commit ee5d004fd0591536a061451eba2b187092e9127c upstream.
The 'event_work' worker used by dm-raid may still be running
when the array is stopped. This can result in an oops.
So flush the workqueue on which it is run after detaching
and before destroying the device.
Reported-by: Heinz Mauelshagen <heinzm@xxxxxxxxxx>
Signed-off-by: NeilBrown <neilb@xxxxxxxx>
Cc: stable@xxxxxxxxxxxxxxx (2.6.38+ please delay 2 weeks after -final release)
Fixes: 9d09e663d550 ("dm: raid456 basic support")
---
Here is the patch rebased on 3.14 as requested.
This same patch applies to 3.10 too.
Thanks.
diff --git a/drivers/md/md.c b/drivers/md/md.c
index 4ad5cc4e63e8..4bb537a35428 100644
--- a/drivers/md/md.c
+++ b/drivers/md/md.c
@@ -5285,6 +5285,8 @@ EXPORT_SYMBOL_GPL(md_stop_writes);
static void __md_stop(struct mddev *mddev)
{
mddev->ready = 0;
+ /* Ensure ->event_work is done */
+ flush_workqueue(md_misc_wq);
mddev->pers->stop(mddev);
if (mddev->pers->sync_request && mddev->to_remove == NULL)
mddev->to_remove = &md_redundancy_group;
Attachment:
signature.asc
Description: PGP signature