From mboxrd@z Thu Jan 1 00:00:00 1970 From: NeilBrown Subject: [md PATCH 5/5] md: allow metadata update while suspending. Date: Tue, 17 Oct 2017 13:46:43 +1100 Message-ID: <150820840357.1646.7614695852624244815.stgit@noble> References: <150820826980.1646.6380214598725492144.stgit@noble> Mime-Version: 1.0 Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <150820826980.1646.6380214598725492144.stgit@noble> Sender: linux-raid-owner@vger.kernel.org To: Shaohua Li Cc: linux-raid@vger.kernel.org List-Id: linux-raid.ids There are various deadlocks that can occur when a thread holds reconfig_mutex and calls ->quiesce(mddev, 1). As some write request block waiting for metadata to be updated (e.g. to record device failure), and as the md thread updates the metadata while the reconfig mutex is held, holding the mutex can stop write requests completing, and this prevents ->quiesce(mddev, 1) from completing. ->quiesce() is now usually called from mddev_suspend(), and it is always called with reconfig_mutex held. So at this time it is safe for the thread to update metadata without explicitly taking the lock. So add 2 new flags, one which says the unlocked updates is allowed, and one which ways it is happening. Then allow it while the quiesce completes, and then wait for it to finish. Reported-and-tested-by: Xiao Ni Signed-off-by: NeilBrown --- drivers/md/md.c | 14 ++++++++++++++ drivers/md/md.h | 6 ++++++ 2 files changed, 20 insertions(+) diff --git a/drivers/md/md.c b/drivers/md/md.c index 60a5022bd9bb..63ecfb063b76 100644 --- a/drivers/md/md.c +++ b/drivers/md/md.c @@ -364,8 +364,12 @@ void mddev_suspend(struct mddev *mddev) return; synchronize_rcu(); wake_up(&mddev->sb_wait); + set_bit(MD_ALLOW_SB_UPDATE, &mddev->flags); + smp_mb__after_atomic(); wait_event(mddev->sb_wait, atomic_read(&mddev->active_io) == 0); mddev->pers->quiesce(mddev, 1); + clear_bit_unlock(MD_ALLOW_SB_UPDATE, &mddev->flags); + wait_event(mddev->sb_wait, !test_bit(MD_UPDATING_SB, &mddev->flags)); del_timer_sync(&mddev->safemode_timer); } @@ -8835,6 +8839,16 @@ void md_check_recovery(struct mddev *mddev) unlock: wake_up(&mddev->sb_wait); mddev_unlock(mddev); + } else if (test_bit(MD_ALLOW_SB_UPDATE, &mddev->flags) && mddev->sb_flags) { + /* Write superblock - thread that called mddev_suspend() + * holds reconfig_mutex for us. + */ + set_bit(MD_UPDATING_SB, &mddev->flags); + smp_mb__after_atomic(); + if (test_bit(MD_ALLOW_SB_UPDATE, &mddev->flags)) + md_update_sb(mddev, 0); + clear_bit_unlock(MD_UPDATING_SB, &mddev->flags); + wake_up(&mddev->sb_wait); } } EXPORT_SYMBOL(md_check_recovery); diff --git a/drivers/md/md.h b/drivers/md/md.h index d8287d3cd1bf..03fc641e5da1 100644 --- a/drivers/md/md.h +++ b/drivers/md/md.h @@ -237,6 +237,12 @@ enum mddev_flags { */ MD_HAS_PPL, /* The raid array has PPL feature set */ MD_HAS_MULTIPLE_PPLS, /* The raid array has multiple PPLs feature set */ + MD_ALLOW_SB_UPDATE, /* md_check_recovery is allowed to update + * the metadata without taking reconfig_mutex. + */ + MD_UPDATING_SB, /* md_check_recovery is updating the metadata + * without explicitly holding reconfig_mutex. + */ }; enum mddev_sb_flags {