From: Andreas Gruenbacher <agruenba@redhat.com>
Date: Mon, 23 Jun 2025 20:20:55 +0000 (+0200)
Subject: gfs2: a minor finish_xmote cleanup
X-Git-Url: http://git.ipfire.org/cgi-bin/gitweb.cgi?a=commitdiff_plain;h=e7ffc0af0e722b4634f997dcf63594b3c65a5884;p=thirdparty%2Fkernel%2Fstable.git

gfs2: a minor finish_xmote cleanup

As a minor clean-up to commit 1fc05c8d8426 ("gfs2: cancel timed-out
glock requests"), when a demote request is in progress in
finish_xmote(), there is no point in waking up the glock holder at the
head of the queue because the reply from dlm cannot be on behalf of that
glock holder.

Signed-off-by: Andreas Gruenbacher <agruenba@redhat.com>
Reviewed-by: Andrew Price <anprice@redhat.com>
---

diff --git a/fs/gfs2/glock.c b/fs/gfs2/glock.c
index 053449a42b496..b6fd1cb17de7b 100644
--- a/fs/gfs2/glock.c
+++ b/fs/gfs2/glock.c
@@ -609,12 +609,11 @@ static void finish_xmote(struct gfs2_glock *gl, unsigned int ret)
 	if (unlikely(gl->gl_state != gl->gl_target)) {
 		struct gfs2_holder *gh = find_first_waiter(gl);
 
-		if (gh && (ret & LM_OUT_CANCELED))
-			gfs2_holder_wake(gh);
 		if (gh && !test_bit(GLF_DEMOTE_IN_PROGRESS, &gl->gl_flags)) {
 			if (ret & LM_OUT_CANCELED) {
 				list_del_init(&gh->gh_list);
 				trace_gfs2_glock_queue(gh, 0);
+				gfs2_holder_wake(gh);
 				gl->gl_target = gl->gl_state;
 				goto out;
 			}