[PATCH 52/72] GFS2: Fix unlock of fcntl locks during withdrawn state

Luis Henriques luis.henriques at canonical.com
Thu Apr 18 09:16:37 UTC 2013 -stable review patch.  If anyone has any objections, please let me know.


From: Steven Whitehouse <swhiteho at redhat.com>

commit c2952d202f710d326ac36a8ea6bd216b20615ec8 upstream.

When withdraw occurs, we need to continue to allow unlocks of fcntl
locks to occur, however these will only be local, since the node has
withdrawn from the cluster. This prevents triggering a VFS level
bug trap due to locks remaining when a file is closed.

Signed-off-by: Steven Whitehouse <swhiteho at redhat.com>
Signed-off-by: Luis Henriques <luis.henriques at canonical.com>
 fs/gfs2/file.c | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/fs/gfs2/file.c b/fs/gfs2/file.c
index a8a81d7..e756a62 100644
--- a/fs/gfs2/file.c
+++ b/fs/gfs2/file.c
@@ -900,8 +900,11 @@ static int gfs2_lock(struct file *file, int cmd, struct file_lock *fl)
 		cmd = F_SETLK;
 		fl->fl_type = F_UNLCK;
-	if (unlikely(test_bit(SDF_SHUTDOWN, &sdp->sd_flags)))
+	if (unlikely(test_bit(SDF_SHUTDOWN, &sdp->sd_flags))) {
+		if (fl->fl_type == F_UNLCK)
+			posix_lock_file_wait(file, fl);
 		return -EIO;
+	}
 	if (IS_GETLK(cmd))
 		return dlm_posix_get(ls->ls_dlm, ip->i_no_addr, file, fl);
 	else if (fl->fl_type == F_UNLCK)

More information about the kernel-team mailing list