[Cluster-devel] [GFS2 PATCH 3/4] gfs2: Fix BUG during unmount after file system withdraw

[Bob Peterson]

Before this patch, when the logd daemon was forced to withdraw, it
would try to request its journal be recovered by another cluster node.
However, in single-user cases with lock_nolock, there are no other
nodes to recover the journal. Function signal_our_withdraw() was
recognizing the lock_nolock situation, but not until after it had
evicted its journal inode. Since the journal descriptor that points
to the inode was never removed from the master list, when the unmount
occurred, it did another iput on the evicted inode, which resulted in
a BUG from fs/inode.c (BUG_ON(inode->i_state & I_CLEAR);).

This patch moves the check for this situation earlier in function
signal_our_withdraw(), which avoids the extra iput, so the unmount
may happen normally.

Signed-off-by: Bob Peterson <rpeterso at redhat.com>
---
 fs/gfs2/util.c | 10 ++++++----
 1 file changed, 6 insertions(+), 4 deletions(-)

diff --git a/fs/gfs2/util.c b/fs/gfs2/util.c
index 9b64d40ab379..aa087a5675af 100644
--- a/fs/gfs2/util.c
+++ b/fs/gfs2/util.c
@@ -119,6 +119,12 @@ static void signal_our_withdraw(struct gfs2_sbd *sdp)
 	if (!sb_rdonly(sdp->sd_vfs))
 		ret = gfs2_make_fs_ro(sdp);
 
+	if (sdp->sd_lockstruct.ls_ops->lm_lock == NULL) { /* lock_nolock */
+		if (!ret)
+			ret = -EIO;
+		clear_bit(SDF_WITHDRAW_RECOVERY, &sdp->sd_flags);
+		goto skip_recovery;
+	}
 	/*
 	 * Drop the glock for our journal so another node can recover it.
 	 */
@@ -159,10 +165,6 @@ static void signal_our_withdraw(struct gfs2_sbd *sdp)
 		wait_on_bit(&gl->gl_flags, GLF_FREEING, TASK_UNINTERRUPTIBLE);
 	}
 
-	if (sdp->sd_lockstruct.ls_ops->lm_lock == NULL) { /* lock_nolock */
-		clear_bit(SDF_WITHDRAW_RECOVERY, &sdp->sd_flags);
-		goto skip_recovery;
-	}
 	/*
 	 * Dequeue the "live" glock, but keep a reference so it's never freed.
 	 */
-- 
2.25.1