Bail from fast_page_fault() if the SPTE is not a shadow-present SPTE.
Functionally, this is not strictly necessary as the !is_access_allowed()
check will eventually reject the fast path, but an early check on
shadow-present skips unnecessary checks and will allow a future patch to
tweak the A/D status auditing to warn if KVM attempts to query A/D bits
without first ensuring the SPTE is a shadow-present SPTE.
Note, is_shadow_present_pte() is quite expensive at this time, i.e. this
might be a net negative in the short term. A future patch will optimize
is_shadow_present_pte() to a single AND operation and remedy the issue.
Signed-off-by: Sean Christopherson <sea...@google.com>
---
arch/x86/kvm/mmu/mmu.c | 3 +++
1 file changed, 3 insertions(+)
diff --git a/arch/x86/kvm/mmu/mmu.c b/arch/x86/kvm/mmu/mmu.c
index d75524bc8423..93b0285e8b38 100644
--- a/arch/x86/kvm/mmu/mmu.c
+++ b/arch/x86/kvm/mmu/mmu.c
@@ -3061,6 +3061,9 @@ static int fast_page_fault(struct kvm_vcpu *vcpu, gpa_t
cr2_or_gpa,
if (!is_shadow_present_pte(spte))
break;
+ if (!is_shadow_present_pte(spte))
+ break;
+
sp = sptep_to_sp(iterator.sptep);
if (!is_last_spte(spte, sp->role.level))
break;
--
2.30.1.766.gb4fecdf3b7-goog
