[PATCH 3.11 021/208] ceph: wake up 'safe' waiters when unregistering request

[Luis Henriques]

3.11.10.3 -stable review patch.  If anyone has any objections, please let me know.

------------------

From: "Yan, Zheng" <zheng.z.yan at intel.com>

commit fc55d2c9448b34218ca58733a6f51fbede09575b upstream.

We also need to wake up 'safe' waiters if error occurs or request
aborted. Otherwise sync(2)/fsync(2) may hang forever.

Signed-off-by: Yan, Zheng <zheng.z.yan at intel.com>
Signed-off-by: Sage Weil <sage at inktank.com>
Signed-off-by: Luis Henriques <luis.henriques at canonical.com>
---
 fs/ceph/mds_client.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/fs/ceph/mds_client.c b/fs/ceph/mds_client.c
index c9c69cd..17b97ff 100644
--- a/fs/ceph/mds_client.c
+++ b/fs/ceph/mds_client.c
@@ -639,6 +639,8 @@ static void __unregister_request(struct ceph_mds_client *mdsc,
 		req->r_unsafe_dir = NULL;
 	}
 
+	complete_all(&req->r_safe_completion);
+
 	ceph_mdsc_put_request(req);
 }
 
@@ -2155,7 +2157,6 @@ static void handle_reply(struct ceph_mds_session *session, struct ceph_msg *msg)
 	if (head->safe) {
 		req->r_got_safe = true;
 		__unregister_request(mdsc, req);
-		complete_all(&req->r_safe_completion);
 
 		if (req->r_got_unsafe) {
 			/*
-- 
1.8.3.2