vhost should set worker to NULL on cgroups attach failure,
so that we won't try to destroy the worker again on close.
Signed-off-by: Michael S. Tsirkin <mst at redhat.com>
---
Dave, no need to bother with this one unless you really want to -
I'll put it on my tree.
This is a follow up to Eric's patch - fixes error handling in a similar way.
drivers/vhost/vhost.c | 1 +
1 files changed, 1 insertions(+), 0 deletions(-)
diff --git a/drivers/vhost/vhost.c b/drivers/vhost/vhost.c
index 7c75dce..666aa36 100644
--- a/drivers/vhost/vhost.c
+++ b/drivers/vhost/vhost.c
@@ -298,6 +298,7 @@ static long vhost_dev_set_owner(struct vhost_dev *dev)
return 0;
err_cgroup:
kthread_stop(worker);
+ dev->worker = NULL;
err_worker:
if (dev->mm)
mmput(dev->mm);
--
1.7.2.rc0.14.g41c1c
