Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
From:   Yu Kuai <yukuai1@huaweicloud.com>
To:     song@kernel.org, xni@redhat.com
Cc:     linux-raid@vger.kernel.org, linux-kernel@vger.kernel.org,
        yukuai3@huawei.com, yukuai1@huaweicloud.com, yi.zhang@huawei.com,
        yangerkun@huawei.com
Subject: [PATCH -next 2/2] md: fix potential hang for mddev_suspend()
Date:   Wed, 30 Aug 2023 17:29:02 +0800
Message-Id: <20230830092902.1236950-3-yukuai1@huaweicloud.com>
In-Reply-To: <20230830092902.1236950-1-yukuai1@huaweicloud.com>
References: <20230830092902.1236950-1-yukuai1@huaweicloud.com>
MIME-Version: 1.0
Content-Transfer-Encoding: 8bit
Precedence: bulk

From: Yu Kuai <yukuai3@huawei.com>

Commit 72adae23a72c ("md: Change active_io to percpu") drop that if
'active_io' is decreased to 0 and array is suspended, 'sb_wait' will be
woke up. This is wrong, however, there is no regression reported and I
think this is probably because 'sb_wait' is used in many scenarios and
it's woke up from other context.

Anyway, fix this potential problem by waking up 'sb_wait' in this case.

Fixes: 72adae23a72c ("md: Change active_io to percpu")
Signed-off-by: Yu Kuai <yukuai3@huawei.com>
---
 drivers/md/md.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/drivers/md/md.c b/drivers/md/md.c
index 0d69b1a2e2d5..1c7eb3cfadb4 100644
--- a/drivers/md/md.c
+++ b/drivers/md/md.c
@@ -399,6 +399,9 @@ static bool md_array_enter(struct mddev *mddev, struct bio *bio)
 static void md_array_exit(struct mddev *mddev)
 {
 	percpu_ref_put(&mddev->active_io);
+	if (percpu_ref_is_zero(&mddev->active_io) &&
+	    wq_has_sleeper(&mddev->sb_wait))
+		wake_up(&mddev->sb_wait);
 }
 
 void md_handle_request(struct mddev *mddev, struct bio *bio)
-- 
2.39.2