---

yaml
---
r: 202379
b: refs/heads/master
c: 9e7b0e7
h: refs/heads/master
i:
  202377: 7223f46
  202375: 32c5797
v: v3

[refs]

@@ -1,2 +1,2 @@
 ---
-refs/heads/master: 5fd5387c89ec99ff6cb82d2477ffeb7211b781c2
+refs/heads/master: 9e7b0e7fba45ca3c6357aeb7091ebc281f1de365

trunk/arch/x86/kvm/paging_tmpl.h

@@ -325,8 +325,32 @@ static u64 *FNAME(fetch)(struct kvm_vcpu *vcpu, gva_t addr,
 			break;
 		}
 
-		if (is_shadow_present_pte(*sptep) && !is_large_pte(*sptep))
-			continue;
+		if (is_shadow_present_pte(*sptep) && !is_large_pte(*sptep)) {
+			struct kvm_mmu_page *child;
+			unsigned direct_access;
+
+			if (level != gw->level)
+				continue;
+
+			/*
+			 * For the direct sp, if the guest pte's dirty bit
+			 * changed form clean to dirty, it will corrupt the
+			 * sp's access: allow writable in the read-only sp,
+			 * so we should update the spte at this point to get
+			 * a new sp with the correct access.
+			 */
+			direct_access = gw->pt_access & gw->pte_access;
+			if (!is_dirty_gpte(gw->ptes[gw->level - 1]))
+				direct_access &= ~ACC_WRITE_MASK;
+
+			child = page_header(*sptep & PT64_BASE_ADDR_MASK);
+			if (child->role.access == direct_access)
+				continue;
+
+			mmu_page_remove_parent_pte(child, sptep);
+			__set_spte(sptep, shadow_trap_nonpresent_pte);
+			kvm_flush_remote_tlbs(vcpu->kvm);
+		}
 
 		if (is_large_pte(*sptep)) {
 			rmap_remove(vcpu->kvm, sptep);