KVM: Fix MMU invalidation bookkeeping in guest_memfd

Acquire mmu_lock and do invalidate_{begin,end}() if and only if there is
at least one memslot that overlaps the to-be-invalidated range.  This
fixes a bug where KVM would leave a danging in-progress invalidation as
the begin() call was unconditional, but the end() was not (only performed
if there was overlap).

Reported-by: Binbin Wu <binbin.wu@linux.intel.com>
Fixes: 1d46f95 ("KVM: Add KVM_CREATE_GUEST_MEMFD ioctl() for guest-specific backing memory")
Link: https://lore.kernel.org/r/20230921203331.3746712-6-seanjc@google.com
Signed-off-by: Sean Christopherson <seanjc@google.com>

Author: sean-jc

virt/kvm/guest_mem.c

@@ -88,14 +88,10 @@ static struct folio *kvm_gmem_get_folio(struct inode *inode, pgoff_t index)
 static void kvm_gmem_invalidate_begin(struct kvm_gmem *gmem, pgoff_t start,
 				      pgoff_t end)
 {
+	bool flush = false, found_memslot = false;
 	struct kvm_memory_slot *slot;
 	struct kvm *kvm = gmem->kvm;
 	unsigned long index;
-	bool flush = false;
-
-	KVM_MMU_LOCK(kvm);
-
-	kvm_mmu_invalidate_begin(kvm);
 
 	xa_for_each_range(&gmem->bindings, index, slot, start, end - 1) {
 		pgoff_t pgoff = slot->gmem.pgoff;
@@ -107,24 +103,33 @@ static void kvm_gmem_invalidate_begin(struct kvm_gmem *gmem, pgoff_t start,
 			.may_block = true,
 		};
 
+		if (!found_memslot) {
+			found_memslot = true;
+
+			KVM_MMU_LOCK(kvm);
+			kvm_mmu_invalidate_begin(kvm);
+		}
+
 		flush |= kvm_mmu_unmap_gfn_range(kvm, &gfn_range);
 	}
 
 	if (flush)
 		kvm_flush_remote_tlbs(kvm);
 
-	KVM_MMU_UNLOCK(kvm);
+	if (found_memslot)
+		KVM_MMU_UNLOCK(kvm);
 }
 
 static void kvm_gmem_invalidate_end(struct kvm_gmem *gmem, pgoff_t start,
 				    pgoff_t end)
 {
 	struct kvm *kvm = gmem->kvm;
 
-	KVM_MMU_LOCK(kvm);
-	if (xa_find(&gmem->bindings, &start, end - 1, XA_PRESENT))
+	if (xa_find(&gmem->bindings, &start, end - 1, XA_PRESENT)) {
+		KVM_MMU_LOCK(kvm);
 		kvm_mmu_invalidate_end(kvm);
-	KVM_MMU_UNLOCK(kvm);
+		KVM_MMU_UNLOCK(kvm);
+	}
 }
 
 static long kvm_gmem_punch_hole(struct inode *inode, loff_t offset, loff_t len)