vhost/test: stop device before reset
authorMichael S. Tsirkin <mst@redhat.com>
Mon, 7 Oct 2019 17:56:59 +0000 (13:56 -0400)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Mon, 27 Jan 2020 13:51:19 +0000 (14:51 +0100)
[ Upstream commit 245cdd9fbd396483d501db83047116e2530f245f ]

When device stop was moved out of reset, test device wasn't updated to
stop before reset, this resulted in a use after free.  Fix by invoking
stop appropriately.

Fixes: b211616d7125 ("vhost: move -net specific code out")
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
drivers/vhost/test.c

index a9be2d8..55090d9 100644 (file)
@@ -162,6 +162,7 @@ static int vhost_test_release(struct inode *inode, struct file *f)
 
        vhost_test_stop(n, &private);
        vhost_test_flush(n);
+       vhost_dev_stop(&n->dev);
        vhost_dev_cleanup(&n->dev);
        /* We do an extra flush before freeing memory,
         * since jobs can re-queue themselves. */
@@ -238,6 +239,7 @@ static long vhost_test_reset_owner(struct vhost_test *n)
        }
        vhost_test_stop(n, &priv);
        vhost_test_flush(n);
+       vhost_dev_stop(&n->dev);
        vhost_dev_reset_owner(&n->dev, umem);
 done:
        mutex_unlock(&n->dev.mutex);