KVM: MMU: delay flush all tlbs on sync_page path

Quote from Avi:
| I don't think we need to flush immediately; set a "tlb dirty" bit somewhere
| that is cleareded when we flush the tlb.  kvm_mmu_notifier_invalidate_page()
| can consult the bit and force a flush if set.

Signed-off-by: Xiao Guangrong <xiaoguangrong@cn.fujitsu.com>
Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com>

3 files changed, 18 insertions, 3 deletions

diff --git a/arch/x86/kvm/paging_tmpl.h b/arch/x86/kvm/paging_tmpl.h
index a43f4cc..2b3d66c 100644
--- a/arch/x86/kvm/paging_tmpl.h
+++ b/arch/x86/kvm/paging_tmpl.h
@@ -746,6 +746,14 @@ static void FNAME(prefetch_page)(struct kvm_vcpu *vcpu,
  * Using the cached information from sp->gfns is safe because:
  * - The spte has a reference to the struct page, so the pfn for a given gfn
  *   can't change unless all sptes pointing to it are nuked first.
+ *
+ * Note:
+ *   We should flush all tlbs if spte is dropped even though guest is
+ *   responsible for it. Since if we don't, kvm_mmu_notifier_invalidate_page
+ *   and kvm_mmu_notifier_invalidate_range_start detect the mapping page isn't
+ *   used by guest then tlbs are not flushed, so guest is allowed to access the
+ *   freed pages.
+ *   And we increase kvm->tlbs_dirty to delay tlbs flush in this case.
  */
 static int FNAME(sync_page)(struct kvm_vcpu *vcpu, struct kvm_mmu_page *sp)
 {
@@ -781,14 +789,14 @@ static int FNAME(sync_page)(struct kvm_vcpu *vcpu, struct kvm_mmu_page *sp)
 		gfn = gpte_to_gfn(gpte);
 
 		if (FNAME(prefetch_invalid_gpte)(vcpu, sp, &sp->spt[i], gpte)) {
-			kvm_flush_remote_tlbs(vcpu->kvm);
+			vcpu->kvm->tlbs_dirty++;
 			continue;
 		}
 
 		if (gfn != sp->gfns[i]) {
 			drop_spte(vcpu->kvm, &sp->spt[i],
 				      shadow_trap_nonpresent_pte);
-			kvm_flush_remote_tlbs(vcpu->kvm);
+			vcpu->kvm->tlbs_dirty++;
 			continue;
 		}
 
diff --git a/include/linux/kvm_host.h b/include/linux/kvm_host.h
index da0794f..ac4e83a 100644
--- a/include/linux/kvm_host.h
+++ b/include/linux/kvm_host.h
@@ -254,6 +254,7 @@ struct kvm {
 	struct mmu_notifier mmu_notifier;
 	unsigned long mmu_notifier_seq;
 	long mmu_notifier_count;
+	long tlbs_dirty;
 #endif
 };
 
@@ -382,6 +383,7 @@ void kvm_vcpu_on_spin(struct kvm_vcpu *vcpu);
 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);
 
diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index 5156d45..ee99b77 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -168,8 +168,12 @@ static bool make_all_cpus_request(struct kvm *kvm, unsigned int req)
 
 void kvm_flush_remote_tlbs(struct kvm *kvm)
 {
+	int dirty_count = kvm->tlbs_dirty;
+
+	smp_mb();
 	if (make_all_cpus_request(kvm, KVM_REQ_TLB_FLUSH))
 		++kvm->stat.remote_tlb_flush;
+	cmpxchg(&kvm->tlbs_dirty, dirty_count, 0);
 }
 
 void kvm_reload_remote_mmus(struct kvm *kvm)
@@ -249,7 +253,7 @@ static void kvm_mmu_notifier_invalidate_page(struct mmu_notifier *mn,
 	idx = srcu_read_lock(&kvm->srcu);
 	spin_lock(&kvm->mmu_lock);
 	kvm->mmu_notifier_seq++;
-	need_tlb_flush = kvm_unmap_hva(kvm, address);
+	need_tlb_flush = kvm_unmap_hva(kvm, address) | kvm->tlbs_dirty;
 	spin_unlock(&kvm->mmu_lock);
 	srcu_read_unlock(&kvm->srcu, idx);
 
@@ -293,6 +297,7 @@ static void kvm_mmu_notifier_invalidate_range_start(struct mmu_notifier *mn,
 	kvm->mmu_notifier_count++;
 	for (; start < end; start += PAGE_SIZE)
 		need_tlb_flush |= kvm_unmap_hva(kvm, start);
+	need_tlb_flush |= kvm->tlbs_dirty;
 	spin_unlock(&kvm->mmu_lock);
 	srcu_read_unlock(&kvm->srcu, idx);