KVM: MMU: Fix setting the accessed bit on non-speculative sptes
authorAvi Kivity <avi@qumranet.com>
Wed, 27 Aug 2008 17:01:04 +0000 (20:01 +0300)
committerAvi Kivity <avi@qumranet.com>
Wed, 15 Oct 2008 08:15:24 +0000 (10:15 +0200)
The accessed bit was accidentally turned on in a random flag word, rather
than, the spte itself, which was lucky, since it used the non-EPT compatible
PT_ACCESSED_MASK.

Fix by turning the bit on in the spte and changing it to use the portable
accessed mask.

Signed-off-by: Avi Kivity <avi@qumranet.com>
arch/x86/kvm/mmu.c

index 853a2889b202aa6746562fb9658260bbe495b489..866d7133cad8cf68d1de4f9432c7b67c4758c32f 100644 (file)
@@ -1192,7 +1192,7 @@ static void mmu_set_spte(struct kvm_vcpu *vcpu, u64 *shadow_pte,
         */
        spte = shadow_base_present_pte | shadow_dirty_mask;
        if (!speculative)
-               pte_access |= PT_ACCESSED_MASK;
+               spte |= shadow_accessed_mask;
        if (!dirty)
                pte_access &= ~ACC_WRITE_MASK;
        if (pte_access & ACC_EXEC_MASK)