KVM: MMU: ignore zapped root pagetables

Mark zapped root pagetables as invalid and ignore such pages during lookup.

This is a problem with the cr3-target feature, where a zapped root table fools
the faulting code into creating a read-only mapping. The result is a lockup
if the instruction can't be emulated.

Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>
Cc: Anthony Liguori <aliguori@us.ibm.com>
Signed-off-by: Avi Kivity <avi@qumranet.com>

arch/x86/kvm/mmu.c

@@ -667,7 +667,8 @@ static struct kvm_mmu_page *kvm_mmu_lookup_page(struct kvm *kvm, gfn_t gfn)
 	index = kvm_page_table_hashfn(gfn);
 	bucket = &kvm->arch.mmu_page_hash[index];
 	hlist_for_each_entry(sp, node, bucket, hash_link)
-		if (sp->gfn == gfn && !sp->role.metaphysical) {
+		if (sp->gfn == gfn && !sp->role.metaphysical
+		    && !sp->role.invalid) {
 			pgprintk("%s: found role %x\n",
 				 __FUNCTION__, sp->role.word);
 			return sp;
@@ -792,8 +793,11 @@ static void kvm_mmu_zap_page(struct kvm *kvm, struct kvm_mmu_page *sp)
 	if (!sp->root_count) {
 		hlist_del(&sp->hash_link);
 		kvm_mmu_free_page(kvm, sp);
-	} else
+	} else {
 		list_move(&sp->link, &kvm->arch.active_mmu_pages);
+		sp->role.invalid = 1;
+		kvm_reload_remote_mmus(kvm);
+	}
 	kvm_mmu_reset_last_pte_updated(kvm);
 }
 
@@ -1073,6 +1077,8 @@ static void mmu_free_roots(struct kvm_vcpu *vcpu)
 
 		sp = page_header(root);
 		--sp->root_count;
+		if (!sp->root_count && sp->role.invalid)
+			kvm_mmu_zap_page(vcpu->kvm, sp);
 		vcpu->arch.mmu.root_hpa = INVALID_PAGE;
 		spin_unlock(&vcpu->kvm->mmu_lock);
 		return;
@@ -1085,6 +1091,8 @@ static void mmu_free_roots(struct kvm_vcpu *vcpu)
 			root &= PT64_BASE_ADDR_MASK;
 			sp = page_header(root);
 			--sp->root_count;
+			if (!sp->root_count && sp->role.invalid)
+				kvm_mmu_zap_page(vcpu->kvm, sp);
 		}
 		vcpu->arch.mmu.pae_root[i] = INVALID_PAGE;
 	}

arch/x86/kvm/x86.c

@@ -2658,6 +2658,10 @@ static int __vcpu_run(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
 		kvm_x86_ops->guest_debug_pre(vcpu);
 
 again:
+	if (vcpu->requests)
+		if (test_and_clear_bit(KVM_REQ_MMU_RELOAD, &vcpu->requests))
+			kvm_mmu_unload(vcpu);
+
 	r = kvm_mmu_reload(vcpu);
 	if (unlikely(r))
 		goto out;
@@ -2689,6 +2693,14 @@ static int __vcpu_run(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
 		goto out;
 	}
 
+	if (vcpu->requests)
+		if (test_bit(KVM_REQ_MMU_RELOAD, &vcpu->requests)) {
+			local_irq_enable();
+			preempt_enable();
+			r = 1;
+			goto out;
+		}
+
 	if (signal_pending(current)) {
 		local_irq_enable();
 		preempt_enable();

include/asm-x86/kvm_host.h

@@ -141,6 +141,7 @@ union kvm_mmu_page_role {
 		unsigned pad_for_nice_hex_output:6;
 		unsigned metaphysical:1;
 		unsigned access:3;
+		unsigned invalid:1;
 	};
 };
 

include/linux/kvm_host.h

@@ -37,6 +37,7 @@
 #define KVM_REQ_TLB_FLUSH          0
 #define KVM_REQ_MIGRATE_TIMER      1
 #define KVM_REQ_REPORT_TPR_ACCESS  2
+#define KVM_REQ_MMU_RELOAD         3
 
 struct kvm_vcpu;
 extern struct kmem_cache *kvm_vcpu_cache;
@@ -190,6 +191,7 @@ void kvm_resched(struct kvm_vcpu *vcpu);
 void kvm_load_guest_fpu(struct kvm_vcpu *vcpu);
 void kvm_put_guest_fpu(struct kvm_vcpu *vcpu);
 void kvm_flush_remote_tlbs(struct kvm *kvm);
+void kvm_reload_remote_mmus(struct kvm *kvm);
 
 long kvm_arch_dev_ioctl(struct file *filp,
 			unsigned int ioctl, unsigned long arg);

virt/kvm/kvm_main.c

@@ -119,6 +119,29 @@ void kvm_flush_remote_tlbs(struct kvm *kvm)
 	smp_call_function_mask(cpus, ack_flush, NULL, 1);
 }
 
+void kvm_reload_remote_mmus(struct kvm *kvm)
+{
+	int i, cpu;
+	cpumask_t cpus;
+	struct kvm_vcpu *vcpu;
+
+	cpus_clear(cpus);
+	for (i = 0; i < KVM_MAX_VCPUS; ++i) {
+		vcpu = kvm->vcpus[i];
+		if (!vcpu)
+			continue;
+		if (test_and_set_bit(KVM_REQ_MMU_RELOAD, &vcpu->requests))
+			continue;
+		cpu = vcpu->cpu;
+		if (cpu != -1 && cpu != raw_smp_processor_id())
+			cpu_set(cpu, cpus);
+	}
+	if (cpus_empty(cpus))
+		return;
+	smp_call_function_mask(cpus, ack_flush, NULL, 1);
+}
+
+
 int kvm_vcpu_init(struct kvm_vcpu *vcpu, struct kvm *kvm, unsigned id)
 {
 	struct page *page;