[PATCH] x86/livepatch: Fix livepatch application when CET is active

[Andrew Cooper <andrew.cooper3@xxxxxxxxxx>]

Right now, trying to apply a livepatch on any system with CET shstk (AMD Zen3
or later, Intel Tiger Lake or Sapphire Rapids and later) fails as follows:

  (XEN) livepatch: lp: Verifying enabled expectations for all functions
  (XEN) common/livepatch.c:1591: livepatch: lp: timeout is 30000000ns
  (XEN) common/livepatch.c:1703: livepatch: lp: CPU28 - IPIing the other 127 
CPUs
  (XEN) livepatch: lp: Applying 1 functions
  (XEN) hi_func: Hi! (called 1 times)
  (XEN) Hook executing.
  (XEN) Assertion 'local_irq_is_enabled() || cpumask_subset(mask, 
cpumask_of(cpu))' failed at arch/x86/smp.c:265
  (XEN) *** DOUBLE FAULT ***
  <many double faults>

The assertion failure is from a global (system wide) TLB flush initiated by
modify_xen_mappings().  I'm not entirely sure when this broke, and I'm not
sure exactly what causes the #DF's, but it doesn't really matter either
because they highlight a latent bug that I'd overlooked with the CET-SS vs
patching work the first place.

While we're careful to arrange for the patching CPU to avoid encountering
non-shstk memory with transient shstk perms, other CPUs can pick these
mappings up too if they need to re-walk for uarch reasons.

Another bug is that for livepatching, we only disable CET if shadow stacks are
in use.  Running on Intel CET systems when Xen is only using CET-IBT will
crash in arch_livepatch_quiesce() when trying to clear CR0.WP with CR4.CET
still active.

Also, we never went and cleared the dirty bits on .rodata.  This would
matter (for the same reason it matters on .text - it becomes a valid target
for WRSS), but we never actually patch .rodata anyway.

Therefore rework how we do patching for both alternatives and livepatches.

Introduce modify_xen_mappings_lite() with a purpose similar to
modify_xen_mappings(), but stripped down to the bare minimum as it's used in
weird contexts.  Leave all complexity to the caller to handle.

Instead of patching by clearing CR0.WP (and having to jump through some
fragile hoops to disable CET in order to do this), just transiently relax the
permissions on .text via l2_identmap[].

The perms are relaxed globally, but is safe enough.  Alternatives run before
we boot APs, and Livepatching runs in a quiesced state where the other CPUs
are not doing anything interesting.

This approach is far more robust.

Signed-off-by: Andrew Cooper <andrew.cooper3@xxxxxxxxxx>
---
CC: Jan Beulich <JBeulich@xxxxxxxx>
CC: Roger Pau Monné <roger.pau@xxxxxxxxxx>
CC: Wei Liu <wl@xxxxxxx>
CC: Konrad Rzeszutek Wilk <konrad.wilk@xxxxxxxxxx>
CC: Ross Lagerwall <ross.lagerwall@xxxxxxxxxx>

Pulling put_pte_flags() out of the loops in modify_xen_mappings_lite() halves
the size of the function.  The code generation of the typesafe pagetable
helpers is terrible, both because of flags needing a 32->64 expand, and
because of _PAGE_NX using cpu_has_nx behind the scene.  We really should
improve how all of this works.
---
 xen/arch/x86/alternative.c       | 45 ++++++++------------
 xen/arch/x86/livepatch.c         | 56 ++++++++++---------------
 xen/arch/x86/mm.c                | 71 ++++++++++++++++++++++++++++++++
 xen/common/virtual_region.c      | 22 +++++++---
 xen/include/xen/mm.h             |  1 +
 xen/include/xen/virtual_region.h |  4 +-
 6 files changed, 132 insertions(+), 67 deletions(-)

diff --git a/xen/arch/x86/alternative.c b/xen/arch/x86/alternative.c
index 2383fa66294c..fc815bc7d627 100644
--- a/xen/arch/x86/alternative.c
+++ b/xen/arch/x86/alternative.c
@@ -382,24 +382,28 @@ static int __init cf_check nmi_apply_alternatives(
      */
     if ( !(alt_done & alt_todo) )
     {
-        unsigned long cr0, cr4;
-
-        cr0 = read_cr0();
-        cr4 = read_cr4();
-
-        if ( cr4 & X86_CR4_CET )
-            write_cr4(cr4 & ~X86_CR4_CET);
-
-        /* Disable WP to allow patching read-only pages. */
-        write_cr0(cr0 & ~X86_CR0_WP);
+        /*
+         * Relax perms on .text to be RWX, so we can modify them.
+         *
+         * This relaxes perms globally, but we run ahead of bringing APs
+         * online, so only have our own TLB to worry about.
+         */
+        modify_xen_mappings_lite(XEN_VIRT_START + MB(2),
+                                 (unsigned long)&__2M_text_end,
+                                 PAGE_HYPERVISOR_RWX);
+        flush_local(FLUSH_TLB_GLOBAL);
 
         _apply_alternatives(__alt_instructions, __alt_instructions_end,
                             alt_done);
 
-        write_cr0(cr0);
-
-        if ( cr4 & X86_CR4_CET )
-            write_cr4(cr4);
+        /*
+         * Reinstate perms on .text to be RW.  This also cleans out the dirty
+         * bits, which matters when CET Shstk is active.
+         */
+        modify_xen_mappings_lite(XEN_VIRT_START + MB(2),
+                                 (unsigned long)&__2M_text_end,
+                                 PAGE_HYPERVISOR_RX);
+        flush_local(FLUSH_TLB_GLOBAL);
 
         alt_done |= alt_todo;
     }
@@ -454,19 +458,6 @@ static void __init _alternative_instructions(bool force)
         panic("Timed out waiting for alternatives self-NMI to hit\n");
 
     set_nmi_callback(saved_nmi_callback);
-
-    /*
-     * When Xen is using shadow stacks, the alternatives clearing CR0.WP and
-     * writing into the mappings set dirty bits, turning the mappings into
-     * shadow stack mappings.
-     *
-     * While we can execute from them, this would also permit them to be the
-     * target of WRSS instructions, so reset the dirty after patching.
-     */
-    if ( cpu_has_xen_shstk )
-        modify_xen_mappings(XEN_VIRT_START + MB(2),
-                            (unsigned long)&__2M_text_end,
-                            PAGE_HYPERVISOR_RX);
 }
 
 void __init alternative_instructions(void)
diff --git a/xen/arch/x86/livepatch.c b/xen/arch/x86/livepatch.c
index f2d783fdc567..a54d991c5f0f 100644
--- a/xen/arch/x86/livepatch.c
+++ b/xen/arch/x86/livepatch.c
@@ -61,46 +61,32 @@ int arch_livepatch_safety_check(void)
 
 int noinline arch_livepatch_quiesce(void)
 {
-    /* If Shadow Stacks are in use, disable CR4.CET so we can modify CR0.WP. */
-    if ( cpu_has_xen_shstk )
-        write_cr4(read_cr4() & ~X86_CR4_CET);
-
-    /* Disable WP to allow changes to read-only pages. */
-    write_cr0(read_cr0() & ~X86_CR0_WP);
+    /*
+     * Relax perms on .text to be RWX, so we can modify them.
+     *
+     * This relaxes perms globally, but all other CPUs are waiting on us.
+     */
+    relax_virtual_region_perms();
+    flush_local(FLUSH_TLB_GLOBAL);
 
     return 0;
 }
 
 void noinline arch_livepatch_revive(void)
 {
-    /* Reinstate WP. */
-    write_cr0(read_cr0() | X86_CR0_WP);
-
-    /* Clobber dirty bits and reinstate CET, if applicable. */
-    if ( IS_ENABLED(CONFIG_XEN_SHSTK) && cpu_has_xen_shstk )
-    {
-        unsigned long tmp;
-
-        reset_virtual_region_perms();
-
-        write_cr4(read_cr4() | X86_CR4_CET);
-
-        /*
-         * Fix up the return address on the shadow stack, which currently
-         * points at arch_livepatch_quiesce()'s caller.
-         *
-         * Note: this is somewhat fragile, and depends on both
-         * arch_livepatch_{quiesce,revive}() being called from the same
-         * function, which is currently the case.
-         *
-         * Any error will result in Xen dying with #CP, and its too late to
-         * recover in any way.
-         */
-        asm volatile ("rdsspq %[ssp];"
-                      "wrssq %[addr], (%[ssp]);"
-                      : [ssp] "=&r" (tmp)
-                      : [addr] "r" (__builtin_return_address(0)));
-    }
+    /*
+     * Reinstate perms on .text to be RX.  This also cleans out the dirty
+     * bits, which matters when CET Shstk is active.
+     *
+     * The other CPUs waiting for us could in principle have re-walked while
+     * we were patching and cached the reduced perms in their TLB.  Therefore,
+     * we need to do a global TLB flush.
+     *
+     * However, we can't use Xen's normal global TLB flush infrastructure, so
+     * delay the TLB flush to arch_livepatch_post_action(), which is called on
+     * all CPUs (including us) on the way out of patching.
+     */
+    tighten_virtual_region_perms();
 }
 
 int arch_livepatch_verify_func(const struct livepatch_func *func)
@@ -197,6 +183,8 @@ void noinline arch_livepatch_revert(const struct 
livepatch_func *func)
  */
 void noinline arch_livepatch_post_action(void)
 {
+    /* See arch_livepatch_revive() */
+    flush_local(FLUSH_TLB_GLOBAL);
 }
 
 static nmi_callback_t *saved_nmi_callback;
diff --git a/xen/arch/x86/mm.c b/xen/arch/x86/mm.c
index 36a07ef77eae..1707bcd2d15c 100644
--- a/xen/arch/x86/mm.c
+++ b/xen/arch/x86/mm.c
@@ -5879,6 +5879,77 @@ int destroy_xen_mappings(unsigned long s, unsigned long 
e)
     return modify_xen_mappings(s, e, _PAGE_NONE);
 }
 
+#if defined(CONFIG_LIVEPATCH) && defined(CONFIG_HAS_ALTERNATIVE)
+/*
+ * Similar to modify_xen_mappings(), but used by the alternatives and
+ * livepatch in weird contexts.  All synchronization, TLB flushing, etc is the
+ * responsibility of the caller, and *MUST* not be introduced here.
+ *
+ * Must be limited to XEN_VIRT_{START,END}, i.e. over l2_xenmap[].
+ * Must be called with preset flags, and over present mappings.
+ * Must be called on leaf page boundaries.
+ */
+void modify_xen_mappings_lite(unsigned long s, unsigned long e, unsigned int 
_nf)
+{
+    unsigned long v = s, fm, nf;
+
+    /* Set of valid PTE bits which may be altered. */
+#define FLAGS_MASK (_PAGE_NX|_PAGE_DIRTY|_PAGE_ACCESSED|_PAGE_RW|_PAGE_PRESENT)
+    _nf &= FLAGS_MASK;
+
+    fm = put_pte_flags(FLAGS_MASK);
+    nf = put_pte_flags(_nf);
+
+    ASSERT(nf & _PAGE_PRESENT);
+    ASSERT(IS_ALIGNED(s, PAGE_SIZE) && s >= XEN_VIRT_START);
+    ASSERT(IS_ALIGNED(e, PAGE_SIZE) && e <= XEN_VIRT_END);
+
+    while ( v < e )
+    {
+        l2_pgentry_t *pl2e = &l2_xenmap[l2_table_offset(v)];
+        l2_pgentry_t l2e = l2e_read_atomic(pl2e);
+        unsigned int l2f = l2e_get_flags(l2e);
+
+        ASSERT(l2f & _PAGE_PRESENT);
+
+        if ( l2e_get_flags(l2e) & _PAGE_PSE )
+        {
+            ASSERT(l1_table_offset(v) == 0);
+
+            l2e_write_atomic(pl2e, l2e_from_intpte((l2e.l2 & ~fm) | nf));
+
+            v += 1UL << L2_PAGETABLE_SHIFT;
+            continue;
+        }
+
+        /* else decend to l1 */
+        {
+            l1_pgentry_t *pl1t = map_l1t_from_l2e(l2e);
+
+            while ( v < e )
+            {
+                l1_pgentry_t *pl1e = &pl1t[l1_table_offset(v)];
+                l1_pgentry_t l1e = l1e_read_atomic(pl1e);
+                unsigned int l1f = l1e_get_flags(l1e);
+
+                ASSERT(l1f & _PAGE_PRESENT);
+
+                l1e_write_atomic(pl1e, l1e_from_intpte((l1e.l1 & ~fm) | nf));
+
+                v += 1UL << L1_PAGETABLE_SHIFT;
+
+                if ( l2_table_offset(v) == 0 )
+                    break;
+            }
+
+            unmap_domain_page(pl1t);
+        }
+    }
+
+#undef FLAGS_MASK
+}
+#endif /* LIVEPATCH || ALTERNATIVE */
+
 void __set_fixmap(
     enum fixed_addresses idx, unsigned long mfn, unsigned long flags)
 {
diff --git a/xen/common/virtual_region.c b/xen/common/virtual_region.c
index 5ecdba9c08ed..ddac5c9147e5 100644
--- a/xen/common/virtual_region.c
+++ b/xen/common/virtual_region.c
@@ -92,16 +92,28 @@ void unregister_virtual_region(struct virtual_region *r)
     remove_virtual_region(r);
 }
 
-#if defined(CONFIG_LIVEPATCH) && defined(CONFIG_XEN_SHSTK)
-void reset_virtual_region_perms(void)
+#if defined(CONFIG_LIVEPATCH) && defined(CONFIG_X86)
+void relax_virtual_region_perms(void)
 {
     const struct virtual_region *region;
 
     rcu_read_lock(&rcu_virtual_region_lock);
     list_for_each_entry_rcu( region, &virtual_region_list, list )
-        modify_xen_mappings((unsigned long)region->start,
-                            ROUNDUP((unsigned long)region->end, PAGE_SIZE),
-                            PAGE_HYPERVISOR_RX);
+        modify_xen_mappings_lite((unsigned long)region->start,
+                                 ROUNDUP((unsigned long)region->end, 
PAGE_SIZE),
+                                 PAGE_HYPERVISOR_RWX);
+    rcu_read_unlock(&rcu_virtual_region_lock);
+}
+
+void tighten_virtual_region_perms(void)
+{
+    const struct virtual_region *region;
+
+    rcu_read_lock(&rcu_virtual_region_lock);
+    list_for_each_entry_rcu( region, &virtual_region_list, list )
+        modify_xen_mappings_lite((unsigned long)region->start,
+                                 ROUNDUP((unsigned long)region->end, 
PAGE_SIZE),
+                                 PAGE_HYPERVISOR_RX);
     rcu_read_unlock(&rcu_virtual_region_lock);
 }
 #endif
diff --git a/xen/include/xen/mm.h b/xen/include/xen/mm.h
index 9d14aed74baa..b0dc3ba9c98d 100644
--- a/xen/include/xen/mm.h
+++ b/xen/include/xen/mm.h
@@ -100,6 +100,7 @@ int map_pages_to_xen(
     unsigned int flags);
 /* Alter the permissions of a range of Xen virtual address space. */
 int modify_xen_mappings(unsigned long s, unsigned long e, unsigned int flags);
+void modify_xen_mappings_lite(unsigned long s, unsigned long e, unsigned int 
flags);
 int destroy_xen_mappings(unsigned long s, unsigned long e);
 /* Retrieve the MFN mapped by VA in Xen virtual address space. */
 mfn_t xen_map_to_mfn(unsigned long va);
diff --git a/xen/include/xen/virtual_region.h b/xen/include/xen/virtual_region.h
index ba408eb87a1a..d05362071135 100644
--- a/xen/include/xen/virtual_region.h
+++ b/xen/include/xen/virtual_region.h
@@ -33,7 +33,9 @@ void setup_virtual_regions(const struct exception_table_entry 
*start,
 void unregister_init_virtual_region(void);
 void register_virtual_region(struct virtual_region *r);
 void unregister_virtual_region(struct virtual_region *r);
-void reset_virtual_region_perms(void);
+
+void relax_virtual_region_perms(void);
+void tighten_virtual_region_perms(void);
 
 #endif /* __XEN_VIRTUAL_REGION_H__ */
 
-- 
2.30.2