diff mbox series

[-next,7/8] md: use new helper to stop sync_thread in __md_stop_writes()

Message ID 20231110172834.3939490-8-yukuai1@huaweicloud.com
State New
Headers
Series md: bugfix and cleanup for sync_thread |

Commit Message

Yu Kuai Nov. 10, 2023, 5:28 p.m. UTC 
  From: Yu Kuai <yukuai3@huawei.com>

md_reap_sync_thread() should only be called when md_do_sync() is done,
for example, holding 'reconfig_mutex' to wait for md_do_sync() to be
done can deadlock(see details in commit 130443d60b1b ("md: refactor
idle/frozen_sync_thread() to fix deadlock")). Hence use new helper to
stop sync_thread.

Signed-off-by: Yu Kuai <yukuai3@huawei.com>
Signed-off-by: Yu Kuai <yukuai1@huaweicloud.com>
---
 drivers/md/md.c | 11 +++++------
 1 file changed, 5 insertions(+), 6 deletions(-)
diff mbox series

Patch

diff --git a/drivers/md/md.c b/drivers/md/md.c
index 7252fae0c989..35f3dd7db369 100644
--- a/drivers/md/md.c
+++ b/drivers/md/md.c
@@ -6325,12 +6325,11 @@  static void md_clean(struct mddev *mddev)
 
 static void __md_stop_writes(struct mddev *mddev)
 {
-	set_bit(MD_RECOVERY_FROZEN, &mddev->recovery);
-	if (work_pending(&mddev->sync_work))
-		flush_workqueue(md_misc_wq);
-	if (mddev->sync_thread) {
-		set_bit(MD_RECOVERY_INTR, &mddev->recovery);
-		md_reap_sync_thread(mddev);
+	if (test_bit(MD_RECOVERY_RUNNING, &mddev->recovery)) {
+		stop_sync_thread(mddev, true, false);
+		mddev_lock_nointr(mddev);
+	} else {
+		set_bit(MD_RECOVERY_FROZEN, &mddev->recovery);
 	}
 
 	del_timer_sync(&mddev->safemode_timer);