Re: [PATCH 1/1] MD/DM RAID: Fix hang due to recent RAID5 locking changes
From: NeilBrown <hidden>
Date: 2013-11-25 00:03:15
Also in:
dm-devel
On Sun, 24 Nov 2013 17:30:43 -0600 Jonathan Brassow [off-list ref] wrote:
When commit 773ca82 was made in v3.12-rc1, it caused RAID4/5/6 devices that were created via device-mapper (dm-raid.c) to hang on creation. This is not necessarily the fault of that commit, but perhaps the way dm-raid.c was setting-up and activating devices. Device-mapper allows I/O and memory allocations in the constructor (i.e. raid_ctr()), but nominal and recovery I/O should not be allowed until a 'resume' is issued (i.e. raid_resume()). It has been problematic (at least in the past) to call mddev_resume before mddev_suspend was called, but this is how DM behaves - CTR then resume. To solve the problem, raid_ctr() was setting up the structures, calling md_run(), and then also calling mddev_suspend(). The stage was then set for raid_resume() to call mddev_resume(). Commit 773ca82 caused a change in behavior during raid5.c:run(). 'setup_conf->grow_stripes->grow_one_stripe' is called which creates the stripe cache and increments 'active_stripes'. 'grow_one_stripe->release_stripe' doesn't actually decrement 'active_stripes' anymore. The side effect of this is that when raid_ctr calls mddev_suspend, it waits for 'active_stripes' to reduce to 0 - which never happens.
Hi Jon,
this sounds like the same bug that is fixed by
commit ad4068de49862b083ac2a15bc50689bb30ce3e44
Author: majianpeng [off-list ref]
Date: Thu Nov 14 15:16:15 2013 +1100
raid5: Use slow_path to release stripe when mddev->thread is null
which is already en-route to 3.12.x. Could you check if it fixes the bug for
you?
Thanks,
NeilBrown
quoted hunk ↗ jump to hunk
You could argue that the MD personalities should be able to handle either a suspend or a resume after 'md_run' is called, but it can't really handle either. To fix this, I've removed the call to mddev_suspend in raid_ctr and I've made the call to the personality's 'quiesce' function within mddev_resume dependent on whether the device is currently suspended. This patch is suitable and recommended for 3.12. Signed-off-by: Jonathan Brassow <redacted> --- drivers/md/dm-raid.c | 1 - drivers/md/md.c | 5 ++++- 2 files changed, 4 insertions(+), 2 deletions(-)diff --git a/drivers/md/dm-raid.c b/drivers/md/dm-raid.c index 4880b69..cdad87c 100644 --- a/drivers/md/dm-raid.c +++ b/drivers/md/dm-raid.c@@ -1249,7 +1249,6 @@ static int raid_ctr(struct dm_target *ti, unsigned argc, char **argv) rs->callbacks.congested_fn = raid_is_congested; dm_table_add_target_callbacks(ti->table, &rs->callbacks); - mddev_suspend(&rs->md); return 0; size_mismatch:diff --git a/drivers/md/md.c b/drivers/md/md.c index 561a65f..383980d 100644 --- a/drivers/md/md.c +++ b/drivers/md/md.c@@ -359,9 +359,12 @@ EXPORT_SYMBOL_GPL(mddev_suspend); void mddev_resume(struct mddev *mddev) { + int should_quiesce = mddev->suspended; + mddev->suspended = 0; wake_up(&mddev->sb_wait); - mddev->pers->quiesce(mddev, 0); + if (should_quiesce) + mddev->pers->quiesce(mddev, 0); set_bit(MD_RECOVERY_NEEDED, &mddev->recovery); md_wakeup_thread(mddev->thread);
Attachments
- signature.asc [application/pgp-signature] 828 bytes