[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-Id: <20230628012931.88911-3-yukuai1@huaweicloud.com>
Date: Wed, 28 Jun 2023 09:29:31 +0800
From: Yu Kuai <yukuai1@...weicloud.com>
To: xni@...hat.com, song@...nel.org, colyli@...e.de
Cc: linux-raid@...r.kernel.org, linux-kernel@...r.kernel.org,
yukuai3@...wei.com, yukuai1@...weicloud.com, yi.zhang@...wei.com,
yangerkun@...wei.com
Subject: [PATCH -next 2/2] md: restore 'noio_flag' for the last mddev_resume()
From: Yu Kuai <yukuai3@...wei.com>
memalloc_noio_save() is called for the first mddev_suspend(), and
repeated mddev_suspend() only increase 'suspended'. However,
memalloc_noio_restore() is also called for the first mddev_resume(),
which means that memory reclaim will be enabled before the last
mddev_resume() is called, while the array is still suspended.
Fix this problem by restore 'noio_flag' for the last mddev_resume().
Fixes: 78f57ef9d50a ("md: use memalloc scope APIs in mddev_suspend()/mddev_resume()")
Signed-off-by: Yu Kuai <yukuai3@...wei.com>
---
drivers/md/md.c | 6 ++++--
1 file changed, 4 insertions(+), 2 deletions(-)
diff --git a/drivers/md/md.c b/drivers/md/md.c
index 962dacfd98cf..a3d98273b295 100644
--- a/drivers/md/md.c
+++ b/drivers/md/md.c
@@ -464,11 +464,13 @@ EXPORT_SYMBOL_GPL(mddev_suspend);
void mddev_resume(struct mddev *mddev)
{
- /* entred the memalloc scope from mddev_suspend() */
- memalloc_noio_restore(mddev->noio_flag);
lockdep_assert_held(&mddev->reconfig_mutex);
if (--mddev->suspended)
return;
+
+ /* entred the memalloc scope from mddev_suspend() */
+ memalloc_noio_restore(mddev->noio_flag);
+
percpu_ref_resurrect(&mddev->active_io);
wake_up(&mddev->sb_wait);
--
2.39.2
Powered by blists - more mailing lists