From: Jeff Layton <jlay...@kernel.org>
commit 28a282616f56990547b9dcd5c6fbd2001344664c upstream.
When ceph_mdsc_do_request returns an error, we can't assume that the
filelock_reply pointer will be set. Only try to fetch fields out of
the r_reply_info when it returns success.
Cc: sta...@vger.kernel.org
Reported-by: Hector Martin <hec...@marcansoft.com>
Signed-off-by: Jeff Layton <jlay...@kernel.org>
Reviewed-by: "Yan, Zheng" <z...@redhat.com>
Signed-off-by: Ilya Dryomov <idryo...@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gre...@linuxfoundation.org>
---
fs/ceph/locks.c | 3 +--
1 file changed, 1 insertion(+), 2 deletions(-)
--- a/fs/ceph/locks.c
+++ b/fs/ceph/locks.c
@@ -111,8 +111,7 @@ static int ceph_lock_message(u8 lock_typ
req->r_wait_for_completion = ceph_lock_wait_for_completion;
err = ceph_mdsc_do_request(mdsc, inode, req);
-
- if (operation == CEPH_MDS_OP_GETFILELOCK) {
+ if (!err && operation == CEPH_MDS_OP_GETFILELOCK) {
fl->fl_pid =
-le64_to_cpu(req->r_reply_info.filelock_reply->pid);
if (CEPH_LOCK_SHARED == req->r_reply_info.filelock_reply->type)
fl->fl_type = F_RDLCK;
