[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[Xen-changelog] [xen stable-4.8] xen/arm64: Add skeleton to harden the branch predictor aliasing attacks



commit 44139fed7c794eb4e47a9bb93061e325bd57fe8c
Author:     Julien Grall <julien.grall@xxxxxxxxxx>
AuthorDate: Tue Jan 16 14:23:36 2018 +0000
Commit:     Stefano Stabellini <sstabellini@xxxxxxxxxx>
CommitDate: Wed Jan 24 13:33:12 2018 -0800

    xen/arm64: Add skeleton to harden the branch predictor aliasing attacks
    
    Aliasing attacked against CPU branch predictors can allow an attacker to
    redirect speculative control flow on some CPUs and potentially divulge
    information from one context to another.
    
    This patch adds initial skeleton code behind a new Kconfig option to
    enable implementation-specific mitigations against these attacks for
    CPUs that are affected.
    
    Most of the mitigations will have to be applied when entering to the
    hypervisor from the guest context. For safety, it is applied at every
    exception entry. So there are potential for optimizing when receiving
    an exception at the same level.
    
    Because the attack is against branch predictor, it is not possible to
    safely use branch instruction before the mitigation is applied.
    Therefore, this has to be done in the vector entry before jump to the
    helper handling a given exception.
    
    On Arm64, each vector can hold 32 instructions. This leave us 31
    instructions for the mitigation. The last one is the branch instruction
    to the helper.
    
    Because a platform may have CPUs with different micro-architectures,
    per-CPU vector table needs to be provided. Realistically, only a few
    different mitigations will be necessary. So provide a small set of
    vector tables. They will be re-used and patch with the mitigations
    on-demand.
    
    This is based on the work done in Linux (see [1]).
    
    This is part of XSA-254.
    
    [1] git://git.kernel.org/pub/scm/linux/kernel/git/arm64/linux.git
    branch ktpi
    
    Signed-off-by: Julien Grall <julien.grall@xxxxxxxxxx>
    Reviewed-by: Stefano Stabellini <sstabellini@xxxxxxxxxx>
    Signed-off-by: Stefano Stabellini <sstabellini@xxxxxxxxxx>
    (cherry picked from commit 4c4fddc166cf528aca49540bcc9ee4f196b01dac)
    (cherry picked from commit fba48eff18c02d716c95b92df804a755620be82e)
    
    Conflicts:
        xen/arch/arm/cpuerrata.c
        xen/include/asm-arm/cpuerrata.h
        xen/include/asm-arm/cpufeature.h
---
 xen/arch/arm/Kconfig             |  20 ++++++
 xen/arch/arm/arm64/Makefile      |   1 +
 xen/arch/arm/arm64/bpi.S         |  64 ++++++++++++++++++
 xen/arch/arm/cpuerrata.c         | 139 +++++++++++++++++++++++++++++++++++++++
 xen/arch/arm/traps.c             |   5 +-
 xen/include/asm-arm/cpufeature.h |   3 +-
 xen/include/asm-arm/processor.h  |   5 +-
 7 files changed, 233 insertions(+), 4 deletions(-)

diff --git a/xen/arch/arm/Kconfig b/xen/arch/arm/Kconfig
index 2e023d1..a71f9c0 100644
--- a/xen/arch/arm/Kconfig
+++ b/xen/arch/arm/Kconfig
@@ -159,6 +159,26 @@ config ARM64_ERRATUM_834220
 
 endmenu
 
+config HARDEN_BRANCH_PREDICTOR
+       bool "Harden the branch predictor against aliasing attacks" if EXPERT
+       default y
+       help
+         Speculation attacks against some high-performance processors rely on
+         being able to manipulate the branch predictor for a victim context by
+         executing aliasing branches in the attacker context.  Such attacks
+         can be partially mitigated against by clearing internal branch
+         predictor state and limiting the prediction logic in some situations.
+
+         This config option will take CPU-specific actions to harden the
+         branch predictor against aliasing attacks and may rely on specific
+         instruction sequences or control bits being set by the system
+         firmware.
+
+         If unsure, say Y.
+
+config ARM64_HARDEN_BRANCH_PREDICTOR
+    def_bool y if ARM_64 && HARDEN_BRANCH_PREDICTOR
+
 source "common/Kconfig"
 
 source "drivers/Kconfig"
diff --git a/xen/arch/arm/arm64/Makefile b/xen/arch/arm/arm64/Makefile
index 149b6b3..aebe5c2 100644
--- a/xen/arch/arm/arm64/Makefile
+++ b/xen/arch/arm/arm64/Makefile
@@ -1,6 +1,7 @@
 subdir-y += lib
 
 obj-y += cache.o
+obj-$(CONFIG_HARDEN_BRANCH_PREDICTOR) += bpi.o
 obj-$(EARLY_PRINTK) += debug.o
 obj-y += domctl.o
 obj-y += domain.o
diff --git a/xen/arch/arm/arm64/bpi.S b/xen/arch/arm/arm64/bpi.S
new file mode 100644
index 0000000..6cc2f17
--- /dev/null
+++ b/xen/arch/arm/arm64/bpi.S
@@ -0,0 +1,64 @@
+/*
+ * Contains CPU specific branch predictor invalidation sequences
+ *
+ * Copyright (C) 2018 ARM Ltd.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ *
+ * This program 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 General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+.macro ventry target
+    .rept 31
+    nop
+    .endr
+    b  \target
+.endm
+
+.macro vectors target
+    ventry \target + 0x000
+    ventry \target + 0x080
+    ventry \target + 0x100
+    ventry \target + 0x180
+
+    ventry \target + 0x200
+    ventry \target + 0x280
+    ventry \target + 0x300
+    ventry \target + 0x380
+
+    ventry \target + 0x400
+    ventry \target + 0x480
+    ventry \target + 0x500
+    ventry \target + 0x580
+
+    ventry \target + 0x600
+    ventry \target + 0x680
+    ventry \target + 0x700
+    ventry \target + 0x780
+.endm
+
+/*
+ * Populate 4 vector tables. This will cover up to 4 different
+ * micro-architectures in a system.
+ */
+    .align     11
+ENTRY(__bp_harden_hyp_vecs_start)
+    .rept 4
+    vectors hyp_traps_vector
+    .endr
+ENTRY(__bp_harden_hyp_vecs_end)
+
+/*
+ * Local variables:
+ * mode: ASM
+ * indent-tabs-mode: nil
+ * End:
+ */
diff --git a/xen/arch/arm/cpuerrata.c b/xen/arch/arm/cpuerrata.c
index 5dad0fe..d0693a2 100644
--- a/xen/arch/arm/cpuerrata.c
+++ b/xen/arch/arm/cpuerrata.c
@@ -1,7 +1,146 @@
 #include <xen/config.h>
+#include <xen/cpumask.h>
+#include <xen/mm.h>
+#include <xen/sizes.h>
+#include <xen/smp.h>
+#include <xen/spinlock.h>
+#include <xen/vmap.h>
+#include <xen/warning.h>
 #include <asm/cpufeature.h>
 #include <asm/cpuerrata.h>
 
+/* Hardening Branch predictor code for Arm64 */
+#ifdef CONFIG_ARM64_HARDEN_BRANCH_PREDICTOR
+
+#define VECTOR_TABLE_SIZE SZ_2K
+
+/*
+ * Number of available table vectors (this should be in-sync with
+ * arch/arm64/bpi.S
+ */
+#define NR_BPI_HYP_VECS 4
+
+extern char __bp_harden_hyp_vecs_start[], __bp_harden_hyp_vecs_end[];
+
+/*
+ * Key for each slot. This is used to find whether a specific workaround
+ * had a slot assigned.
+ *
+ * The key is virtual address of the vector workaround
+ */
+static uintptr_t bp_harden_slot_key[NR_BPI_HYP_VECS];
+
+/*
+ * [hyp_vec_start, hyp_vec_end[ corresponds to the first 31 instructions
+ * of each vector. The last (i.e 32th) instruction is used to branch to
+ * the original entry.
+ *
+ * Those instructions will be copied on each vector to harden them.
+ */
+static bool copy_hyp_vect_bpi(unsigned int slot, const char *hyp_vec_start,
+                              const char *hyp_vec_end)
+{
+    void *dst_remapped;
+    const void *dst = __bp_harden_hyp_vecs_start + slot * VECTOR_TABLE_SIZE;
+    unsigned int i;
+    mfn_t dst_mfn = _mfn(virt_to_mfn(dst));
+
+    BUG_ON(((hyp_vec_end - hyp_vec_start) / 4) > 31);
+
+    /*
+     * Vectors are part of the text that are mapped read-only. So re-map
+     * the vector table to be able to update vectors.
+     */
+    dst_remapped = __vmap(&dst_mfn,
+                          1UL << get_order_from_bytes(VECTOR_TABLE_SIZE),
+                          1, 1, PAGE_HYPERVISOR, VMAP_DEFAULT);
+    if ( !dst_remapped )
+        return false;
+
+    dst_remapped += (vaddr_t)dst & ~PAGE_MASK;
+
+    for ( i = 0; i < VECTOR_TABLE_SIZE; i += 0x80 )
+    {
+        memcpy(dst_remapped + i, hyp_vec_start, hyp_vec_end - hyp_vec_start);
+    }
+
+    clean_dcache_va_range(dst_remapped, VECTOR_TABLE_SIZE);
+    invalidate_icache();
+
+    vunmap(dst_remapped);
+
+    return true;
+}
+
+static bool __maybe_unused
+install_bp_hardening_vec(const struct arm_cpu_capabilities *entry,
+                         const char *hyp_vec_start,
+                         const char *hyp_vec_end)
+{
+    static int last_slot = -1;
+    static DEFINE_SPINLOCK(bp_lock);
+    unsigned int i, slot = -1;
+    bool ret = true;
+
+    /*
+     * Enable callbacks are called on every CPU based on the
+     * capabilities. So double-check whether the CPU matches the
+     * entry.
+     */
+    if ( !entry->matches(entry) )
+        return true;
+
+    /*
+     * No need to install hardened vector when the processor has
+     * ID_AA64PRF0_EL1.CSV2 set.
+     */
+    if ( cpu_data[smp_processor_id()].pfr64.csv2 )
+        return true;
+
+    spin_lock(&bp_lock);
+
+    /*
+     * Look up whether the hardening vector had a slot already
+     * assigned.
+     */
+    for ( i = 0; i < 4; i++ )
+    {
+        if ( bp_harden_slot_key[i] == (uintptr_t)hyp_vec_start )
+        {
+            slot = i;
+            break;
+        }
+    }
+
+    if ( slot == -1 )
+    {
+        last_slot++;
+        /* Check we don't overrun the number of slots available. */
+        BUG_ON(NR_BPI_HYP_VECS <= last_slot);
+
+        slot = last_slot;
+        ret = copy_hyp_vect_bpi(slot, hyp_vec_start, hyp_vec_end);
+
+        /* Only update the slot if the copy succeeded. */
+        if ( ret )
+            bp_harden_slot_key[slot] = (uintptr_t)hyp_vec_start;
+    }
+
+    if ( ret )
+    {
+        /* Install the new vector table. */
+        WRITE_SYSREG((vaddr_t)(__bp_harden_hyp_vecs_start + slot * 
VECTOR_TABLE_SIZE),
+                     VBAR_EL2);
+        isb();
+    }
+
+    spin_unlock(&bp_lock);
+
+    return ret;
+}
+
+#endif /* CONFIG_ARM64_HARDEN_BRANCH_PREDICTOR */
+
 #define MIDR_RANGE(model, min, max)     \
     .matches = is_affected_midr_range,  \
     .midr_model = model,                \
diff --git a/xen/arch/arm/traps.c b/xen/arch/arm/traps.c
index bc4cbbb..d578a5c 100644
--- a/xen/arch/arm/traps.c
+++ b/xen/arch/arm/traps.c
@@ -136,7 +136,10 @@ presmp_initcall(vwfi_init);
 
 void init_traps(void)
 {
-    /* Setup Hyp vector base */
+    /*
+     * Setup Hyp vector base. Note they might get updated with the
+     * branch predictor hardening.
+     */
     WRITE_SYSREG((vaddr_t)hyp_traps_vector, VBAR_EL2);
 
     /* Trap Debug and Performance Monitor accesses */
diff --git a/xen/include/asm-arm/cpufeature.h b/xen/include/asm-arm/cpufeature.h
index 45d57cb..e26584f 100644
--- a/xen/include/asm-arm/cpufeature.h
+++ b/xen/include/asm-arm/cpufeature.h
@@ -40,8 +40,9 @@
 #define ARM32_WORKAROUND_766422 2
 #define ARM64_WORKAROUND_834220 3
 #define LIVEPATCH_FEATURE   4
+#define ARM_HARDEN_BRANCH_PREDICTOR 5
 
-#define ARM_NCAPS           5
+#define ARM_NCAPS           6
 
 #ifndef __ASSEMBLY__
 
diff --git a/xen/include/asm-arm/processor.h b/xen/include/asm-arm/processor.h
index 2c9c193..51ecbd2 100644
--- a/xen/include/asm-arm/processor.h
+++ b/xen/include/asm-arm/processor.h
@@ -318,8 +318,9 @@ struct cpuinfo_arm {
             unsigned long fp:4;   /* Floating Point */
             unsigned long simd:4; /* Advanced SIMD */
             unsigned long gic:4;  /* GIC support */
-            unsigned long __res0:4;
-            unsigned long __res1;
+            unsigned long __res0:28;
+            unsigned long csv2:4;
+            unsigned long __res1:4;
         };
     } pfr64;
 
--
generated by git-patchbot for /home/xen/git/xen.git#stable-4.8

_______________________________________________
Xen-changelog mailing list
Xen-changelog@xxxxxxxxxxxxxxxxxxxx
https://lists.xenproject.org/xen-changelog

 


Rackspace

Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.