]> err.no Git - linux-2.6/commitdiff
[PATCH] md: possible fix for unplug problem
authorNeilBrown <neilb@suse.de>
Mon, 10 Jul 2006 11:44:14 +0000 (04:44 -0700)
committerLinus Torvalds <torvalds@g5.osdl.org>
Mon, 10 Jul 2006 20:24:16 +0000 (13:24 -0700)
I have reports of a problem with raid5 which turns out to be because the raid5
device gets stuck in a 'plugged' state.  This shouldn't be able to happen as
3msec after it gets plugged it should get unplugged.  However it happens
none-the-less.  This patch fixes the problem and is a reasonable thing to do,
though it might hurt performance slightly in some cases.

Until I can find the real problem, we should probably have this workaround in
place.

Signed-off-by: Neil Brown <neilb@suse.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
drivers/md/raid5.c

index 7433871f4b3a45e866c6a7d2508bfdaaa28f30e4..5764387c69896fe4057d0cd5502d3c052629f0d6 100644 (file)
@@ -270,7 +270,7 @@ static struct stripe_head *get_active_stripe(raid5_conf_t *conf, sector_t sector
                                                     < (conf->max_nr_stripes *3/4)
                                                     || !conf->inactive_blocked),
                                                    conf->device_lock,
-                                                   unplug_slaves(conf->mddev)
+                                                   raid5_unplug_device(conf->mddev->queue)
                                        );
                                conf->inactive_blocked = 0;
                        } else