted ? HVM_HCALL_preempted : HVM_HCALL_completed; } +enum mc_disposition hvm_do_multicall_call(struct mc_state *state) +{ + struct vcpu *curr = current; + hypercall_fn_t *func = NULL; + + if ( hvm_guest_x86_mode(curr) == 8 ) + { + struct multicall_entry *call = &state->call; + + if ( call->op < ARRAY_SIZE(hvm_hypercall_table) ) + func = array_access_nospec(hvm_hypercall_table, call->op).native; + if ( func ) + call->result = func(call->args[0], call->args[1], call->args[2], + call->args[3], call->args[4], call->args[5]); + else + call->result = -ENOSYS; + } + else + { + struct compat_multicall_entry *call = &state->compat_call; + + if ( call->op < ARRAY_SIZE(hvm_hypercall_table) ) + func = array_access_nospec(hvm_hypercall_table, call->op).compat; + if ( func ) + call->result = func(call->args[0], call->args[1], call->args[2], + call->args[3], call->args[4], call->args[5]); + else + call->result = -ENOSYS; + } + + return !hvm_get_cpl(curr) ? mc_continue : mc_preempt; +} + /* * Local variables: * mode: C diff --git a/xen/arch/x86/hypercall.c b/xen/arch/x86/hypercall.c index 73c1bcb147..2370d31d3f 100644 --- a/xen/arch/x86/hypercall.c +++ b/xen/arch/x86/hypercall.c @@ -20,6 +20,7 @@ */ #include +#include #ifdef CONFIG_COMPAT #define ARGS(x, n) \ @@ -273,13 +274,18 @@ int hypercall_xlat_continuation(unsigned int *id, unsigned int nr, return rc; } -#ifndef CONFIG_PV -/* Stub for arch_do_multicall_call */ -enum mc_disposition arch_do_multicall_call(struct mc_state *mc) +enum mc_disposition arch_do_multicall_call(struct mc_state *state) { + const struct domain *currd = current->domain; + + if ( is_pv_domain(currd) ) + return pv_do_multicall_call(state); + + if ( is_hvm_domain(currd) ) + return hvm_do_multicall_call(state); + return mc_exit; } -#endif /* * Local variables: diff --git a/xen/arch/x86/pv/hypercall.c b/xen/arch/x86/pv/hypercall.c index d573f74aa1..200b8a05b0 100644 --- a/xen/arch/x86/pv/hypercall.c +++ b/xen/arch/x86/pv/hypercall.c @@ -23,6 +23,7 @@ #include #include #include +#include #include #ifdef CONFIG_PV32 @@ -245,7 +246,7 @@ void pv_hypercall(struct cpu_user_regs *regs) perfc_incr(hypercalls); } -enum mc_disposition arch_do_multicall_call(struct mc_state *state) +enum mc_disposition pv_do_multicall_call(struct mc_state *state) { struct vcpu *curr = current; unsigned long op; diff --git a/xen/include/asm-x86/multicall.h b/xen/include/asm-x86/multicall.h new file mode 100644 index 0000000000..7e1d4c121a --- /dev/null +++ b/xen/include/asm-x86/multicall.h @@ -0,0 +1,12 @@ +/****************************************************************************** + * asm-x86/multicall.h + */ + +#ifndef __ASM_X86_MULTICALL_H__ +#define __ASM_X86_MULTICALL_H__ + +#include + +typeof(arch_do_multicall_call) pv_do_multicall_call, hvm_do_multicall_call; + +#endif /* __ASM_X86_MULTICALL_H__ */ -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 14:44:17 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 14:44:17 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146853.270355 (Exim 4.92) (envelope-from ) id 1lwQaX-00062q-GQ; Thu, 24 Jun 2021 14:44:17 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146853.270355; Thu, 24 Jun 2021 14:44:17 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwQaX-00062i-Cv; Thu, 24 Jun 2021 14:44:17 +0000 Received: by outflank-mailman (input) for mailman id 146853; Thu, 24 Jun 2021 14:44:16 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwQaW-00062S-0R for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 14:44:16 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwQaV-0002BG-VH for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 14:44:15 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwQaV-0003uQ-UE for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 14:44:15 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=CvtwVmRpT2r7t2MfkIpEHq7DCXxYf9trjF2P3gU9K8s=; b=gUswChZrTZGU4WGwo07DV4P/JZ J6CBu72NhwPfVYRcJzY3yPLVRn+Qx6xK6p5sNaD7otEWOj6DqM0mYf6+2Psy41xyiM44dxEnG6P1Y YKVZt4BhacLBC+gEAXC4eGtNr9YZI/qeCxa3jTmTOzkRvQH5x49z1Hp5R6DK0Y+TuA2c=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] libxencall: osdep_hypercall() should return long Message-Id: Date: Thu, 24 Jun 2021 14:44:15 +0000 commit 6409210a5f51860cb17b5e0f97debe164dab26d7 Author: Jan Beulich AuthorDate: Thu Jun 24 16:38:37 2021 +0200 Commit: Jan Beulich CommitDate: Thu Jun 24 16:38:37 2021 +0200 libxencall: osdep_hypercall() should return long Some hypercalls, memory-op in particular, can return values requiring more than 31 bits to represent. Hence the underlying layers need to make sure they won't truncate such values. (Note that for Solaris the function also gets renamed, to match the other OSes.) Due to them merely propagating ioctl()'s return value, this change is benign on Linux and Solaris. IOW there's an actual effect here only for the BSDs and MiniOS, but even then further adjustments are needed at the xencall() level. Signed-off-by: Jan Beulich Acked-by: Andrew Cooper Acked-by: Ian Jackson --- tools/libs/call/freebsd.c | 2 +- tools/libs/call/linux.c | 2 +- tools/libs/call/minios.c | 2 +- tools/libs/call/netbsd.c | 2 +- tools/libs/call/private.h | 2 +- tools/libs/call/solaris.c | 2 +- 6 files changed, 6 insertions(+), 6 deletions(-) diff --git a/tools/libs/call/freebsd.c b/tools/libs/call/freebsd.c index 28bfd852c1..6d2e8704b3 100644 --- a/tools/libs/call/freebsd.c +++ b/tools/libs/call/freebsd.c @@ -62,7 +62,7 @@ int osdep_xencall_close(xencall_handle *xcall) return close(fd); } -int osdep_hypercall(xencall_handle *xcall, privcmd_hypercall_t *hypercall) +long osdep_hypercall(xencall_handle *xcall, privcmd_hypercall_t *hypercall) { int fd = xcall->fd; int ret; diff --git a/tools/libs/call/linux.c b/tools/libs/call/linux.c index 51fa4899eb..6d588e6bea 100644 --- a/tools/libs/call/linux.c +++ b/tools/libs/call/linux.c @@ -80,7 +80,7 @@ int osdep_xencall_close(xencall_handle *xcall) return 0; } -int osdep_hypercall(xencall_handle *xcall, privcmd_hypercall_t *hypercall) +long osdep_hypercall(xencall_handle *xcall, privcmd_hypercall_t *hypercall) { return ioctl(xcall->fd, IOCTL_PRIVCMD_HYPERCALL, hypercall); } diff --git a/tools/libs/call/minios.c b/tools/libs/call/minios.c index 9f7a96995f..3f33b83578 100644 --- a/tools/libs/call/minios.c +++ b/tools/libs/call/minios.c @@ -38,7 +38,7 @@ int osdep_xencall_close(xencall_handle *xcall) return 0; } -int osdep_hypercall(xencall_handle *xcall, privcmd_hypercall_t *hypercall) +long osdep_hypercall(xencall_handle *xcall, privcmd_hypercall_t *hypercall) { multicall_entry_t call; int i, ret; diff --git a/tools/libs/call/netbsd.c b/tools/libs/call/netbsd.c index 4dcc2919ba..91ae68357e 100644 --- a/tools/libs/call/netbsd.c +++ b/tools/libs/call/netbsd.c @@ -96,7 +96,7 @@ void osdep_free_pages(xencall_handle *xcall, void *ptr, size_t npages) free(ptr); } -int osdep_hypercall(xencall_handle *xcall, privcmd_hypercall_t *hypercall) +long osdep_hypercall(xencall_handle *xcall, privcmd_hypercall_t *hypercall) { int fd = xcall->fd; int error = ioctl(fd, IOCTL_PRIVCMD_HYPERCALL, hypercall); diff --git a/tools/libs/call/private.h b/tools/libs/call/private.h index 7944ac5baf..9c3aa432ef 100644 --- a/tools/libs/call/private.h +++ b/tools/libs/call/private.h @@ -55,7 +55,7 @@ struct xencall_handle { int osdep_xencall_open(xencall_handle *xcall); int osdep_xencall_close(xencall_handle *xcall); -int osdep_hypercall(xencall_handle *xcall, privcmd_hypercall_t *hypercall); +long osdep_hypercall(xencall_handle *xcall, privcmd_hypercall_t *hypercall); void *osdep_alloc_pages(xencall_handle *xcall, size_t nr_pages); void osdep_free_pages(xencall_handle *xcall, void *p, size_t nr_pages); diff --git a/tools/libs/call/solaris.c b/tools/libs/call/solaris.c index c63b6a329a..304262bd05 100644 --- a/tools/libs/call/solaris.c +++ b/tools/libs/call/solaris.c @@ -80,7 +80,7 @@ void osdep_free_hypercall_buffer(xencall_handle *xcall, void *ptr, free(ptr); } -int do_xen_hypercall(xencall_handle *xcall, privcmd_hypercall_t *hypercall) +long osdep_hypercall(xencall_handle *xcall, privcmd_hypercall_t *hypercall) { int fd = xcall->fd; return ioctl(fd, IOCTL_PRIVCMD_HYPERCALL, hypercall); -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 14:44:27 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 14:44:27 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146854.270358 (Exim 4.92) (envelope-from ) id 1lwQah-00066U-HW; Thu, 24 Jun 2021 14:44:27 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146854.270358; Thu, 24 Jun 2021 14:44:27 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwQah-00066M-EW; Thu, 24 Jun 2021 14:44:27 +0000 Received: by outflank-mailman (input) for mailman id 146854; Thu, 24 Jun 2021 14:44:26 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwQag-000660-3d for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 14:44:26 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwQag-0002BU-2T for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 14:44:26 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwQag-0003vT-1h for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 14:44:26 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=mjrTbeOPmAwQDfaLSJD6jHqBsbPtbGdF/XMB9712OcA=; b=7F/QtvvCe9I6TzwUVL4NQqa6V3 z19wiigveFfuYJY7cTytTLPn1TuwRk1bzaWHL+USt+coGz5PfW/yKwwkx6r9wXktqxQFQgbPfDO/Q HNGOoTkmPdLuLKaQZwbiIDr4tJZhkidImxvdfNBePZwViVxlhVjc3BtcqU3zXAtYEqx4=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] libxencall: introduce variant of xencall2() returning long Message-Id: Date: Thu, 24 Jun 2021 14:44:26 +0000 commit bef64f2c0019f828824a67f918604fe40768b1de Author: Jan Beulich AuthorDate: Thu Jun 24 16:39:02 2021 +0200 Commit: Jan Beulich CommitDate: Thu Jun 24 16:39:02 2021 +0200 libxencall: introduce variant of xencall2() returning long Some hypercalls, memory-op in particular, can return values requiring more than 31 bits to represent. Hence the underlying layers need to make sure they won't truncate such values. Signed-off-by: Jan Beulich Acked-by: Ian Jackson Acked-by: Andrew Cooper --- tools/include/xencall.h | 4 ++++ tools/libs/call/core.c | 11 +++++++++++ tools/libs/call/libxencall.map | 5 +++++ 3 files changed, 20 insertions(+) diff --git a/tools/include/xencall.h b/tools/include/xencall.h index 2d0c42ad5e..fc95ed0fe5 100644 --- a/tools/include/xencall.h +++ b/tools/include/xencall.h @@ -113,6 +113,10 @@ int xencall5(xencall_handle *xcall, unsigned int op, uint64_t arg1, uint64_t arg2, uint64_t arg3, uint64_t arg4, uint64_t arg5); +/* Variant(s) of the above, as needed, returning "long" instead of "int". */ +long xencall2L(xencall_handle *xcall, unsigned int op, + uint64_t arg1, uint64_t arg2); + /* * Allocate and free memory which is suitable for use as a pointer * argument to a hypercall. diff --git a/tools/libs/call/core.c b/tools/libs/call/core.c index 57d3a33e6b..02c4f8e1ae 100644 --- a/tools/libs/call/core.c +++ b/tools/libs/call/core.c @@ -127,6 +127,17 @@ int xencall2(xencall_handle *xcall, unsigned int op, return osdep_hypercall(xcall, &call); } +long xencall2L(xencall_handle *xcall, unsigned int op, + uint64_t arg1, uint64_t arg2) +{ + privcmd_hypercall_t call = { + .op = op, + .arg = { arg1, arg2 }, + }; + + return osdep_hypercall(xcall, &call); +} + int xencall3(xencall_handle *xcall, unsigned int op, uint64_t arg1, uint64_t arg2, uint64_t arg3) { diff --git a/tools/libs/call/libxencall.map b/tools/libs/call/libxencall.map index 6922b96511..ad7518b0e0 100644 --- a/tools/libs/call/libxencall.map +++ b/tools/libs/call/libxencall.map @@ -27,3 +27,8 @@ VERS_1.2 { global: xencall_fd; } VERS_1.1; + +VERS_1.3 { + global: + xencall2L; +} VERS_1.2; -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 14:44:37 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 14:44:37 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146855.270361 (Exim 4.92) (envelope-from ) id 1lwQar-00069C-J3; Thu, 24 Jun 2021 14:44:37 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146855.270361; Thu, 24 Jun 2021 14:44:37 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwQar-000694-GF; Thu, 24 Jun 2021 14:44:37 +0000 Received: by outflank-mailman (input) for mailman id 146855; Thu, 24 Jun 2021 14:44:36 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwQaq-00068t-6e for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 14:44:36 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwQaq-0002Bh-5W for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 14:44:36 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwQaq-0003wN-4q for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 14:44:36 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=WE5Azbxj1bwL23jl5WlyMGHC9ShDgZ6XxGoblAJgSeE=; b=2OWYLTniUKqN4XPrE7ODSbldrO Ln5Fh8epf7jXZ6nPEycXkEIKfkisKhg/DROie5W4XdfrdcVrvQsS3ogHUd4ptZcAPjqxKUNNLD65u iUqKRzLQYuNcfa/k8M1iukYMp9x5Tx8L62FkNDfKeC66BGXUrVUuncQAJ8luLw97InfQ=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] libxc: use multicall for memory-op on Linux (and Solaris) Message-Id: Date: Thu, 24 Jun 2021 14:44:36 +0000 commit 6f02d1ea4a109a32f346a5d1de63a42b291c354c Author: Jan Beulich AuthorDate: Thu Jun 24 16:39:26 2021 +0200 Commit: Jan Beulich CommitDate: Thu Jun 24 16:39:26 2021 +0200 libxc: use multicall for memory-op on Linux (and Solaris) Some sub-functions, XENMEM_maximum_gpfn and XENMEM_maximum_ram_page in particular, can return values requiring more than 31 bits to represent. Hence we cannot issue the hypercall directly when the return value of ioctl() is used to propagate this value. This is the case for Linux and Solaris (and hence needs changing), while the BSDs avoid using the return value for dual purposes altogether, and MiniOS already wraps all hypercalls in a multicall. Suggested-by: Jürgen Groß Signed-off-by: Jan Beulich Acked-by: Ian Jackson Acked-by: Andrew Cooper --- tools/libs/ctrl/xc_private.c | 43 +++++++++++++++++++++++++++++++++++++++++-- 1 file changed, 41 insertions(+), 2 deletions(-) diff --git a/tools/libs/ctrl/xc_private.c b/tools/libs/ctrl/xc_private.c index d94f846686..c0422662f0 100644 --- a/tools/libs/ctrl/xc_private.c +++ b/tools/libs/ctrl/xc_private.c @@ -337,8 +337,47 @@ long do_memory_op(xc_interface *xch, int cmd, void *arg, size_t len) goto out1; } - ret = xencall2(xch->xcall, __HYPERVISOR_memory_op, - cmd, HYPERCALL_BUFFER_AS_ARG(arg)); +#if defined(__linux__) || defined(__sun__) + /* + * Some sub-ops return values which don't fit in "int". On platforms + * without a specific hypercall return value field in the privcmd + * interface structure, issue the request as a single-element multicall, + * to be able to capture the full return value. + */ + if ( sizeof(long) > sizeof(int) ) + { + multicall_entry_t multicall = { + .op = __HYPERVISOR_memory_op, + .args[0] = cmd, + .args[1] = HYPERCALL_BUFFER_AS_ARG(arg), + }, *call = &multicall; + DECLARE_HYPERCALL_BOUNCE(call, sizeof(*call), + XC_HYPERCALL_BUFFER_BOUNCE_BOTH); + + if ( xc_hypercall_bounce_pre(xch, call) ) + { + PERROR("Could not bounce buffer for memory_op hypercall"); + goto out1; + } + + ret = do_multicall_op(xch, HYPERCALL_BUFFER(call), 1); + + xc_hypercall_bounce_post(xch, call); + + if ( !ret ) + { + ret = multicall.result; + if ( multicall.result > ~0xfffUL ) + { + errno = -ret; + ret = -1; + } + } + } + else +#endif + ret = xencall2L(xch->xcall, __HYPERVISOR_memory_op, + cmd, HYPERCALL_BUFFER_AS_ARG(arg)); xc_hypercall_bounce_post(xch, arg); out1: -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 14:44:47 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 14:44:47 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146856.270368 (Exim 4.92) (envelope-from ) id 1lwQb1-0006CT-Lt; Thu, 24 Jun 2021 14:44:47 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146856.270368; Thu, 24 Jun 2021 14:44:47 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwQb1-0006CL-Hp; Thu, 24 Jun 2021 14:44:47 +0000 Received: by outflank-mailman (input) for mailman id 146856; Thu, 24 Jun 2021 14:44:46 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwQb0-0006Bt-9T for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 14:44:46 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwQb0-0002Bs-8P for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 14:44:46 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwQb0-0003x4-7k for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 14:44:46 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=v7rqvxwik0f8s2oyfsugY2F0aynP6LREtl5WxQEPoa8=; b=KcX/l/ZdKN2ytzWhnrQ9Myy5NN VGsbA9AtYsruZH9TzweanR6kph1h7wBHDgl1Nn+M6WTnQjlstSLW8GF60DSSwFlGv3CtugkMGWEha IptiY6GVXRuR7vaOhtmK16HS6Ya6UXT97EpQ2eKIz6gnP+/2cli6oz+NOEkc0C8lU2h8=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] libxencall: drop bogus mentioning of xencall6() Message-Id: Date: Thu, 24 Jun 2021 14:44:46 +0000 commit 7ffbed8681a0ab203535ab1e13c48f18ecdba907 Author: Jan Beulich AuthorDate: Thu Jun 24 16:39:55 2021 +0200 Commit: Jan Beulich CommitDate: Thu Jun 24 16:39:55 2021 +0200 libxencall: drop bogus mentioning of xencall6() There's no xencall6(), so the version script also shouldn't mention it. If such a function would ever appear, it shouldn't land in version 1.0. No change to the generated binary, nor abi-dumper's view of the object. Signed-off-by: Jan Beulich Acked-by: Ian Jackson Reviewed-by: Andrew Cooper --- tools/libs/call/libxencall.map | 1 - 1 file changed, 1 deletion(-) diff --git a/tools/libs/call/libxencall.map b/tools/libs/call/libxencall.map index ad7518b0e0..d18a3174e9 100644 --- a/tools/libs/call/libxencall.map +++ b/tools/libs/call/libxencall.map @@ -9,7 +9,6 @@ VERS_1.0 { xencall3; xencall4; xencall5; - xencall6; xencall_alloc_buffer; xencall_free_buffer; -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 14:44:57 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 14:44:57 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146857.270370 (Exim 4.92) (envelope-from ) id 1lwQbB-0006G5-Om; Thu, 24 Jun 2021 14:44:57 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146857.270370; Thu, 24 Jun 2021 14:44:57 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwQbB-0006Fx-Lo; Thu, 24 Jun 2021 14:44:57 +0000 Received: by outflank-mailman (input) for mailman id 146857; Thu, 24 Jun 2021 14:44:56 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwQbA-0006Fa-Cv for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 14:44:56 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwQbA-0002C3-Bu for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 14:44:56 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwQbA-0003y2-B8 for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 14:44:56 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=EGr8u+tFAIMt4F++Uhw/2Hijs2j2Gqxw2PHWj5FPuEE=; b=M4PY1QX3q0PEWX82ym5ozGpspm H0m9I6iYBRXtGsgwHZ+ihOvZFbFz2OPscFFmEQQEiHmc2TnMW10njvbFX1ZWy8z6hqKcFB7ki9pjZ vDUta4jm9e+PWUzt+XLqsRYiKdzVAne6x9Pg/ydnh5wFW1QGeSnXMQlLJ7I2WCCuBEGw=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] libxc: make xc_domain_maximum_gpfn() endianness-agnostic Message-Id: Date: Thu, 24 Jun 2021 14:44:56 +0000 commit 190facdf877c8fc298dafa584f69e7aff8606c35 Author: Jan Beulich AuthorDate: Thu Jun 24 16:40:57 2021 +0200 Commit: Jan Beulich CommitDate: Thu Jun 24 16:40:57 2021 +0200 libxc: make xc_domain_maximum_gpfn() endianness-agnostic libxc generally uses uint32_t to represent domain IDs. This is fine as long as addresses of such variables aren't taken, to then pass into hypercalls: To the hypervisor, a domain ID is a 16-bit value. Introduce a wrapper struct to deal with the issue. (On architectures with arguments passed in registers, an intermediate variable would have been created by the compiler already anyway, just one of the wrong type.) The public interface change is both source and binary compatible for the architectures we currently support. Signed-off-by: Jan Beulich Acked-by: Ian Jackson Reviewed-by: Andrew Cooper --- tools/libs/ctrl/xc_domain.c | 3 ++- xen/common/memory.c | 9 ++++++--- xen/include/public/memory.h | 11 +++++++++-- 3 files changed, 17 insertions(+), 6 deletions(-) diff --git a/tools/libs/ctrl/xc_domain.c b/tools/libs/ctrl/xc_domain.c index 7d118848f1..3803871538 100644 --- a/tools/libs/ctrl/xc_domain.c +++ b/tools/libs/ctrl/xc_domain.c @@ -856,7 +856,8 @@ int xc_domain_get_tsc_info(xc_interface *xch, int xc_domain_maximum_gpfn(xc_interface *xch, uint32_t domid, xen_pfn_t *gpfns) { - long rc = do_memory_op(xch, XENMEM_maximum_gpfn, &domid, sizeof(domid)); + struct xen_memory_domain dom = { .domid = domid }; + long rc = do_memory_op(xch, XENMEM_maximum_gpfn, &dom, sizeof(dom)); if ( rc >= 0 ) { diff --git a/xen/common/memory.c b/xen/common/memory.c index 72a6b70cb5..ba212f971c 100644 --- a/xen/common/memory.c +++ b/xen/common/memory.c @@ -1351,7 +1351,6 @@ long do_memory_op(unsigned long cmd, XEN_GUEST_HANDLE_PARAM(void) arg) long rc; struct xen_memory_reservation reservation; struct memop_args args; - domid_t domid; unsigned long start_extent = cmd >> MEMOP_EXTENT_SHIFT; int op = cmd & MEMOP_CMD_MASK; @@ -1452,13 +1451,16 @@ long do_memory_op(unsigned long cmd, XEN_GUEST_HANDLE_PARAM(void) arg) case XENMEM_current_reservation: case XENMEM_maximum_reservation: case XENMEM_maximum_gpfn: + { + struct xen_memory_domain domain; + if ( unlikely(start_extent) ) return -EINVAL; - if ( copy_from_guest(&domid, arg, 1) ) + if ( copy_from_guest(&domain, arg, 1) ) return -EFAULT; - d = rcu_lock_domain_by_any_id(domid); + d = rcu_lock_domain_by_any_id(domain.domid); if ( d == NULL ) return -ESRCH; @@ -1486,6 +1488,7 @@ long do_memory_op(unsigned long cmd, XEN_GUEST_HANDLE_PARAM(void) arg) rcu_unlock_domain(d); break; + } case XENMEM_add_to_physmap: { diff --git a/xen/include/public/memory.h b/xen/include/public/memory.h index 50e73eef98..383a9468c3 100644 --- a/xen/include/public/memory.h +++ b/xen/include/public/memory.h @@ -148,16 +148,23 @@ DEFINE_XEN_GUEST_HANDLE(xen_memory_exchange_t); */ #define XENMEM_maximum_ram_page 2 +struct xen_memory_domain { + /* [IN] Domain information is being queried for. */ + domid_t domid; +}; + /* * Returns the current or maximum memory reservation, in pages, of the * specified domain (may be DOMID_SELF). Returns -ve errcode on failure. - * arg == addr of domid_t. + * arg == addr of struct xen_memory_domain. */ #define XENMEM_current_reservation 3 #define XENMEM_maximum_reservation 4 /* - * Returns the maximum GPFN in use by the guest, or -ve errcode on failure. + * Returns the maximum GFN in use by the specified domain (may be DOMID_SELF). + * Returns -ve errcode on failure. + * arg == addr of struct xen_memory_domain. */ #define XENMEM_maximum_gpfn 14 -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Thu Jun 24 15:00:08 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Thu, 24 Jun 2021 15:00:08 +0000 Received: from list by lists.xenproject.org with outflank-mailman.146873.270408 (Exim 4.92) (envelope-from ) id 1lwQpp-00016w-K0; Thu, 24 Jun 2021 15:00:05 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 146873.270408; Thu, 24 Jun 2021 15:00:05 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwQpp-00016o-Fk; Thu, 24 Jun 2021 15:00:05 +0000 Received: by outflank-mailman (input) for mailman id 146873; Thu, 24 Jun 2021 15:00:04 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwQpo-00010K-O7 for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 15:00:04 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwQpo-0002VH-N6 for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 15:00:04 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwQpo-0005D2-MF for xen-changelog@lists.xenproject.org; Thu, 24 Jun 2021 15:00:04 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=6sjqbEBbZDi6eIvaEjlZMnAfp6AI59LyUFFIhTJutYU=; b=5E7nYYUlsHnF4ZTwv6H8E/i3oZ ioX9woV91J6TcoaYSCeax7bYUhv24WBSmfzP+6t4oLrHmmUN3f+IoWpTTXJPeU9xq7s589s8crTnL 3aJbcNDBhJL+1Pd2zuilsBnxLxiB5V9V5adTSESWuCHLivW3MNYj0wQrBJKee7t2v14c=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] tools/xenstored: Remove redundant check in socket_can_process() Message-Id: Date: Thu, 24 Jun 2021 15:00:04 +0000 commit e87d8f60fa9b6eaa6a2357545a96e4fff05dbef0 Author: Julien Grall AuthorDate: Thu Jun 24 15:55:03 2021 +0100 Commit: Julien Grall CommitDate: Thu Jun 24 15:55:49 2021 +0100 tools/xenstored: Remove redundant check in socket_can_process() Commit 3adfb50315d9 ("tools/xenstored: Introduce a wrapper for conn->funcs->can_{read, write}") consolidated the check !conn->is_ignored in two new wrappers. This means the check in socket_can_process() is now redundant. In fact it should have been removed in orignal commit (as it was done for the domain helpers). Reported-by: Raphael Ning Reviewed-by: Juergen Gross --- tools/xenstore/xenstored_core.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/tools/xenstore/xenstored_core.c b/tools/xenstore/xenstored_core.c index 9ffd2ac66d..cf7297a96c 100644 --- a/tools/xenstore/xenstored_core.c +++ b/tools/xenstore/xenstored_core.c @@ -1752,7 +1752,7 @@ static bool socket_can_process(struct connection *conn, int mask) return false; } - return (fds[conn->pollfd_idx].revents & mask) && !conn->is_ignored; + return (fds[conn->pollfd_idx].revents & mask); } static bool socket_can_write(struct connection *conn) -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Fri Jun 25 09:44:10 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Fri, 25 Jun 2021 09:44:10 +0000 Received: from list by lists.xenproject.org with outflank-mailman.147092.270891 (Exim 4.92) (envelope-from ) id 1lwiNZ-0006YC-Ik; Fri, 25 Jun 2021 09:44:05 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 147092.270891; Fri, 25 Jun 2021 09:44:05 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwiNZ-0006Y2-F4; Fri, 25 Jun 2021 09:44:05 +0000 Received: by outflank-mailman (input) for mailman id 147092; Fri, 25 Jun 2021 09:44:04 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwiNY-0006XP-BZ for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 09:44:04 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwiNY-0007ic-Al for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 09:44:04 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwiNY-0002y9-A1 for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 09:44:04 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=VzBp7uwVpHUK50FOr4IcDWjDXp1rvu5Vr9V8tei+ipc=; b=MSErlMIsBBmRLiawB4Gt/L1/Qy 11veUTnvo0yZW/xIrEKNyQ5J7IjlLL2otr32Bd/X5ij2AW9Ye6dDYR5YeKFS5S3Yy4xHk44dtW07I Krw9SMz3bC2juk+RQYw4YLAOEX4laMy21Lm4jMSnjM8E+eIfArGW8E7LpACLhLdRzqTM=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] tools/xenstored: Correctly read the requests header from the stream Message-Id: Date: Fri, 25 Jun 2021 09:44:04 +0000 commit 8a9b94982b0e0928ad874907f5a5b005944ab7cf Author: Julien Grall AuthorDate: Fri Jun 25 07:45:22 2021 +0100 Commit: Julien Grall CommitDate: Fri Jun 25 10:40:06 2021 +0100 tools/xenstored: Correctly read the requests header from the stream Commit c0fe360f42 ("tools/xenstored: Extend restore code to handle multiple input buffer") extend the read_buffered_state() to support multiple input buffers. Unfortunately, the commit didn't go far enough and still used sc->data (start of the buffers) for retrieving the header. This would lead to read the wrong headers for second and follow-up commands. Use data in place for sc->data for the source of the memcpy()s. Fixes: c0fe360f42 ("tools/xenstored: Extend restore code to handle multiple input buffer") Reported-by: Raphael Ning Signed-off-by: Julien Grall Reviewed-by: Juergen Gross --- tools/xenstore/xenstored_core.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/tools/xenstore/xenstored_core.c b/tools/xenstore/xenstored_core.c index cf7297a96c..16c856730c 100644 --- a/tools/xenstore/xenstored_core.c +++ b/tools/xenstore/xenstored_core.c @@ -2717,11 +2717,11 @@ void read_state_buffered_data(const void *ctx, struct connection *conn, len = sc->data_in_len - (data - sc->data); if (len < sizeof(bdata->hdr)) { bdata->inhdr = true; - memcpy(&bdata->hdr, sc->data, len); + memcpy(&bdata->hdr, data, len); bdata->used = len; } else { bdata->inhdr = false; - memcpy(&bdata->hdr, sc->data, sizeof(bdata->hdr)); + memcpy(&bdata->hdr, data, sizeof(bdata->hdr)); if (bdata->hdr.msg.len <= DEFAULT_BUFFER_SIZE) bdata->buffer = bdata->default_buffer; else -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Fri Jun 25 11:44:10 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Fri, 25 Jun 2021 11:44:10 +0000 Received: from list by lists.xenproject.org with outflank-mailman.147130.270984 (Exim 4.92) (envelope-from ) id 1lwkFi-0002J4-8M; Fri, 25 Jun 2021 11:44:06 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 147130.270984; Fri, 25 Jun 2021 11:44:06 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwkFi-0002Iw-5H; Fri, 25 Jun 2021 11:44:06 +0000 Received: by outflank-mailman (input) for mailman id 147130; Fri, 25 Jun 2021 11:44:05 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwkFg-0002In-Ur for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 11:44:04 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwkFg-0001Lw-Tz for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 11:44:04 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwkFg-0004Np-T3 for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 11:44:04 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=SHi6Cjt96JBnEo/aXA2t1S3H3uMB1e6bLnN9zrQCvOk=; b=dmBnqPE8Mb0+RBUTx56DuYVREf yqr2+o6BuiCzGHEQt98sy5vx8LClxXn4cNsihpkRjbGMbq/YO5/6bh7gCNJtKSkmFaRkpaxbl4psg HxVkthnmzl2JHWHhVOx/v+EuCCUcCT1x3kDiQ7ESApiRdSlNPJMVtsO5K1NAG5vAYqv4=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] libxencall: Bump SONAME following new functionality Message-Id: Date: Fri, 25 Jun 2021 11:44:04 +0000 commit 01a2d001dea2219c9702afbe2d6fd1b0af539203 Author: Andrew Cooper AuthorDate: Thu Jun 24 18:49:14 2021 +0100 Commit: Andrew Cooper CommitDate: Fri Jun 25 12:38:18 2021 +0100 libxencall: Bump SONAME following new functionality Fixes: bef64f2c00 ("libxencall: introduce variant of xencall2() returning long") Signed-off-by: Andrew Cooper Reviewed-by: Ian Jackson --- tools/libs/call/Makefile | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/tools/libs/call/Makefile b/tools/libs/call/Makefile index 4ed201b3b3..93d404b79e 100644 --- a/tools/libs/call/Makefile +++ b/tools/libs/call/Makefile @@ -2,7 +2,7 @@ XEN_ROOT = $(CURDIR)/../../.. include $(XEN_ROOT)/tools/Rules.mk MAJOR = 1 -MINOR = 2 +MINOR = 3 SRCS-y += core.c buffer.c SRCS-$(CONFIG_Linux) += linux.c -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Fri Jun 25 12:11:08 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Fri, 25 Jun 2021 12:11:08 +0000 Received: from list by lists.xenproject.org with outflank-mailman.147147.271019 (Exim 4.92) (envelope-from ) id 1lwkfp-0006Ku-Ax; Fri, 25 Jun 2021 12:11:05 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 147147.271019; Fri, 25 Jun 2021 12:11:05 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwkfp-0006Km-7x; Fri, 25 Jun 2021 12:11:05 +0000 Received: by outflank-mailman (input) for mailman id 147147; Fri, 25 Jun 2021 12:11:04 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwkfo-0006Kg-Iu for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 12:11:04 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwkfo-0001o3-FM for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 12:11:04 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwkfo-0006cH-EH for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 12:11:04 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=dFR3KUbaT0edKNYAP1ffw5RMCtX2MkE5wLJ0F5XChoI=; b=6wXvXxNTpjgXh9wkb57CO6zLIA y/AgNi7EfP6TMbHRgv+T3EhuScQ60l4aKHtK1A4fKnpF0NxnTzFHlhI5U3g+AlLK+bjXc6uHNTih2 vyWZXtcgEB0IiOaAuXw1tpkyxu8IdvytsmfNuUXLuhtf3NeF/zUgnWMN5No3DbyB9UpM=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] IOMMU/PCI: don't let domain cleanup continue when device de-assignment failed Message-Id: Date: Fri, 25 Jun 2021 12:11:04 +0000 commit f591755823a7e94fc6b4b8ddce71f0421a94fa09 Author: Jan Beulich AuthorDate: Fri Jun 25 14:06:55 2021 +0200 Commit: Jan Beulich CommitDate: Fri Jun 25 14:06:55 2021 +0200 IOMMU/PCI: don't let domain cleanup continue when device de-assignment failed Failure here could in principle mean the device may still be issuing DMA requests, which would continue to be translated by the page tables the device entry currently points at. With this we cannot allow the subsequent cleanup step of freeing the page tables to occur, to prevent use-after-free issues. We would need to accept, for the time being, that in such a case the remaining domain resources will all be leaked, and the domain will continue to exist as a zombie. However, with flushes no longer timing out (and with proper timeout detection for device I/O TLB flushing yet to be implemented), there's no way anymore for failures to occur, except due to bugs elsewhere. Hence the change here is merely a "just in case" one. In order to continue the loop in spite of an error, we can't use pci_get_pdev_by_domain() anymore. I have no idea why it was used here in the first place, instead of the cheaper list iteration. Signed-off-by: Jan Beulich Reviewed-by: Paul Durrant --- xen/drivers/passthrough/pci.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/xen/drivers/passthrough/pci.c b/xen/drivers/passthrough/pci.c index 199ce08612..fc4fa2e5c3 100644 --- a/xen/drivers/passthrough/pci.c +++ b/xen/drivers/passthrough/pci.c @@ -894,7 +894,7 @@ static int deassign_device(struct domain *d, uint16_t seg, uint8_t bus, int pci_release_devices(struct domain *d) { - struct pci_dev *pdev; + struct pci_dev *pdev, *tmp; u8 bus, devfn; int ret; @@ -905,15 +905,15 @@ int pci_release_devices(struct domain *d) pcidevs_unlock(); return ret; } - while ( (pdev = pci_get_pdev_by_domain(d, -1, -1, -1)) ) + list_for_each_entry_safe ( pdev, tmp, &d->pdev_list, domain_list ) { bus = pdev->bus; devfn = pdev->devfn; - deassign_device(d, pdev->seg, bus, devfn); + ret = deassign_device(d, pdev->seg, bus, devfn) ?: ret; } pcidevs_unlock(); - return 0; + return ret; } #define PCI_CLASS_BRIDGE_HOST 0x0600 -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Fri Jun 25 22:11:14 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Fri, 25 Jun 2021 22:11:14 +0000 Received: from list by lists.xenproject.org with outflank-mailman.147445.271804 (Exim 4.92) (envelope-from ) id 1lwu2U-0002b1-DI; Fri, 25 Jun 2021 22:11:06 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 147445.271804; Fri, 25 Jun 2021 22:11:06 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwu2U-0002at-AR; Fri, 25 Jun 2021 22:11:06 +0000 Received: by outflank-mailman (input) for mailman id 147445; Fri, 25 Jun 2021 22:11:05 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwu2T-0002an-Gh for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 22:11:05 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwu2T-00052A-Fm for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 22:11:05 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwu2T-0007sF-El for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 22:11:05 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=9r8QtTTbN9G01ATZ4HKl25jHCpX9VJ//l1KyqHiJSVg=; b=Cc5DbnayERdiRNF1Cl9YJSJy0A jK1WRStyMdjWMTPvAzSMbBsdOfpYbdZ5QZ0mXZ57lA382iAqpAfEjNF1DicjbJfb3Ei8ba68HSYMr XxsOh5exC1MYWc340mVcZdx2BHLkofLbrSUKjXYkmArrQcmUFLJvCsbWCVDKXSomSAOg=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] xen/arm: smmuv1: Fixed stream matching register allocation Message-Id: Date: Fri, 25 Jun 2021 22:11:05 +0000 commit 93fe1817e7ca8c80f93c098a191e1cc3483190df Author: Rahul Singh AuthorDate: Fri Jun 25 17:37:26 2021 +0100 Commit: Stefano Stabellini CommitDate: Fri Jun 25 14:53:48 2021 -0700 xen/arm: smmuv1: Fixed stream matching register allocation SMR allocation should be based on the number of supported stream matching register for each SMMU device. Issue introduced by commit 5e08586afbb90b2e2d56c175c07db77a4afa873c when backported the patches from Linux to XEN to fix the stream match conflict issue when two devices have the same stream-id. Acked-by: Stefano Stabellini Tested-by: Stefano Stabellini Signed-off-by: Rahul Singh --- xen/drivers/passthrough/arm/smmu.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/xen/drivers/passthrough/arm/smmu.c b/xen/drivers/passthrough/arm/smmu.c index 1a68c2ab3b..05c2d238b3 100644 --- a/xen/drivers/passthrough/arm/smmu.c +++ b/xen/drivers/passthrough/arm/smmu.c @@ -149,6 +149,7 @@ typedef enum irqreturn irqreturn_t; #define kzalloc(size, flags) _xzalloc(size, sizeof(void *)) #define devm_kzalloc(dev, size, flags) _xzalloc(size, sizeof(void *)) #define kmalloc_array(size, n, flags) _xmalloc_array(size, sizeof(void *), n) +#define kzalloc_array(size, n, flags) _xzalloc_array(size, sizeof(void *), n) static void __iomem *devm_ioremap_resource(struct device *dev, struct resource *res) @@ -2221,7 +2222,7 @@ static int arm_smmu_device_cfg_probe(struct arm_smmu_device *smmu) smmu->smr_mask_mask = smr >> SMR_MASK_SHIFT; /* Zero-initialised to mark as invalid */ - smmu->smrs = devm_kzalloc(smmu->dev, sizeof(*smmu->smrs), GFP_KERNEL); + smmu->smrs = kzalloc_array(sizeof(*smmu->smrs), size, GFP_KERNEL); if (!smmu->smrs) return -ENOMEM; -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Fri Jun 25 22:11:16 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Fri, 25 Jun 2021 22:11:16 +0000 Received: from list by lists.xenproject.org with outflank-mailman.147446.271808 (Exim 4.92) (envelope-from ) id 1lwu2e-0002cu-FW; Fri, 25 Jun 2021 22:11:16 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 147446.271808; Fri, 25 Jun 2021 22:11:16 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwu2e-0002cl-CA; Fri, 25 Jun 2021 22:11:16 +0000 Received: by outflank-mailman (input) for mailman id 147446; Fri, 25 Jun 2021 22:11:15 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwu2d-0002cZ-Jh for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 22:11:15 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwu2d-00052E-In for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 22:11:15 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwu2d-0007t4-I1 for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 22:11:15 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=4mNQ/3duFq2ha7ixDrIbHWjYBV60H7vcaPDmjApJ6MU=; b=LRdLRfrMhuI4N+QDR70SM2hS2N NT7ui3sDZLB1fQrpp9yBhnThMe5R3ldeZJy5qf7yj3XU9JnRCExfRbJ+zEu7aNhfs/0v27WSQLMos XR+AZllvvy+dze9slD2hNOIdGmaopZlic+/XYBA55r6bcm3NZ3jHFcosDkk+W0ILt+fk=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] xen/arm: smmuv1: Set privileged attr to 'default' Message-Id: Date: Fri, 25 Jun 2021 22:11:15 +0000 commit bb11edcec1a953ce590da797b0d005cd60f21e83 Author: Rahul Singh AuthorDate: Fri Jun 25 17:37:27 2021 +0100 Commit: Stefano Stabellini CommitDate: Fri Jun 25 14:54:17 2021 -0700 xen/arm: smmuv1: Set privileged attr to 'default' Backport commit e19898077cfb642fe151ba22981e795c74d9e114 "iommu/arm-smmu: Set privileged attribute to 'default' instead of 'unprivileged'" Original commit message: Currently the driver sets all the device transactions privileges to UNPRIVILEGED, but there are cases where the iommu masters wants to isolate privileged supervisor and unprivileged user. So don't override the privileged setting to unprivileged, instead set it to default as incoming and let it be controlled by the pagetable settings. Acked-by: Will Deacon Signed-off-by: Sricharan R Signed-off-by: Will Deacon Signed-off-by: Rahul Singh Acked-by: Stefano Stabellini Tested-by: Stefano Stabellini --- xen/drivers/passthrough/arm/smmu.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/xen/drivers/passthrough/arm/smmu.c b/xen/drivers/passthrough/arm/smmu.c index 05c2d238b3..da2cd457d7 100644 --- a/xen/drivers/passthrough/arm/smmu.c +++ b/xen/drivers/passthrough/arm/smmu.c @@ -1567,7 +1567,7 @@ static int arm_smmu_domain_add_master(struct arm_smmu_domain *smmu_domain, continue; s2cr[idx].type = type ; - s2cr[idx].privcfg = S2CR_PRIVCFG_UNPRIV; + s2cr[idx].privcfg = S2CR_PRIVCFG_DEFAULT; s2cr[idx].cbndx = cbndx; arm_smmu_write_s2cr(smmu, idx); } -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Fri Jun 25 23:33:10 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Fri, 25 Jun 2021 23:33:10 +0000 Received: from list by lists.xenproject.org with outflank-mailman.147459.271848 (Exim 4.92) (envelope-from ) id 1lwvJn-0001AM-DL; Fri, 25 Jun 2021 23:33:03 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 147459.271848; Fri, 25 Jun 2021 23:33:03 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvJn-0001AE-AU; Fri, 25 Jun 2021 23:33:03 +0000 Received: by outflank-mailman (input) for mailman id 147459; Fri, 25 Jun 2021 23:33:02 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvJm-0001A8-HA for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:33:02 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvJm-0006OK-Dp for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:33:02 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwvJm-0006W9-Cp for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:33:02 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=HDarlu80FdGlLP8kHxyH86n/0AxRshHJztYFnyzOVMU=; b=Cr+OsyUzLtKKp9XCd93Tvgq7L4 HPe5Pgs+UTuUGPi8EhHMK6vS+JitloNVy/XEoxydSFrWi7mNEtpKll4HIV5LyUYx1VRqMmfaYM2gc VdG6mXuaMtSFnjsxtsYNb2DcJdWZdUFBlZvjQLbZPGpXGjOyC86OGiiIBS24L5/Q6XR4=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] MAINTAINERS: Add myself as reviewers for tools/xenstore Message-Id: Date: Fri, 25 Jun 2021 23:33:02 +0000 commit 82b1b19469cb5595c589898c26b5f33f37f90963 Author: Julien Grall AuthorDate: Thu Jun 24 09:02:56 2021 +0100 Commit: Julien Grall CommitDate: Thu Jun 24 09:02:56 2021 +0100 MAINTAINERS: Add myself as reviewers for tools/xenstore I would like to help reviewing Xenstored patches. It is more convenient to find them if I am CCed. Signed-off-by: Julien Grall Acked-by: Juergen Gross --- MAINTAINERS | 1 + 1 file changed, 1 insertion(+) diff --git a/MAINTAINERS b/MAINTAINERS index 39750bb75d..dd8c011456 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -628,6 +628,7 @@ XENSTORE M: Ian Jackson M: Wei Liu M: Juergen Gross +R: Julien Grall S: Supported F: tools/xenstore/ -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Fri Jun 25 23:33:13 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Fri, 25 Jun 2021 23:33:13 +0000 Received: from list by lists.xenproject.org with outflank-mailman.147460.271852 (Exim 4.92) (envelope-from ) id 1lwvJx-0001Bz-F3; Fri, 25 Jun 2021 23:33:13 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 147460.271852; Fri, 25 Jun 2021 23:33:13 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvJx-0001Br-C2; Fri, 25 Jun 2021 23:33:13 +0000 Received: by outflank-mailman (input) for mailman id 147460; Fri, 25 Jun 2021 23:33:12 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvJw-0001Bh-Ik for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:33:12 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvJw-0006OT-Hr for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:33:12 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwvJw-0006Wz-GL for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:33:12 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=zTc8Wh8EunRksKQ1gSYtyp9UsiaZXolVOE2Yl2zTfqI=; b=HtOR3C1hfEqmh03lBBUAvi8lj5 IwTcC80fxjEQRfL0t4ofFfMxA25U2X3wbELk19u9lWLW5vANI0q3BtgufKej8mOrcQJxjdSs5OTz4 pa9ZK92ARd7L1eqgu60f+NXuzVu3Ch+yWLfSvvFBieBVvHaj6T0Awm7IVuLLSZjhflL8=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] tools/xenstored: Introduce lu_get_connection() and use it Message-Id: Date: Fri, 25 Jun 2021 23:33:12 +0000 commit 3ce271ef535cfe3c0839fdea56f33abf5fdbf23c Author: Julien Grall AuthorDate: Thu Jun 24 09:02:59 2021 +0100 Commit: Julien Grall CommitDate: Thu Jun 24 09:02:59 2021 +0100 tools/xenstored: Introduce lu_get_connection() and use it At the moment, dump_state_buffered_data() is taking two connections in parameters (one is the connection to dump, the other is the connection used to request LU). The naming doesn't help to distinguish (c vs conn) them and this already lead to several mistake while modifying the function. To remove the confusion, introduce an help lu_get_connection() that will return the connection used to request LU and use it in place of the existing parameter. Signed-off-by: Julien Grall Reviewed-by: Luca Fancellu Reviewed-by: Juergen Gross --- tools/xenstore/xenstored_control.c | 13 ++++++++++++- tools/xenstore/xenstored_control.h | 2 ++ tools/xenstore/xenstored_core.c | 7 +++---- tools/xenstore/xenstored_core.h | 1 - tools/xenstore/xenstored_domain.c | 6 +++--- tools/xenstore/xenstored_domain.h | 2 +- 6 files changed, 21 insertions(+), 10 deletions(-) diff --git a/tools/xenstore/xenstored_control.c b/tools/xenstore/xenstored_control.c index 0d57f9f940..d08a2b9614 100644 --- a/tools/xenstore/xenstored_control.c +++ b/tools/xenstore/xenstored_control.c @@ -104,6 +104,17 @@ static const char *lu_begin(struct connection *conn) return NULL; } + +struct connection *lu_get_connection(void) +{ + return lu_status ? lu_status->conn : NULL; +} + +#else +struct connection *lu_get_connection(void) +{ + return NULL; +} #endif struct cmd_s { @@ -516,7 +527,7 @@ static const char *lu_dump_state(const void *ctx, struct connection *conn) ret = dump_state_global(fp); if (ret) goto out; - ret = dump_state_connections(fp, conn); + ret = dump_state_connections(fp); if (ret) goto out; ret = dump_state_special_nodes(fp); diff --git a/tools/xenstore/xenstored_control.h b/tools/xenstore/xenstored_control.h index aac61f0590..6842b8d887 100644 --- a/tools/xenstore/xenstored_control.h +++ b/tools/xenstore/xenstored_control.h @@ -18,3 +18,5 @@ int do_control(struct connection *conn, struct buffered_data *in); void lu_read_state(void); + +struct connection *lu_get_connection(void); diff --git a/tools/xenstore/xenstored_core.c b/tools/xenstore/xenstored_core.c index 883a1a582a..607187361d 100644 --- a/tools/xenstore/xenstored_core.c +++ b/tools/xenstore/xenstored_core.c @@ -2369,14 +2369,13 @@ const char *dump_state_global(FILE *fp) /* Called twice: first with fp == NULL to get length, then for writing data. */ const char *dump_state_buffered_data(FILE *fp, const struct connection *c, - const struct connection *conn, struct xs_state_connection *sc) { unsigned int len = 0, used; struct buffered_data *out, *in = c->in; bool partial = true; - if (in && c != conn) { + if (in && c != lu_get_connection()) { len = in->inhdr ? in->used : sizeof(in->hdr); if (fp && fwrite(&in->hdr, len, 1, fp) != 1) return "Dump read data error"; @@ -2416,8 +2415,8 @@ const char *dump_state_buffered_data(FILE *fp, const struct connection *c, } /* Add "OK" for live-update command. */ - if (c == conn) { - struct xsd_sockmsg msg = conn->in->hdr.msg; + if (c == lu_get_connection()) { + struct xsd_sockmsg msg = c->in->hdr.msg; msg.len = sizeof("OK"); if (fp && fwrite(&msg, sizeof(msg), 1, fp) != 1) diff --git a/tools/xenstore/xenstored_core.h b/tools/xenstore/xenstored_core.h index bb36111ecc..89ce155e75 100644 --- a/tools/xenstore/xenstored_core.h +++ b/tools/xenstore/xenstored_core.h @@ -269,7 +269,6 @@ void set_tdb_key(const char *name, TDB_DATA *key); const char *dump_state_global(FILE *fp); const char *dump_state_buffered_data(FILE *fp, const struct connection *c, - const struct connection *conn, struct xs_state_connection *sc); const char *dump_state_nodes(FILE *fp, const void *ctx); const char *dump_state_node_perms(FILE *fp, const struct xs_permissions *perms, diff --git a/tools/xenstore/xenstored_domain.c b/tools/xenstore/xenstored_domain.c index 322b0dbca4..6d8d29cbe4 100644 --- a/tools/xenstore/xenstored_domain.c +++ b/tools/xenstore/xenstored_domain.c @@ -1183,7 +1183,7 @@ void wrl_apply_debit_trans_commit(struct connection *conn) wrl_apply_debit_actual(conn->domain); } -const char *dump_state_connections(FILE *fp, struct connection *conn) +const char *dump_state_connections(FILE *fp) { const char *ret = NULL; unsigned int conn_id = 1; @@ -1209,7 +1209,7 @@ const char *dump_state_connections(FILE *fp, struct connection *conn) sc.spec.socket_fd = c->fd; } - ret = dump_state_buffered_data(NULL, c, conn, &sc); + ret = dump_state_buffered_data(NULL, c, &sc); if (ret) return ret; head.length += sc.data_in_len + sc.data_out_len; @@ -1219,7 +1219,7 @@ const char *dump_state_connections(FILE *fp, struct connection *conn) if (fwrite(&sc, offsetof(struct xs_state_connection, data), 1, fp) != 1) return "Dump connection state error"; - ret = dump_state_buffered_data(fp, c, conn, NULL); + ret = dump_state_buffered_data(fp, c, NULL); if (ret) return ret; ret = dump_state_align(fp); diff --git a/tools/xenstore/xenstored_domain.h b/tools/xenstore/xenstored_domain.h index cc5147d7e7..62ee471ea6 100644 --- a/tools/xenstore/xenstored_domain.h +++ b/tools/xenstore/xenstored_domain.h @@ -101,7 +101,7 @@ void wrl_log_periodic(struct wrl_timestampt now); void wrl_apply_debit_direct(struct connection *conn); void wrl_apply_debit_trans_commit(struct connection *conn); -const char *dump_state_connections(FILE *fp, struct connection *conn); +const char *dump_state_connections(FILE *fp); const char *dump_state_special_nodes(FILE *fp); void read_state_connection(const void *ctx, const void *state); -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Fri Jun 25 23:33:23 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Fri, 25 Jun 2021 23:33:23 +0000 Received: from list by lists.xenproject.org with outflank-mailman.147461.271856 (Exim 4.92) (envelope-from ) id 1lwvK7-0001FT-GT; Fri, 25 Jun 2021 23:33:23 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 147461.271856; Fri, 25 Jun 2021 23:33:23 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvK7-0001Ev-Dc; Fri, 25 Jun 2021 23:33:23 +0000 Received: by outflank-mailman (input) for mailman id 147461; Fri, 25 Jun 2021 23:33:22 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvK6-0001Ea-ML for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:33:22 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvK6-0006Oe-LW for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:33:22 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwvK6-0006Xl-KL for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:33:22 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=Z8Hs1tsKMKK0hy2lkk2VW16RAMDmXdayYAuMmYdSWWI=; b=DdJILrnRPjGJhkeyBQLJ6Ui/1j vGLysUu1j/FLTVTrXqG5xQmYISVes/H8lyY5WW0aeq6QA/Pa4CRPAUBXIR69DmzCq/r+RDG+UqEFL 3ES0JbJfqx26tHpKanKEoBWujoG+LESNbRHFgAlogBuZOvUunh4TFc4bOqn+tRW0jtEA=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] tools/xenstore: Don't assume conn->in points to the LU request Message-Id: Date: Fri, 25 Jun 2021 23:33:22 +0000 commit 65f19ed62aa1efa75c0dfccbf5d0514daa85be18 Author: Julien Grall AuthorDate: Thu Jun 24 09:06:58 2021 +0100 Commit: Julien Grall CommitDate: Thu Jun 24 09:06:58 2021 +0100 tools/xenstore: Don't assume conn->in points to the LU request call_delayed() is currently assuming that conn->in is NULL when handling delayed request. However, the connection is not paused. Therefore new request can be processed and conn->in may be non-NULL if we have only received a partial request. Furthermore, as we overwrite conn->in, the current partial request will not be transferred. This will result to corrupt the connection. Rather than updating conn->in, stash the LU request in lu_status and let each callback for delayed request to update conn->in when necessary. To keep a sane interface, the code to write the "OK" response the LU request is moved in xenstored_core.c. Fixes: c5ca1404b4 ("tools/xenstore: add support for delaying execution of a xenstore request") Fixes: ed6eebf17d ("tools/xenstore: dump the xenstore state for live update") Signed-off-by: Julien Grall Reviewed-by: Juergen Gross --- tools/xenstore/xenstored_control.c | 41 ++++++++++++++++++++++++++++++++++++-- tools/xenstore/xenstored_control.h | 3 +++ tools/xenstore/xenstored_core.c | 21 ++++++------------- 3 files changed, 48 insertions(+), 17 deletions(-) diff --git a/tools/xenstore/xenstored_control.c b/tools/xenstore/xenstored_control.c index d08a2b9614..7acc2d134f 100644 --- a/tools/xenstore/xenstored_control.c +++ b/tools/xenstore/xenstored_control.c @@ -50,6 +50,9 @@ struct live_update { /* For verification the correct connection is acting. */ struct connection *conn; + /* Pointer to the command used to request LU */ + struct buffered_data *in; + #ifdef __MINIOS__ void *kernel; unsigned int kernel_size; @@ -100,6 +103,7 @@ static const char *lu_begin(struct connection *conn) if (!lu_status) return "Allocation failure."; lu_status->conn = conn; + lu_status->in = conn->in; talloc_set_destructor(lu_status, lu_destroy); return NULL; @@ -110,11 +114,34 @@ struct connection *lu_get_connection(void) return lu_status ? lu_status->conn : NULL; } +unsigned int lu_write_response(FILE *fp) +{ + struct xsd_sockmsg msg; + + assert(lu_status); + + msg = lu_status->in->hdr.msg; + + msg.len = sizeof("OK"); + if (fp && fwrite(&msg, sizeof(msg), 1, fp) != 1) + return 0; + if (fp && fwrite("OK", msg.len, 1, fp) != 1) + return 0; + + return sizeof(msg) + msg.len; +} + #else struct connection *lu_get_connection(void) { return NULL; } + +unsigned int lu_write_response(FILE *fp) +{ + /* Unsupported */ + return 0; +} #endif struct cmd_s { @@ -658,6 +685,8 @@ static bool do_lu_start(struct delayed_request *req) { time_t now = time(NULL); const char *ret; + struct buffered_data *saved_in; + struct connection *conn = lu_status->conn; if (!lu_check_lu_allowed()) { if (now < lu_status->started_at + lu_status->timeout) @@ -668,8 +697,9 @@ static bool do_lu_start(struct delayed_request *req) } } + assert(req->in == lu_status->in); /* Dump out internal state, including "OK" for live update. */ - ret = lu_dump_state(req->in, lu_status->conn); + ret = lu_dump_state(req->in, conn); if (!ret) { /* Perform the activation of new binary. */ ret = lu_activate_binary(req->in); @@ -677,7 +707,14 @@ static bool do_lu_start(struct delayed_request *req) /* We will reach this point only in case of failure. */ out: - send_reply(lu_status->conn, XS_CONTROL, ret, strlen(ret) + 1); + /* + * send_reply() will send the response for conn->in. Save the current + * conn->in and restore it afterwards. + */ + saved_in = conn->in; + conn->in = req->in; + send_reply(conn, XS_CONTROL, ret, strlen(ret) + 1); + conn->in = saved_in; talloc_free(lu_status); return true; diff --git a/tools/xenstore/xenstored_control.h b/tools/xenstore/xenstored_control.h index 6842b8d887..27d7f19e4b 100644 --- a/tools/xenstore/xenstored_control.h +++ b/tools/xenstore/xenstored_control.h @@ -20,3 +20,6 @@ int do_control(struct connection *conn, struct buffered_data *in); void lu_read_state(void); struct connection *lu_get_connection(void); + +/* Write the "OK" response for the live-update command */ +unsigned int lu_write_response(FILE *fp); diff --git a/tools/xenstore/xenstored_core.c b/tools/xenstore/xenstored_core.c index 607187361d..4b6509b90d 100644 --- a/tools/xenstore/xenstored_core.c +++ b/tools/xenstore/xenstored_core.c @@ -270,17 +270,12 @@ static int undelay_request(void *_req) return 0; } -static void call_delayed(struct connection *conn, struct delayed_request *req) +static void call_delayed(struct delayed_request *req) { - assert(conn->in == NULL); - conn->in = req->in; - if (req->func(req)) { undelay_request(req); talloc_set_destructor(req, NULL); } - - conn->in = NULL; } int delay_request(struct connection *conn, struct buffered_data *in, @@ -2342,7 +2337,7 @@ int main(int argc, char *argv[]) list_for_each_entry_safe(req, tmp, &conn->delayed, list) - call_delayed(conn, req); + call_delayed(req); } } @@ -2375,7 +2370,7 @@ const char *dump_state_buffered_data(FILE *fp, const struct connection *c, struct buffered_data *out, *in = c->in; bool partial = true; - if (in && c != lu_get_connection()) { + if (in) { len = in->inhdr ? in->used : sizeof(in->hdr); if (fp && fwrite(&in->hdr, len, 1, fp) != 1) return "Dump read data error"; @@ -2416,16 +2411,12 @@ const char *dump_state_buffered_data(FILE *fp, const struct connection *c, /* Add "OK" for live-update command. */ if (c == lu_get_connection()) { - struct xsd_sockmsg msg = c->in->hdr.msg; + unsigned int rc = lu_write_response(fp); - msg.len = sizeof("OK"); - if (fp && fwrite(&msg, sizeof(msg), 1, fp) != 1) + if (!rc) return "Dump buffered data error"; - len += sizeof(msg); - if (fp && fwrite("OK", msg.len, 1, fp) != 1) - return "Dump buffered data error"; - len += msg.len; + len += rc; } if (sc) -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Fri Jun 25 23:33:33 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Fri, 25 Jun 2021 23:33:33 +0000 Received: from list by lists.xenproject.org with outflank-mailman.147462.271861 (Exim 4.92) (envelope-from ) id 1lwvKH-0001Jb-Kg; Fri, 25 Jun 2021 23:33:33 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 147462.271861; Fri, 25 Jun 2021 23:33:33 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvKH-0001JP-GX; Fri, 25 Jun 2021 23:33:33 +0000 Received: by outflank-mailman (input) for mailman id 147462; Fri, 25 Jun 2021 23:33:32 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvKG-0001JE-Pc for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:33:32 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvKG-0006P4-Oi for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:33:32 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwvKG-0006YO-Nw for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:33:32 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=c8NhGntfElWR8dsK+eW+UcRJixNKRRNxNRg49H+bov0=; b=c/+gkBkbQtCdz4gNmC2VyLYCv8 CW2b51QXth1zbk9QOVIhN1iY06qIfjFQYy9BVuz3dWr0uueX1x0gqUTCYpko2q/tBvXdm0x3075rW M3+G7cBx2tjEaPVpzk2tftDLIjLS+tJpRWwwuo4IEXEdDrzCKcUOll4CW0K0ohGKQuac=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] tools/xenstored: Limit the number of requests a connection can delay Message-Id: Date: Fri, 25 Jun 2021 23:33:32 +0000 commit 2aa9e00bd3d0431fc99d4baaee2cddbf40bbee61 Author: Julien Grall AuthorDate: Thu Jun 24 09:07:30 2021 +0100 Commit: Julien Grall CommitDate: Thu Jun 24 09:07:30 2021 +0100 tools/xenstored: Limit the number of requests a connection can delay Currently, only liveupdate request can be delayed. The request can only be performed by a privileged connection (e.g. dom0). So it is fine to have no limits. In a follow-up patch we will want to delay request for unprivileged connection as well. So it is best to apply a limit. For now and for simplicity, only a single request can be delayed for a given unprivileged connection. Take the opportunity to tweak the prototype and provide a way to bypass the quota check. This would be useful when the function is called from the restore code. Signed-off-by: Julien Grall Reviewed-by: Luca Fancellu Reviewed-by: Juergen Gross --- tools/xenstore/xenstored_control.c | 2 +- tools/xenstore/xenstored_core.c | 11 ++++++++++- tools/xenstore/xenstored_core.h | 3 ++- 3 files changed, 13 insertions(+), 3 deletions(-) diff --git a/tools/xenstore/xenstored_control.c b/tools/xenstore/xenstored_control.c index 7acc2d134f..1c24d4869e 100644 --- a/tools/xenstore/xenstored_control.c +++ b/tools/xenstore/xenstored_control.c @@ -737,7 +737,7 @@ static const char *lu_start(const void *ctx, struct connection *conn, lu_status->timeout = to; lu_status->started_at = time(NULL); - errno = delay_request(conn, conn->in, do_lu_start, NULL); + errno = delay_request(conn, conn->in, do_lu_start, NULL, false); return NULL; } diff --git a/tools/xenstore/xenstored_core.c b/tools/xenstore/xenstored_core.c index 4b6509b90d..268b0187e6 100644 --- a/tools/xenstore/xenstored_core.c +++ b/tools/xenstore/xenstored_core.c @@ -279,10 +279,19 @@ static void call_delayed(struct delayed_request *req) } int delay_request(struct connection *conn, struct buffered_data *in, - bool (*func)(struct delayed_request *), void *data) + bool (*func)(struct delayed_request *), void *data, + bool no_quota_check) { struct delayed_request *req; + /* + * Only allow one request can be delayed for an unprivileged + * connection. + */ + if (!no_quota_check && domain_is_unprivileged(conn) && + !list_empty(&conn->delayed)) + return ENOSPC; + req = talloc(in, struct delayed_request); if (!req) return ENOMEM; diff --git a/tools/xenstore/xenstored_core.h b/tools/xenstore/xenstored_core.h index 89ce155e75..34839b34f6 100644 --- a/tools/xenstore/xenstored_core.h +++ b/tools/xenstore/xenstored_core.h @@ -213,7 +213,8 @@ char *get_parent(const void *ctx, const char *node); /* Delay a request. */ int delay_request(struct connection *conn, struct buffered_data *in, - bool (*func)(struct delayed_request *), void *data); + bool (*func)(struct delayed_request *), void *data, + bool no_quota_check); /* Tracing infrastructure. */ void trace_create(const void *data, const char *type); -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Fri Jun 25 23:33:43 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Fri, 25 Jun 2021 23:33:43 +0000 Received: from list by lists.xenproject.org with outflank-mailman.147463.271864 (Exim 4.92) (envelope-from ) id 1lwvKR-0001Mv-Ku; Fri, 25 Jun 2021 23:33:43 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 147463.271864; Fri, 25 Jun 2021 23:33:43 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvKR-0001Mn-IC; Fri, 25 Jun 2021 23:33:43 +0000 Received: by outflank-mailman (input) for mailman id 147463; Fri, 25 Jun 2021 23:33:42 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvKQ-0001Mf-Sb for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:33:42 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvKQ-0006PR-Rn for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:33:42 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwvKQ-0006ZH-Qx for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:33:42 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=i/p8qVZ0nPmaqBKCQ2QBwPxv/qOf3T+6IHNEzKG5Dmc=; b=rdByARcBkbq4BQgPV21qLgFJUL NXrJWOyfm4vu/J185pZU0UC9f1c306PlPLcygC/tStA9m1xz3W9ThBEyDKf9G5zVXjgj570gPoRg+ LW/mbY8/W4qAzJgGTLtILb133CuOWAGKIF19SNnuhCs7Jwji5XTEyJpZykoqdI1DngIA=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] tools/xenstored: xenstored_core.h should include fcntl.h Message-Id: Date: Fri, 25 Jun 2021 23:33:42 +0000 commit a8002238b49c3423c7081dfbac3580651f5242ef Author: Julien Grall AuthorDate: Thu Jun 24 09:07:52 2021 +0100 Commit: Julien Grall CommitDate: Thu Jun 24 09:07:52 2021 +0100 tools/xenstored: xenstored_core.h should include fcntl.h xenstored_core.h will consider live-udpate is not supported if O_CLOEXEC doesn't exist. However, the header doesn't include the one defining O_CLOEXEC (i.e. fcntl.h). This means that depending on the header included, some source file will think Live-Update is not supported. I am not aware of any issue with the existing. Therefore this is just a latent bug so far. Prevent any potential issue by including fcntl.h in xenstored_core.h Fixes: cd831ee438 ("tools/xenstore: handle CLOEXEC flag for local files and pipes") Signed-off-by: Julien Grall Reviewed-by: Luca Fancellu Reviewed-by: Juergen Gross --- tools/xenstore/xenstored_core.h | 1 + 1 file changed, 1 insertion(+) diff --git a/tools/xenstore/xenstored_core.h b/tools/xenstore/xenstored_core.h index 34839b34f6..dac5171569 100644 --- a/tools/xenstore/xenstored_core.h +++ b/tools/xenstore/xenstored_core.h @@ -24,6 +24,7 @@ #include #include +#include #include #include #include -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Fri Jun 25 23:33:53 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Fri, 25 Jun 2021 23:33:53 +0000 Received: from list by lists.xenproject.org with outflank-mailman.147464.271868 (Exim 4.92) (envelope-from ) id 1lwvKb-0001QK-Md; Fri, 25 Jun 2021 23:33:53 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 147464.271868; Fri, 25 Jun 2021 23:33:53 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvKb-0001QA-Ji; Fri, 25 Jun 2021 23:33:53 +0000 Received: by outflank-mailman (input) for mailman id 147464; Fri, 25 Jun 2021 23:33:53 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvKa-0001Q1-Vn for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:33:52 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvKa-0006Pd-Uv for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:33:52 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwvKa-0006a0-U8 for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:33:52 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=u8W799cwc2xJ33huxxGY13ms/r9KXFa5ongo/nuvNZk=; b=uLHIt35dFKTNLJT2bAy0IcFMdo L1tzo1jT4vzGwwXfokJxQOUua/BGX7DIfN9ILnmbRyIQgUjMKVQcB84/L2WpC7KRgAHwhSYuoD0kJ yhAh3SwQkE9L757FYZnalArkvhcO+sOG3lixihCuYpazXsws+Gc1cG3sv2tPlqO1+zG8=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] tools/xenstored: Introduce a wrapper for conn->funcs->can_{read, write} Message-Id: Date: Fri, 25 Jun 2021 23:33:52 +0000 commit 3adfb50315d9b5643bec7a8a5378625f15986200 Author: Julien Grall AuthorDate: Thu Jun 24 09:08:42 2021 +0100 Commit: Julien Grall CommitDate: Thu Jun 24 09:08:42 2021 +0100 tools/xenstored: Introduce a wrapper for conn->funcs->can_{read, write} Currently, the callbacks can_read and can_write are called directly. This doesn't allow us to add generic check and therefore requires duplication. At the moment, one check that could benefit to be common is whether the connection should ignored. The position is slightly different between domain and socket because for the latter we want to check the state of the file descriptor first. In follow-up patches, there will be more potential generic checks. This patch provides wrappers to read/write a connection and move the check ->is_ignored after the callback for everyone. This also requires to replace the direct call to domain_can_read() and domain_can_write() with the new wrapper. At the same time, both functions can now be static. Note that the implementations need to be moved earlier in the file xenstored_domain.c to avoid declaring the prototype. Signed-off-by: Julien Grall Reviewed-by: Luca Fancellu Reviewed-by: Juergen Gross --- tools/xenstore/xenstored_core.c | 18 ++++++++++++++---- tools/xenstore/xenstored_domain.c | 40 +++++++++++++++++---------------------- tools/xenstore/xenstored_domain.h | 4 ---- 3 files changed, 31 insertions(+), 31 deletions(-) diff --git a/tools/xenstore/xenstored_core.c b/tools/xenstore/xenstored_core.c index 268b0187e6..4c0d6e1874 100644 --- a/tools/xenstore/xenstored_core.c +++ b/tools/xenstore/xenstored_core.c @@ -334,6 +334,16 @@ static int destroy_conn(void *_conn) return 0; } +static bool conn_can_read(struct connection *conn) +{ + return conn->funcs->can_read(conn) && !conn->is_ignored; +} + +static bool conn_can_write(struct connection *conn) +{ + return conn->funcs->can_write(conn) && !conn->is_ignored; +} + /* This function returns index inside the array if succeed, -1 if fail */ static int set_fd(int fd, short events) { @@ -396,8 +406,8 @@ static void initialize_fds(int *p_sock_pollfd_idx, int *ptimeout) list_for_each_entry(conn, &connections, list) { if (conn->domain) { wrl_check_timeout(conn->domain, now, ptimeout); - if (domain_can_read(conn) || - (domain_can_write(conn) && + if (conn_can_read(conn) || + (conn_can_write(conn) && !list_empty(&conn->out_list))) *ptimeout = 0; } else { @@ -2325,14 +2335,14 @@ int main(int argc, char *argv[]) if (&next->list != &connections) talloc_increase_ref_count(next); - if (conn->funcs->can_read(conn)) + if (conn_can_read(conn)) handle_input(conn); if (talloc_free(conn) == 0) continue; talloc_increase_ref_count(conn); - if (conn->funcs->can_write(conn)) + if (conn_can_write(conn)) handle_output(conn); if (talloc_free(conn) == 0) continue; diff --git a/tools/xenstore/xenstored_domain.c b/tools/xenstore/xenstored_domain.c index 6d8d29cbe4..47e9107c14 100644 --- a/tools/xenstore/xenstored_domain.c +++ b/tools/xenstore/xenstored_domain.c @@ -172,6 +172,23 @@ static int readchn(struct connection *conn, void *data, unsigned int len) return len; } +static bool domain_can_write(struct connection *conn) +{ + struct xenstore_domain_interface *intf = conn->domain->interface; + + return ((intf->rsp_prod - intf->rsp_cons) != XENSTORE_RING_SIZE); +} + +static bool domain_can_read(struct connection *conn) +{ + struct xenstore_domain_interface *intf = conn->domain->interface; + + if (domain_is_unprivileged(conn) && conn->domain->wrl_credit < 0) + return false; + + return (intf->req_cons != intf->req_prod); +} + static const struct interface_funcs domain_funcs = { .write = writechn, .read = readchn, @@ -290,19 +307,6 @@ void handle_event(void) barf_perror("Failed to write to event fd"); } -bool domain_can_read(struct connection *conn) -{ - struct xenstore_domain_interface *intf = conn->domain->interface; - - if (domain_is_unprivileged(conn) && conn->domain->wrl_credit < 0) - return false; - - if (conn->is_ignored) - return false; - - return (intf->req_cons != intf->req_prod); -} - static bool domid_is_unprivileged(unsigned int domid) { return domid != 0 && domid != priv_domid; @@ -314,16 +318,6 @@ bool domain_is_unprivileged(struct connection *conn) domid_is_unprivileged(conn->domain->domid); } -bool domain_can_write(struct connection *conn) -{ - struct xenstore_domain_interface *intf = conn->domain->interface; - - if (conn->is_ignored) - return false; - - return ((intf->rsp_prod - intf->rsp_cons) != XENSTORE_RING_SIZE); -} - static char *talloc_domain_path(void *context, unsigned int domid) { return talloc_asprintf(context, "/local/domain/%u", domid); diff --git a/tools/xenstore/xenstored_domain.h b/tools/xenstore/xenstored_domain.h index 62ee471ea6..1e929b8f8c 100644 --- a/tools/xenstore/xenstored_domain.h +++ b/tools/xenstore/xenstored_domain.h @@ -51,10 +51,6 @@ void domain_deinit(void); /* Returns the implicit path of a connection (only domains have this) */ const char *get_implicit_path(const struct connection *conn); -/* Can connection attached to domain read/write. */ -bool domain_can_read(struct connection *conn); -bool domain_can_write(struct connection *conn); - bool domain_is_unprivileged(struct connection *conn); /* Remove node permissions for no longer existing domains. */ -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Fri Jun 25 23:34:03 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Fri, 25 Jun 2021 23:34:03 +0000 Received: from list by lists.xenproject.org with outflank-mailman.147465.271872 (Exim 4.92) (envelope-from ) id 1lwvKl-0001TB-OH; Fri, 25 Jun 2021 23:34:03 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 147465.271872; Fri, 25 Jun 2021 23:34:03 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvKl-0001T3-LE; Fri, 25 Jun 2021 23:34:03 +0000 Received: by outflank-mailman (input) for mailman id 147465; Fri, 25 Jun 2021 23:34:03 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvKl-0001Sx-2k for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:34:03 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvKl-0006Q2-1p for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:34:03 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwvKl-0006ar-0v for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:34:03 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=DSDPq+e8cgVCPnwxfaPlNlm4jpWkD/VK0pcAYsqZJJo=; b=giBOikcsjeWUu1UjHF1IsFwE9u jVTcJdGAymsj9kc0sfRAq+6Jta5cktiBtMRDj7tbXRJhQ09Dis4x491wETpLxeDL9hy0OM9gqC+qA D+ksPxyw4vj8p5HNXaHPXia3xindksT3f5soJgfEjD0IQtsAattBUPRo53pGlUigFRy0=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] tools/xenstored: delay_request: don't assume conn->in == in Message-Id: Date: Fri, 25 Jun 2021 23:34:03 +0000 commit 361f1deb42ec32ea97df7e79f8df12bfc2d0e049 Author: Julien Grall AuthorDate: Thu Jun 24 09:08:56 2021 +0100 Commit: Julien Grall CommitDate: Thu Jun 24 09:08:56 2021 +0100 tools/xenstored: delay_request: don't assume conn->in == in delay_request() is currently assuming that the request delayed is always conn->in. This is currently correct, but it is a call for a latent bug as the function allows the caller to specify any request. To prevent any future surprise, check if the request delayed is the current one. Fixes: c5ca1404b4 ("tools/xenstore: add support for delaying execution of a xenstore request") Signed-off-by: Julien Grall Reviewed-by: Luca Fancellu Reviewed-by: Juergen Gross --- tools/xenstore/xenstored_core.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/tools/xenstore/xenstored_core.c b/tools/xenstore/xenstored_core.c index 4c0d6e1874..3286eb55d3 100644 --- a/tools/xenstore/xenstored_core.c +++ b/tools/xenstore/xenstored_core.c @@ -306,7 +306,9 @@ int delay_request(struct connection *conn, struct buffered_data *in, delayed_requests++; list_add(&req->list, &conn->delayed); - conn->in = NULL; + /* Unlink the request from conn if this is the current one */ + if (conn->in == in) + conn->in = NULL; return 0; } -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Fri Jun 25 23:34:13 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Fri, 25 Jun 2021 23:34:13 +0000 Received: from list by lists.xenproject.org with outflank-mailman.147466.271876 (Exim 4.92) (envelope-from ) id 1lwvKv-0001WC-Pf; Fri, 25 Jun 2021 23:34:13 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 147466.271876; Fri, 25 Jun 2021 23:34:13 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvKv-0001W4-Mm; Fri, 25 Jun 2021 23:34:13 +0000 Received: by outflank-mailman (input) for mailman id 147466; Fri, 25 Jun 2021 23:34:13 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvKv-0001Vx-64 for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:34:13 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvKv-0006QD-5L for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:34:13 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwvKv-0006bw-48 for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:34:13 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=OGbhpWcN7ipZPzFf/n0Xrm647xnMo7szU6BRxDJBCLo=; b=GDYxZXiz7rhPA6gxmuthtAtk1a ylQySI29+zhIHc2+MifVcXDLQTc7dtbZWX9Dj/d0dKKRIXNXaIfEEX/urVz3fgxPcRSZ1rlAp3Y3R lH+WNAVzRDVmGaIiIo1qLrgfkk9yASSUHuKgxk73/RQM0jpENA6er09ahYc+8LiuVsUU=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] tools/xenstored: Extend restore code to handle multiple input buffer Message-Id: Date: Fri, 25 Jun 2021 23:34:13 +0000 commit c0fe360f42245da6869362584b1f115a71d36ad6 Author: Julien Grall AuthorDate: Thu Jun 24 11:41:00 2021 +0100 Commit: Julien Grall CommitDate: Thu Jun 24 11:41:00 2021 +0100 tools/xenstored: Extend restore code to handle multiple input buffer Currently, the restore code is considering the stream will contain at most one in-flight request per connection. In a follow-up changes, we will want to transfer multiple in-flight requests. The function read_state_buffered() is now extended to restore multiple in-flight request. Complete requests will be queued as delayed requests, if there a partial request (only the last one can) then it will used as the current in-flight request. Note that we want to bypass the quota check for delayed requests as the new Xenstore may have a lower limit. Lastly, there is no need to change the specification as there was no restriction on the number of in-flight requests preserved. Signed-off-by: Julien Grall Reviewed-by: Luca Fancellu Reviewed-by: Juergen Gross --- tools/xenstore/xenstored_core.c | 56 +++++++++++++++++++++++++++++++++++------ 1 file changed, 48 insertions(+), 8 deletions(-) diff --git a/tools/xenstore/xenstored_core.c b/tools/xenstore/xenstored_core.c index 3286eb55d3..a83593b566 100644 --- a/tools/xenstore/xenstored_core.c +++ b/tools/xenstore/xenstored_core.c @@ -1486,6 +1486,10 @@ static void process_message(struct connection *conn, struct buffered_data *in) enum xsd_sockmsg_type type = in->hdr.msg.type; int ret; + /* At least send_error() and send_reply() expects conn->in == in */ + assert(conn->in == in); + trace_io(conn, in, 0); + if ((unsigned int)type >= XS_TYPE_COUNT || !wire_funcs[type].func) { eprintf("Client unknown operation %i", type); send_error(conn, ENOSYS); @@ -1515,6 +1519,23 @@ static void process_message(struct connection *conn, struct buffered_data *in) conn->transaction = NULL; } +static bool process_delayed_message(struct delayed_request *req) +{ + struct connection *conn = req->data; + struct buffered_data *saved_in = conn->in; + + /* + * Part of process_message() expects conn->in to contains the + * processed response. So save the current conn->in and restore it + * afterwards. + */ + conn->in = req->in; + process_message(req->data, req->in); + conn->in = saved_in; + + return true; +} + static void consider_message(struct connection *conn) { if (verbose) @@ -1582,7 +1603,6 @@ static void handle_input(struct connection *conn) if (in->used != in->hdr.msg.len) return; - trace_io(conn, in, 0); consider_message(conn); return; @@ -2611,14 +2631,20 @@ void read_state_buffered_data(const void *ctx, struct connection *conn, unsigned int len; bool partial = sc->data_resp_len; - if (sc->data_in_len) { + for (data = sc->data; data < sc->data + sc->data_in_len; data += len) { bdata = new_buffer(conn); if (!bdata) barf("error restoring read data"); - if (sc->data_in_len < sizeof(bdata->hdr)) { + + /* + * We don't know yet if there is more than one message + * to process. So the len is the size of the leftover data. + */ + len = sc->data_in_len - (data - sc->data); + if (len < sizeof(bdata->hdr)) { bdata->inhdr = true; - memcpy(&bdata->hdr, sc->data, sc->data_in_len); - bdata->used = sc->data_in_len; + memcpy(&bdata->hdr, sc->data, len); + bdata->used = len; } else { bdata->inhdr = false; memcpy(&bdata->hdr, sc->data, sizeof(bdata->hdr)); @@ -2629,12 +2655,26 @@ void read_state_buffered_data(const void *ctx, struct connection *conn, bdata->hdr.msg.len); if (!bdata->buffer) barf("Error allocating in buffer"); - bdata->used = sc->data_in_len - sizeof(bdata->hdr); - memcpy(bdata->buffer, sc->data + sizeof(bdata->hdr), + bdata->used = min_t(unsigned int, + len - sizeof(bdata->hdr), + bdata->hdr.msg.len); + memcpy(bdata->buffer, data + sizeof(bdata->hdr), bdata->used); + /* Update len to match the size of the message. */ + len = bdata->used + sizeof(bdata->hdr); } - conn->in = bdata; + /* + * If the message is not complete, then it means this was + * the current processed message. All the other messages + * will be queued to be handled after restoring. + */ + if (bdata->inhdr || bdata->used != bdata->hdr.msg.len) { + assert(conn->in == NULL); + conn->in = bdata; + } else if (delay_request(conn, bdata, process_delayed_message, + conn, true)) + barf("Unable to delay the request"); } for (data = sc->data + sc->data_in_len; -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Fri Jun 25 23:34:24 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Fri, 25 Jun 2021 23:34:24 +0000 Received: from list by lists.xenproject.org with outflank-mailman.147467.271880 (Exim 4.92) (envelope-from ) id 1lwvL6-0001Yq-RM; Fri, 25 Jun 2021 23:34:24 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 147467.271880; Fri, 25 Jun 2021 23:34:24 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvL6-0001Yi-OK; Fri, 25 Jun 2021 23:34:24 +0000 Received: by outflank-mailman (input) for mailman id 147467; Fri, 25 Jun 2021 23:34:23 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvL5-0001YW-9Y for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:34:23 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvL5-0006QO-8g for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:34:23 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwvL5-0006cw-7n for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:34:23 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=VwaFdZO62641DocpeD8qmrj4O2Sw/5ZqMYLj9+wBS2s=; b=D0DvWGXR2D351/VrbzP8QwGlVq yS/ZEtauVSc/v2yHKE840KGuYZ5ll1D+rohrFajJH7Uq49CplmRH7XRAiUweSCDPRzOWn5LKyXpRm H45uCmFTclOkbsgah88j7/GlGXQrjr5M+ScUKtL/VQQBdsnQ58ZD4PVKwWl6Mlm4uxJE=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] iommu/arm: ipmmu-vmsa: Add compatible for Renesas R-Car M3-W+ SoC Message-Id: Date: Fri, 25 Jun 2021 23:34:23 +0000 commit ae8d47ede97707af07e772a164a475aa5f447708 Author: Oleksandr Tyshchenko AuthorDate: Mon Jun 14 22:18:12 2021 +0300 Commit: Julien Grall CommitDate: Thu Jun 24 11:43:53 2021 +0100 iommu/arm: ipmmu-vmsa: Add compatible for Renesas R-Car M3-W+ SoC The "renesas,r8a77961" string identifies M3-W+ (aka M3-W ES3.0) instead of "renesas,r8a7796" since Linux commit: "9c9f7891093b02eb64ca4e1c7ab776a4296c058f soc: renesas: Identify R-Car M3-W+". Add new compatible to the Xen driver. Signed-off-by: Oleksandr Tyshchenko Acked-by: Julien Grall --- xen/drivers/passthrough/arm/ipmmu-vmsa.c | 1 + 1 file changed, 1 insertion(+) diff --git a/xen/drivers/passthrough/arm/ipmmu-vmsa.c b/xen/drivers/passthrough/arm/ipmmu-vmsa.c index 8b8e3a00ba..1255b0d4fe 100644 --- a/xen/drivers/passthrough/arm/ipmmu-vmsa.c +++ b/xen/drivers/passthrough/arm/ipmmu-vmsa.c @@ -1316,6 +1316,7 @@ static const struct dt_device_match ipmmu_dt_match[] __initconst = DT_MATCH_COMPATIBLE("renesas,ipmmu-r8a7795"), DT_MATCH_COMPATIBLE("renesas,ipmmu-r8a77965"), DT_MATCH_COMPATIBLE("renesas,ipmmu-r8a7796"), + DT_MATCH_COMPATIBLE("renesas,ipmmu-r8a77961"), { /* sentinel */ }, }; -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Fri Jun 25 23:34:34 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Fri, 25 Jun 2021 23:34:34 +0000 Received: from list by lists.xenproject.org with outflank-mailman.147468.271884 (Exim 4.92) (envelope-from ) id 1lwvLG-0001cX-U5; Fri, 25 Jun 2021 23:34:34 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 147468.271884; Fri, 25 Jun 2021 23:34:34 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvLG-0001cP-R7; Fri, 25 Jun 2021 23:34:34 +0000 Received: by outflank-mailman (input) for mailman id 147468; Fri, 25 Jun 2021 23:34:33 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvLF-0001cD-Cg for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:34:33 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvLF-0006Qt-Bz for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:34:33 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwvLF-0006dh-B1 for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:34:33 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=dkyhHc0dCwqNLpXIfZWn/ruMx6Q+Yr8Ao6XuuCtQdr0=; b=FIuTOawlUHe2y2X8RnY0J/8ODV aCeUjh97imXqLJ3vgoEXnwWptRCg/SgLIklCXf5WtM/6VvYfgrPxiXUiGLS50CHwjvPXoYsMj0OxH 80hALpbzOc6inB/2AeNkz8I+5Tbezp78fl/jSl4EOjdMN5qF+YInYk/PZMYpY5pOP5YI=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] maintainers: adding new reviewer for xsm Message-Id: Date: Fri, 25 Jun 2021 23:34:33 +0000 commit 76a0aa9c4d7a9fc6fee1158fd9df82ae9b8b605d Author: Daniel P. Smith AuthorDate: Thu Jun 17 19:49:55 2021 -0400 Commit: Julien Grall CommitDate: Thu Jun 24 11:45:32 2021 +0100 maintainers: adding new reviewer for xsm Would like to add myself as a reviewer for XSM. Signed-off-by: Daniel P. Smith Acked-by: Stefano Stabellini Acked-by: Andrew Cooper --- MAINTAINERS | 1 + 1 file changed, 1 insertion(+) diff --git a/MAINTAINERS b/MAINTAINERS index dd8c011456..8a52a03969 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -641,6 +641,7 @@ F: xen/include/xen/trace.h XSM/FLASK M: Daniel De Graaf +R: Daniel P. Smith S: Supported F: tools/flask/ F: xen/include/xsm/ -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Fri Jun 25 23:34:45 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Fri, 25 Jun 2021 23:34:45 +0000 Received: from list by lists.xenproject.org with outflank-mailman.147471.271902 (Exim 4.92) (envelope-from ) id 1lwvLR-0001wr-BM; Fri, 25 Jun 2021 23:34:45 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 147471.271902; Fri, 25 Jun 2021 23:34:45 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvLR-0001wj-8B; Fri, 25 Jun 2021 23:34:45 +0000 Received: by outflank-mailman (input) for mailman id 147471; Fri, 25 Jun 2021 23:34:43 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvLP-0001p6-G5 for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:34:43 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvLP-0006RA-FB for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:34:43 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwvLP-0006et-EJ for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:34:43 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=IqWteitSeBHuRo9jH2/AaFoqy2iT6a2yYjXkXz85jd8=; b=WxrCd/+S6WJIir0S+HnGhOwUsR 01EQ/MOHBCZousWYBNN6cYTL4nIrjlCKhjs68WU7+T7EB0bW6AzYf5Hu3MfpA9Y86HCZm+BtxUWoW 91A83LQctGFevES1OrkSpvJGwqJ6Ljemnzqdv8R9JAk4jIf6+2DziZNEd1yjInqK5tXM=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] tools/xenstored: Dump delayed requests Message-Id: Date: Fri, 25 Jun 2021 23:34:43 +0000 commit 54f73e6fc6ef3fbd287cd030fb70d43f8f92244d Author: Julien Grall AuthorDate: Thu Jun 24 12:12:54 2021 +0100 Commit: Julien Grall CommitDate: Thu Jun 24 12:12:54 2021 +0100 tools/xenstored: Dump delayed requests Currently, only Live-Update request can be delayed. In a follow-up, we will want to delay more requests (e.g. transaction start). Therefore we want to preserve delayed requests across Live-Update. Delayed requests are just complete "in" buffer. So the code is refactored to allow sharing the code to dump "in" buffer. Signed-off-by: Julien Grall Reviewed-by: Luca Fancellu Reviewed-by: Juergen Gross --- tools/xenstore/xenstored_core.c | 47 ++++++++++++++++++++++++++++++++--------- 1 file changed, 37 insertions(+), 10 deletions(-) diff --git a/tools/xenstore/xenstored_core.c b/tools/xenstore/xenstored_core.c index a83593b566..0cb46d8c64 100644 --- a/tools/xenstore/xenstored_core.c +++ b/tools/xenstore/xenstored_core.c @@ -2403,25 +2403,52 @@ const char *dump_state_global(FILE *fp) return NULL; } +static const char *dump_input_buffered_data(FILE *fp, + const struct buffered_data *in, + unsigned int *total_len) +{ + unsigned int hlen = in->inhdr ? in->used : sizeof(in->hdr); + + *total_len += hlen; + if (fp && fwrite(&in->hdr, hlen, 1, fp) != 1) + return "Dump read data error"; + if (!in->inhdr && in->used) { + *total_len += in->used; + if (fp && fwrite(in->buffer, in->used, 1, fp) != 1) + return "Dump read data error"; + } + + return NULL; +} + /* Called twice: first with fp == NULL to get length, then for writing data. */ const char *dump_state_buffered_data(FILE *fp, const struct connection *c, struct xs_state_connection *sc) { unsigned int len = 0, used; - struct buffered_data *out, *in = c->in; + struct buffered_data *out; bool partial = true; + struct delayed_request *req; + const char *ret; - if (in) { - len = in->inhdr ? in->used : sizeof(in->hdr); - if (fp && fwrite(&in->hdr, len, 1, fp) != 1) - return "Dump read data error"; - if (!in->inhdr && in->used) { - len += in->used; - if (fp && fwrite(in->buffer, in->used, 1, fp) != 1) - return "Dump read data error"; - } + /* Dump any command that was delayed */ + list_for_each_entry(req, &c->delayed, list) { + /* + * We only want to preserve commands that weren't processed at + * all. All the other delayed requests (such as do_lu_start()) + * must be processed before Live-Update. + */ + if (req->func != process_delayed_message) + continue; + + assert(!req->in->inhdr); + if ((ret = dump_input_buffered_data(fp, req->in, &len))) + return ret; } + if (c->in && (ret = dump_input_buffered_data(fp, c->in, &len))) + return ret; + if (sc) { sc->data_in_len = len; sc->data_resp_len = 0; -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Fri Jun 25 23:34:55 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Fri, 25 Jun 2021 23:34:55 +0000 Received: from list by lists.xenproject.org with outflank-mailman.147473.271911 (Exim 4.92) (envelope-from ) id 1lwvLb-00024G-ET; Fri, 25 Jun 2021 23:34:55 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 147473.271911; Fri, 25 Jun 2021 23:34:55 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvLb-000245-Aa; Fri, 25 Jun 2021 23:34:55 +0000 Received: by outflank-mailman (input) for mailman id 147473; Fri, 25 Jun 2021 23:34:53 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvLZ-00022Y-JC for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:34:53 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvLZ-0006RU-IP for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:34:53 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwvLZ-0006g2-HZ for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:34:53 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=XuOtSRZwv5clvQwIbjGtyASWeFuST1FNNL11junjTVo=; b=m56LOiKGAzzV4l15MkG9IWpiLz iTiX8V1Imn4VgbYjPsRYNjlD7kjUPgSdksWm0UnrZt7ywU+DFPk1x7xEp7s7VXZd+KYa+g+yotqDT yMwb8cOfSnTOJH9fz22eU9nrRfWEqAKs0WcIxkgZmn7etRtcxlVrydZxKzBJHWW+jD+4=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] tools/xenstored: Delay new transaction while Live-Update is pending Message-Id: Date: Fri, 25 Jun 2021 23:34:53 +0000 commit c8be420025ada0f90e98ae0408226e29017757df Author: Julien Grall AuthorDate: Thu Jun 24 12:14:56 2021 +0100 Commit: Julien Grall CommitDate: Thu Jun 24 12:14:56 2021 +0100 tools/xenstored: Delay new transaction while Live-Update is pending At the moment, Live-Update will, by default, not proceed if there are in-flight transactions. It is possible force it by passing -F but this will break any connection with in-flight transactions. There are PV drivers out that may never terminate some transaction. On host running such guest, we would need to use -F. Unfortunately, this also risks to break well-behaving guests (and even dom0) because Live-Update will happen as soon as the timeout is hit. Ideally, we would want to preserve transactions but this requires some work and a lot of testing to be able to use it in production. As a stop gap, we want to limit the damage of -F. This patch will delay any transactions that are started after Live-Update has been requested. If the request cannot be delayed, the connection will be stalled to avoid loosing requests. If the connection has already a pending transaction before Live-Update, then new transaction will not be delayed. This is to avoid the connection to stall. With this stop gap in place, domains with long running transactions will still break when using -F, but other domains which starts a transaction in the middle of Live-Update will continue to work. Signed-off-by: Julien Grall Reviewed-by: Luca Fancellu Reviewed-by: Juergen Gross --- tools/xenstore/xenstored_control.c | 10 ++++++++ tools/xenstore/xenstored_control.h | 2 ++ tools/xenstore/xenstored_core.c | 49 +++++++++++++++++++++++++++++++++++++- tools/xenstore/xenstored_core.h | 3 +++ 4 files changed, 63 insertions(+), 1 deletion(-) diff --git a/tools/xenstore/xenstored_control.c b/tools/xenstore/xenstored_control.c index 1c24d4869e..a045f102a4 100644 --- a/tools/xenstore/xenstored_control.c +++ b/tools/xenstore/xenstored_control.c @@ -131,6 +131,11 @@ unsigned int lu_write_response(FILE *fp) return sizeof(msg) + msg.len; } +bool lu_is_pending(void) +{ + return lu_status != NULL; +} + #else struct connection *lu_get_connection(void) { @@ -142,6 +147,11 @@ unsigned int lu_write_response(FILE *fp) /* Unsupported */ return 0; } + +bool lu_is_pending(void) +{ + return false; +} #endif struct cmd_s { diff --git a/tools/xenstore/xenstored_control.h b/tools/xenstore/xenstored_control.h index 27d7f19e4b..98b6fbcea2 100644 --- a/tools/xenstore/xenstored_control.h +++ b/tools/xenstore/xenstored_control.h @@ -23,3 +23,5 @@ struct connection *lu_get_connection(void); /* Write the "OK" response for the live-update command */ unsigned int lu_write_response(FILE *fp); + +bool lu_is_pending(void); diff --git a/tools/xenstore/xenstored_core.c b/tools/xenstore/xenstored_core.c index 0cb46d8c64..9ffd2ac66d 100644 --- a/tools/xenstore/xenstored_core.c +++ b/tools/xenstore/xenstored_core.c @@ -338,7 +338,20 @@ static int destroy_conn(void *_conn) static bool conn_can_read(struct connection *conn) { - return conn->funcs->can_read(conn) && !conn->is_ignored; + if (!conn->funcs->can_read(conn)) + return false; + + if (conn->is_ignored) + return false; + + /* + * For stalled connection, we want to process the pending + * command as soon as live-update has aborted. + */ + if (conn->is_stalled) + return !lu_is_pending(); + + return true; } static bool conn_can_write(struct connection *conn) @@ -417,6 +430,12 @@ static void initialize_fds(int *p_sock_pollfd_idx, int *ptimeout) if (!list_empty(&conn->out_list)) events |= POLLOUT; conn->pollfd_idx = set_fd(conn->fd, events); + /* + * For stalled connection, we want to process the + * pending command as soon as live-update has aborted. + */ + if (conn->is_stalled && !lu_is_pending()) + *ptimeout = 0; } } } @@ -1524,6 +1543,9 @@ static bool process_delayed_message(struct delayed_request *req) struct connection *conn = req->data; struct buffered_data *saved_in = conn->in; + if (lu_is_pending()) + return false; + /* * Part of process_message() expects conn->in to contains the * processed response. So save the current conn->in and restore it @@ -1543,6 +1565,30 @@ static void consider_message(struct connection *conn) sockmsg_string(conn->in->hdr.msg.type), conn->in->hdr.msg.len, conn); + conn->is_stalled = false; + /* + * Currently, Live-Update is not supported if there is active + * transactions. In order to reduce the number of retry, delay + * any new request to start a transaction if Live-Update is pending + * and there are no transactions in-flight. + * + * If we can't delay the request, then mark the connection as + * stalled. This will ignore new requests until Live-Update happened + * or it was aborted. + */ + if (lu_is_pending() && conn->transaction_started == 0 && + conn->in->hdr.msg.type == XS_TRANSACTION_START) { + trace("Delaying transaction start for connection %p req_id %u\n", + conn, conn->in->hdr.msg.req_id); + + if (delay_request(conn, conn->in, process_delayed_message, + conn, false) != 0) { + trace("Stalling connection %p\n", conn); + conn->is_stalled = true; + } + return; + } + process_message(conn, conn->in); assert(conn->in == NULL); @@ -1629,6 +1675,7 @@ struct connection *new_connection(const struct interface_funcs *funcs) new->pollfd_idx = -1; new->funcs = funcs; new->is_ignored = false; + new->is_stalled = false; new->transaction_started = 0; INIT_LIST_HEAD(&new->out_list); INIT_LIST_HEAD(&new->watches); diff --git a/tools/xenstore/xenstored_core.h b/tools/xenstore/xenstored_core.h index dac5171569..258f6ff382 100644 --- a/tools/xenstore/xenstored_core.h +++ b/tools/xenstore/xenstored_core.h @@ -110,6 +110,9 @@ struct connection /* Is this connection ignored? */ bool is_ignored; + /* Is the connection stalled? */ + bool is_stalled; + /* Buffered incoming data. */ struct buffered_data *in; -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Fri Jun 25 23:35:05 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Fri, 25 Jun 2021 23:35:05 +0000 Received: from list by lists.xenproject.org with outflank-mailman.147475.271913 (Exim 4.92) (envelope-from ) id 1lwvLl-0002Gv-FN; Fri, 25 Jun 2021 23:35:05 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 147475.271913; Fri, 25 Jun 2021 23:35:05 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvLl-0002Go-C9; Fri, 25 Jun 2021 23:35:05 +0000 Received: by outflank-mailman (input) for mailman id 147475; Fri, 25 Jun 2021 23:35:03 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvLj-0002Fd-MA for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:35:03 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvLj-0006Rq-LO for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:35:03 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwvLj-0006hK-Kg for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:35:03 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=CkXNdhVM5x2HGUOx0nbyNSSECkHIuMjnthcX/a6ad+s=; b=P4E6qkw8ctxIcdoeaTpOKcU7m8 gI3WhsNKviVDxsJxpUZalR11MQVvrE17JSALtQvZ4A6V4CIISXfZkTpyF0SUvIt+Ps4FbkGec/+9T kaaht/G1URNzKTaEwrkLT7l2yTErQKdKy3T2C6ENLcGRsp9ZA/JtnRRLnBHvsyoGB1tw=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] tools/xenstored: Don't crash xenstored when Live-Update is cancelled Message-Id: Date: Fri, 25 Jun 2021 23:35:03 +0000 commit 90bafdba8ebf41c9af31b5c725a938da2a75d292 Author: Julien GralL AuthorDate: Thu Jun 24 12:15:49 2021 +0100 Commit: Julien Grall CommitDate: Thu Jun 24 12:15:49 2021 +0100 tools/xenstored: Don't crash xenstored when Live-Update is cancelled As Live-Update is asynchronous, it is possible to receive a request to cancel it (either on the same connection or from a different one). Currently, this will crash xenstored because do_lu_start() assumes lu_status will be valid. This is not the case when Live-Update has been cancelled. This will result to dereference a NULL pointer and crash Xenstored. Rework do_lu_start() to check if lu_status is NULL and return an error in this case. Fixes: af216a99fb ("tools/xenstore: add the basic framework for doing the live update") Signed-off-by: Julien Grall Reviewed-by: Luca Fancellu Reviewed-by: Juergen Gross --- tools/xenstore/xenstored_control.c | 15 +++++++++++++-- 1 file changed, 13 insertions(+), 2 deletions(-) diff --git a/tools/xenstore/xenstored_control.c b/tools/xenstore/xenstored_control.c index a045f102a4..a1b1bd5a71 100644 --- a/tools/xenstore/xenstored_control.c +++ b/tools/xenstore/xenstored_control.c @@ -696,7 +696,18 @@ static bool do_lu_start(struct delayed_request *req) time_t now = time(NULL); const char *ret; struct buffered_data *saved_in; - struct connection *conn = lu_status->conn; + struct connection *conn = req->data; + + /* + * Cancellation may have been requested asynchronously. In this + * case, lu_status will be NULL. + */ + if (!lu_status) { + ret = "Cancellation was requested"; + goto out; + } + + assert(lu_status->conn == conn); if (!lu_check_lu_allowed()) { if (now < lu_status->started_at + lu_status->timeout) @@ -747,7 +758,7 @@ static const char *lu_start(const void *ctx, struct connection *conn, lu_status->timeout = to; lu_status->started_at = time(NULL); - errno = delay_request(conn, conn->in, do_lu_start, NULL, false); + errno = delay_request(conn, conn->in, do_lu_start, conn, false); return NULL; } -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Fri Jun 25 23:35:15 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Fri, 25 Jun 2021 23:35:15 +0000 Received: from list by lists.xenproject.org with outflank-mailman.147478.271917 (Exim 4.92) (envelope-from ) id 1lwvLv-0002PK-H7; Fri, 25 Jun 2021 23:35:15 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 147478.271917; Fri, 25 Jun 2021 23:35:15 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvLv-0002PC-E2; Fri, 25 Jun 2021 23:35:15 +0000 Received: by outflank-mailman (input) for mailman id 147478; Fri, 25 Jun 2021 23:35:13 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvLt-0002Op-PK for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:35:13 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvLt-0006S4-Oa for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:35:13 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwvLt-0006iN-Nr for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:35:13 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=Zf0VJ754gtICrlCKlYL5ZRq8JC51HV+MN6Tz2ZLE5T0=; b=KViUSvT28Od4nK1hMWzjN9DiWM EsH3WOsXgFEEI/Pw/9T4OhtGUgjGpBvZCYOHNJCWdF3n0ZXmEj8zdj2yMkK1FzhOSZ1GyVIi/41k1 s3RPHxabs7foAmc8bjcjTOaOIb/fES1kuSQPC60bhnjHCY9fFtx/OTokQ9XTYO6jfG9c=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] VT-d: undo device mappings upon error Message-Id: Date: Fri, 25 Jun 2021 23:35:13 +0000 commit f3401d65d9f0dce508c3d7da55de4a093d748ae1 Author: Jan Beulich AuthorDate: Thu Jun 24 16:28:25 2021 +0200 Commit: Jan Beulich CommitDate: Thu Jun 24 16:28:25 2021 +0200 VT-d: undo device mappings upon error When - flushes (supposedly not possible anymore after XSA-373), - secondary mappings for legacy PCI devices behind bridges, - secondary mappings for chipset quirks, or - find_upstream_bridge() invocations fail, the successfully established device mappings should not be left around. Further, when (parts of) unmapping fail, simply returning an error is typically not enough. Crash the domain instead in such cases, arranging for domain cleanup to continue in a best effort manner despite such failures. Finally make domain_context_unmap()'s error behavior consistent in the legacy PCI device case: Don't bail from the function in one special case, but always just exit the switch statement. Signed-off-by: Jan Beulich Reviewed-by: Paul Durrant Reviewed-by: Kevin Tian --- xen/drivers/passthrough/vtd/iommu.c | 57 +++++++++++++++++++++++++++++++------ 1 file changed, 48 insertions(+), 9 deletions(-) diff --git a/xen/drivers/passthrough/vtd/iommu.c b/xen/drivers/passthrough/vtd/iommu.c index b7aa15e5ef..d9062422d5 100644 --- a/xen/drivers/passthrough/vtd/iommu.c +++ b/xen/drivers/passthrough/vtd/iommu.c @@ -1442,9 +1442,15 @@ int domain_context_mapping_one( if ( !seg && !rc ) rc = me_wifi_quirk(domain, bus, devfn, MAP_ME_PHANTOM_FUNC); + if ( rc ) + domain_context_unmap_one(domain, iommu, bus, devfn); + return rc; } +static int domain_context_unmap(struct domain *d, uint8_t devfn, + struct pci_dev *pdev); + static int domain_context_mapping(struct domain *domain, u8 devfn, struct pci_dev *pdev) { @@ -1505,16 +1511,21 @@ static int domain_context_mapping(struct domain *domain, u8 devfn, if ( ret ) break; - if ( find_upstream_bridge(seg, &bus, &devfn, &secbus) < 1 ) - break; + if ( (ret = find_upstream_bridge(seg, &bus, &devfn, &secbus)) < 1 ) + { + if ( !ret ) + break; + ret = -ENXIO; + } /* * Mapping a bridge should, if anything, pass the struct pci_dev of * that bridge. Since bridges don't normally get assigned to guests, * their owner would be the wrong one. Pass NULL instead. */ - ret = domain_context_mapping_one(domain, drhd->iommu, bus, devfn, - NULL); + if ( ret >= 0 ) + ret = domain_context_mapping_one(domain, drhd->iommu, bus, devfn, + NULL); /* * Devices behind PCIe-to-PCI/PCIx bridge may generate different @@ -1531,6 +1542,9 @@ static int domain_context_mapping(struct domain *domain, u8 devfn, ret = domain_context_mapping_one(domain, drhd->iommu, secbus, 0, NULL); + if ( ret ) + domain_context_unmap(domain, devfn, pdev); + break; default: @@ -1609,6 +1623,19 @@ int domain_context_unmap_one( if ( !iommu->drhd->segment && !rc ) rc = me_wifi_quirk(domain, bus, devfn, UNMAP_ME_PHANTOM_FUNC); + if ( rc && !is_hardware_domain(domain) && domain != dom_io ) + { + if ( domain->is_dying ) + { + printk(XENLOG_ERR "%pd: error %d unmapping %04x:%02x:%02x.%u\n", + domain, rc, iommu->drhd->segment, bus, + PCI_SLOT(devfn), PCI_FUNC(devfn)); + rc = 0; /* Make upper layers continue in a best effort manner. */ + } + else + domain_crash(domain); + } + return rc; } @@ -1661,17 +1688,29 @@ static int domain_context_unmap(struct domain *domain, u8 devfn, tmp_bus = bus; tmp_devfn = devfn; - if ( find_upstream_bridge(seg, &tmp_bus, &tmp_devfn, &secbus) < 1 ) + if ( (ret = find_upstream_bridge(seg, &tmp_bus, &tmp_devfn, + &secbus)) < 1 ) + { + if ( ret ) + { + ret = -ENXIO; + if ( !domain->is_dying && + !is_hardware_domain(domain) && domain != dom_io ) + { + domain_crash(domain); + /* Make upper layers continue in a best effort manner. */ + ret = 0; + } + } break; + } /* PCIe to PCI/PCIx bridge */ if ( pdev_type(seg, tmp_bus, tmp_devfn) == DEV_TYPE_PCIe2PCI_BRIDGE ) { ret = domain_context_unmap_one(domain, iommu, tmp_bus, tmp_devfn); - if ( ret ) - return ret; - - ret = domain_context_unmap_one(domain, iommu, secbus, 0); + if ( !ret ) + ret = domain_context_unmap_one(domain, iommu, secbus, 0); } else /* Legacy PCI bridge */ ret = domain_context_unmap_one(domain, iommu, tmp_bus, tmp_devfn); -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Fri Jun 25 23:35:25 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Fri, 25 Jun 2021 23:35:25 +0000 Received: from list by lists.xenproject.org with outflank-mailman.147480.271921 (Exim 4.92) (envelope-from ) id 1lwvM5-0002U1-J9; Fri, 25 Jun 2021 23:35:25 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 147480.271921; Fri, 25 Jun 2021 23:35:25 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvM5-0002Tt-Fb; Fri, 25 Jun 2021 23:35:25 +0000 Received: by outflank-mailman (input) for mailman id 147480; Fri, 25 Jun 2021 23:35:23 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvM3-0002Tb-Sd for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:35:23 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvM3-0006SI-Rw for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:35:23 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwvM3-0006jZ-R2 for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:35:23 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=kJTIRWU2AG9X/kdQkJb0kkVgcBCoPtzbtjywf7jxtyE=; b=TLgkFkXQcSQgDdAsp2sPIDQca9 StCXo0mJA54/oq6y9HooDkmjogs9dDZhTnOhTIv1Q7cIb6mEp3PEL2G2O7FPYfLtjTITdwkZ0nyPd aZGAOsoWcN6auOP4GpL7kSYAhORCtHnp7LFkXt25YReWfBttdnBEJJ4gQk/IaoTFIdiE=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] VT-d: adjust domid map updating when unmapping context Message-Id: Date: Fri, 25 Jun 2021 23:35:23 +0000 commit 32655880057ce2829f962d46916ea6cec60f98d3 Author: Jan Beulich AuthorDate: Thu Jun 24 16:29:13 2021 +0200 Commit: Jan Beulich CommitDate: Thu Jun 24 16:29:13 2021 +0200 VT-d: adjust domid map updating when unmapping context When an earlier error occurred, cleaning up the domid mapping data is wrong, as references likely still exist. The only exception to this is when the actual unmapping worked, but some flush failed (supposedly impossible after XSA-373). The guest will get crashed in such a case though, so add fallback cleanup to domain destruction to cover this case. This in turn makes it desirable to silence the dprintk() in domain_iommu_domid(). Note that no error will be returned anymore when the lookup fails - in the common case lookup failure would already have caused domain_context_unmap_one() to fail, yet even from a more general perspective it doesn't look right to fail domain_context_unmap() in such a case when this was the last device, but not when any earlier unmap was otherwise successful. Signed-off-by: Jan Beulich Reviewed-by: Kevin Tian --- xen/drivers/passthrough/vtd/iommu.c | 39 +++++++++++++++++++++++-------------- 1 file changed, 24 insertions(+), 15 deletions(-) diff --git a/xen/drivers/passthrough/vtd/iommu.c b/xen/drivers/passthrough/vtd/iommu.c index d9062422d5..0342daaa3c 100644 --- a/xen/drivers/passthrough/vtd/iommu.c +++ b/xen/drivers/passthrough/vtd/iommu.c @@ -80,9 +80,11 @@ static int domain_iommu_domid(struct domain *d, i = find_next_bit(iommu->domid_bitmap, nr_dom, i+1); } - dprintk(XENLOG_ERR VTDPREFIX, - "Cannot get valid iommu domid: domid=%d iommu->index=%d\n", - d->domain_id, iommu->index); + if ( !d->is_dying ) + dprintk(XENLOG_ERR VTDPREFIX, + "Cannot get valid iommu %u domid: %pd\n", + iommu->index, d); + return -1; } @@ -147,6 +149,17 @@ static int context_get_domain_id(struct context_entry *context, return domid; } +static void cleanup_domid_map(struct domain *domain, struct vtd_iommu *iommu) +{ + int iommu_domid = domain_iommu_domid(domain, iommu); + + if ( iommu_domid >= 0 ) + { + clear_bit(iommu_domid, iommu->domid_bitmap); + iommu->domid_map[iommu_domid] = 0; + } +} + static void sync_cache(const void *addr, unsigned int size) { static unsigned long clflush_size = 0; @@ -1724,6 +1737,9 @@ static int domain_context_unmap(struct domain *domain, u8 devfn, goto out; } + if ( ret ) + goto out; + /* * if no other devices under the same iommu owned by this domain, * clear iommu in iommu_bitmap and clear domain_id in domid_bitmp @@ -1743,19 +1759,8 @@ static int domain_context_unmap(struct domain *domain, u8 devfn, if ( found == 0 ) { - int iommu_domid; - clear_bit(iommu->index, &dom_iommu(domain)->arch.vtd.iommu_bitmap); - - iommu_domid = domain_iommu_domid(domain, iommu); - if ( iommu_domid == -1 ) - { - ret = -EINVAL; - goto out; - } - - clear_bit(iommu_domid, iommu->domid_bitmap); - iommu->domid_map[iommu_domid] = 0; + cleanup_domid_map(domain, iommu); } out: @@ -1775,6 +1780,7 @@ static void iommu_domain_teardown(struct domain *d) { struct domain_iommu *hd = dom_iommu(d); struct mapped_rmrr *mrmrr, *tmp; + const struct acpi_drhd_unit *drhd; if ( list_empty(&acpi_drhd_units) ) return; @@ -1786,6 +1792,9 @@ static void iommu_domain_teardown(struct domain *d) } ASSERT(!hd->arch.vtd.pgd_maddr); + + for_each_drhd_unit ( drhd ) + cleanup_domid_map(d, drhd->iommu); } static int __must_check intel_iommu_map_page(struct domain *d, dfn_t dfn, -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Fri Jun 25 23:35:35 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Fri, 25 Jun 2021 23:35:35 +0000 Received: from list by lists.xenproject.org with outflank-mailman.147481.271925 (Exim 4.92) (envelope-from ) id 1lwvMF-0002Xa-MK; Fri, 25 Jun 2021 23:35:35 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 147481.271925; Fri, 25 Jun 2021 23:35:35 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvMF-0002XP-JL; Fri, 25 Jun 2021 23:35:35 +0000 Received: by outflank-mailman (input) for mailman id 147481; Fri, 25 Jun 2021 23:35:34 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvMD-0002X1-Vw for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:35:33 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvMD-0006So-VE for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:35:33 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwvMD-0006kc-US for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:35:33 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=JRnusCcsgMCOdoEj+uGiRPPCss2jVIf1W7VmKkuyZVk=; b=MQPV+D+azOGULsum2fLIuL2/Mp 4mjj09POPssZCMd32GBT9V9QKiugVMdhGparVqCE5RbwjCmAhclRei/SAQG631kE9zRB5BoRlVLFO wLsoZ+cBsh+zivxzBr64kr2lclm4il85QYkvIDL0in4Z2Chj4pQDovdWg69hp+dgIWCk=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] VT-d: clear_fault_bits() should clear all fault bits Message-Id: Date: Fri, 25 Jun 2021 23:35:33 +0000 commit 80589800ae62fce43fd3921e8fbd362465fe5ba3 Author: Jan Beulich AuthorDate: Thu Jun 24 16:29:42 2021 +0200 Commit: Jan Beulich CommitDate: Thu Jun 24 16:29:42 2021 +0200 VT-d: clear_fault_bits() should clear all fault bits If there is any way for one fault to be left set in the recording registers, there's no reason there couldn't also be multiple ones. If PPF set set (being the OR or all F fields), simply loop over the entire range of fault recording registers, clearing F everywhere. Since PPF is a r/o bit, also remove it from DMA_FSTS_FAULTS (arguably the constant's name is ambiguous as well). Signed-off-by: Jan Beulich Reviewed-by: Kevin Tian --- xen/drivers/passthrough/vtd/iommu.c | 16 +++++++++++++--- xen/drivers/passthrough/vtd/iommu.h | 5 ++--- 2 files changed, 15 insertions(+), 6 deletions(-) diff --git a/xen/drivers/passthrough/vtd/iommu.c b/xen/drivers/passthrough/vtd/iommu.c index 0342daaa3c..f3524a1def 100644 --- a/xen/drivers/passthrough/vtd/iommu.c +++ b/xen/drivers/passthrough/vtd/iommu.c @@ -2094,13 +2094,23 @@ static int __hwdom_init setup_hwdom_device(u8 devfn, struct pci_dev *pdev) void clear_fault_bits(struct vtd_iommu *iommu) { - u64 val; unsigned long flags; spin_lock_irqsave(&iommu->register_lock, flags); - val = dmar_readq(iommu->reg, cap_fault_reg_offset(iommu->cap) + 8); - dmar_writeq(iommu->reg, cap_fault_reg_offset(iommu->cap) + 8, val); + + if ( dmar_readl(iommu->reg, DMAR_FSTS_REG) & DMA_FSTS_PPF ) + { + unsigned int reg = cap_fault_reg_offset(iommu->cap); + unsigned int end = reg + cap_num_fault_regs(iommu->cap); + + do { + dmar_writel(iommu->reg, reg + 12, DMA_FRCD_F); + reg += PRIMARY_FAULT_REG_LEN; + } while ( reg < end ); + } + dmar_writel(iommu->reg, DMAR_FSTS_REG, DMA_FSTS_FAULTS); + spin_unlock_irqrestore(&iommu->register_lock, flags); } diff --git a/xen/drivers/passthrough/vtd/iommu.h b/xen/drivers/passthrough/vtd/iommu.h index cd516ac5f8..02390cb84b 100644 --- a/xen/drivers/passthrough/vtd/iommu.h +++ b/xen/drivers/passthrough/vtd/iommu.h @@ -174,9 +174,8 @@ #define DMA_FSTS_IQE (1u << 4) #define DMA_FSTS_ICE (1u << 5) #define DMA_FSTS_ITE (1u << 6) -#define DMA_FSTS_FAULTS (DMA_FSTS_PFO | DMA_FSTS_PPF | DMA_FSTS_AFO | \ - DMA_FSTS_APF | DMA_FSTS_IQE | DMA_FSTS_ICE | \ - DMA_FSTS_ITE) +#define DMA_FSTS_FAULTS (DMA_FSTS_PFO | DMA_FSTS_AFO | DMA_FSTS_APF | \ + DMA_FSTS_IQE | DMA_FSTS_ICE | DMA_FSTS_ITE) #define dma_fsts_fault_record_index(s) (((s) >> 8) & 0xff) /* FRCD_REG, 32 bits access */ -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Fri Jun 25 23:35:45 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Fri, 25 Jun 2021 23:35:45 +0000 Received: from list by lists.xenproject.org with outflank-mailman.147482.271929 (Exim 4.92) (envelope-from ) id 1lwvMP-0002b9-Nk; Fri, 25 Jun 2021 23:35:45 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 147482.271929; Fri, 25 Jun 2021 23:35:45 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvMP-0002b3-Km; Fri, 25 Jun 2021 23:35:45 +0000 Received: by outflank-mailman (input) for mailman id 147482; Fri, 25 Jun 2021 23:35:44 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvMO-0002ai-2f for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:35:44 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvMO-0006Sy-1w for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:35:44 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwvMO-0006m5-1J for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:35:44 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=DF3s8TuV2wC2yEgVlEOr5CR4Z70STC8jcpHFDsrGook=; b=aUFYc6voXTJJozuKJ4JKFX5CQb zOOXcs9hsffpoFbWXoRCXgu5F/0poTesnc61V7lTyu4cM/Mg5MuGPwVQM9BmDBCHLzK26fodvdQgl Tu2EhmalcezlKNCP6lcNHt0Siw/SDAf59FR/WWy0E1mUe6xEVdKmZTqzU3KOPirqQPSk=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] VT-d: don't lose errors when flushing TLBs on multiple IOMMUs Message-Id: Date: Fri, 25 Jun 2021 23:35:44 +0000 commit e7059776f9755b989a992d229c68c3d7778412be Author: Jan Beulich AuthorDate: Thu Jun 24 16:30:06 2021 +0200 Commit: Jan Beulich CommitDate: Thu Jun 24 16:30:06 2021 +0200 VT-d: don't lose errors when flushing TLBs on multiple IOMMUs While no longer an immediate problem with flushes no longer timing out, errors (if any) get properly reported by iommu_flush_iotlb_{dsi,psi}(). Overwriting such an error with, perhaps, a success indicator received from another IOMMU will misguide callers. Record the first error, but don't bail from the loop (such that further necessary invalidation gets carried out on a best effort basis). Signed-off-by: Jan Beulich Reviewed-by: Kevin Tian --- xen/drivers/passthrough/vtd/iommu.c | 11 ++++++----- 1 file changed, 6 insertions(+), 5 deletions(-) diff --git a/xen/drivers/passthrough/vtd/iommu.c b/xen/drivers/passthrough/vtd/iommu.c index f3524a1def..0f10951ce9 100644 --- a/xen/drivers/passthrough/vtd/iommu.c +++ b/xen/drivers/passthrough/vtd/iommu.c @@ -643,7 +643,7 @@ static int __must_check iommu_flush_iotlb(struct domain *d, dfn_t dfn, struct vtd_iommu *iommu; bool_t flush_dev_iotlb; int iommu_domid; - int rc = 0; + int ret = 0; /* * No need pcideves_lock here because we have flush @@ -651,6 +651,8 @@ static int __must_check iommu_flush_iotlb(struct domain *d, dfn_t dfn, */ for_each_drhd_unit ( drhd ) { + int rc; + iommu = drhd->iommu; if ( !test_bit(iommu->index, &hd->arch.vtd.iommu_bitmap) ) @@ -673,13 +675,12 @@ static int __must_check iommu_flush_iotlb(struct domain *d, dfn_t dfn, flush_dev_iotlb); if ( rc > 0 ) - { iommu_flush_write_buffer(iommu); - rc = 0; - } + else if ( !ret ) + ret = rc; } - return rc; + return ret; } static int __must_check iommu_flush_iotlb_pages(struct domain *d, -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Fri Jun 25 23:35:54 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Fri, 25 Jun 2021 23:35:54 +0000 Received: from list by lists.xenproject.org with outflank-mailman.147483.271933 (Exim 4.92) (envelope-from ) id 1lwvMY-0002e7-PH; Fri, 25 Jun 2021 23:35:54 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 147483.271933; Fri, 25 Jun 2021 23:35:54 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvMY-0002dz-MG; Fri, 25 Jun 2021 23:35:54 +0000 Received: by outflank-mailman (input) for mailman id 147483; Fri, 25 Jun 2021 23:35:54 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvMY-0002ds-66 for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:35:54 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvMY-0006TF-5L for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:35:54 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwvMY-0006ns-4Z for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:35:54 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=VDl7ClclBjt/LYWqXnwd+trLTNqmSEp/v8sH3kg/Y0Q=; b=yXfXlDjvrXjsHcbYhPdshAzCL2 8o1yKjXb3Td60IK73UMrPJ0+3EG6drjTr+r00zx4P5pq3gtOXbN/qBMZFL9Oodg0FFeIyBKVZqHna tkwjsXMMqxV4s1dnWAbdXuVUl2uj695+CubgaOlBkwwDXb92cpGg4HSGoKYhclGWIE7k=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] VT-d: centralize mapping of QI entries Message-Id: Date: Fri, 25 Jun 2021 23:35:54 +0000 commit a0eb19723d9a6006bf465f0af60e229206d39333 Author: Jan Beulich AuthorDate: Thu Jun 24 16:30:32 2021 +0200 Commit: Jan Beulich CommitDate: Thu Jun 24 16:30:32 2021 +0200 VT-d: centralize mapping of QI entries Introduce a helper function to reduce redundancy. Take the opportunity to express the logic without using the somewhat odd QINVAL_ENTRY_ORDER. Also take the opportunity to uniformly unmap after updating queue tail and dropping the lock (like was done so far only by queue_invalidate_context_sync()). Signed-off-by: Jan Beulich Reviewed-by: Kevin Tian --- xen/drivers/passthrough/vtd/qinval.c | 64 ++++++++++++++++-------------------- 1 file changed, 29 insertions(+), 35 deletions(-) diff --git a/xen/drivers/passthrough/vtd/qinval.c b/xen/drivers/passthrough/vtd/qinval.c index 311bf107ed..2e4a55a869 100644 --- a/xen/drivers/passthrough/vtd/qinval.c +++ b/xen/drivers/passthrough/vtd/qinval.c @@ -69,6 +69,16 @@ static void qinval_update_qtail(struct vtd_iommu *iommu, unsigned int index) dmar_writel(iommu->reg, DMAR_IQT_REG, val << QINVAL_INDEX_SHIFT); } +static struct qinval_entry *qi_map_entry(const struct vtd_iommu *iommu, + unsigned int index) +{ + paddr_t base = iommu->qinval_maddr + + ((index * sizeof(struct qinval_entry)) & PAGE_MASK); + struct qinval_entry *entries = map_vtd_domain_page(base); + + return &entries[index % (PAGE_SIZE / sizeof(*entries))]; +} + static int __must_check queue_invalidate_context_sync(struct vtd_iommu *iommu, u16 did, u16 source_id, u8 function_mask, @@ -76,15 +86,11 @@ static int __must_check queue_invalidate_context_sync(struct vtd_iommu *iommu, { unsigned long flags; unsigned int index; - u64 entry_base; - struct qinval_entry *qinval_entry, *qinval_entries; + struct qinval_entry *qinval_entry; spin_lock_irqsave(&iommu->register_lock, flags); index = qinval_next_index(iommu); - entry_base = iommu->qinval_maddr + - ((index >> QINVAL_ENTRY_ORDER) << PAGE_SHIFT); - qinval_entries = map_vtd_domain_page(entry_base); - qinval_entry = &qinval_entries[index % (1 << QINVAL_ENTRY_ORDER)]; + qinval_entry = qi_map_entry(iommu, index); qinval_entry->q.cc_inv_dsc.lo.type = TYPE_INVAL_CONTEXT; qinval_entry->q.cc_inv_dsc.lo.granu = granu; @@ -98,7 +104,7 @@ static int __must_check queue_invalidate_context_sync(struct vtd_iommu *iommu, qinval_update_qtail(iommu, index); spin_unlock_irqrestore(&iommu->register_lock, flags); - unmap_vtd_domain_page(qinval_entries); + unmap_vtd_domain_page(qinval_entry); return invalidate_sync(iommu); } @@ -110,15 +116,11 @@ static int __must_check queue_invalidate_iotlb_sync(struct vtd_iommu *iommu, { unsigned long flags; unsigned int index; - u64 entry_base; - struct qinval_entry *qinval_entry, *qinval_entries; + struct qinval_entry *qinval_entry; spin_lock_irqsave(&iommu->register_lock, flags); index = qinval_next_index(iommu); - entry_base = iommu->qinval_maddr + - ((index >> QINVAL_ENTRY_ORDER) << PAGE_SHIFT); - qinval_entries = map_vtd_domain_page(entry_base); - qinval_entry = &qinval_entries[index % (1 << QINVAL_ENTRY_ORDER)]; + qinval_entry = qi_map_entry(iommu, index); qinval_entry->q.iotlb_inv_dsc.lo.type = TYPE_INVAL_IOTLB; qinval_entry->q.iotlb_inv_dsc.lo.granu = granu; @@ -133,10 +135,11 @@ static int __must_check queue_invalidate_iotlb_sync(struct vtd_iommu *iommu, qinval_entry->q.iotlb_inv_dsc.hi.res_1 = 0; qinval_entry->q.iotlb_inv_dsc.hi.addr = addr >> PAGE_SHIFT_4K; - unmap_vtd_domain_page(qinval_entries); qinval_update_qtail(iommu, index); spin_unlock_irqrestore(&iommu->register_lock, flags); + unmap_vtd_domain_page(qinval_entry); + return invalidate_sync(iommu); } @@ -147,17 +150,13 @@ static int __must_check queue_invalidate_wait(struct vtd_iommu *iommu, static DEFINE_PER_CPU(uint32_t, poll_slot); unsigned int index; unsigned long flags; - u64 entry_base; - struct qinval_entry *qinval_entry, *qinval_entries; + struct qinval_entry *qinval_entry; uint32_t *this_poll_slot = &this_cpu(poll_slot); spin_lock_irqsave(&iommu->register_lock, flags); ACCESS_ONCE(*this_poll_slot) = QINVAL_STAT_INIT; index = qinval_next_index(iommu); - entry_base = iommu->qinval_maddr + - ((index >> QINVAL_ENTRY_ORDER) << PAGE_SHIFT); - qinval_entries = map_vtd_domain_page(entry_base); - qinval_entry = &qinval_entries[index % (1 << QINVAL_ENTRY_ORDER)]; + qinval_entry = qi_map_entry(iommu, index); qinval_entry->q.inv_wait_dsc.lo.type = TYPE_INVAL_WAIT; qinval_entry->q.inv_wait_dsc.lo.iflag = iflag; @@ -167,10 +166,11 @@ static int __must_check queue_invalidate_wait(struct vtd_iommu *iommu, qinval_entry->q.inv_wait_dsc.lo.sdata = QINVAL_STAT_DONE; qinval_entry->q.inv_wait_dsc.hi.saddr = virt_to_maddr(this_poll_slot); - unmap_vtd_domain_page(qinval_entries); qinval_update_qtail(iommu, index); spin_unlock_irqrestore(&iommu->register_lock, flags); + unmap_vtd_domain_page(qinval_entry); + /* Now we don't support interrupt method */ if ( sw ) { @@ -246,16 +246,12 @@ int qinval_device_iotlb_sync(struct vtd_iommu *iommu, struct pci_dev *pdev, { unsigned long flags; unsigned int index; - u64 entry_base; - struct qinval_entry *qinval_entry, *qinval_entries; + struct qinval_entry *qinval_entry; ASSERT(pdev); spin_lock_irqsave(&iommu->register_lock, flags); index = qinval_next_index(iommu); - entry_base = iommu->qinval_maddr + - ((index >> QINVAL_ENTRY_ORDER) << PAGE_SHIFT); - qinval_entries = map_vtd_domain_page(entry_base); - qinval_entry = &qinval_entries[index % (1 << QINVAL_ENTRY_ORDER)]; + qinval_entry = qi_map_entry(iommu, index); qinval_entry->q.dev_iotlb_inv_dsc.lo.type = TYPE_INVAL_DEVICE_IOTLB; qinval_entry->q.dev_iotlb_inv_dsc.lo.res_1 = 0; @@ -268,10 +264,11 @@ int qinval_device_iotlb_sync(struct vtd_iommu *iommu, struct pci_dev *pdev, qinval_entry->q.dev_iotlb_inv_dsc.hi.res_1 = 0; qinval_entry->q.dev_iotlb_inv_dsc.hi.addr = addr >> PAGE_SHIFT_4K; - unmap_vtd_domain_page(qinval_entries); qinval_update_qtail(iommu, index); spin_unlock_irqrestore(&iommu->register_lock, flags); + unmap_vtd_domain_page(qinval_entry); + return dev_invalidate_sync(iommu, pdev, did); } @@ -280,16 +277,12 @@ static int __must_check queue_invalidate_iec_sync(struct vtd_iommu *iommu, { unsigned long flags; unsigned int index; - u64 entry_base; - struct qinval_entry *qinval_entry, *qinval_entries; + struct qinval_entry *qinval_entry; int ret; spin_lock_irqsave(&iommu->register_lock, flags); index = qinval_next_index(iommu); - entry_base = iommu->qinval_maddr + - ((index >> QINVAL_ENTRY_ORDER) << PAGE_SHIFT); - qinval_entries = map_vtd_domain_page(entry_base); - qinval_entry = &qinval_entries[index % (1 << QINVAL_ENTRY_ORDER)]; + qinval_entry = qi_map_entry(iommu, index); qinval_entry->q.iec_inv_dsc.lo.type = TYPE_INVAL_IEC; qinval_entry->q.iec_inv_dsc.lo.granu = granu; @@ -299,10 +292,11 @@ static int __must_check queue_invalidate_iec_sync(struct vtd_iommu *iommu, qinval_entry->q.iec_inv_dsc.lo.res_2 = 0; qinval_entry->q.iec_inv_dsc.hi.res = 0; - unmap_vtd_domain_page(qinval_entries); qinval_update_qtail(iommu, index); spin_unlock_irqrestore(&iommu->register_lock, flags); + unmap_vtd_domain_page(qinval_entry); + ret = invalidate_sync(iommu); /* -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Fri Jun 25 23:36:04 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Fri, 25 Jun 2021 23:36:04 +0000 Received: from list by lists.xenproject.org with outflank-mailman.147484.271937 (Exim 4.92) (envelope-from ) id 1lwvMi-0002hT-R9; Fri, 25 Jun 2021 23:36:04 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 147484.271937; Fri, 25 Jun 2021 23:36:04 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvMi-0002hL-Nu; Fri, 25 Jun 2021 23:36:04 +0000 Received: by outflank-mailman (input) for mailman id 147484; Fri, 25 Jun 2021 23:36:04 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvMi-0002hF-99 for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:36:04 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvMi-0006Td-8N for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:36:04 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwvMi-0006pI-7m for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:36:04 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=4bIx3h/nKAvpZO0j4NkFm0PpJpGPHkFbH/YnWWz96XA=; b=flWb5P5oqJp5GRMqSsNDa3k1OZ HditiMlXwe9t9Ub6dcH+u00i6kEWBnCM7JmOqwp2K8kDY+z1blKHB4z9sSTnmeAv3NH0Ncbvk2oTS 9ca9si7zofiZMS5oNg+Ba5sCjozfkjkkq+ilpTq1Py11a40CAaEq+2onFcVm8bwngA5E=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] VT-d: drop/move a few QI related constants Message-Id: Date: Fri, 25 Jun 2021 23:36:04 +0000 commit f7db924309c25b18951c058f80d3e5979114b055 Author: Jan Beulich AuthorDate: Thu Jun 24 16:30:51 2021 +0200 Commit: Jan Beulich CommitDate: Thu Jun 24 16:30:51 2021 +0200 VT-d: drop/move a few QI related constants Replace uses of QINVAL_ENTRY_ORDER and QINVAL_INDEX_SHIFT, such that the constants can be dropped. Move the remaining QINVAL_* ones to the single source file using them. Signed-off-by: Jan Beulich Reviewed-by: Kevin Tian --- xen/drivers/passthrough/vtd/iommu.h | 11 ----------- xen/drivers/passthrough/vtd/qinval.c | 24 ++++++++++++++++-------- 2 files changed, 16 insertions(+), 19 deletions(-) diff --git a/xen/drivers/passthrough/vtd/iommu.h b/xen/drivers/passthrough/vtd/iommu.h index 02390cb84b..ee28ecac6d 100644 --- a/xen/drivers/passthrough/vtd/iommu.h +++ b/xen/drivers/passthrough/vtd/iommu.h @@ -451,17 +451,6 @@ struct qinval_entry { }q; }; -/* Each entry is 16 bytes, so 2^8 entries per page */ -#define QINVAL_ENTRY_ORDER ( PAGE_SHIFT - 4 ) -#define QINVAL_MAX_ENTRY_NR (1u << (7 + QINVAL_ENTRY_ORDER)) - -/* Status data flag */ -#define QINVAL_STAT_INIT 0 -#define QINVAL_STAT_DONE 1 - -/* Queue invalidation head/tail shift */ -#define QINVAL_INDEX_SHIFT 4 - #define TYPE_INVAL_CONTEXT 0x1 #define TYPE_INVAL_IOTLB 0x2 #define TYPE_INVAL_DEVICE_IOTLB 0x3 diff --git a/xen/drivers/passthrough/vtd/qinval.c b/xen/drivers/passthrough/vtd/qinval.c index 2e4a55a869..b0e3672231 100644 --- a/xen/drivers/passthrough/vtd/qinval.c +++ b/xen/drivers/passthrough/vtd/qinval.c @@ -29,6 +29,13 @@ #include "extern.h" #include "../ats.h" +/* Each entry is 16 bytes, and there can be up to 2^7 pages. */ +#define QINVAL_MAX_ENTRY_NR (1u << (7 + PAGE_SHIFT_4K - 4)) + +/* Status data flag */ +#define QINVAL_STAT_INIT 0 +#define QINVAL_STAT_DONE 1 + static unsigned int __read_mostly qi_pg_order; static unsigned int __read_mostly qi_entry_nr; @@ -45,11 +52,11 @@ static unsigned int qinval_next_index(struct vtd_iommu *iommu) { unsigned int tail = dmar_readl(iommu->reg, DMAR_IQT_REG); - tail >>= QINVAL_INDEX_SHIFT; + tail /= sizeof(struct qinval_entry); /* (tail+1 == head) indicates a full queue, wait for HW */ while ( ((tail + 1) & (qi_entry_nr - 1)) == - (dmar_readl(iommu->reg, DMAR_IQH_REG) >> QINVAL_INDEX_SHIFT) ) + (dmar_readl(iommu->reg, DMAR_IQH_REG) / sizeof(struct qinval_entry)) ) { printk_once(XENLOG_ERR VTDPREFIX " IOMMU#%u: no QI slot available\n", iommu->index); @@ -66,7 +73,7 @@ static void qinval_update_qtail(struct vtd_iommu *iommu, unsigned int index) /* Need hold register lock when update tail */ ASSERT( spin_is_locked(&iommu->register_lock) ); val = (index + 1) & (qi_entry_nr - 1); - dmar_writel(iommu->reg, DMAR_IQT_REG, val << QINVAL_INDEX_SHIFT); + dmar_writel(iommu->reg, DMAR_IQT_REG, val * sizeof(struct qinval_entry)); } static struct qinval_entry *qi_map_entry(const struct vtd_iommu *iommu, @@ -413,17 +420,18 @@ int enable_qinval(struct vtd_iommu *iommu) * only one entry left. */ BUILD_BUG_ON(CONFIG_NR_CPUS * 2 >= QINVAL_MAX_ENTRY_NR); - qi_pg_order = get_order_from_bytes((num_present_cpus() * 2 + 1) << - (PAGE_SHIFT - - QINVAL_ENTRY_ORDER)); - qi_entry_nr = 1u << (qi_pg_order + QINVAL_ENTRY_ORDER); + qi_pg_order = get_order_from_bytes((num_present_cpus() * 2 + 1) * + sizeof(struct qinval_entry)); + qi_entry_nr = (PAGE_SIZE << qi_pg_order) / + sizeof(struct qinval_entry); dprintk(XENLOG_INFO VTDPREFIX, "QI: using %u-entry ring(s)\n", qi_entry_nr); } iommu->qinval_maddr = - alloc_pgtable_maddr(qi_entry_nr >> QINVAL_ENTRY_ORDER, + alloc_pgtable_maddr(PFN_DOWN(qi_entry_nr * + sizeof(struct qinval_entry)), iommu->node); if ( iommu->qinval_maddr == 0 ) { -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Fri Jun 25 23:36:14 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Fri, 25 Jun 2021 23:36:14 +0000 Received: from list by lists.xenproject.org with outflank-mailman.147485.271941 (Exim 4.92) (envelope-from ) id 1lwvMs-0002kp-Sf; Fri, 25 Jun 2021 23:36:14 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 147485.271941; Fri, 25 Jun 2021 23:36:14 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvMs-0002kh-PW; Fri, 25 Jun 2021 23:36:14 +0000 Received: by outflank-mailman (input) for mailman id 147485; Fri, 25 Jun 2021 23:36:14 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvMs-0002kZ-C5 for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:36:14 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvMs-0006To-BN for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:36:14 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwvMs-0006qd-Ag for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:36:14 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=mo41Fb3dfFYumktpH7v0l8MwBNOQ1TKQqAhc9VNCR2Q=; b=wnfsx+v0dtJxT2+07TC8n46/3B SdbNEg/ImLa4SwInf6d3P2M7lII9PhpeqsqORDku0wVApCh65v6ISO4gwy3qMm4xsO5XpcCrZHZNd 6Svx+MYwaBtS7KMofyWpd/7SXpTvHlYf9gncRRHsZqnwTM6v8pFrUw/MijozuyXep0uE=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] x86/HVM: wire up multicalls Message-Id: Date: Fri, 25 Jun 2021 23:36:14 +0000 commit 198a2bc6f149ca41e16d90fc73de2c81affe4490 Author: Jan Beulich AuthorDate: Thu Jun 24 16:35:39 2021 +0200 Commit: Jan Beulich CommitDate: Thu Jun 24 16:35:39 2021 +0200 x86/HVM: wire up multicalls To be able to use them from, in particular, the tool stack, they need to be supported for all guest types. Note that xc_resource_op() already does, so would not work without this on PVH Dom0. Signed-off-by: Jan Beulich Begrudingly acked-by: Andrew Cooper Acked-by: Ian Jackson --- xen/arch/x86/hvm/hypercall.c | 35 +++++++++++++++++++++++++++++++++++ xen/arch/x86/hypercall.c | 14 ++++++++++---- xen/arch/x86/pv/hypercall.c | 3 ++- xen/include/asm-x86/multicall.h | 12 ++++++++++++ 4 files changed, 59 insertions(+), 5 deletions(-) diff --git a/xen/arch/x86/hvm/hypercall.c b/xen/arch/x86/hvm/hypercall.c index 2d27fcbd2c..9e1f7249f3 100644 --- a/xen/arch/x86/hvm/hypercall.c +++ b/xen/arch/x86/hvm/hypercall.c @@ -26,6 +26,7 @@ #include #include #include +#include #include #include @@ -125,6 +126,7 @@ static const struct { hypercall_fn_t *native, *compat; } hvm_hypercall_table[] = { HVM_CALL(memory_op), + COMPAT_CALL(multicall), #ifdef CONFIG_GRANT_TABLE HVM_CALL(grant_table_op), #endif @@ -334,6 +336,39 @@ int hvm_hypercall(struct cpu_user_regs *regs) return curr->hcall_preempted ? HVM_HCALL_preempted : HVM_HCALL_completed; } +enum mc_disposition hvm_do_multicall_call(struct mc_state *state) +{ + struct vcpu *curr = current; + hypercall_fn_t *func = NULL; + + if ( hvm_guest_x86_mode(curr) == 8 ) + { + struct multicall_entry *call = &state->call; + + if ( call->op < ARRAY_SIZE(hvm_hypercall_table) ) + func = array_access_nospec(hvm_hypercall_table, call->op).native; + if ( func ) + call->result = func(call->args[0], call->args[1], call->args[2], + call->args[3], call->args[4], call->args[5]); + else + call->result = -ENOSYS; + } + else + { + struct compat_multicall_entry *call = &state->compat_call; + + if ( call->op < ARRAY_SIZE(hvm_hypercall_table) ) + func = array_access_nospec(hvm_hypercall_table, call->op).compat; + if ( func ) + call->result = func(call->args[0], call->args[1], call->args[2], + call->args[3], call->args[4], call->args[5]); + else + call->result = -ENOSYS; + } + + return !hvm_get_cpl(curr) ? mc_continue : mc_preempt; +} + /* * Local variables: * mode: C diff --git a/xen/arch/x86/hypercall.c b/xen/arch/x86/hypercall.c index 73c1bcb147..2370d31d3f 100644 --- a/xen/arch/x86/hypercall.c +++ b/xen/arch/x86/hypercall.c @@ -20,6 +20,7 @@ */ #include +#include #ifdef CONFIG_COMPAT #define ARGS(x, n) \ @@ -273,13 +274,18 @@ int hypercall_xlat_continuation(unsigned int *id, unsigned int nr, return rc; } -#ifndef CONFIG_PV -/* Stub for arch_do_multicall_call */ -enum mc_disposition arch_do_multicall_call(struct mc_state *mc) +enum mc_disposition arch_do_multicall_call(struct mc_state *state) { + const struct domain *currd = current->domain; + + if ( is_pv_domain(currd) ) + return pv_do_multicall_call(state); + + if ( is_hvm_domain(currd) ) + return hvm_do_multicall_call(state); + return mc_exit; } -#endif /* * Local variables: diff --git a/xen/arch/x86/pv/hypercall.c b/xen/arch/x86/pv/hypercall.c index d573f74aa1..200b8a05b0 100644 --- a/xen/arch/x86/pv/hypercall.c +++ b/xen/arch/x86/pv/hypercall.c @@ -23,6 +23,7 @@ #include #include #include +#include #include #ifdef CONFIG_PV32 @@ -245,7 +246,7 @@ void pv_hypercall(struct cpu_user_regs *regs) perfc_incr(hypercalls); } -enum mc_disposition arch_do_multicall_call(struct mc_state *state) +enum mc_disposition pv_do_multicall_call(struct mc_state *state) { struct vcpu *curr = current; unsigned long op; diff --git a/xen/include/asm-x86/multicall.h b/xen/include/asm-x86/multicall.h new file mode 100644 index 0000000000..7e1d4c121a --- /dev/null +++ b/xen/include/asm-x86/multicall.h @@ -0,0 +1,12 @@ +/****************************************************************************** + * asm-x86/multicall.h + */ + +#ifndef __ASM_X86_MULTICALL_H__ +#define __ASM_X86_MULTICALL_H__ + +#include + +typeof(arch_do_multicall_call) pv_do_multicall_call, hvm_do_multicall_call; + +#endif /* __ASM_X86_MULTICALL_H__ */ -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Fri Jun 25 23:36:24 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Fri, 25 Jun 2021 23:36:24 +0000 Received: from list by lists.xenproject.org with outflank-mailman.147486.271945 (Exim 4.92) (envelope-from ) id 1lwvN2-0002oy-VP; Fri, 25 Jun 2021 23:36:24 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 147486.271945; Fri, 25 Jun 2021 23:36:24 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvN2-0002oq-SU; Fri, 25 Jun 2021 23:36:24 +0000 Received: by outflank-mailman (input) for mailman id 147486; Fri, 25 Jun 2021 23:36:24 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvN2-0002oh-FA for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:36:24 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvN2-0006Tz-EL for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:36:24 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwvN2-0006ri-Df for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:36:24 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=dzWqgaylvLv9WntqQIKw56WRHV6CGtEiEXp81tiLcpM=; b=lM3USHxfeNbZlFZoRDmmYVX6Cc XhHxj5BxkV050tqq8Dc0GwjF53Y6+jNF8lF8Bc4t5Rak7csrC/KUMXG9Wna94UVGHvlNtFXv/jGOR 9Kizz/HNjnEorRe7zRNFnKloNNJMHLNYXPltRURVdfmQQ095n2pbOAfwWPMsF8K7fUoM=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] libxencall: osdep_hypercall() should return long Message-Id: Date: Fri, 25 Jun 2021 23:36:24 +0000 commit 6409210a5f51860cb17b5e0f97debe164dab26d7 Author: Jan Beulich AuthorDate: Thu Jun 24 16:38:37 2021 +0200 Commit: Jan Beulich CommitDate: Thu Jun 24 16:38:37 2021 +0200 libxencall: osdep_hypercall() should return long Some hypercalls, memory-op in particular, can return values requiring more than 31 bits to represent. Hence the underlying layers need to make sure they won't truncate such values. (Note that for Solaris the function also gets renamed, to match the other OSes.) Due to them merely propagating ioctl()'s return value, this change is benign on Linux and Solaris. IOW there's an actual effect here only for the BSDs and MiniOS, but even then further adjustments are needed at the xencall() level. Signed-off-by: Jan Beulich Acked-by: Andrew Cooper Acked-by: Ian Jackson --- tools/libs/call/freebsd.c | 2 +- tools/libs/call/linux.c | 2 +- tools/libs/call/minios.c | 2 +- tools/libs/call/netbsd.c | 2 +- tools/libs/call/private.h | 2 +- tools/libs/call/solaris.c | 2 +- 6 files changed, 6 insertions(+), 6 deletions(-) diff --git a/tools/libs/call/freebsd.c b/tools/libs/call/freebsd.c index 28bfd852c1..6d2e8704b3 100644 --- a/tools/libs/call/freebsd.c +++ b/tools/libs/call/freebsd.c @@ -62,7 +62,7 @@ int osdep_xencall_close(xencall_handle *xcall) return close(fd); } -int osdep_hypercall(xencall_handle *xcall, privcmd_hypercall_t *hypercall) +long osdep_hypercall(xencall_handle *xcall, privcmd_hypercall_t *hypercall) { int fd = xcall->fd; int ret; diff --git a/tools/libs/call/linux.c b/tools/libs/call/linux.c index 51fa4899eb..6d588e6bea 100644 --- a/tools/libs/call/linux.c +++ b/tools/libs/call/linux.c @@ -80,7 +80,7 @@ int osdep_xencall_close(xencall_handle *xcall) return 0; } -int osdep_hypercall(xencall_handle *xcall, privcmd_hypercall_t *hypercall) +long osdep_hypercall(xencall_handle *xcall, privcmd_hypercall_t *hypercall) { return ioctl(xcall->fd, IOCTL_PRIVCMD_HYPERCALL, hypercall); } diff --git a/tools/libs/call/minios.c b/tools/libs/call/minios.c index 9f7a96995f..3f33b83578 100644 --- a/tools/libs/call/minios.c +++ b/tools/libs/call/minios.c @@ -38,7 +38,7 @@ int osdep_xencall_close(xencall_handle *xcall) return 0; } -int osdep_hypercall(xencall_handle *xcall, privcmd_hypercall_t *hypercall) +long osdep_hypercall(xencall_handle *xcall, privcmd_hypercall_t *hypercall) { multicall_entry_t call; int i, ret; diff --git a/tools/libs/call/netbsd.c b/tools/libs/call/netbsd.c index 4dcc2919ba..91ae68357e 100644 --- a/tools/libs/call/netbsd.c +++ b/tools/libs/call/netbsd.c @@ -96,7 +96,7 @@ void osdep_free_pages(xencall_handle *xcall, void *ptr, size_t npages) free(ptr); } -int osdep_hypercall(xencall_handle *xcall, privcmd_hypercall_t *hypercall) +long osdep_hypercall(xencall_handle *xcall, privcmd_hypercall_t *hypercall) { int fd = xcall->fd; int error = ioctl(fd, IOCTL_PRIVCMD_HYPERCALL, hypercall); diff --git a/tools/libs/call/private.h b/tools/libs/call/private.h index 7944ac5baf..9c3aa432ef 100644 --- a/tools/libs/call/private.h +++ b/tools/libs/call/private.h @@ -55,7 +55,7 @@ struct xencall_handle { int osdep_xencall_open(xencall_handle *xcall); int osdep_xencall_close(xencall_handle *xcall); -int osdep_hypercall(xencall_handle *xcall, privcmd_hypercall_t *hypercall); +long osdep_hypercall(xencall_handle *xcall, privcmd_hypercall_t *hypercall); void *osdep_alloc_pages(xencall_handle *xcall, size_t nr_pages); void osdep_free_pages(xencall_handle *xcall, void *p, size_t nr_pages); diff --git a/tools/libs/call/solaris.c b/tools/libs/call/solaris.c index c63b6a329a..304262bd05 100644 --- a/tools/libs/call/solaris.c +++ b/tools/libs/call/solaris.c @@ -80,7 +80,7 @@ void osdep_free_hypercall_buffer(xencall_handle *xcall, void *ptr, free(ptr); } -int do_xen_hypercall(xencall_handle *xcall, privcmd_hypercall_t *hypercall) +long osdep_hypercall(xencall_handle *xcall, privcmd_hypercall_t *hypercall) { int fd = xcall->fd; return ioctl(fd, IOCTL_PRIVCMD_HYPERCALL, hypercall); -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Fri Jun 25 23:36:36 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Fri, 25 Jun 2021 23:36:36 +0000 Received: from list by lists.xenproject.org with outflank-mailman.147487.271949 (Exim 4.92) (envelope-from ) id 1lwvNE-0002sH-1P; Fri, 25 Jun 2021 23:36:36 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 147487.271949; Fri, 25 Jun 2021 23:36:36 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvND-0002s5-UC; Fri, 25 Jun 2021 23:36:35 +0000 Received: by outflank-mailman (input) for mailman id 147487; Fri, 25 Jun 2021 23:36:34 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvNC-0002rg-Ie for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:36:34 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvNC-0006UV-Hv for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:36:34 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwvNC-0006tF-Gv for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:36:34 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=HDGU1ZURb7jSxTX0V9hMDk0S/TNWJKynypoTcajJuqk=; b=u5pJI+UZqWeeiQicueslmrufiw 19qTQTNvCXGIDQyBy02fOTBALcL7n9CwycyKiXVp+iG9a67McrbzYEQQ155BznvYm7am1iTbeIzDm UKBPCXBvpe+d4rDN4/uFcoquc1lZuJfPsDt1VMpA3xOxmKSPD4pDYb3E4pIITca8yhto=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] libxencall: introduce variant of xencall2() returning long Message-Id: Date: Fri, 25 Jun 2021 23:36:34 +0000 commit bef64f2c0019f828824a67f918604fe40768b1de Author: Jan Beulich AuthorDate: Thu Jun 24 16:39:02 2021 +0200 Commit: Jan Beulich CommitDate: Thu Jun 24 16:39:02 2021 +0200 libxencall: introduce variant of xencall2() returning long Some hypercalls, memory-op in particular, can return values requiring more than 31 bits to represent. Hence the underlying layers need to make sure they won't truncate such values. Signed-off-by: Jan Beulich Acked-by: Ian Jackson Acked-by: Andrew Cooper --- tools/include/xencall.h | 4 ++++ tools/libs/call/core.c | 11 +++++++++++ tools/libs/call/libxencall.map | 5 +++++ 3 files changed, 20 insertions(+) diff --git a/tools/include/xencall.h b/tools/include/xencall.h index 2d0c42ad5e..fc95ed0fe5 100644 --- a/tools/include/xencall.h +++ b/tools/include/xencall.h @@ -113,6 +113,10 @@ int xencall5(xencall_handle *xcall, unsigned int op, uint64_t arg1, uint64_t arg2, uint64_t arg3, uint64_t arg4, uint64_t arg5); +/* Variant(s) of the above, as needed, returning "long" instead of "int". */ +long xencall2L(xencall_handle *xcall, unsigned int op, + uint64_t arg1, uint64_t arg2); + /* * Allocate and free memory which is suitable for use as a pointer * argument to a hypercall. diff --git a/tools/libs/call/core.c b/tools/libs/call/core.c index 57d3a33e6b..02c4f8e1ae 100644 --- a/tools/libs/call/core.c +++ b/tools/libs/call/core.c @@ -127,6 +127,17 @@ int xencall2(xencall_handle *xcall, unsigned int op, return osdep_hypercall(xcall, &call); } +long xencall2L(xencall_handle *xcall, unsigned int op, + uint64_t arg1, uint64_t arg2) +{ + privcmd_hypercall_t call = { + .op = op, + .arg = { arg1, arg2 }, + }; + + return osdep_hypercall(xcall, &call); +} + int xencall3(xencall_handle *xcall, unsigned int op, uint64_t arg1, uint64_t arg2, uint64_t arg3) { diff --git a/tools/libs/call/libxencall.map b/tools/libs/call/libxencall.map index 6922b96511..ad7518b0e0 100644 --- a/tools/libs/call/libxencall.map +++ b/tools/libs/call/libxencall.map @@ -27,3 +27,8 @@ VERS_1.2 { global: xencall_fd; } VERS_1.1; + +VERS_1.3 { + global: + xencall2L; +} VERS_1.2; -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Fri Jun 25 23:36:46 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Fri, 25 Jun 2021 23:36:46 +0000 Received: from list by lists.xenproject.org with outflank-mailman.147488.271953 (Exim 4.92) (envelope-from ) id 1lwvNO-0002ve-2n; Fri, 25 Jun 2021 23:36:46 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 147488.271953; Fri, 25 Jun 2021 23:36:46 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvNN-0002vS-Vj; Fri, 25 Jun 2021 23:36:45 +0000 Received: by outflank-mailman (input) for mailman id 147488; Fri, 25 Jun 2021 23:36:44 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvNM-0002vG-Ld for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:36:44 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvNM-0006Ug-Ky for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:36:44 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwvNM-0006uY-KD for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:36:44 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=1/zwGNzt5PiFLlkkS95G8N7rz0Hos6gTaZRzzAUZd5c=; b=yt9PiXImpPpU3RSG4YirHPaDTq FszFOkGuEaQZmvrg7E/NnrOeMW2TafJyKG3uBgcg2JAgyI5+wmR1tPlEvgttcYFHE46dzn9jqDa2C iU8nIJajzQsmp/c8LTEak32YTym9/XwwxuKGhze+YlWhr/WU8zyh9kJLEj/tLrmxjNL4=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] libxc: use multicall for memory-op on Linux (and Solaris) Message-Id: Date: Fri, 25 Jun 2021 23:36:44 +0000 commit 6f02d1ea4a109a32f346a5d1de63a42b291c354c Author: Jan Beulich AuthorDate: Thu Jun 24 16:39:26 2021 +0200 Commit: Jan Beulich CommitDate: Thu Jun 24 16:39:26 2021 +0200 libxc: use multicall for memory-op on Linux (and Solaris) Some sub-functions, XENMEM_maximum_gpfn and XENMEM_maximum_ram_page in particular, can return values requiring more than 31 bits to represent. Hence we cannot issue the hypercall directly when the return value of ioctl() is used to propagate this value. This is the case for Linux and Solaris (and hence needs changing), while the BSDs avoid using the return value for dual purposes altogether, and MiniOS already wraps all hypercalls in a multicall. Suggested-by: Jürgen Groß Signed-off-by: Jan Beulich Acked-by: Ian Jackson Acked-by: Andrew Cooper --- tools/libs/ctrl/xc_private.c | 43 +++++++++++++++++++++++++++++++++++++++++-- 1 file changed, 41 insertions(+), 2 deletions(-) diff --git a/tools/libs/ctrl/xc_private.c b/tools/libs/ctrl/xc_private.c index d94f846686..c0422662f0 100644 --- a/tools/libs/ctrl/xc_private.c +++ b/tools/libs/ctrl/xc_private.c @@ -337,8 +337,47 @@ long do_memory_op(xc_interface *xch, int cmd, void *arg, size_t len) goto out1; } - ret = xencall2(xch->xcall, __HYPERVISOR_memory_op, - cmd, HYPERCALL_BUFFER_AS_ARG(arg)); +#if defined(__linux__) || defined(__sun__) + /* + * Some sub-ops return values which don't fit in "int". On platforms + * without a specific hypercall return value field in the privcmd + * interface structure, issue the request as a single-element multicall, + * to be able to capture the full return value. + */ + if ( sizeof(long) > sizeof(int) ) + { + multicall_entry_t multicall = { + .op = __HYPERVISOR_memory_op, + .args[0] = cmd, + .args[1] = HYPERCALL_BUFFER_AS_ARG(arg), + }, *call = &multicall; + DECLARE_HYPERCALL_BOUNCE(call, sizeof(*call), + XC_HYPERCALL_BUFFER_BOUNCE_BOTH); + + if ( xc_hypercall_bounce_pre(xch, call) ) + { + PERROR("Could not bounce buffer for memory_op hypercall"); + goto out1; + } + + ret = do_multicall_op(xch, HYPERCALL_BUFFER(call), 1); + + xc_hypercall_bounce_post(xch, call); + + if ( !ret ) + { + ret = multicall.result; + if ( multicall.result > ~0xfffUL ) + { + errno = -ret; + ret = -1; + } + } + } + else +#endif + ret = xencall2L(xch->xcall, __HYPERVISOR_memory_op, + cmd, HYPERCALL_BUFFER_AS_ARG(arg)); xc_hypercall_bounce_post(xch, arg); out1: -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Fri Jun 25 23:36:56 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Fri, 25 Jun 2021 23:36:56 +0000 Received: from list by lists.xenproject.org with outflank-mailman.147489.271956 (Exim 4.92) (envelope-from ) id 1lwvNY-0002yW-3t; Fri, 25 Jun 2021 23:36:56 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 147489.271956; Fri, 25 Jun 2021 23:36:56 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvNY-0002yO-0u; Fri, 25 Jun 2021 23:36:56 +0000 Received: by outflank-mailman (input) for mailman id 147489; Fri, 25 Jun 2021 23:36:54 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvNW-0002y5-OX for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:36:54 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvNW-0006Us-Nr for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:36:54 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwvNW-0006vk-ND for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:36:54 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=rjzarlXHzPCeEiuJBbZ0gIcDRexHiqyvuxt5nS89yZ0=; b=JrC4g/C5seU0Pl5Kf9G6xU4sK4 y/Jn3WggIlFS14yVtTdHiqvC+GgIS/TyduGAYGRUF0+l+bnECyS7ApgagKwKGouWd5QZ/8dZAek1d rRsMQKSfK1a1r6BS0vu8+uRLZ/mziPrR/QKGKR2GxZuQkpU4ImGAJGspQ+6r/BBYh3C8=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] libxencall: drop bogus mentioning of xencall6() Message-Id: Date: Fri, 25 Jun 2021 23:36:54 +0000 commit 7ffbed8681a0ab203535ab1e13c48f18ecdba907 Author: Jan Beulich AuthorDate: Thu Jun 24 16:39:55 2021 +0200 Commit: Jan Beulich CommitDate: Thu Jun 24 16:39:55 2021 +0200 libxencall: drop bogus mentioning of xencall6() There's no xencall6(), so the version script also shouldn't mention it. If such a function would ever appear, it shouldn't land in version 1.0. No change to the generated binary, nor abi-dumper's view of the object. Signed-off-by: Jan Beulich Acked-by: Ian Jackson Reviewed-by: Andrew Cooper --- tools/libs/call/libxencall.map | 1 - 1 file changed, 1 deletion(-) diff --git a/tools/libs/call/libxencall.map b/tools/libs/call/libxencall.map index ad7518b0e0..d18a3174e9 100644 --- a/tools/libs/call/libxencall.map +++ b/tools/libs/call/libxencall.map @@ -9,7 +9,6 @@ VERS_1.0 { xencall3; xencall4; xencall5; - xencall6; xencall_alloc_buffer; xencall_free_buffer; -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Fri Jun 25 23:37:06 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Fri, 25 Jun 2021 23:37:06 +0000 Received: from list by lists.xenproject.org with outflank-mailman.147490.271960 (Exim 4.92) (envelope-from ) id 1lwvNi-00031m-5G; Fri, 25 Jun 2021 23:37:06 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 147490.271960; Fri, 25 Jun 2021 23:37:06 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvNi-00031e-2I; Fri, 25 Jun 2021 23:37:06 +0000 Received: by outflank-mailman (input) for mailman id 147490; Fri, 25 Jun 2021 23:37:04 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvNg-00031P-Rh for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:37:04 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvNg-0006VP-Qy for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:37:04 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwvNg-0006x7-QI for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:37:04 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=swNg/dtdf3qROH6L1rnSOx7BDscp1ruAQaqbPX4SoqE=; b=a/muTmdD5YTiPr9rvwyxI0JjfS iahr/4on450kiFYSgoOIcRWmS8QVy4+Z3T1K9DHxxAT/tZY/GbZjecj9OPPo4wLWA1qujREgETK6j pD3lTHYqDXeQR6knMBw9Z98oDW4tqwAX3kHYO04J47/vGnkWMYpWBdwDm3GK5imuN2Kk=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] libxc: make xc_domain_maximum_gpfn() endianness-agnostic Message-Id: Date: Fri, 25 Jun 2021 23:37:04 +0000 commit 190facdf877c8fc298dafa584f69e7aff8606c35 Author: Jan Beulich AuthorDate: Thu Jun 24 16:40:57 2021 +0200 Commit: Jan Beulich CommitDate: Thu Jun 24 16:40:57 2021 +0200 libxc: make xc_domain_maximum_gpfn() endianness-agnostic libxc generally uses uint32_t to represent domain IDs. This is fine as long as addresses of such variables aren't taken, to then pass into hypercalls: To the hypervisor, a domain ID is a 16-bit value. Introduce a wrapper struct to deal with the issue. (On architectures with arguments passed in registers, an intermediate variable would have been created by the compiler already anyway, just one of the wrong type.) The public interface change is both source and binary compatible for the architectures we currently support. Signed-off-by: Jan Beulich Acked-by: Ian Jackson Reviewed-by: Andrew Cooper --- tools/libs/ctrl/xc_domain.c | 3 ++- xen/common/memory.c | 9 ++++++--- xen/include/public/memory.h | 11 +++++++++-- 3 files changed, 17 insertions(+), 6 deletions(-) diff --git a/tools/libs/ctrl/xc_domain.c b/tools/libs/ctrl/xc_domain.c index 7d118848f1..3803871538 100644 --- a/tools/libs/ctrl/xc_domain.c +++ b/tools/libs/ctrl/xc_domain.c @@ -856,7 +856,8 @@ int xc_domain_get_tsc_info(xc_interface *xch, int xc_domain_maximum_gpfn(xc_interface *xch, uint32_t domid, xen_pfn_t *gpfns) { - long rc = do_memory_op(xch, XENMEM_maximum_gpfn, &domid, sizeof(domid)); + struct xen_memory_domain dom = { .domid = domid }; + long rc = do_memory_op(xch, XENMEM_maximum_gpfn, &dom, sizeof(dom)); if ( rc >= 0 ) { diff --git a/xen/common/memory.c b/xen/common/memory.c index 72a6b70cb5..ba212f971c 100644 --- a/xen/common/memory.c +++ b/xen/common/memory.c @@ -1351,7 +1351,6 @@ long do_memory_op(unsigned long cmd, XEN_GUEST_HANDLE_PARAM(void) arg) long rc; struct xen_memory_reservation reservation; struct memop_args args; - domid_t domid; unsigned long start_extent = cmd >> MEMOP_EXTENT_SHIFT; int op = cmd & MEMOP_CMD_MASK; @@ -1452,13 +1451,16 @@ long do_memory_op(unsigned long cmd, XEN_GUEST_HANDLE_PARAM(void) arg) case XENMEM_current_reservation: case XENMEM_maximum_reservation: case XENMEM_maximum_gpfn: + { + struct xen_memory_domain domain; + if ( unlikely(start_extent) ) return -EINVAL; - if ( copy_from_guest(&domid, arg, 1) ) + if ( copy_from_guest(&domain, arg, 1) ) return -EFAULT; - d = rcu_lock_domain_by_any_id(domid); + d = rcu_lock_domain_by_any_id(domain.domid); if ( d == NULL ) return -ESRCH; @@ -1486,6 +1488,7 @@ long do_memory_op(unsigned long cmd, XEN_GUEST_HANDLE_PARAM(void) arg) rcu_unlock_domain(d); break; + } case XENMEM_add_to_physmap: { diff --git a/xen/include/public/memory.h b/xen/include/public/memory.h index 50e73eef98..383a9468c3 100644 --- a/xen/include/public/memory.h +++ b/xen/include/public/memory.h @@ -148,16 +148,23 @@ DEFINE_XEN_GUEST_HANDLE(xen_memory_exchange_t); */ #define XENMEM_maximum_ram_page 2 +struct xen_memory_domain { + /* [IN] Domain information is being queried for. */ + domid_t domid; +}; + /* * Returns the current or maximum memory reservation, in pages, of the * specified domain (may be DOMID_SELF). Returns -ve errcode on failure. - * arg == addr of domid_t. + * arg == addr of struct xen_memory_domain. */ #define XENMEM_current_reservation 3 #define XENMEM_maximum_reservation 4 /* - * Returns the maximum GPFN in use by the guest, or -ve errcode on failure. + * Returns the maximum GFN in use by the specified domain (may be DOMID_SELF). + * Returns -ve errcode on failure. + * arg == addr of struct xen_memory_domain. */ #define XENMEM_maximum_gpfn 14 -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Fri Jun 25 23:37:16 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Fri, 25 Jun 2021 23:37:16 +0000 Received: from list by lists.xenproject.org with outflank-mailman.147491.271965 (Exim 4.92) (envelope-from ) id 1lwvNs-00034t-6y; Fri, 25 Jun 2021 23:37:16 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 147491.271965; Fri, 25 Jun 2021 23:37:16 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvNs-00034l-3n; Fri, 25 Jun 2021 23:37:16 +0000 Received: by outflank-mailman (input) for mailman id 147491; Fri, 25 Jun 2021 23:37:15 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvNq-00034a-Uu for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:37:14 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lwvNq-0006Va-UD for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:37:14 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lwvNq-00073Z-TO for xen-changelog@lists.xenproject.org; Fri, 25 Jun 2021 23:37:14 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=LLfL/fsBLHZNQwiA+EeA+rIHstFrBVz0iJQgoChyD2I=; b=TBiJrF54a+vUtbI8qtrOYPVphl qydmHSmZXN/1jytBR3uALnufPmhPv2wVAuHsBj1KZwyobAtvaIJeQbRLW2peXcHrsVzpEOjkhGDLx IUzSOP7SIHmCsHCUonMKJ3H5w6UJ8hAVEmuWYlqWiVHlkCoD9FxikUCcc7nwQZgANotA=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] tools/xenstored: Remove redundant check in socket_can_process() Message-Id: Date: Fri, 25 Jun 2021 23:37:14 +0000 commit e87d8f60fa9b6eaa6a2357545a96e4fff05dbef0 Author: Julien Grall AuthorDate: Thu Jun 24 15:55:03 2021 +0100 Commit: Julien Grall CommitDate: Thu Jun 24 15:55:49 2021 +0100 tools/xenstored: Remove redundant check in socket_can_process() Commit 3adfb50315d9 ("tools/xenstored: Introduce a wrapper for conn->funcs->can_{read, write}") consolidated the check !conn->is_ignored in two new wrappers. This means the check in socket_can_process() is now redundant. In fact it should have been removed in orignal commit (as it was done for the domain helpers). Reported-by: Raphael Ning Reviewed-by: Juergen Gross --- tools/xenstore/xenstored_core.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/tools/xenstore/xenstored_core.c b/tools/xenstore/xenstored_core.c index 9ffd2ac66d..cf7297a96c 100644 --- a/tools/xenstore/xenstored_core.c +++ b/tools/xenstore/xenstored_core.c @@ -1752,7 +1752,7 @@ static bool socket_can_process(struct connection *conn, int mask) return false; } - return (fds[conn->pollfd_idx].revents & mask) && !conn->is_ignored; + return (fds[conn->pollfd_idx].revents & mask); } static bool socket_can_write(struct connection *conn) -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Sat Jun 26 14:00:10 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Sat, 26 Jun 2021 14:00:10 +0000 Received: from list by lists.xenproject.org with outflank-mailman.147563.272149 (Exim 4.92) (envelope-from ) id 1lx8qr-0004YT-0r; Sat, 26 Jun 2021 14:00:05 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 147563.272149; Sat, 26 Jun 2021 14:00:04 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lx8qq-0004Y1-TS; Sat, 26 Jun 2021 14:00:04 +0000 Received: by outflank-mailman (input) for mailman id 147563; Sat, 26 Jun 2021 14:00:02 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lx8qo-0004Mt-QC for xen-changelog@lists.xenproject.org; Sat, 26 Jun 2021 14:00:02 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lx8qo-0006YV-PI for xen-changelog@lists.xenproject.org; Sat, 26 Jun 2021 14:00:02 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lx8qo-0002oA-NU for xen-changelog@lists.xenproject.org; Sat, 26 Jun 2021 14:00:02 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=xxsiOVHip4RSTYiDGvZHvBEALo7UC5P1yZn+8TL1dkE=; b=HOdCT3KndPSCl0392xiP998S3t XyjTLuOJyF2VSUKLSlnnOy1St6pBCfQso/Js6hf0zMKW5WXtp75AHYQSYzFR4lZNlpDoabH9ur/hw r7QmA2KNglBMNeq5+hVf/oWxcIIgwSVMWjOza8pYk8kYmBePPx8iQE3FV5dIlCfVJnJE=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] tools/xenstored: Correctly read the requests header from the stream Message-Id: Date: Sat, 26 Jun 2021 14:00:02 +0000 commit 8a9b94982b0e0928ad874907f5a5b005944ab7cf Author: Julien Grall AuthorDate: Fri Jun 25 07:45:22 2021 +0100 Commit: Julien Grall CommitDate: Fri Jun 25 10:40:06 2021 +0100 tools/xenstored: Correctly read the requests header from the stream Commit c0fe360f42 ("tools/xenstored: Extend restore code to handle multiple input buffer") extend the read_buffered_state() to support multiple input buffers. Unfortunately, the commit didn't go far enough and still used sc->data (start of the buffers) for retrieving the header. This would lead to read the wrong headers for second and follow-up commands. Use data in place for sc->data for the source of the memcpy()s. Fixes: c0fe360f42 ("tools/xenstored: Extend restore code to handle multiple input buffer") Reported-by: Raphael Ning Signed-off-by: Julien Grall Reviewed-by: Juergen Gross --- tools/xenstore/xenstored_core.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/tools/xenstore/xenstored_core.c b/tools/xenstore/xenstored_core.c index cf7297a96c..16c856730c 100644 --- a/tools/xenstore/xenstored_core.c +++ b/tools/xenstore/xenstored_core.c @@ -2717,11 +2717,11 @@ void read_state_buffered_data(const void *ctx, struct connection *conn, len = sc->data_in_len - (data - sc->data); if (len < sizeof(bdata->hdr)) { bdata->inhdr = true; - memcpy(&bdata->hdr, sc->data, len); + memcpy(&bdata->hdr, data, len); bdata->used = len; } else { bdata->inhdr = false; - memcpy(&bdata->hdr, sc->data, sizeof(bdata->hdr)); + memcpy(&bdata->hdr, data, sizeof(bdata->hdr)); if (bdata->hdr.msg.len <= DEFAULT_BUFFER_SIZE) bdata->buffer = bdata->default_buffer; else -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Sat Jun 26 14:00:15 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Sat, 26 Jun 2021 14:00:15 +0000 Received: from list by lists.xenproject.org with outflank-mailman.147565.272153 (Exim 4.92) (envelope-from ) id 1lx8r1-0004ul-2X; Sat, 26 Jun 2021 14:00:15 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 147565.272153; Sat, 26 Jun 2021 14:00:15 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lx8r0-0004uc-Vh; Sat, 26 Jun 2021 14:00:14 +0000 Received: by outflank-mailman (input) for mailman id 147565; Sat, 26 Jun 2021 14:00:12 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lx8qy-0004uB-TH for xen-changelog@lists.xenproject.org; Sat, 26 Jun 2021 14:00:12 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lx8qy-0006YZ-SQ for xen-changelog@lists.xenproject.org; Sat, 26 Jun 2021 14:00:12 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lx8qy-0002pe-RT for xen-changelog@lists.xenproject.org; Sat, 26 Jun 2021 14:00:12 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=eDDDGW+WDhIBnpG8tYQrdw07CjmJyZKwKENBC+fpCBM=; b=GZcThwBFFn39PO6r9Xcp/vssai k7WmRqM7uC1Shv21GX8rmRNlOKVk2E3yut2QocG9wjesnSG7xZ3Buknu3RqkAm5XRF9LEYYWj8OZW DNNCCSUzrpY/mUbdCLs4IunfAoyBLke3YhQNLtuHl7u0AvlOTAgVc+HVuvO7ZzUxl5vQ=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] libxencall: Bump SONAME following new functionality Message-Id: Date: Sat, 26 Jun 2021 14:00:12 +0000 commit 01a2d001dea2219c9702afbe2d6fd1b0af539203 Author: Andrew Cooper AuthorDate: Thu Jun 24 18:49:14 2021 +0100 Commit: Andrew Cooper CommitDate: Fri Jun 25 12:38:18 2021 +0100 libxencall: Bump SONAME following new functionality Fixes: bef64f2c00 ("libxencall: introduce variant of xencall2() returning long") Signed-off-by: Andrew Cooper Reviewed-by: Ian Jackson --- tools/libs/call/Makefile | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/tools/libs/call/Makefile b/tools/libs/call/Makefile index 4ed201b3b3..93d404b79e 100644 --- a/tools/libs/call/Makefile +++ b/tools/libs/call/Makefile @@ -2,7 +2,7 @@ XEN_ROOT = $(CURDIR)/../../.. include $(XEN_ROOT)/tools/Rules.mk MAJOR = 1 -MINOR = 2 +MINOR = 3 SRCS-y += core.c buffer.c SRCS-$(CONFIG_Linux) += linux.c -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Sat Jun 26 14:00:25 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Sat, 26 Jun 2021 14:00:25 +0000 Received: from list by lists.xenproject.org with outflank-mailman.147566.272157 (Exim 4.92) (envelope-from ) id 1lx8rB-0004xw-4F; Sat, 26 Jun 2021 14:00:25 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 147566.272157; Sat, 26 Jun 2021 14:00:25 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lx8rB-0004xp-0z; Sat, 26 Jun 2021 14:00:25 +0000 Received: by outflank-mailman (input) for mailman id 147566; Sat, 26 Jun 2021 14:00:23 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lx8r9-0004xN-0O for xen-changelog@lists.xenproject.org; Sat, 26 Jun 2021 14:00:23 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lx8r8-0006Yt-Vk for xen-changelog@lists.xenproject.org; Sat, 26 Jun 2021 14:00:22 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lx8r8-0002qm-Uj for xen-changelog@lists.xenproject.org; Sat, 26 Jun 2021 14:00:22 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=VaIqPsZkYz0C35FKJm8rGiTV4o2EbuCYQqzffS5H3m0=; b=wHqMvS15zjgup8Xig33b5svjve 6iTGtBqeIu2DmuaSCK8xJpPrNwqRBMvnrhjbSSViKQEiXtqaYrSpogUpBAc2EjCRTXQnKdeakYWoD 0hDFYqkXmV8NYxz/B07RTVXJZNy6sTI9aOl/Jjyc87Ukjyisq21ZSqsh6w2hZMDkK6Sw=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] IOMMU/PCI: don't let domain cleanup continue when device de-assignment failed Message-Id: Date: Sat, 26 Jun 2021 14:00:22 +0000 commit f591755823a7e94fc6b4b8ddce71f0421a94fa09 Author: Jan Beulich AuthorDate: Fri Jun 25 14:06:55 2021 +0200 Commit: Jan Beulich CommitDate: Fri Jun 25 14:06:55 2021 +0200 IOMMU/PCI: don't let domain cleanup continue when device de-assignment failed Failure here could in principle mean the device may still be issuing DMA requests, which would continue to be translated by the page tables the device entry currently points at. With this we cannot allow the subsequent cleanup step of freeing the page tables to occur, to prevent use-after-free issues. We would need to accept, for the time being, that in such a case the remaining domain resources will all be leaked, and the domain will continue to exist as a zombie. However, with flushes no longer timing out (and with proper timeout detection for device I/O TLB flushing yet to be implemented), there's no way anymore for failures to occur, except due to bugs elsewhere. Hence the change here is merely a "just in case" one. In order to continue the loop in spite of an error, we can't use pci_get_pdev_by_domain() anymore. I have no idea why it was used here in the first place, instead of the cheaper list iteration. Signed-off-by: Jan Beulich Reviewed-by: Paul Durrant --- xen/drivers/passthrough/pci.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/xen/drivers/passthrough/pci.c b/xen/drivers/passthrough/pci.c index 199ce08612..fc4fa2e5c3 100644 --- a/xen/drivers/passthrough/pci.c +++ b/xen/drivers/passthrough/pci.c @@ -894,7 +894,7 @@ static int deassign_device(struct domain *d, uint16_t seg, uint8_t bus, int pci_release_devices(struct domain *d) { - struct pci_dev *pdev; + struct pci_dev *pdev, *tmp; u8 bus, devfn; int ret; @@ -905,15 +905,15 @@ int pci_release_devices(struct domain *d) pcidevs_unlock(); return ret; } - while ( (pdev = pci_get_pdev_by_domain(d, -1, -1, -1)) ) + list_for_each_entry_safe ( pdev, tmp, &d->pdev_list, domain_list ) { bus = pdev->bus; devfn = pdev->devfn; - deassign_device(d, pdev->seg, bus, devfn); + ret = deassign_device(d, pdev->seg, bus, devfn) ?: ret; } pcidevs_unlock(); - return 0; + return ret; } #define PCI_CLASS_BRIDGE_HOST 0x0600 -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Sun Jun 27 01:33:10 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Sun, 27 Jun 2021 01:33:10 +0000 Received: from list by lists.xenproject.org with outflank-mailman.147627.272307 (Exim 4.92) (envelope-from ) id 1lxJfU-00036l-4L; Sun, 27 Jun 2021 01:33:04 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 147627.272307; Sun, 27 Jun 2021 01:33:04 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lxJfT-00036W-Vb; Sun, 27 Jun 2021 01:33:03 +0000 Received: by outflank-mailman (input) for mailman id 147627; Sun, 27 Jun 2021 01:33:02 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lxJfS-00036Q-PT for xen-changelog@lists.xenproject.org; Sun, 27 Jun 2021 01:33:02 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lxJfS-0002ug-M6 for xen-changelog@lists.xenproject.org; Sun, 27 Jun 2021 01:33:02 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lxJfS-0006MK-L8 for xen-changelog@lists.xenproject.org; Sun, 27 Jun 2021 01:33:02 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=1oSU93fNwFd6ymXd33GKgerlUxHSjvVDgJ1jVNuV7dA=; b=fPv17eNJJxT5sgcOgkoyI3PHsT E62P2ryR8UZdGN0nOubexmJ4nmOrXnIsl3MEHoRMMu1IjgJL7YKEwhHfFWImze8VSZDae/khDd2IF B8T0vDKEuAgD4gxqV6bgix06a8h4WMeAF4rAIePyz9vYlXTV2A4r7z5/lXgbgLeym9bI=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] xen/arm: smmuv1: Fixed stream matching register allocation Message-Id: Date: Sun, 27 Jun 2021 01:33:02 +0000 commit 93fe1817e7ca8c80f93c098a191e1cc3483190df Author: Rahul Singh AuthorDate: Fri Jun 25 17:37:26 2021 +0100 Commit: Stefano Stabellini CommitDate: Fri Jun 25 14:53:48 2021 -0700 xen/arm: smmuv1: Fixed stream matching register allocation SMR allocation should be based on the number of supported stream matching register for each SMMU device. Issue introduced by commit 5e08586afbb90b2e2d56c175c07db77a4afa873c when backported the patches from Linux to XEN to fix the stream match conflict issue when two devices have the same stream-id. Acked-by: Stefano Stabellini Tested-by: Stefano Stabellini Signed-off-by: Rahul Singh --- xen/drivers/passthrough/arm/smmu.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/xen/drivers/passthrough/arm/smmu.c b/xen/drivers/passthrough/arm/smmu.c index 1a68c2ab3b..05c2d238b3 100644 --- a/xen/drivers/passthrough/arm/smmu.c +++ b/xen/drivers/passthrough/arm/smmu.c @@ -149,6 +149,7 @@ typedef enum irqreturn irqreturn_t; #define kzalloc(size, flags) _xzalloc(size, sizeof(void *)) #define devm_kzalloc(dev, size, flags) _xzalloc(size, sizeof(void *)) #define kmalloc_array(size, n, flags) _xmalloc_array(size, sizeof(void *), n) +#define kzalloc_array(size, n, flags) _xzalloc_array(size, sizeof(void *), n) static void __iomem *devm_ioremap_resource(struct device *dev, struct resource *res) @@ -2221,7 +2222,7 @@ static int arm_smmu_device_cfg_probe(struct arm_smmu_device *smmu) smmu->smr_mask_mask = smr >> SMR_MASK_SHIFT; /* Zero-initialised to mark as invalid */ - smmu->smrs = devm_kzalloc(smmu->dev, sizeof(*smmu->smrs), GFP_KERNEL); + smmu->smrs = kzalloc_array(sizeof(*smmu->smrs), size, GFP_KERNEL); if (!smmu->smrs) return -ENOMEM; -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Sun Jun 27 01:33:14 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Sun, 27 Jun 2021 01:33:14 +0000 Received: from list by lists.xenproject.org with outflank-mailman.147628.272311 (Exim 4.92) (envelope-from ) id 1lxJfe-00038X-5V; Sun, 27 Jun 2021 01:33:14 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 147628.272311; Sun, 27 Jun 2021 01:33:14 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lxJfe-00038N-10; Sun, 27 Jun 2021 01:33:14 +0000 Received: by outflank-mailman (input) for mailman id 147628; Sun, 27 Jun 2021 01:33:12 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lxJfc-00038F-QE for xen-changelog@lists.xenproject.org; Sun, 27 Jun 2021 01:33:12 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lxJfc-0002ul-PN for xen-changelog@lists.xenproject.org; Sun, 27 Jun 2021 01:33:12 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lxJfc-0006NH-OI for xen-changelog@lists.xenproject.org; Sun, 27 Jun 2021 01:33:12 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=WVfVZ9N7VSLnO/buzX3Ex/0qDbhlIP2l/aHeveDNcYw=; b=kT158KFLDwuDYldl+DjsfMnVV2 r3HwnA7tzQrwlrpd7hcndU/RXWYiPPAtOwqbB9/uOAE5IwCo0X1lCkis3n4JzXn+bon44TEpNByD4 5rPi7R3gLxutasQkDtsJgZxrCyjcn5P3sAwGSqTNjTSgYLuZn9dl49IyN5TT8Q/kAFBU=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] xen/arm: smmuv1: Set privileged attr to 'default' Message-Id: Date: Sun, 27 Jun 2021 01:33:12 +0000 commit bb11edcec1a953ce590da797b0d005cd60f21e83 Author: Rahul Singh AuthorDate: Fri Jun 25 17:37:27 2021 +0100 Commit: Stefano Stabellini CommitDate: Fri Jun 25 14:54:17 2021 -0700 xen/arm: smmuv1: Set privileged attr to 'default' Backport commit e19898077cfb642fe151ba22981e795c74d9e114 "iommu/arm-smmu: Set privileged attribute to 'default' instead of 'unprivileged'" Original commit message: Currently the driver sets all the device transactions privileges to UNPRIVILEGED, but there are cases where the iommu masters wants to isolate privileged supervisor and unprivileged user. So don't override the privileged setting to unprivileged, instead set it to default as incoming and let it be controlled by the pagetable settings. Acked-by: Will Deacon Signed-off-by: Sricharan R Signed-off-by: Will Deacon Signed-off-by: Rahul Singh Acked-by: Stefano Stabellini Tested-by: Stefano Stabellini --- xen/drivers/passthrough/arm/smmu.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/xen/drivers/passthrough/arm/smmu.c b/xen/drivers/passthrough/arm/smmu.c index 05c2d238b3..da2cd457d7 100644 --- a/xen/drivers/passthrough/arm/smmu.c +++ b/xen/drivers/passthrough/arm/smmu.c @@ -1567,7 +1567,7 @@ static int arm_smmu_domain_add_master(struct arm_smmu_domain *smmu_domain, continue; s2cr[idx].type = type ; - s2cr[idx].privcfg = S2CR_PRIVCFG_UNPRIV; + s2cr[idx].privcfg = S2CR_PRIVCFG_DEFAULT; s2cr[idx].cbndx = cbndx; arm_smmu_write_s2cr(smmu, idx); } -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Mon Jun 28 10:22:12 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Mon, 28 Jun 2021 10:22:12 +0000 Received: from list by lists.xenproject.org with outflank-mailman.147778.272738 (Exim 4.92) (envelope-from ) id 1lxoP0-0000as-Oo; Mon, 28 Jun 2021 10:22:06 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 147778.272738; Mon, 28 Jun 2021 10:22:06 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lxoP0-0000ak-LX; Mon, 28 Jun 2021 10:22:06 +0000 Received: by outflank-mailman (input) for mailman id 147778; Mon, 28 Jun 2021 10:22:05 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lxoOz-0000ae-86 for xen-changelog@lists.xenproject.org; Mon, 28 Jun 2021 10:22:05 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lxoOz-0006yP-6N for xen-changelog@lists.xenproject.org; Mon, 28 Jun 2021 10:22:05 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lxoOz-0000Z2-5D for xen-changelog@lists.xenproject.org; Mon, 28 Jun 2021 10:22:05 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=BF4qHpXZ0jqRZyG8rOrw+e6I1BioqA2FTntYP4dFcSI=; b=bYpguPbN9FsH9AF4l20b7AHGtM U9QIyv8rN0vhbd/vCTjMH1jQXy9b779CiNpWoO5jP6PEaHTZmoUdThq55ycdjbdSlXCQFIamyN0JP oaJmbdeQGR6gb7QbtaldwPaNTvRrhG8CeT41gT+WYgOonulZPlE+7KLzUfjIoXm8KJc8=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] Replace FSF street address with canonical URL (again) Message-Id: Date: Mon, 28 Jun 2021 10:22:05 +0000 commit c636a5fe59575d84778f676ca1728fbd1a7c7104 Author: Andrew Cooper AuthorDate: Fri Jun 25 14:35:02 2021 +0100 Commit: Andrew Cooper CommitDate: Mon Jun 28 11:17:14 2021 +0100 Replace FSF street address with canonical URL (again) As recommended in http://www.gnu.org/licenses/gpl-howto.en.html. Exactly as per changeset 443701ef0c7ff3 - Some errors have crept back in in the past 6 years. Signed-off-by: Andrew Cooper Acked-by: Julien Grall --- tools/libs/guest/xg_dom_hvmloader.c | 3 +-- xen/arch/arm/acpi/boot.c | 3 +-- xen/common/argo.c | 3 +-- xen/include/asm-arm/acpi.h | 3 +-- xen/include/xen/argo.h | 3 +-- xen/include/xen/rbtree.h | 3 +-- xen/lib/rbtree.c | 3 +-- 7 files changed, 7 insertions(+), 14 deletions(-) diff --git a/tools/libs/guest/xg_dom_hvmloader.c b/tools/libs/guest/xg_dom_hvmloader.c index ae50d98011..39e1e5f579 100644 --- a/tools/libs/guest/xg_dom_hvmloader.c +++ b/tools/libs/guest/xg_dom_hvmloader.c @@ -14,8 +14,7 @@ * Lesser General Public License for more details. * * You should have received a copy of the GNU Lesser General Public - * License along with this library; if not, write to the Free Software - * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA + * License along with this library; If not, see . * */ diff --git a/xen/arch/arm/acpi/boot.c b/xen/arch/arm/acpi/boot.c index 7ea2990cb8..db5085e15d 100644 --- a/xen/arch/arm/acpi/boot.c +++ b/xen/arch/arm/acpi/boot.c @@ -19,8 +19,7 @@ * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License - * along with this program; if not, write to the Free Software - * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + * along with this program; If not, see . * * ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ */ diff --git a/xen/common/argo.c b/xen/common/argo.c index 49be715f63..eaea7ba888 100644 --- a/xen/common/argo.c +++ b/xen/common/argo.c @@ -12,8 +12,7 @@ * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License - * along with this program; if not, write to the Free Software - * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + * along with this program; If not, see . */ #include diff --git a/xen/include/asm-arm/acpi.h b/xen/include/asm-arm/acpi.h index b52ae2d6ef..e53973e054 100644 --- a/xen/include/asm-arm/acpi.h +++ b/xen/include/asm-arm/acpi.h @@ -14,8 +14,7 @@ * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License - * along with this program; if not, write to the Free Software - * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + * along with this program; If not, see . * * ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ */ diff --git a/xen/include/xen/argo.h b/xen/include/xen/argo.h index 2ba7e5c0c0..fd4cfdd55c 100644 --- a/xen/include/xen/argo.h +++ b/xen/include/xen/argo.h @@ -9,8 +9,7 @@ * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License - * along with this program; if not, write to the Free Software - * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + * along with this program; If not, see . */ #ifndef __XEN_ARGO_H__ diff --git a/xen/include/xen/rbtree.h b/xen/include/xen/rbtree.h index 1b72590e4e..77bf57d4ab 100644 --- a/xen/include/xen/rbtree.h +++ b/xen/include/xen/rbtree.h @@ -13,8 +13,7 @@ GNU General Public License for more details. You should have received a copy of the GNU General Public License - along with this program; if not, write to the Free Software - Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + along with this program; If not, see . linux/include/linux/rbtree.h diff --git a/xen/lib/rbtree.c b/xen/lib/rbtree.c index 95e045d524..85a4f20313 100644 --- a/xen/lib/rbtree.c +++ b/xen/lib/rbtree.c @@ -15,8 +15,7 @@ GNU General Public License for more details. You should have received a copy of the GNU General Public License - along with this program; if not, write to the Free Software - Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + along with this program; If not, see . linux/lib/rbtree.c */ -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Tue Jun 29 02:22:09 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 29 Jun 2021 02:22:09 +0000 Received: from list by lists.xenproject.org with outflank-mailman.147964.273210 (Exim 4.92) (envelope-from ) id 1ly3O0-0004yz-1u; Tue, 29 Jun 2021 02:22:04 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 147964.273210; Tue, 29 Jun 2021 02:22:04 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1ly3Nz-0004yq-Uk; Tue, 29 Jun 2021 02:22:03 +0000 Received: by outflank-mailman (input) for mailman id 147964; Tue, 29 Jun 2021 02:22:02 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1ly3Ny-0004yk-8F for xen-changelog@lists.xenproject.org; Tue, 29 Jun 2021 02:22:02 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1ly3Ny-0000Be-7Q for xen-changelog@lists.xenproject.org; Tue, 29 Jun 2021 02:22:02 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1ly3Ny-00034R-5t for xen-changelog@lists.xenproject.org; Tue, 29 Jun 2021 02:22:02 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=J/IvZVZev+RCCUhKJBN6+W/bqIq5rC2cLdHBcJ9guMQ=; b=qRR4MPztLC/s+g1mWxw+O3brG0 Zx2kgxUKpRgMqGXJZH/aT46y3WO94cZv30Km/3FllvYxE2BSUU/gohLgeda0X1NGE3my/i44M/9MW xg94GQFRrIt7CizpClRo5nmaudl1RVYZJsK1pIWtE2NXe7LSbLaTPLMKFUK/EKeoQXdk=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] Replace FSF street address with canonical URL (again) Message-Id: Date: Tue, 29 Jun 2021 02:22:02 +0000 commit c636a5fe59575d84778f676ca1728fbd1a7c7104 Author: Andrew Cooper AuthorDate: Fri Jun 25 14:35:02 2021 +0100 Commit: Andrew Cooper CommitDate: Mon Jun 28 11:17:14 2021 +0100 Replace FSF street address with canonical URL (again) As recommended in http://www.gnu.org/licenses/gpl-howto.en.html. Exactly as per changeset 443701ef0c7ff3 - Some errors have crept back in in the past 6 years. Signed-off-by: Andrew Cooper Acked-by: Julien Grall --- tools/libs/guest/xg_dom_hvmloader.c | 3 +-- xen/arch/arm/acpi/boot.c | 3 +-- xen/common/argo.c | 3 +-- xen/include/asm-arm/acpi.h | 3 +-- xen/include/xen/argo.h | 3 +-- xen/include/xen/rbtree.h | 3 +-- xen/lib/rbtree.c | 3 +-- 7 files changed, 7 insertions(+), 14 deletions(-) diff --git a/tools/libs/guest/xg_dom_hvmloader.c b/tools/libs/guest/xg_dom_hvmloader.c index ae50d98011..39e1e5f579 100644 --- a/tools/libs/guest/xg_dom_hvmloader.c +++ b/tools/libs/guest/xg_dom_hvmloader.c @@ -14,8 +14,7 @@ * Lesser General Public License for more details. * * You should have received a copy of the GNU Lesser General Public - * License along with this library; if not, write to the Free Software - * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA + * License along with this library; If not, see . * */ diff --git a/xen/arch/arm/acpi/boot.c b/xen/arch/arm/acpi/boot.c index 7ea2990cb8..db5085e15d 100644 --- a/xen/arch/arm/acpi/boot.c +++ b/xen/arch/arm/acpi/boot.c @@ -19,8 +19,7 @@ * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License - * along with this program; if not, write to the Free Software - * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + * along with this program; If not, see . * * ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ */ diff --git a/xen/common/argo.c b/xen/common/argo.c index 49be715f63..eaea7ba888 100644 --- a/xen/common/argo.c +++ b/xen/common/argo.c @@ -12,8 +12,7 @@ * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License - * along with this program; if not, write to the Free Software - * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + * along with this program; If not, see . */ #include diff --git a/xen/include/asm-arm/acpi.h b/xen/include/asm-arm/acpi.h index b52ae2d6ef..e53973e054 100644 --- a/xen/include/asm-arm/acpi.h +++ b/xen/include/asm-arm/acpi.h @@ -14,8 +14,7 @@ * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License - * along with this program; if not, write to the Free Software - * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + * along with this program; If not, see . * * ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ */ diff --git a/xen/include/xen/argo.h b/xen/include/xen/argo.h index 2ba7e5c0c0..fd4cfdd55c 100644 --- a/xen/include/xen/argo.h +++ b/xen/include/xen/argo.h @@ -9,8 +9,7 @@ * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License - * along with this program; if not, write to the Free Software - * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + * along with this program; If not, see . */ #ifndef __XEN_ARGO_H__ diff --git a/xen/include/xen/rbtree.h b/xen/include/xen/rbtree.h index 1b72590e4e..77bf57d4ab 100644 --- a/xen/include/xen/rbtree.h +++ b/xen/include/xen/rbtree.h @@ -13,8 +13,7 @@ GNU General Public License for more details. You should have received a copy of the GNU General Public License - along with this program; if not, write to the Free Software - Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + along with this program; If not, see . linux/include/linux/rbtree.h diff --git a/xen/lib/rbtree.c b/xen/lib/rbtree.c index 95e045d524..85a4f20313 100644 --- a/xen/lib/rbtree.c +++ b/xen/lib/rbtree.c @@ -15,8 +15,7 @@ GNU General Public License for more details. You should have received a copy of the GNU General Public License - along with this program; if not, write to the Free Software - Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + along with this program; If not, see . linux/lib/rbtree.c */ -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Tue Jun 29 09:11:12 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 29 Jun 2021 09:11:12 +0000 Received: from list by lists.xenproject.org with outflank-mailman.147989.273292 (Exim 4.92) (envelope-from ) id 1ly9lq-0004Ao-Ao; Tue, 29 Jun 2021 09:11:06 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 147989.273292; Tue, 29 Jun 2021 09:11:06 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1ly9lq-0004Ae-7T; Tue, 29 Jun 2021 09:11:06 +0000 Received: by outflank-mailman (input) for mailman id 147989; Tue, 29 Jun 2021 09:11:04 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1ly9lo-0004AW-N9 for xen-changelog@lists.xenproject.org; Tue, 29 Jun 2021 09:11:04 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1ly9lo-00087d-Kq for xen-changelog@lists.xenproject.org; Tue, 29 Jun 2021 09:11:04 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1ly9lo-0002Yu-JK for xen-changelog@lists.xenproject.org; Tue, 29 Jun 2021 09:11:04 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=TaBFTw12igy8dCMtPDqgB64oWWlYR43mkYC7jxm5DAs=; b=6QKHN9V/QNlZwdG2tFGZ5WVdoQ DvDm/E7DJj+5gcLaBGEjrj+lQKCA+eDKgzVQTmLaqzvSEIybUW6fIN20pXaPbP8weJru5uKqVLcI1 BhVYoq2YPs8S3Ra0t5i1XydMTFakPZMjzFkmH9/YJ8ynMpUCXjoGhb7e5Rgl7v8/WO+A=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] fully replace mfn_to_gmfn() Message-Id: Date: Tue, 29 Jun 2021 09:11:04 +0000 commit f24ec0ed220a43e19059c3fe4f803be458930a1d Author: Jan Beulich AuthorDate: Tue Jun 29 11:00:51 2021 +0200 Commit: Jan Beulich CommitDate: Tue Jun 29 11:00:51 2021 +0200 fully replace mfn_to_gmfn() Convert the two remaining uses as well as Arm's stub to the properly named and type-safe mfn_to_gfn(), dropping x86's definition (where we already have mfn_to_gfn()). Signed-off-by: Jan Beulich Acked-by: Julien Grall Acked-by: Andrew Cooper --- xen/common/domctl.c | 3 ++- xen/common/memory.c | 8 ++++---- xen/include/asm-arm/mm.h | 3 +-- xen/include/asm-x86/mm.h | 5 ----- 4 files changed, 7 insertions(+), 12 deletions(-) diff --git a/xen/common/domctl.c b/xen/common/domctl.c index ef202c2b8c..12d6144d28 100644 --- a/xen/common/domctl.c +++ b/xen/common/domctl.c @@ -111,7 +111,8 @@ void getdomaininfo(struct domain *d, struct xen_domctl_getdomaininfo *info) info->outstanding_pages = d->outstanding_pages; info->shr_pages = atomic_read(&d->shr_pages); info->paged_pages = atomic_read(&d->paged_pages); - info->shared_info_frame = mfn_to_gmfn(d, virt_to_mfn(d->shared_info)); + info->shared_info_frame = + gfn_x(mfn_to_gfn(d, _mfn(virt_to_mfn(d->shared_info)))); BUG_ON(SHARED_M2P(info->shared_info_frame)); info->cpupool = cpupool_get_id(d); diff --git a/xen/common/memory.c b/xen/common/memory.c index ba212f971c..e07bd9a5ea 100644 --- a/xen/common/memory.c +++ b/xen/common/memory.c @@ -714,13 +714,13 @@ static long memory_exchange(XEN_GUEST_HANDLE_PARAM(xen_memory_exchange_t) arg) */ while ( (page = page_list_remove_head(&in_chunk_list)) ) { - unsigned long gfn; + gfn_t gfn; mfn = page_to_mfn(page); - gfn = mfn_to_gmfn(d, mfn_x(mfn)); + gfn = mfn_to_gfn(d, mfn); /* Pages were unshared above */ - BUG_ON(SHARED_M2P(gfn)); - if ( guest_physmap_remove_page(d, _gfn(gfn), mfn, 0) ) + BUG_ON(SHARED_M2P(gfn_x(gfn))); + if ( guest_physmap_remove_page(d, gfn, mfn, 0) ) domain_crash(d); free_domheap_page(page); } diff --git a/xen/include/asm-arm/mm.h b/xen/include/asm-arm/mm.h index 0b7de3102e..ded74d29da 100644 --- a/xen/include/asm-arm/mm.h +++ b/xen/include/asm-arm/mm.h @@ -328,8 +328,7 @@ struct page_info *get_page_from_gva(struct vcpu *v, vaddr_t va, /* Xen always owns P2M on ARM */ #define set_gpfn_from_mfn(mfn, pfn) do { (void) (mfn), (void)(pfn); } while (0) -#define mfn_to_gmfn(_d, mfn) (mfn) - +#define mfn_to_gfn(d, mfn) ((void)(d), _gfn(mfn_x(mfn))) /* Arch-specific portion of memory_op hypercall. */ long arch_memory_op(int op, XEN_GUEST_HANDLE_PARAM(void) arg); diff --git a/xen/include/asm-x86/mm.h b/xen/include/asm-x86/mm.h index 80b0ccffaa..cb90527499 100644 --- a/xen/include/asm-x86/mm.h +++ b/xen/include/asm-x86/mm.h @@ -527,11 +527,6 @@ extern struct rangeset *mmio_ro_ranges; #define get_gpfn_from_mfn(mfn) (machine_to_phys_mapping[(mfn)]) -#define mfn_to_gmfn(_d, mfn) \ - ( (paging_mode_translate(_d)) \ - ? get_gpfn_from_mfn(mfn) \ - : (mfn) ) - #define compat_pfn_to_cr3(pfn) (((unsigned)(pfn) << 12) | ((unsigned)(pfn) >> 20)) #define compat_cr3_to_pfn(cr3) (((unsigned)(cr3) >> 12) | ((unsigned)(cr3) << 20)) -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Tue Jun 29 09:11:16 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 29 Jun 2021 09:11:16 +0000 Received: from list by lists.xenproject.org with outflank-mailman.147990.273296 (Exim 4.92) (envelope-from ) id 1ly9m0-0004D9-Bu; Tue, 29 Jun 2021 09:11:16 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 147990.273296; Tue, 29 Jun 2021 09:11:16 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1ly9m0-0004D1-91; Tue, 29 Jun 2021 09:11:16 +0000 Received: by outflank-mailman (input) for mailman id 147990; Tue, 29 Jun 2021 09:11:14 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1ly9ly-0004Co-Ox for xen-changelog@lists.xenproject.org; Tue, 29 Jun 2021 09:11:14 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1ly9ly-00087h-O5 for xen-changelog@lists.xenproject.org; Tue, 29 Jun 2021 09:11:14 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1ly9ly-0002aD-N3 for xen-changelog@lists.xenproject.org; Tue, 29 Jun 2021 09:11:14 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=sZ/ITv/Hhh1yclxCuSMZfdFhhzidPvBRUOEXzgVVJtU=; b=IqSjt7CdxPQUXXN6tUJOXN2o1q elZVFSh3WP6+3cCWZWyAULwTx4i2ojNH1a55qvbH2piTxB+nM/2vdn/0Yj33lpyF3G2G94POrRSyP o5aMAABEyAVitftUx02ck56o5T7cm/yDoNe9pn9loVW47wA8cwo8xbOUwyl9LUsTc3Jg=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] x86/paging: deal with log-dirty stats overflow Message-Id: Date: Tue, 29 Jun 2021 09:11:14 +0000 commit 17e91570c5a442d0500b05c157966ac212a1f12f Author: Jan Beulich AuthorDate: Tue Jun 29 11:02:35 2021 +0200 Commit: Jan Beulich CommitDate: Tue Jun 29 11:02:35 2021 +0200 x86/paging: deal with log-dirty stats overflow While the precise values are unlikely of interest once they exceed 4 billion (allowing us to leave alone the domctl struct), we still shouldn't wrap or truncate the actual values. It is in particular problematic if the truncated values were zero (causing libxenguest to skip an iteration altogether) or a very small value (leading to premature exiting of the pre-copy phase). Change the internal fields to unsigned long, and suitably saturate for copying to guest context. Signed-off-by: Jan Beulich Acked-by: Andrew Cooper --- xen/arch/x86/mm/paging.c | 8 +++++--- xen/include/asm-x86/domain.h | 4 ++-- 2 files changed, 7 insertions(+), 5 deletions(-) diff --git a/xen/arch/x86/mm/paging.c b/xen/arch/x86/mm/paging.c index c304c24526..dd6b2bdf6f 100644 --- a/xen/arch/x86/mm/paging.c +++ b/xen/arch/x86/mm/paging.c @@ -446,14 +446,16 @@ static int paging_log_dirty_op(struct domain *d, clean = (sc->op == XEN_DOMCTL_SHADOW_OP_CLEAN); - PAGING_DEBUG(LOGDIRTY, "log-dirty %s: dom %u faults=%u dirty=%u\n", + PAGING_DEBUG(LOGDIRTY, "log-dirty %s: dom %u faults=%lu dirty=%lu\n", (clean) ? "clean" : "peek", d->domain_id, d->arch.paging.log_dirty.fault_count, d->arch.paging.log_dirty.dirty_count); - sc->stats.fault_count = d->arch.paging.log_dirty.fault_count; - sc->stats.dirty_count = d->arch.paging.log_dirty.dirty_count; + sc->stats.fault_count = min(d->arch.paging.log_dirty.fault_count, + UINT32_MAX + 0UL); + sc->stats.dirty_count = min(d->arch.paging.log_dirty.dirty_count, + UINT32_MAX + 0UL); if ( guest_handle_is_null(sc->dirty_bitmap) ) /* caller may have wanted just to clean the state or access stats. */ diff --git a/xen/include/asm-x86/domain.h b/xen/include/asm-x86/domain.h index 7213d184b0..92d54de0b9 100644 --- a/xen/include/asm-x86/domain.h +++ b/xen/include/asm-x86/domain.h @@ -190,8 +190,8 @@ struct log_dirty_domain { unsigned int failed_allocs; /* log-dirty mode stats */ - unsigned int fault_count; - unsigned int dirty_count; + unsigned long fault_count; + unsigned long dirty_count; /* functions which are paging mode specific */ const struct log_dirty_ops { -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Tue Jun 29 09:11:26 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 29 Jun 2021 09:11:26 +0000 Received: from list by lists.xenproject.org with outflank-mailman.147991.273300 (Exim 4.92) (envelope-from ) id 1ly9mA-0004Gx-Dc; Tue, 29 Jun 2021 09:11:26 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 147991.273300; Tue, 29 Jun 2021 09:11:26 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1ly9mA-0004Gp-AW; Tue, 29 Jun 2021 09:11:26 +0000 Received: by outflank-mailman (input) for mailman id 147991; Tue, 29 Jun 2021 09:11:24 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1ly9m8-0004GW-S2 for xen-changelog@lists.xenproject.org; Tue, 29 Jun 2021 09:11:24 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1ly9m8-00088B-RE for xen-changelog@lists.xenproject.org; Tue, 29 Jun 2021 09:11:24 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1ly9m8-0002b3-QL for xen-changelog@lists.xenproject.org; Tue, 29 Jun 2021 09:11:24 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=SCSkn+ACJ4qWAUJvooF1QbK389mGbbMMLZSQAvhj+pE=; b=Y7XnniZitFGxdxKt8zXw4bVxE2 eCTuXkNKcPoPJOePk9vvSEmV00amrK5tGQMUlDU9IIGGJHXT7HL0F0/oSEnOPR5PPT5TYMsMxRYww yVS32jB2MSdALgGlHkdRhbcHPvO5P7U4pJO3vFz9U60RpEZefVRawqokZ8j0BR0dmBYU=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] x86/mm: pull a sanity check earlier in xenmem_add_to_physmap_one() Message-Id: Date: Tue, 29 Jun 2021 09:11:24 +0000 commit f8582da0417660269bec69e399f8667f761e886b Author: Jan Beulich AuthorDate: Tue Jun 29 11:03:29 2021 +0200 Commit: Jan Beulich CommitDate: Tue Jun 29 11:03:29 2021 +0200 x86/mm: pull a sanity check earlier in xenmem_add_to_physmap_one() We should try to limit the failure reasons after we've started making changes. Signed-off-by: Jan Beulich Acked-by: Andrew Cooper --- xen/arch/x86/mm/p2m.c | 16 +++++++++------- 1 file changed, 9 insertions(+), 7 deletions(-) diff --git a/xen/arch/x86/mm/p2m.c b/xen/arch/x86/mm/p2m.c index dbb1cbeb59..7b3cf7e9fc 100644 --- a/xen/arch/x86/mm/p2m.c +++ b/xen/arch/x86/mm/p2m.c @@ -2756,6 +2756,15 @@ int xenmem_add_to_physmap_one( goto put_both; } + /* XENMAPSPACE_gmfn: Check if the MFN is associated with another GFN. */ + old_gpfn = get_gpfn_from_mfn(mfn_x(mfn)); + ASSERT(!SHARED_M2P(old_gpfn)); + if ( space == XENMAPSPACE_gmfn && old_gpfn != gfn ) + { + rc = -EXDEV; + goto put_both; + } + /* Remove previously mapped page if it was present. */ prev_mfn = get_gfn(d, gfn_x(gpfn), &p2mt); if ( mfn_valid(prev_mfn) ) @@ -2774,13 +2783,6 @@ int xenmem_add_to_physmap_one( goto put_both; /* Unmap from old location, if any. */ - old_gpfn = get_gpfn_from_mfn(mfn_x(mfn)); - ASSERT(!SHARED_M2P(old_gpfn)); - if ( space == XENMAPSPACE_gmfn && old_gpfn != gfn ) - { - rc = -EXDEV; - goto put_both; - } if ( old_gpfn != INVALID_M2P_ENTRY ) rc = guest_physmap_remove_page(d, _gfn(old_gpfn), mfn, PAGE_ORDER_4K); -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Tue Jun 29 10:44:12 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 29 Jun 2021 10:44:12 +0000 Received: from list by lists.xenproject.org with outflank-mailman.147999.273315 (Exim 4.92) (envelope-from ) id 1lyBDq-0004iE-CQ; Tue, 29 Jun 2021 10:44:06 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 147999.273315; Tue, 29 Jun 2021 10:44:06 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lyBDq-0004i6-9a; Tue, 29 Jun 2021 10:44:06 +0000 Received: by outflank-mailman (input) for mailman id 147999; Tue, 29 Jun 2021 10:44:05 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lyBDp-0004hx-JN for xen-changelog@lists.xenproject.org; Tue, 29 Jun 2021 10:44:05 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lyBDp-0001Hy-Hu for xen-changelog@lists.xenproject.org; Tue, 29 Jun 2021 10:44:05 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lyBDp-0001lI-Gx for xen-changelog@lists.xenproject.org; Tue, 29 Jun 2021 10:44:05 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=dDiOWqg3CAw1GJa1JfqQtrRG6H61w2j2SGmrwK8Xq6k=; b=xkgXonbu7bKr/4UCK+3Kk6N3+a WnfWsyRePuJcYW1zOzHcKNYdFqH/zJPpDyuU9AUjQycydYKB85Ibldd0lw9XiXIpdZjW+FUR9U/9s 2ZkgEyBZ4Hb5bhPJvCwa37kWMWiREj51gFV+QjyFYbBjmGhScQUQwRlzwMo9RO5u5FCE=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] x86emul: avoid using _PRE_EFLAGS() in a few cases Message-Id: Date: Tue, 29 Jun 2021 10:44:05 +0000 commit f004fd7cbbba9f3d58b399d5becfe4e8c5230fb6 Author: Jan Beulich AuthorDate: Tue Jun 29 12:33:37 2021 +0200 Commit: Jan Beulich CommitDate: Tue Jun 29 12:33:37 2021 +0200 x86emul: avoid using _PRE_EFLAGS() in a few cases The macro expanding to quite a few insns, replace its use by simply clearing the status flags when the to be executed insn doesn't depend on their initial state, in cases where this is easily possible. (There are more cases where the uses are hidden inside macros, and where some of the users of the macros want guest flags put in place before running the insn, i.e. the macros can't be updated as easily.) Signed-off-by: Jan Beulich Reviewed-by: Andrew Cooper --- xen/arch/x86/x86_emulate/x86_emulate.c | 9 ++++++--- 1 file changed, 6 insertions(+), 3 deletions(-) diff --git a/xen/arch/x86/x86_emulate/x86_emulate.c b/xen/arch/x86/x86_emulate/x86_emulate.c index 45828b726b..c9a9e031cf 100644 --- a/xen/arch/x86/x86_emulate/x86_emulate.c +++ b/xen/arch/x86/x86_emulate/x86_emulate.c @@ -6864,7 +6864,8 @@ x86_emulate( } opc[2] = 0xc3; - invoke_stub(_PRE_EFLAGS("[eflags]", "[mask]", "[tmp]"), + _regs.eflags &= ~EFLAGS_MASK; + invoke_stub("", _POST_EFLAGS("[eflags]", "[mask]", "[tmp]"), [eflags] "+g" (_regs.eflags), [tmp] "=&r" (dummy), "+m" (*mmvalp) @@ -8112,7 +8113,8 @@ x86_emulate( opc[2] = 0xc3; copy_VEX(opc, vex); - invoke_stub(_PRE_EFLAGS("[eflags]", "[mask]", "[tmp]"), + _regs.eflags &= ~EFLAGS_MASK; + invoke_stub("", _POST_EFLAGS("[eflags]", "[mask]", "[tmp]"), [eflags] "+g" (_regs.eflags), "=a" (dst.val), [tmp] "=&r" (dummy) @@ -11699,13 +11701,14 @@ int x86_emul_rmw( break; case rmw_xadd: + *eflags &= ~EFLAGS_MASK; switch ( state->op_bytes ) { unsigned long dummy; #define XADD(sz, cst, mod) \ case sz: \ - asm ( _PRE_EFLAGS("[efl]", "[msk]", "[tmp]") \ + asm ( "" \ COND_LOCK(xadd) " %"#mod"[reg], %[mem]; " \ _POST_EFLAGS("[efl]", "[msk]", "[tmp]") \ : [reg] "+" #cst (state->ea.val), \ -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Tue Jun 29 10:44:16 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 29 Jun 2021 10:44:16 +0000 Received: from list by lists.xenproject.org with outflank-mailman.148000.273319 (Exim 4.92) (envelope-from ) id 1lyBE0-0004jx-E8; Tue, 29 Jun 2021 10:44:16 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 148000.273319; Tue, 29 Jun 2021 10:44:16 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lyBE0-0004jp-B8; Tue, 29 Jun 2021 10:44:16 +0000 Received: by outflank-mailman (input) for mailman id 148000; Tue, 29 Jun 2021 10:44:15 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lyBDz-0004jj-Lz for xen-changelog@lists.xenproject.org; Tue, 29 Jun 2021 10:44:15 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lyBDz-0001I3-L5 for xen-changelog@lists.xenproject.org; Tue, 29 Jun 2021 10:44:15 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lyBDz-0001lx-K2 for xen-changelog@lists.xenproject.org; Tue, 29 Jun 2021 10:44:15 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=loYYDztazJ54xx+yZhy60BzmlRLFjUVcKCiTyQgb/pQ=; b=ry5WYIkaPbjM1tX1JAtJfzRB3k bBn7GA3cnU728ZcQd8TVNVRVorBDHQJkmtbzN4D6Oa7UjUXy/p2Xlf0Q24KmACZELaYBoGVi4VEnU rdJD8cyRN4bWYohT3IeG/1i7SiAhFAOLJi5Xo+oFuizn6ATQc988OrGD2u749BO8a4Cg=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] AMD/IOMMU: redo awaiting of command completion Message-Id: Date: Tue, 29 Jun 2021 10:44:15 +0000 commit b175fc0c1274df34ff8bd89d603c88b014b22f10 Author: Jan Beulich AuthorDate: Tue Jun 29 12:34:37 2021 +0200 Commit: Jan Beulich CommitDate: Tue Jun 29 12:34:37 2021 +0200 AMD/IOMMU: redo awaiting of command completion The present abuse of the completion interrupt does not only stand in the way of, down the road, using it for its actual purpose, but also requires holding the IOMMU lock while waiting for command completion, limiting parallelism and keeping interrupts off for non-negligible periods of time. Have the IOMMU do an ordinary memory write instead of signaling an otherwise disabled interrupt (by just updating a status register bit). Since IOMMU_COMP_WAIT_I_FLAG_SHIFT is now unused and IOMMU_COMP_WAIT_[FS]_FLAG_SHIFT already were, drop all three of them while at it. Signed-off-by: Jan Beulich Reviewed-by: Paul Durrant --- xen/drivers/passthrough/amd/iommu-defs.h | 3 --- xen/drivers/passthrough/amd/iommu_cmd.c | 30 ++++++++++++++++-------------- 2 files changed, 16 insertions(+), 17 deletions(-) diff --git a/xen/drivers/passthrough/amd/iommu-defs.h b/xen/drivers/passthrough/amd/iommu-defs.h index 0c97db2a32..774234dfd2 100644 --- a/xen/drivers/passthrough/amd/iommu-defs.h +++ b/xen/drivers/passthrough/amd/iommu-defs.h @@ -178,11 +178,8 @@ struct amd_iommu_dte { #define IOMMU_COMP_WAIT_DATA_BUFFER_SIZE 8 #define IOMMU_COMP_WAIT_DATA_BUFFER_ALIGNMENT 8 #define IOMMU_COMP_WAIT_S_FLAG_MASK 0x00000001 -#define IOMMU_COMP_WAIT_S_FLAG_SHIFT 0 #define IOMMU_COMP_WAIT_I_FLAG_MASK 0x00000002 -#define IOMMU_COMP_WAIT_I_FLAG_SHIFT 1 #define IOMMU_COMP_WAIT_F_FLAG_MASK 0x00000004 -#define IOMMU_COMP_WAIT_F_FLAG_SHIFT 2 #define IOMMU_COMP_WAIT_ADDR_LOW_MASK 0xFFFFFFF8 #define IOMMU_COMP_WAIT_ADDR_LOW_SHIFT 3 #define IOMMU_COMP_WAIT_ADDR_HIGH_MASK 0x000FFFFF diff --git a/xen/drivers/passthrough/amd/iommu_cmd.c b/xen/drivers/passthrough/amd/iommu_cmd.c index 92a688dc9d..e8541276ca 100644 --- a/xen/drivers/passthrough/amd/iommu_cmd.c +++ b/xen/drivers/passthrough/amd/iommu_cmd.c @@ -20,6 +20,9 @@ #include "iommu.h" #include "../ats.h" +#define CMD_COMPLETION_INIT 0 +#define CMD_COMPLETION_DONE 1 + static void send_iommu_command(struct amd_iommu *iommu, const uint32_t cmd[4]) { @@ -49,28 +52,27 @@ static void send_iommu_command(struct amd_iommu *iommu, static void flush_command_buffer(struct amd_iommu *iommu, unsigned int timeout_base) { - uint32_t cmd[4]; + static DEFINE_PER_CPU(uint64_t, poll_slot); + uint64_t *this_poll_slot = &this_cpu(poll_slot); + paddr_t addr = virt_to_maddr(this_poll_slot); + /* send a COMPLETION_WAIT command to flush command buffer */ + uint32_t cmd[4] = { + addr | MASK_INSR(IOMMU_CONTROL_ENABLED, + IOMMU_COMP_WAIT_S_FLAG_MASK), + (addr >> 32) | MASK_INSR(IOMMU_CMD_COMPLETION_WAIT, + IOMMU_CMD_OPCODE_MASK), + CMD_COMPLETION_DONE + }; s_time_t start, timeout; static unsigned int __read_mostly threshold = 1; - /* RW1C 'ComWaitInt' in status register */ - writel(IOMMU_STATUS_COMP_WAIT_INT, - iommu->mmio_base + IOMMU_STATUS_MMIO_OFFSET); + ACCESS_ONCE(*this_poll_slot) = CMD_COMPLETION_INIT; - /* send an empty COMPLETION_WAIT command to flush command buffer */ - cmd[3] = cmd[2] = 0; - set_field_in_reg_u32(IOMMU_CMD_COMPLETION_WAIT, 0, - IOMMU_CMD_OPCODE_MASK, - IOMMU_CMD_OPCODE_SHIFT, &cmd[1]); - set_field_in_reg_u32(IOMMU_CONTROL_ENABLED, 0, - IOMMU_COMP_WAIT_I_FLAG_MASK, - IOMMU_COMP_WAIT_I_FLAG_SHIFT, &cmd[0]); send_iommu_command(iommu, cmd); start = NOW(); timeout = start + (timeout_base ?: 100) * MILLISECS(threshold); - while ( !(readl(iommu->mmio_base + IOMMU_STATUS_MMIO_OFFSET) & - IOMMU_STATUS_COMP_WAIT_INT) ) + while ( ACCESS_ONCE(*this_poll_slot) != CMD_COMPLETION_DONE ) { if ( timeout && NOW() > timeout ) { -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Tue Jun 29 10:44:26 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Tue, 29 Jun 2021 10:44:26 +0000 Received: from list by lists.xenproject.org with outflank-mailman.148001.273323 (Exim 4.92) (envelope-from ) id 1lyBEA-0004n6-Gt; Tue, 29 Jun 2021 10:44:26 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 148001.273323; Tue, 29 Jun 2021 10:44:26 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lyBEA-0004my-Dz; Tue, 29 Jun 2021 10:44:26 +0000 Received: by outflank-mailman (input) for mailman id 148001; Tue, 29 Jun 2021 10:44:25 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lyBE9-0004ms-PI for xen-changelog@lists.xenproject.org; Tue, 29 Jun 2021 10:44:25 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lyBE9-0001Ia-OR for xen-changelog@lists.xenproject.org; Tue, 29 Jun 2021 10:44:25 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lyBE9-0001mp-NM for xen-changelog@lists.xenproject.org; Tue, 29 Jun 2021 10:44:25 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=+WAwd776p/YaFnwdxeHPMIlh5hAjTElMpD626kEgpz0=; b=k2fseNPt3t/JDf90XxUwWGrTUq S/mwz8J+EVeInoGwmltQcJ2rXezq7JoIfHEn9RNQar+rQyVirzmGReGfULLRHuh9bHn615MtrnzXQ ZlcIoAuUgT6/h7yy1+pJouNqmrPtULzn1LGdqyI8cHkRLRoWv4RbvueFPAcuUbV6JdGk=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen staging] AMD/IOMMU: re-work locking around sending of commands Message-Id: Date: Tue, 29 Jun 2021 10:44:25 +0000 commit f95b7b37cfc6d4613721df9357090d14712013c0 Author: Jan Beulich AuthorDate: Tue Jun 29 12:35:12 2021 +0200 Commit: Jan Beulich CommitDate: Tue Jun 29 12:35:12 2021 +0200 AMD/IOMMU: re-work locking around sending of commands It appears unhelpful to me for flush_command_buffer() to block all progress elsewhere for the given IOMMU by holding its lock while waiting for command completion. There's no real need for callers of that function or of send_iommu_command() to hold the lock. Contain all command sending related locking to the latter function. Signed-off-by: Jan Beulich Reviewed-by: Paul Durrant --- xen/drivers/passthrough/amd/iommu_cmd.c | 23 +++++------------------ xen/drivers/passthrough/amd/iommu_guest.c | 3 ++- xen/drivers/passthrough/amd/iommu_init.c | 15 +++++++-------- xen/drivers/passthrough/amd/iommu_intr.c | 6 ------ xen/drivers/passthrough/amd/pci_amd_iommu.c | 15 ++++++++++----- 5 files changed, 24 insertions(+), 38 deletions(-) diff --git a/xen/drivers/passthrough/amd/iommu_cmd.c b/xen/drivers/passthrough/amd/iommu_cmd.c index e8541276ca..4db8e1651f 100644 --- a/xen/drivers/passthrough/amd/iommu_cmd.c +++ b/xen/drivers/passthrough/amd/iommu_cmd.c @@ -27,6 +27,9 @@ static void send_iommu_command(struct amd_iommu *iommu, const uint32_t cmd[4]) { uint32_t tail; + unsigned long flags; + + spin_lock_irqsave(&iommu->lock, flags); tail = iommu->cmd_buffer.tail + sizeof(cmd_entry_t); if ( tail == iommu->cmd_buffer.size ) @@ -47,6 +50,8 @@ static void send_iommu_command(struct amd_iommu *iommu, iommu->cmd_buffer.tail = tail; writel(tail, iommu->mmio_base + IOMMU_CMD_BUFFER_TAIL_OFFSET); + + spin_unlock_irqrestore(&iommu->lock, flags); } static void flush_command_buffer(struct amd_iommu *iommu, @@ -273,7 +278,6 @@ static void invalidate_iommu_all(struct amd_iommu *iommu) void amd_iommu_flush_iotlb(u8 devfn, const struct pci_dev *pdev, daddr_t daddr, unsigned int order) { - unsigned long flags; struct amd_iommu *iommu; unsigned int req_id, queueid, maxpend; @@ -300,10 +304,8 @@ void amd_iommu_flush_iotlb(u8 devfn, const struct pci_dev *pdev, maxpend = pdev->ats.queue_depth & 0xff; /* send INVALIDATE_IOTLB_PAGES command */ - spin_lock_irqsave(&iommu->lock, flags); invalidate_iotlb_pages(iommu, maxpend, 0, queueid, daddr, req_id, order); flush_command_buffer(iommu, iommu_dev_iotlb_timeout); - spin_unlock_irqrestore(&iommu->lock, flags); } static void amd_iommu_flush_all_iotlbs(struct domain *d, daddr_t daddr, @@ -330,17 +332,14 @@ static void amd_iommu_flush_all_iotlbs(struct domain *d, daddr_t daddr, static void _amd_iommu_flush_pages(struct domain *d, daddr_t daddr, unsigned int order) { - unsigned long flags; struct amd_iommu *iommu; unsigned int dom_id = d->domain_id; /* send INVALIDATE_IOMMU_PAGES command */ for_each_amd_iommu ( iommu ) { - spin_lock_irqsave(&iommu->lock, flags); invalidate_iommu_pages(iommu, daddr, dom_id, order); flush_command_buffer(iommu, 0); - spin_unlock_irqrestore(&iommu->lock, flags); } if ( ats_enabled ) @@ -360,37 +359,25 @@ void amd_iommu_flush_pages(struct domain *d, void amd_iommu_flush_device(struct amd_iommu *iommu, uint16_t bdf) { - ASSERT( spin_is_locked(&iommu->lock) ); - invalidate_dev_table_entry(iommu, bdf); flush_command_buffer(iommu, 0); } void amd_iommu_flush_intremap(struct amd_iommu *iommu, uint16_t bdf) { - ASSERT( spin_is_locked(&iommu->lock) ); - invalidate_interrupt_table(iommu, bdf); flush_command_buffer(iommu, 0); } void amd_iommu_flush_all_caches(struct amd_iommu *iommu) { - ASSERT( spin_is_locked(&iommu->lock) ); - invalidate_iommu_all(iommu); flush_command_buffer(iommu, 0); } void amd_iommu_send_guest_cmd(struct amd_iommu *iommu, u32 cmd[]) { - unsigned long flags; - - spin_lock_irqsave(&iommu->lock, flags); - send_iommu_command(iommu, cmd); /* TBD: Timeout selection may require peeking into cmd[]. */ flush_command_buffer(iommu, 0); - - spin_unlock_irqrestore(&iommu->lock, flags); } diff --git a/xen/drivers/passthrough/amd/iommu_guest.c b/xen/drivers/passthrough/amd/iommu_guest.c index 00c5ccd7b5..85828490ff 100644 --- a/xen/drivers/passthrough/amd/iommu_guest.c +++ b/xen/drivers/passthrough/amd/iommu_guest.c @@ -449,9 +449,10 @@ static int do_invalidate_dte(struct domain *d, cmd_entry_t *cmd) spin_lock_irqsave(&iommu->lock, flags); dte_set_gcr3_table(mdte, hdom_id, gcr3_mfn << PAGE_SHIFT, gv, glx); - amd_iommu_flush_device(iommu, req_id); spin_unlock_irqrestore(&iommu->lock, flags); + amd_iommu_flush_device(iommu, req_id); + return 0; } diff --git a/xen/drivers/passthrough/amd/iommu_init.c b/xen/drivers/passthrough/amd/iommu_init.c index bb52c181f8..312f335c1d 100644 --- a/xen/drivers/passthrough/amd/iommu_init.c +++ b/xen/drivers/passthrough/amd/iommu_init.c @@ -871,7 +871,10 @@ static void enable_iommu(struct amd_iommu *iommu) spin_lock_irqsave(&iommu->lock, flags); if ( unlikely(iommu->enabled) ) - goto out; + { + spin_unlock_irqrestore(&iommu->lock, flags); + return; + } amd_iommu_erratum_746_workaround(iommu); @@ -921,13 +924,12 @@ static void enable_iommu(struct amd_iommu *iommu) set_iommu_translation_control(iommu, IOMMU_CONTROL_ENABLED); - if ( iommu->features.flds.ia_sup ) - amd_iommu_flush_all_caches(iommu); - iommu->enabled = 1; - out: spin_unlock_irqrestore(&iommu->lock, flags); + + if ( iommu->features.flds.ia_sup ) + amd_iommu_flush_all_caches(iommu); } static void disable_iommu(struct amd_iommu *iommu) @@ -1544,7 +1546,6 @@ static int _invalidate_all_devices( { unsigned int bdf; u16 req_id; - unsigned long flags; struct amd_iommu *iommu; for ( bdf = 0; bdf < ivrs_bdf_entries; bdf++ ) @@ -1553,10 +1554,8 @@ static int _invalidate_all_devices( req_id = ivrs_mappings[bdf].dte_requestor_id; if ( iommu ) { - spin_lock_irqsave(&iommu->lock, flags); amd_iommu_flush_device(iommu, req_id); amd_iommu_flush_intremap(iommu, req_id); - spin_unlock_irqrestore(&iommu->lock, flags); } } diff --git a/xen/drivers/passthrough/amd/iommu_intr.c b/xen/drivers/passthrough/amd/iommu_intr.c index d78bb209bc..bf00370574 100644 --- a/xen/drivers/passthrough/amd/iommu_intr.c +++ b/xen/drivers/passthrough/amd/iommu_intr.c @@ -310,9 +310,7 @@ static int update_intremap_entry_from_ioapic( entry.ptr32->flds.remap_en = false; spin_unlock(lock); - spin_lock(&iommu->lock); amd_iommu_flush_intremap(iommu, req_id); - spin_unlock(&iommu->lock); spin_lock(lock); } @@ -527,11 +525,9 @@ static int update_intremap_entry_from_msi_msg( if ( iommu->enabled ) { - spin_lock_irqsave(&iommu->lock, flags); amd_iommu_flush_intremap(iommu, req_id); if ( alias_id != req_id ) amd_iommu_flush_intremap(iommu, alias_id); - spin_unlock_irqrestore(&iommu->lock, flags); } return 0; @@ -567,11 +563,9 @@ static int update_intremap_entry_from_msi_msg( entry.ptr32->flds.remap_en = false; spin_unlock(lock); - spin_lock(&iommu->lock); amd_iommu_flush_intremap(iommu, req_id); if ( alias_id != req_id ) amd_iommu_flush_intremap(iommu, alias_id); - spin_unlock(&iommu->lock); spin_lock(lock); } diff --git a/xen/drivers/passthrough/amd/pci_amd_iommu.c b/xen/drivers/passthrough/amd/pci_amd_iommu.c index 085fe2f577..2dce505b97 100644 --- a/xen/drivers/passthrough/amd/pci_amd_iommu.c +++ b/xen/drivers/passthrough/amd/pci_amd_iommu.c @@ -129,6 +129,8 @@ static void amd_iommu_setup_domain_device( iommu_has_cap(iommu, PCI_CAP_IOTLB_SHIFT) ) dte->i = ats_enabled; + spin_unlock_irqrestore(&iommu->lock, flags); + amd_iommu_flush_device(iommu, req_id); AMD_IOMMU_DEBUG("Setup I/O page table: device id = %#x, type = %#x, " @@ -138,8 +140,8 @@ static void amd_iommu_setup_domain_device( page_to_maddr(hd->arch.amd.root_table), domain->domain_id, hd->arch.amd.paging_mode); } - - spin_unlock_irqrestore(&iommu->lock, flags); + else + spin_unlock_irqrestore(&iommu->lock, flags); ASSERT(pcidevs_locked()); @@ -307,6 +309,8 @@ static void amd_iommu_disable_domain_device(const struct domain *domain, smp_wmb(); dte->v = true; + spin_unlock_irqrestore(&iommu->lock, flags); + amd_iommu_flush_device(iommu, req_id); AMD_IOMMU_DEBUG("Disable: device id = %#x, " @@ -314,7 +318,8 @@ static void amd_iommu_disable_domain_device(const struct domain *domain, req_id, domain->domain_id, dom_iommu(domain)->arch.amd.paging_mode); } - spin_unlock_irqrestore(&iommu->lock, flags); + else + spin_unlock_irqrestore(&iommu->lock, flags); ASSERT(pcidevs_locked()); @@ -455,9 +460,9 @@ static int amd_iommu_add_device(u8 devfn, struct pci_dev *pdev) iommu->dev_table.buffer + (bdf * IOMMU_DEV_TABLE_ENTRY_SIZE), ivrs_mappings[bdf].intremap_table, iommu, iommu_intremap); - amd_iommu_flush_device(iommu, bdf); - spin_unlock_irqrestore(&iommu->lock, flags); + + amd_iommu_flush_device(iommu, bdf); } amd_iommu_setup_domain_device(pdev->domain, iommu, devfn, pdev); -- generated by git-patchbot for /home/xen/git/xen.git#staging From xen-changelog-bounces@lists.xenproject.org Wed Jun 30 02:44:15 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Wed, 30 Jun 2021 02:44:15 +0000 Received: from list by lists.xenproject.org with outflank-mailman.148113.273598 (Exim 4.92) (envelope-from ) id 1lyQCq-0001Ww-82; Wed, 30 Jun 2021 02:44:04 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 148113.273598; Wed, 30 Jun 2021 02:44:04 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lyQCq-0001Wn-4l; Wed, 30 Jun 2021 02:44:04 +0000 Received: by outflank-mailman (input) for mailman id 148113; Wed, 30 Jun 2021 02:44:02 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lyQCo-0001Wh-TS for xen-changelog@lists.xenproject.org; Wed, 30 Jun 2021 02:44:02 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lyQCo-0002kS-PD for xen-changelog@lists.xenproject.org; Wed, 30 Jun 2021 02:44:02 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lyQCo-00033H-No for xen-changelog@lists.xenproject.org; Wed, 30 Jun 2021 02:44:02 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=u37na/1vkrmbqnF47IMqRaAZJecfpWQxE0p6vcv/l0A=; b=fHOTwOU29OTebTWtUTJMR0wEoO LKbDyymFi5QEG/X28lfQvwW++/WDt60qIsqmCK8qqh+SM/TME7r+G3UMWyPZho9E9Epce52VmZams yL3zl4/QFiciPMCdgvLkA0MgcQ7DcN3UM8/cTU2UzRAL+Z3tz6l/8LvUETUoFDMUzxsE=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] fully replace mfn_to_gmfn() Message-Id: Date: Wed, 30 Jun 2021 02:44:02 +0000 commit f24ec0ed220a43e19059c3fe4f803be458930a1d Author: Jan Beulich AuthorDate: Tue Jun 29 11:00:51 2021 +0200 Commit: Jan Beulich CommitDate: Tue Jun 29 11:00:51 2021 +0200 fully replace mfn_to_gmfn() Convert the two remaining uses as well as Arm's stub to the properly named and type-safe mfn_to_gfn(), dropping x86's definition (where we already have mfn_to_gfn()). Signed-off-by: Jan Beulich Acked-by: Julien Grall Acked-by: Andrew Cooper --- xen/common/domctl.c | 3 ++- xen/common/memory.c | 8 ++++---- xen/include/asm-arm/mm.h | 3 +-- xen/include/asm-x86/mm.h | 5 ----- 4 files changed, 7 insertions(+), 12 deletions(-) diff --git a/xen/common/domctl.c b/xen/common/domctl.c index ef202c2b8c..12d6144d28 100644 --- a/xen/common/domctl.c +++ b/xen/common/domctl.c @@ -111,7 +111,8 @@ void getdomaininfo(struct domain *d, struct xen_domctl_getdomaininfo *info) info->outstanding_pages = d->outstanding_pages; info->shr_pages = atomic_read(&d->shr_pages); info->paged_pages = atomic_read(&d->paged_pages); - info->shared_info_frame = mfn_to_gmfn(d, virt_to_mfn(d->shared_info)); + info->shared_info_frame = + gfn_x(mfn_to_gfn(d, _mfn(virt_to_mfn(d->shared_info)))); BUG_ON(SHARED_M2P(info->shared_info_frame)); info->cpupool = cpupool_get_id(d); diff --git a/xen/common/memory.c b/xen/common/memory.c index ba212f971c..e07bd9a5ea 100644 --- a/xen/common/memory.c +++ b/xen/common/memory.c @@ -714,13 +714,13 @@ static long memory_exchange(XEN_GUEST_HANDLE_PARAM(xen_memory_exchange_t) arg) */ while ( (page = page_list_remove_head(&in_chunk_list)) ) { - unsigned long gfn; + gfn_t gfn; mfn = page_to_mfn(page); - gfn = mfn_to_gmfn(d, mfn_x(mfn)); + gfn = mfn_to_gfn(d, mfn); /* Pages were unshared above */ - BUG_ON(SHARED_M2P(gfn)); - if ( guest_physmap_remove_page(d, _gfn(gfn), mfn, 0) ) + BUG_ON(SHARED_M2P(gfn_x(gfn))); + if ( guest_physmap_remove_page(d, gfn, mfn, 0) ) domain_crash(d); free_domheap_page(page); } diff --git a/xen/include/asm-arm/mm.h b/xen/include/asm-arm/mm.h index 0b7de3102e..ded74d29da 100644 --- a/xen/include/asm-arm/mm.h +++ b/xen/include/asm-arm/mm.h @@ -328,8 +328,7 @@ struct page_info *get_page_from_gva(struct vcpu *v, vaddr_t va, /* Xen always owns P2M on ARM */ #define set_gpfn_from_mfn(mfn, pfn) do { (void) (mfn), (void)(pfn); } while (0) -#define mfn_to_gmfn(_d, mfn) (mfn) - +#define mfn_to_gfn(d, mfn) ((void)(d), _gfn(mfn_x(mfn))) /* Arch-specific portion of memory_op hypercall. */ long arch_memory_op(int op, XEN_GUEST_HANDLE_PARAM(void) arg); diff --git a/xen/include/asm-x86/mm.h b/xen/include/asm-x86/mm.h index 80b0ccffaa..cb90527499 100644 --- a/xen/include/asm-x86/mm.h +++ b/xen/include/asm-x86/mm.h @@ -527,11 +527,6 @@ extern struct rangeset *mmio_ro_ranges; #define get_gpfn_from_mfn(mfn) (machine_to_phys_mapping[(mfn)]) -#define mfn_to_gmfn(_d, mfn) \ - ( (paging_mode_translate(_d)) \ - ? get_gpfn_from_mfn(mfn) \ - : (mfn) ) - #define compat_pfn_to_cr3(pfn) (((unsigned)(pfn) << 12) | ((unsigned)(pfn) >> 20)) #define compat_cr3_to_pfn(cr3) (((unsigned)(cr3) >> 12) | ((unsigned)(cr3) << 20)) -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Wed Jun 30 02:44:15 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Wed, 30 Jun 2021 02:44:15 +0000 Received: from list by lists.xenproject.org with outflank-mailman.148114.273602 (Exim 4.92) (envelope-from ) id 1lyQCz-0001YP-9N; Wed, 30 Jun 2021 02:44:13 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 148114.273602; Wed, 30 Jun 2021 02:44:13 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lyQCz-0001YH-6K; Wed, 30 Jun 2021 02:44:13 +0000 Received: by outflank-mailman (input) for mailman id 148114; Wed, 30 Jun 2021 02:44:12 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lyQCy-0001YB-TH for xen-changelog@lists.xenproject.org; Wed, 30 Jun 2021 02:44:12 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lyQCy-0002kW-SR for xen-changelog@lists.xenproject.org; Wed, 30 Jun 2021 02:44:12 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lyQCy-000343-RT for xen-changelog@lists.xenproject.org; Wed, 30 Jun 2021 02:44:12 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=hdcqGz+GDmklczCSOqTkXe9HBOx6Ks8CdmyGSowfkMw=; b=lsOb3JeCbYJBzRjP2PhJ7a6Aw7 +w7/loJkewbrqN6TJTT+7QxNhYswV6T1qtO0w4zNQ3hT01/WqM1NSGRcHW71uA3Z5kIDENnCGDaD9 OTzeDpdk8vUT21Fcl+nKlI8f9L9UqYQK+1hzkByfVCwPEgTOretdRhaQbMykiRMtRXgg=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] x86/paging: deal with log-dirty stats overflow Message-Id: Date: Wed, 30 Jun 2021 02:44:12 +0000 commit 17e91570c5a442d0500b05c157966ac212a1f12f Author: Jan Beulich AuthorDate: Tue Jun 29 11:02:35 2021 +0200 Commit: Jan Beulich CommitDate: Tue Jun 29 11:02:35 2021 +0200 x86/paging: deal with log-dirty stats overflow While the precise values are unlikely of interest once they exceed 4 billion (allowing us to leave alone the domctl struct), we still shouldn't wrap or truncate the actual values. It is in particular problematic if the truncated values were zero (causing libxenguest to skip an iteration altogether) or a very small value (leading to premature exiting of the pre-copy phase). Change the internal fields to unsigned long, and suitably saturate for copying to guest context. Signed-off-by: Jan Beulich Acked-by: Andrew Cooper --- xen/arch/x86/mm/paging.c | 8 +++++--- xen/include/asm-x86/domain.h | 4 ++-- 2 files changed, 7 insertions(+), 5 deletions(-) diff --git a/xen/arch/x86/mm/paging.c b/xen/arch/x86/mm/paging.c index c304c24526..dd6b2bdf6f 100644 --- a/xen/arch/x86/mm/paging.c +++ b/xen/arch/x86/mm/paging.c @@ -446,14 +446,16 @@ static int paging_log_dirty_op(struct domain *d, clean = (sc->op == XEN_DOMCTL_SHADOW_OP_CLEAN); - PAGING_DEBUG(LOGDIRTY, "log-dirty %s: dom %u faults=%u dirty=%u\n", + PAGING_DEBUG(LOGDIRTY, "log-dirty %s: dom %u faults=%lu dirty=%lu\n", (clean) ? "clean" : "peek", d->domain_id, d->arch.paging.log_dirty.fault_count, d->arch.paging.log_dirty.dirty_count); - sc->stats.fault_count = d->arch.paging.log_dirty.fault_count; - sc->stats.dirty_count = d->arch.paging.log_dirty.dirty_count; + sc->stats.fault_count = min(d->arch.paging.log_dirty.fault_count, + UINT32_MAX + 0UL); + sc->stats.dirty_count = min(d->arch.paging.log_dirty.dirty_count, + UINT32_MAX + 0UL); if ( guest_handle_is_null(sc->dirty_bitmap) ) /* caller may have wanted just to clean the state or access stats. */ diff --git a/xen/include/asm-x86/domain.h b/xen/include/asm-x86/domain.h index 7213d184b0..92d54de0b9 100644 --- a/xen/include/asm-x86/domain.h +++ b/xen/include/asm-x86/domain.h @@ -190,8 +190,8 @@ struct log_dirty_domain { unsigned int failed_allocs; /* log-dirty mode stats */ - unsigned int fault_count; - unsigned int dirty_count; + unsigned long fault_count; + unsigned long dirty_count; /* functions which are paging mode specific */ const struct log_dirty_ops { -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Wed Jun 30 02:44:24 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Wed, 30 Jun 2021 02:44:24 +0000 Received: from list by lists.xenproject.org with outflank-mailman.148115.273605 (Exim 4.92) (envelope-from ) id 1lyQDA-0001bT-Ag; Wed, 30 Jun 2021 02:44:24 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 148115.273605; Wed, 30 Jun 2021 02:44:24 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lyQDA-0001bL-7r; Wed, 30 Jun 2021 02:44:24 +0000 Received: by outflank-mailman (input) for mailman id 148115; Wed, 30 Jun 2021 02:44:23 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lyQD9-0001b6-0o for xen-changelog@lists.xenproject.org; Wed, 30 Jun 2021 02:44:23 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lyQD8-0002ki-WA for xen-changelog@lists.xenproject.org; Wed, 30 Jun 2021 02:44:22 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lyQD8-00034Y-Ul for xen-changelog@lists.xenproject.org; Wed, 30 Jun 2021 02:44:22 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=BMfXeZA86f0WrQ/9T9fAuNuOYyKD7CNhBu6RirN1AmU=; b=p+sAOx0nh2tLV8ZLt5qGtMK8zu EN6Y161WtBbkgyO06hNkMkuy+17F5yy0uFQeJEJJHdtuidggECnE35PKnKOC3eSM1OEPzLQGy9YXJ ii1btwQ8hcmU708vtZlVSV+VcpwtjXACvcv7ta48MmB0mZiREfk/29pWShzztix4mk4s=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] x86/mm: pull a sanity check earlier in xenmem_add_to_physmap_one() Message-Id: Date: Wed, 30 Jun 2021 02:44:22 +0000 commit f8582da0417660269bec69e399f8667f761e886b Author: Jan Beulich AuthorDate: Tue Jun 29 11:03:29 2021 +0200 Commit: Jan Beulich CommitDate: Tue Jun 29 11:03:29 2021 +0200 x86/mm: pull a sanity check earlier in xenmem_add_to_physmap_one() We should try to limit the failure reasons after we've started making changes. Signed-off-by: Jan Beulich Acked-by: Andrew Cooper --- xen/arch/x86/mm/p2m.c | 16 +++++++++------- 1 file changed, 9 insertions(+), 7 deletions(-) diff --git a/xen/arch/x86/mm/p2m.c b/xen/arch/x86/mm/p2m.c index dbb1cbeb59..7b3cf7e9fc 100644 --- a/xen/arch/x86/mm/p2m.c +++ b/xen/arch/x86/mm/p2m.c @@ -2756,6 +2756,15 @@ int xenmem_add_to_physmap_one( goto put_both; } + /* XENMAPSPACE_gmfn: Check if the MFN is associated with another GFN. */ + old_gpfn = get_gpfn_from_mfn(mfn_x(mfn)); + ASSERT(!SHARED_M2P(old_gpfn)); + if ( space == XENMAPSPACE_gmfn && old_gpfn != gfn ) + { + rc = -EXDEV; + goto put_both; + } + /* Remove previously mapped page if it was present. */ prev_mfn = get_gfn(d, gfn_x(gpfn), &p2mt); if ( mfn_valid(prev_mfn) ) @@ -2774,13 +2783,6 @@ int xenmem_add_to_physmap_one( goto put_both; /* Unmap from old location, if any. */ - old_gpfn = get_gpfn_from_mfn(mfn_x(mfn)); - ASSERT(!SHARED_M2P(old_gpfn)); - if ( space == XENMAPSPACE_gmfn && old_gpfn != gfn ) - { - rc = -EXDEV; - goto put_both; - } if ( old_gpfn != INVALID_M2P_ENTRY ) rc = guest_physmap_remove_page(d, _gfn(old_gpfn), mfn, PAGE_ORDER_4K); -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Wed Jun 30 15:33:09 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Wed, 30 Jun 2021 15:33:09 +0000 Received: from list by lists.xenproject.org with outflank-mailman.148201.273865 (Exim 4.92) (envelope-from ) id 1lycD2-0004yX-LZ; Wed, 30 Jun 2021 15:33:04 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 148201.273865; Wed, 30 Jun 2021 15:33:04 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lycD2-0004yO-IU; Wed, 30 Jun 2021 15:33:04 +0000 Received: by outflank-mailman (input) for mailman id 148201; Wed, 30 Jun 2021 15:33:02 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lycD0-0004yI-S5 for xen-changelog@lists.xenproject.org; Wed, 30 Jun 2021 15:33:02 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lycD0-0008EJ-R2 for xen-changelog@lists.xenproject.org; Wed, 30 Jun 2021 15:33:02 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lycD0-0006JT-Pu for xen-changelog@lists.xenproject.org; Wed, 30 Jun 2021 15:33:02 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=XCLBRe0RzVbxfPKH0PIDq/9Ozo7E+vqHqYzk8p39FLA=; b=zUfnQWqrgrR1RuMheoxi9kPcqy MY9xH3S0puZVVEk1wHKWcndrcXnc3HZVGvm3lI6YgMGiROpHyNC/dQiae14+U3SZYyAjQWEZqv+YL qZla37otBwaR4Q8aggX3NvT4530L3xYZgPYYcMMOZE/v6y+c3yGY1KHh/UegAmA1U9Cc=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] x86emul: avoid using _PRE_EFLAGS() in a few cases Message-Id: Date: Wed, 30 Jun 2021 15:33:02 +0000 commit f004fd7cbbba9f3d58b399d5becfe4e8c5230fb6 Author: Jan Beulich AuthorDate: Tue Jun 29 12:33:37 2021 +0200 Commit: Jan Beulich CommitDate: Tue Jun 29 12:33:37 2021 +0200 x86emul: avoid using _PRE_EFLAGS() in a few cases The macro expanding to quite a few insns, replace its use by simply clearing the status flags when the to be executed insn doesn't depend on their initial state, in cases where this is easily possible. (There are more cases where the uses are hidden inside macros, and where some of the users of the macros want guest flags put in place before running the insn, i.e. the macros can't be updated as easily.) Signed-off-by: Jan Beulich Reviewed-by: Andrew Cooper --- xen/arch/x86/x86_emulate/x86_emulate.c | 9 ++++++--- 1 file changed, 6 insertions(+), 3 deletions(-) diff --git a/xen/arch/x86/x86_emulate/x86_emulate.c b/xen/arch/x86/x86_emulate/x86_emulate.c index 45828b726b..c9a9e031cf 100644 --- a/xen/arch/x86/x86_emulate/x86_emulate.c +++ b/xen/arch/x86/x86_emulate/x86_emulate.c @@ -6864,7 +6864,8 @@ x86_emulate( } opc[2] = 0xc3; - invoke_stub(_PRE_EFLAGS("[eflags]", "[mask]", "[tmp]"), + _regs.eflags &= ~EFLAGS_MASK; + invoke_stub("", _POST_EFLAGS("[eflags]", "[mask]", "[tmp]"), [eflags] "+g" (_regs.eflags), [tmp] "=&r" (dummy), "+m" (*mmvalp) @@ -8112,7 +8113,8 @@ x86_emulate( opc[2] = 0xc3; copy_VEX(opc, vex); - invoke_stub(_PRE_EFLAGS("[eflags]", "[mask]", "[tmp]"), + _regs.eflags &= ~EFLAGS_MASK; + invoke_stub("", _POST_EFLAGS("[eflags]", "[mask]", "[tmp]"), [eflags] "+g" (_regs.eflags), "=a" (dst.val), [tmp] "=&r" (dummy) @@ -11699,13 +11701,14 @@ int x86_emul_rmw( break; case rmw_xadd: + *eflags &= ~EFLAGS_MASK; switch ( state->op_bytes ) { unsigned long dummy; #define XADD(sz, cst, mod) \ case sz: \ - asm ( _PRE_EFLAGS("[efl]", "[msk]", "[tmp]") \ + asm ( "" \ COND_LOCK(xadd) " %"#mod"[reg], %[mem]; " \ _POST_EFLAGS("[efl]", "[msk]", "[tmp]") \ : [reg] "+" #cst (state->ea.val), \ -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Wed Jun 30 15:33:14 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Wed, 30 Jun 2021 15:33:14 +0000 Received: from list by lists.xenproject.org with outflank-mailman.148202.273869 (Exim 4.92) (envelope-from ) id 1lycDC-00050a-NA; Wed, 30 Jun 2021 15:33:14 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 148202.273869; Wed, 30 Jun 2021 15:33:14 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lycDC-00050S-K2; Wed, 30 Jun 2021 15:33:14 +0000 Received: by outflank-mailman (input) for mailman id 148202; Wed, 30 Jun 2021 15:33:13 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lycDA-00050E-V4 for xen-changelog@lists.xenproject.org; Wed, 30 Jun 2021 15:33:12 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lycDA-0008ES-UE for xen-changelog@lists.xenproject.org; Wed, 30 Jun 2021 15:33:12 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lycDA-0006KU-TO for xen-changelog@lists.xenproject.org; Wed, 30 Jun 2021 15:33:12 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=V+F+gjDDtY1e04p493WTpbJCrqbKj5U6LAgIEh00Jsw=; b=uByik/CGBFMVaDdqJUOa/t5zzA 6fzPtDqKQVz488k1XtXbBV9L1X+04OTaHjbQNVjyrCVnE31qwhG+05KOpxw94YSdPzbHM51eN/1qI xg6uSliUKR3/SKr+gWVDYXhd7ds4y+sJwNCI64sSkDnV/R0mX35dEdQxK6lcUTAEzU0c=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] AMD/IOMMU: redo awaiting of command completion Message-Id: Date: Wed, 30 Jun 2021 15:33:12 +0000 commit b175fc0c1274df34ff8bd89d603c88b014b22f10 Author: Jan Beulich AuthorDate: Tue Jun 29 12:34:37 2021 +0200 Commit: Jan Beulich CommitDate: Tue Jun 29 12:34:37 2021 +0200 AMD/IOMMU: redo awaiting of command completion The present abuse of the completion interrupt does not only stand in the way of, down the road, using it for its actual purpose, but also requires holding the IOMMU lock while waiting for command completion, limiting parallelism and keeping interrupts off for non-negligible periods of time. Have the IOMMU do an ordinary memory write instead of signaling an otherwise disabled interrupt (by just updating a status register bit). Since IOMMU_COMP_WAIT_I_FLAG_SHIFT is now unused and IOMMU_COMP_WAIT_[FS]_FLAG_SHIFT already were, drop all three of them while at it. Signed-off-by: Jan Beulich Reviewed-by: Paul Durrant --- xen/drivers/passthrough/amd/iommu-defs.h | 3 --- xen/drivers/passthrough/amd/iommu_cmd.c | 30 ++++++++++++++++-------------- 2 files changed, 16 insertions(+), 17 deletions(-) diff --git a/xen/drivers/passthrough/amd/iommu-defs.h b/xen/drivers/passthrough/amd/iommu-defs.h index 0c97db2a32..774234dfd2 100644 --- a/xen/drivers/passthrough/amd/iommu-defs.h +++ b/xen/drivers/passthrough/amd/iommu-defs.h @@ -178,11 +178,8 @@ struct amd_iommu_dte { #define IOMMU_COMP_WAIT_DATA_BUFFER_SIZE 8 #define IOMMU_COMP_WAIT_DATA_BUFFER_ALIGNMENT 8 #define IOMMU_COMP_WAIT_S_FLAG_MASK 0x00000001 -#define IOMMU_COMP_WAIT_S_FLAG_SHIFT 0 #define IOMMU_COMP_WAIT_I_FLAG_MASK 0x00000002 -#define IOMMU_COMP_WAIT_I_FLAG_SHIFT 1 #define IOMMU_COMP_WAIT_F_FLAG_MASK 0x00000004 -#define IOMMU_COMP_WAIT_F_FLAG_SHIFT 2 #define IOMMU_COMP_WAIT_ADDR_LOW_MASK 0xFFFFFFF8 #define IOMMU_COMP_WAIT_ADDR_LOW_SHIFT 3 #define IOMMU_COMP_WAIT_ADDR_HIGH_MASK 0x000FFFFF diff --git a/xen/drivers/passthrough/amd/iommu_cmd.c b/xen/drivers/passthrough/amd/iommu_cmd.c index 92a688dc9d..e8541276ca 100644 --- a/xen/drivers/passthrough/amd/iommu_cmd.c +++ b/xen/drivers/passthrough/amd/iommu_cmd.c @@ -20,6 +20,9 @@ #include "iommu.h" #include "../ats.h" +#define CMD_COMPLETION_INIT 0 +#define CMD_COMPLETION_DONE 1 + static void send_iommu_command(struct amd_iommu *iommu, const uint32_t cmd[4]) { @@ -49,28 +52,27 @@ static void send_iommu_command(struct amd_iommu *iommu, static void flush_command_buffer(struct amd_iommu *iommu, unsigned int timeout_base) { - uint32_t cmd[4]; + static DEFINE_PER_CPU(uint64_t, poll_slot); + uint64_t *this_poll_slot = &this_cpu(poll_slot); + paddr_t addr = virt_to_maddr(this_poll_slot); + /* send a COMPLETION_WAIT command to flush command buffer */ + uint32_t cmd[4] = { + addr | MASK_INSR(IOMMU_CONTROL_ENABLED, + IOMMU_COMP_WAIT_S_FLAG_MASK), + (addr >> 32) | MASK_INSR(IOMMU_CMD_COMPLETION_WAIT, + IOMMU_CMD_OPCODE_MASK), + CMD_COMPLETION_DONE + }; s_time_t start, timeout; static unsigned int __read_mostly threshold = 1; - /* RW1C 'ComWaitInt' in status register */ - writel(IOMMU_STATUS_COMP_WAIT_INT, - iommu->mmio_base + IOMMU_STATUS_MMIO_OFFSET); + ACCESS_ONCE(*this_poll_slot) = CMD_COMPLETION_INIT; - /* send an empty COMPLETION_WAIT command to flush command buffer */ - cmd[3] = cmd[2] = 0; - set_field_in_reg_u32(IOMMU_CMD_COMPLETION_WAIT, 0, - IOMMU_CMD_OPCODE_MASK, - IOMMU_CMD_OPCODE_SHIFT, &cmd[1]); - set_field_in_reg_u32(IOMMU_CONTROL_ENABLED, 0, - IOMMU_COMP_WAIT_I_FLAG_MASK, - IOMMU_COMP_WAIT_I_FLAG_SHIFT, &cmd[0]); send_iommu_command(iommu, cmd); start = NOW(); timeout = start + (timeout_base ?: 100) * MILLISECS(threshold); - while ( !(readl(iommu->mmio_base + IOMMU_STATUS_MMIO_OFFSET) & - IOMMU_STATUS_COMP_WAIT_INT) ) + while ( ACCESS_ONCE(*this_poll_slot) != CMD_COMPLETION_DONE ) { if ( timeout && NOW() > timeout ) { -- generated by git-patchbot for /home/xen/git/xen.git#master From xen-changelog-bounces@lists.xenproject.org Wed Jun 30 15:33:24 2021 Return-path: Envelope-to: archives@lists.xen.org Delivery-date: Wed, 30 Jun 2021 15:33:24 +0000 Received: from list by lists.xenproject.org with outflank-mailman.148203.273873 (Exim 4.92) (envelope-from ) id 1lycDM-00053X-OY; Wed, 30 Jun 2021 15:33:24 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 148203.273873; Wed, 30 Jun 2021 15:33:24 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lycDM-00053P-Ld; Wed, 30 Jun 2021 15:33:24 +0000 Received: by outflank-mailman (input) for mailman id 148203; Wed, 30 Jun 2021 15:33:23 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lycDL-000533-2a for xen-changelog@lists.xenproject.org; Wed, 30 Jun 2021 15:33:23 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1lycDL-0008Ec-1H for xen-changelog@lists.xenproject.org; Wed, 30 Jun 2021 15:33:23 +0000 Received: from xen by xenbits.xenproject.org with local (Exim 4.92) (envelope-from ) id 1lycDL-0006LN-0P for xen-changelog@lists.xenproject.org; Wed, 30 Jun 2021 15:33:23 +0000 X-BeenThere: xen-changelog@lists.xenproject.org List-Id: "Change log for Mercurial \(receive only\)" List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-changelog-bounces@lists.xenproject.org Precedence: list Sender: "Xen-changelog" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Date:Message-Id:Subject:Reply-To:To:From; bh=AU845nHJTeEDanqz6jwTR8ViMUm72tJLXCPOjkp7pS4=; b=LofrfaauGnM6bYlGDXxK1MbshY QtON9V6tjYkv6IaRX1wTp79b50jd0M1cHVOa3iBb3eQHpNrsrj04i8UYm4P1jsDoUaffFjXMiVZpg qCMrJl9NGLkFwwzckHP+iJPvAZK17ujyHizW4+HxljVUBTdoF2rJk9vVF+VtQ3AFFqdg=; From: patchbot@xen.org To: xen-changelog@lists.xenproject.org Reply-To: xen-devel@lists.xenproject.org Subject: [xen master] AMD/IOMMU: re-work locking around sending of commands Message-Id: Date: Wed, 30 Jun 2021 15:33:23 +0000 commit f95b7b37cfc6d4613721df9357090d14712013c0 Author: Jan Beulich AuthorDate: Tue Jun 29 12:35:12 2021 +0200 Commit: Jan Beulich CommitDate: Tue Jun 29 12:35:12 2021 +0200 AMD/IOMMU: re-work locking around sending of commands It appears unhelpful to me for flush_command_buffer() to block all progress elsewhere for the given IOMMU by holding its lock while waiting for command completion. There's no real need for callers of that function or of send_iommu_command() to hold the lock. Contain all command sending related locking to the latter function. Signed-off-by: Jan Beulich Reviewed-by: Paul Durrant --- xen/drivers/passthrough/amd/iommu_cmd.c | 23 +++++------------------ xen/drivers/passthrough/amd/iommu_guest.c | 3 ++- xen/drivers/passthrough/amd/iommu_init.c | 15 +++++++-------- xen/drivers/passthrough/amd/iommu_intr.c | 6 ------ xen/drivers/passthrough/amd/pci_amd_iommu.c | 15 ++++++++++----- 5 files changed, 24 insertions(+), 38 deletions(-) diff --git a/xen/drivers/passthrough/amd/iommu_cmd.c b/xen/drivers/passthrough/amd/iommu_cmd.c index e8541276ca..4db8e1651f 100644 --- a/xen/drivers/passthrough/amd/iommu_cmd.c +++ b/xen/drivers/passthrough/amd/iommu_cmd.c @@ -27,6 +27,9 @@ static void send_iommu_command(struct amd_iommu *iommu, const uint32_t cmd[4]) { uint32_t tail; + unsigned long flags; + + spin_lock_irqsave(&iommu->lock, flags); tail = iommu->cmd_buffer.tail + sizeof(cmd_entry_t); if ( tail == iommu->cmd_buffer.size ) @@ -47,6 +50,8 @@ static void send_iommu_command(struct amd_iommu *iommu, iommu->cmd_buffer.tail = tail; writel(tail, iommu->mmio_base + IOMMU_CMD_BUFFER_TAIL_OFFSET); + + spin_unlock_irqrestore(&iommu->lock, flags); } static void flush_command_buffer(struct amd_iommu *iommu, @@ -273,7 +278,6 @@ static void invalidate_iommu_all(struct amd_iommu *iommu) void amd_iommu_flush_iotlb(u8 devfn, const struct pci_dev *pdev, daddr_t daddr, unsigned int order) { - unsigned long flags; struct amd_iommu *iommu; unsigned int req_id, queueid, maxpend; @@ -300,10 +304,8 @@ void amd_iommu_flush_iotlb(u8 devfn, const struct pci_dev *pdev, maxpend = pdev->ats.queue_depth & 0xff; /* send INVALIDATE_IOTLB_PAGES command */ - spin_lock_irqsave(&iommu->lock, flags); invalidate_iotlb_pages(iommu, maxpend, 0, queueid, daddr, req_id, order); flush_command_buffer(iommu, iommu_dev_iotlb_timeout); - spin_unlock_irqrestore(&iommu->lock, flags); } static void amd_iommu_flush_all_iotlbs(struct domain *d, daddr_t daddr, @@ -330,17 +332,14 @@ static void amd_iommu_flush_all_iotlbs(struct domain *d, daddr_t daddr, static void _amd_iommu_flush_pages(struct domain *d, daddr_t daddr, unsigned int order) { - unsigned long flags; struct amd_iommu *iommu; unsigned int dom_id = d->domain_id; /* send INVALIDATE_IOMMU_PAGES command */ for_each_amd_iommu ( iommu ) { - spin_lock_irqsave(&iommu->lock, flags); invalidate_iommu_pages(iommu, daddr, dom_id, order); flush_command_buffer(iommu, 0); - spin_unlock_irqrestore(&iommu->lock, flags); } if ( ats_enabled ) @@ -360,37 +359,25 @@ void amd_iommu_flush_pages(struct domain *d, void amd_iommu_flush_device(struct amd_iommu *iommu, uint16_t bdf) { - ASSERT( spin_is_locked(&iommu->lock) ); - invalidate_dev_table_entry(iommu, bdf); flush_command_buffer(iommu, 0); } void amd_iommu_flush_intremap(struct amd_iommu *iommu, uint16_t bdf) { - ASSERT( spin_is_locked(&iommu->lock) ); - invalidate_interrupt_table(iommu, bdf); flush_command_buffer(iommu, 0); } void amd_iommu_flush_all_caches(struct amd_iommu *iommu) { - ASSERT( spin_is_locked(&iommu->lock) ); - invalidate_iommu_all(iommu); flush_command_buffer(iommu, 0); } void amd_iommu_send_guest_cmd(struct amd_iommu *iommu, u32 cmd[]) { - unsigned long flags; - - spin_lock_irqsave(&iommu->lock, flags); - send_iommu_command(iommu, cmd); /* TBD: Timeout selection may require peeking into cmd[]. */ flush_command_buffer(iommu, 0); - - spin_unlock_irqrestore(&iommu->lock, flags); } diff --git a/xen/drivers/passthrough/amd/iommu_guest.c b/xen/drivers/passthrough/amd/iommu_guest.c index 00c5ccd7b5..85828490ff 100644 --- a/xen/drivers/passthrough/amd/iommu_guest.c +++ b/xen/drivers/passthrough/amd/iommu_guest.c @@ -449,9 +449,10 @@ static int do_invalidate_dte(struct domain *d, cmd_entry_t *cmd) spin_lock_irqsave(&iommu->lock, flags); dte_set_gcr3_table(mdte, hdom_id, gcr3_mfn << PAGE_SHIFT, gv, glx); - amd_iommu_flush_device(iommu, req_id); spin_unlock_irqrestore(&iommu->lock, flags); + amd_iommu_flush_device(iommu, req_id); + return 0; } diff --git a/xen/drivers/passthrough/amd/iommu_init.c b/xen/drivers/passthrough/amd/iommu_init.c index bb52c181f8..312f335c1d 100644 --- a/xen/drivers/passthrough/amd/iommu_init.c +++ b/xen/drivers/passthrough/amd/iommu_init.c @@ -871,7 +871,10 @@ static void enable_iommu(struct amd_iommu *iommu) spin_lock_irqsave(&iommu->lock, flags); if ( unlikely(iommu->enabled) ) - goto out; + { + spin_unlock_irqrestore(&iommu->lock, flags); + return; + } amd_iommu_erratum_746_workaround(iommu); @@ -921,13 +924,12 @@ static void enable_iommu(struct amd_iommu *iommu) set_iommu_translation_control(iommu, IOMMU_CONTROL_ENABLED); - if ( iommu->features.flds.ia_sup ) - amd_iommu_flush_all_caches(iommu); - iommu->enabled = 1; - out: spin_unlock_irqrestore(&iommu->lock, flags); + + if ( iommu->features.flds.ia_sup ) + amd_iommu_flush_all_caches(iommu); } static void disable_iommu(struct amd_iommu *iommu) @@ -1544,7 +1546,6 @@ static int _invalidate_all_devices( { unsigned int bdf; u16 req_id; - unsigned long flags; struct amd_iommu *iommu; for ( bdf = 0; bdf < ivrs_bdf_entries; bdf++ ) @@ -1553,10 +1554,8 @@ static int _invalidate_all_devices( req_id = ivrs_mappings[bdf].dte_requestor_id; if ( iommu ) { - spin_lock_irqsave(&iommu->lock, flags); amd_iommu_flush_device(iommu, req_id); amd_iommu_flush_intremap(iommu, req_id); - spin_unlock_irqrestore(&iommu->lock, flags); } } diff --git a/xen/drivers/passthrough/amd/iommu_intr.c b/xen/drivers/passthrough/amd/iommu_intr.c index d78bb209bc..bf00370574 100644 --- a/xen/drivers/passthrough/amd/iommu_intr.c +++ b/xen/drivers/passthrough/amd/iommu_intr.c @@ -310,9 +310,7 @@ static int update_intremap_entry_from_ioapic( entry.ptr32->flds.remap_en = false; spin_unlock(lock); - spin_lock(&iommu->lock); amd_iommu_flush_intremap(iommu, req_id); - spin_unlock(&iommu->lock); spin_lock(lock); } @@ -527,11 +525,9 @@ static int update_intremap_entry_from_msi_msg( if ( iommu->enabled ) { - spin_lock_irqsave(&iommu->lock, flags); amd_iommu_flush_intremap(iommu, req_id); if ( alias_id != req_id ) amd_iommu_flush_intremap(iommu, alias_id); - spin_unlock_irqrestore(&iommu->lock, flags); } return 0; @@ -567,11 +563,9 @@ static int update_intremap_entry_from_msi_msg( entry.ptr32->flds.remap_en = false; spin_unlock(lock); - spin_lock(&iommu->lock); amd_iommu_flush_intremap(iommu, req_id); if ( alias_id != req_id ) amd_iommu_flush_intremap(iommu, alias_id); - spin_unlock(&iommu->lock); spin_lock(lock); } diff --git a/xen/drivers/passthrough/amd/pci_amd_iommu.c b/xen/drivers/passthrough/amd/pci_amd_iommu.c index 085fe2f577..2dce505b97 100644 --- a/xen/drivers/passthrough/amd/pci_amd_iommu.c +++ b/xen/drivers/passthrough/amd/pci_amd_iommu.c @@ -129,6 +129,8 @@ static void amd_iommu_setup_domain_device( iommu_has_cap(iommu, PCI_CAP_IOTLB_SHIFT) ) dte->i = ats_enabled; + spin_unlock_irqrestore(&iommu->lock, flags); + amd_iommu_flush_device(iommu, req_id); AMD_IOMMU_DEBUG("Setup I/O page table: device id = %#x, type = %#x, " @@ -138,8 +140,8 @@ static void amd_iommu_setup_domain_device( page_to_maddr(hd->arch.amd.root_table), domain->domain_id, hd->arch.amd.paging_mode); } - - spin_unlock_irqrestore(&iommu->lock, flags); + else + spin_unlock_irqrestore(&iommu->lock, flags); ASSERT(pcidevs_locked()); @@ -307,6 +309,8 @@ static void amd_iommu_disable_domain_device(const struct domain *domain, smp_wmb(); dte->v = true; + spin_unlock_irqrestore(&iommu->lock, flags); + amd_iommu_flush_device(iommu, req_id); AMD_IOMMU_DEBUG("Disable: device id = %#x, " @@ -314,7 +318,8 @@ static void amd_iommu_disable_domain_device(const struct domain *domain, req_id, domain->domain_id, dom_iommu(domain)->arch.amd.paging_mode); } - spin_unlock_irqrestore(&iommu->lock, flags); + else + spin_unlock_irqrestore(&iommu->lock, flags); ASSERT(pcidevs_locked()); @@ -455,9 +460,9 @@ static int amd_iommu_add_device(u8 devfn, struct pci_dev *pdev) iommu->dev_table.buffer + (bdf * IOMMU_DEV_TABLE_ENTRY_SIZE), ivrs_mappings[bdf].intremap_table, iommu, iommu_intremap); - amd_iommu_flush_device(iommu, bdf); - spin_unlock_irqrestore(&iommu->lock, flags); + + amd_iommu_flush_device(iommu, bdf); } amd_iommu_setup_domain_device(pdev->domain, iommu, devfn, pdev); -- generated by git-patchbot for /home/xen/git/xen.git#master