kvm: free resources after canceling async_pf
authorRadim Krčmář <rkrcmar@redhat.com>
Wed, 4 Sep 2013 20:32:23 +0000 (22:32 +0200)
committerBen Hutchings <ben@decadent.org.uk>
Mon, 9 Jun 2014 12:29:03 +0000 (13:29 +0100)
commit 28b441e24088081c1e213139d1303b451a34a4f4 upstream.

When we cancel 'async_pf_execute()', we should behave as if the work was
never scheduled in 'kvm_setup_async_pf()'.
Fixes a bug when we can't unload module because the vm wasn't destroyed.

Signed-off-by: Radim Krčmář <rkrcmar@redhat.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Gleb Natapov <gleb@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
virt/kvm/async_pf.c

index 74268b4c2ee167932d514281bf6ecc3187462a3e..1758e32863cf26cd76d6a0c85632dfa74afbf2b3 100644 (file)
@@ -101,8 +101,11 @@ void kvm_clear_async_pf_completion_queue(struct kvm_vcpu *vcpu)
                                   typeof(*work), queue);
                cancel_work_sync(&work->work);
                list_del(&work->queue);
-               if (!work->done) /* work was canceled */
+               if (!work->done) { /* work was canceled */
+                       mmdrop(work->mm);
+                       kvm_put_kvm(vcpu->kvm); /* == work->vcpu->kvm */
                        kmem_cache_free(async_pf_cache, work);
+               }
        }
 
        spin_lock(&vcpu->async_pf.lock);