[v2] md raid: fix hang when stopping arrays with metadata through dm-raid

[PATCH V2] md raid: fix hang when stopping arrays with metadata through dm-raid

Posted by Heinz Mauelshagen 4 months, 2 weeks ago

When using device-mapper's dm-raid target, stopping a RAID array can cause the
system to hang under specific conditions.

This occurs when:

- A dm-raid managed device tree is suspended from top to bottom
   (the top-level RAID device is suspended first, followed by its
    underlying metadata and data devices)

- The top-level RAID device is then removed

Removing the top-level device triggers a hang in the following sequence: the dm-raid
destructor calls md_stop(), which tries to flush the write-intent bitmap by writing
to the metadata sub-devices. However, these devices are already suspended, making
them unable to complete the write operations and causing an indefinite block.

Fix:

- Prevent bitmap flushing when md_stop() is called from dm-raid destructor context
  and avoid a quiescing/unquescing cycle which could also cause I/O

- Still allow write-intent bitmap flushing when called from dm-raid suspend context

This ensures that RAID array teardown can complete successfully even when the
underlying devices are in a suspended state.

This second patch uses md_is_rdwr() to distinguish between suspend and
destructor paths as elaborated on above.

Signed-off-by: Heinz Mauelshagen <heinzm@redhat.com>
---
 drivers/md/md.c | 12 +++++++-----
 1 file changed, 7 insertions(+), 5 deletions(-)

diff --git a/drivers/md/md.c b/drivers/md/md.c
index 4e033c26fdd4..78408d2f78fc 100644
--- a/drivers/md/md.c
+++ b/drivers/md/md.c
@@ -6541,12 +6541,14 @@ static void __md_stop_writes(struct mddev *mddev)
 {
 	timer_delete_sync(&mddev->safemode_timer);
 
-	if (mddev->pers && mddev->pers->quiesce) {
-		mddev->pers->quiesce(mddev, 1);
-		mddev->pers->quiesce(mddev, 0);
-	}
+	if (md_is_rdwr(mddev) || !mddev_is_dm(mddev)) {
+		if (mddev->pers && mddev->pers->quiesce) {
+			mddev->pers->quiesce(mddev, 1);
+			mddev->pers->quiesce(mddev, 0);
+		}
 
-	mddev->bitmap_ops->flush(mddev);
+		mddev->bitmap_ops->flush(mddev);
+	}
 
 	if (md_is_rdwr(mddev) &&
 	    ((!mddev->in_sync && !mddev_is_clustered(mddev)) ||
-- 
2.51.0

Fwd: [PATCH V2] md raid: fix hang when stopping arrays with metadata through dm-raid

Posted by Heinz Mauelshagen 3 weeks, 3 days ago

Resending, as it seems to have gotten lost.

---------- Forwarded message ---------
From: Heinz Mauelshagen <heinzm@redhat.com>
Date: Tue, Sep 23, 2025 at 2:58 PM
Subject: [PATCH V2] md raid: fix hang when stopping arrays with
metadata through dm-raid
To: <yukuai1@huaweicloud.com>, <song@kernel.org>
Cc: <linux-raid@vger.kernel.org>, <linux-kernel@vger.kernel.org>,
Heinz Mauelshagen <heinzm@redhat.com>


When using device-mapper's dm-raid target, stopping a RAID array can cause the
system to hang under specific conditions.

This occurs when:

- A dm-raid managed device tree is suspended from top to bottom
   (the top-level RAID device is suspended first, followed by its
    underlying metadata and data devices)

- The top-level RAID device is then removed

Removing the top-level device triggers a hang in the following
sequence: the dm-raid
destructor calls md_stop(), which tries to flush the write-intent
bitmap by writing
to the metadata sub-devices. However, these devices are already
suspended, making
them unable to complete the write operations and causing an indefinite block.

Fix:

- Prevent bitmap flushing when md_stop() is called from dm-raid
destructor context
  and avoid a quiescing/unquescing cycle which could also cause I/O

- Still allow write-intent bitmap flushing when called from dm-raid
suspend context

This ensures that RAID array teardown can complete successfully even when the
underlying devices are in a suspended state.

This second patch uses md_is_rdwr() to distinguish between suspend and
destructor paths as elaborated on above.

Signed-off-by: Heinz Mauelshagen <heinzm@redhat.com>
---
 drivers/md/md.c | 12 +++++++-----
 1 file changed, 7 insertions(+), 5 deletions(-)

diff --git a/drivers/md/md.c b/drivers/md/md.c
index 4e033c26fdd4..78408d2f78fc 100644
--- a/drivers/md/md.c
+++ b/drivers/md/md.c
@@ -6541,12 +6541,14 @@ static void __md_stop_writes(struct mddev *mddev)
 {
        timer_delete_sync(&mddev->safemode_timer);

-       if (mddev->pers && mddev->pers->quiesce) {
-               mddev->pers->quiesce(mddev, 1);
-               mddev->pers->quiesce(mddev, 0);
-       }
+       if (md_is_rdwr(mddev) || !mddev_is_dm(mddev)) {
+               if (mddev->pers && mddev->pers->quiesce) {
+                       mddev->pers->quiesce(mddev, 1);
+                       mddev->pers->quiesce(mddev, 0);
+               }

-       mddev->bitmap_ops->flush(mddev);
+               mddev->bitmap_ops->flush(mddev);
+       }

        if (md_is_rdwr(mddev) &&
            ((!mddev->in_sync && !mddev_is_clustered(mddev)) ||
--
2.51.0

Re: Fwd: [PATCH V2] md raid: fix hang when stopping arrays with metadata through dm-raid

Posted by Yu Kuai 1 week, 5 days ago

在 2026/1/15 1:52, Heinz Mauelshagen 写道:

> hen using device-mapper's dm-raid target, stopping a RAID array can cause the
> system to hang under specific conditions.
>
> This occurs when:
>
> - A dm-raid managed device tree is suspended from top to bottom
>     (the top-level RAID device is suspended first, followed by its
>      underlying metadata and data devices)
>
> - The top-level RAID device is then removed
>
> Removing the top-level device triggers a hang in the following
> sequence: the dm-raid
> destructor calls md_stop(), which tries to flush the write-intent
> bitmap by writing
> to the metadata sub-devices. However, these devices are already
> suspended, making
> them unable to complete the write operations and causing an indefinite block.
>
> Fix:
>
> - Prevent bitmap flushing when md_stop() is called from dm-raid
> destructor context
>    and avoid a quiescing/unquescing cycle which could also cause I/O
>
> - Still allow write-intent bitmap flushing when called from dm-raid
> suspend context
>
> This ensures that RAID array teardown can complete successfully even when the
> underlying devices are in a suspended state.
>
> This second patch uses md_is_rdwr() to distinguish between suspend and
> destructor paths as elaborated on above.
>
> Signed-off-by: Heinz Mauelshagen<heinzm@redhat.com>
> ---
>   drivers/md/md.c | 12 +++++++-----
>   1 file changed, 7 insertions(+), 5 deletions(-)

There are conflicts now, applied with some rebasing:

diff --git a/drivers/md/md.c b/drivers/md/md.c
index 606f616190d7..59cd303548de 100644
--- a/drivers/md/md.c
+++ b/drivers/md/md.c
@@ -6851,13 +6851,15 @@ static void __md_stop_writes(struct mddev *mddev)
  {
         timer_delete_sync(&mddev->safemode_timer);

-       if (mddev->pers && mddev->pers->quiesce) {
-               mddev->pers->quiesce(mddev, 1);
-               mddev->pers->quiesce(mddev, 0);
-       }
+       if (md_is_rdwr(mddev) || !mddev_is_dm(mddev)) {
+               if (mddev->pers && mddev->pers->quiesce) {
+                       mddev->pers->quiesce(mddev, 1);
+                       mddev->pers->quiesce(mddev, 0);
+               }

-       if (md_bitmap_enabled(mddev, true))
-               mddev->bitmap_ops->flush(mddev);
+               if (md_bitmap_enabled(mddev, true))
+                       mddev->bitmap_ops->flush(mddev);
+       }

         if (md_is_rdwr(mddev) &&
             ((!mddev->in_sync && !mddev_is_clustered(mddev)) ||

-- 
Thansk,
Kuai