2022-03-04 00:45:04

by Eric Dumazet

[permalink] [raw]
Subject: [PATCH] md: use msleep() in md_notify_reboot()

From: Eric Dumazet <[email protected]>

Calling mdelay(1000) from process context, even while a reboot
is in progress, does not make sense.

Using msleep() allows other threads to make progress.

Signed-off-by: Eric Dumazet <[email protected]>
Cc: [email protected]
---
drivers/md/md.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/md/md.c b/drivers/md/md.c
index 4d38bd7dadd604acdc9832d62c81c913abddc07d..330c53e212b82dd58670962fc9204cf4eec484f4 100644
--- a/drivers/md/md.c
+++ b/drivers/md/md.c
@@ -9583,7 +9583,7 @@ static int md_notify_reboot(struct notifier_block *this,
* driver, we do want to have a safe RAID driver ...
*/
if (need_delay)
- mdelay(1000*1);
+ msleep(1000);

return NOTIFY_DONE;
}
--
2.35.1.616.g0bdcbb4464-goog


2022-03-04 09:13:59

by Song Liu

[permalink] [raw]
Subject: Re: [PATCH] md: use msleep() in md_notify_reboot()

On Thu, Mar 3, 2022 at 3:19 PM Eric Dumazet <[email protected]> wrote:
>
> From: Eric Dumazet <[email protected]>
>
> Calling mdelay(1000) from process context, even while a reboot
> is in progress, does not make sense.
>
> Using msleep() allows other threads to make progress.
>
> Signed-off-by: Eric Dumazet <[email protected]>
> Cc: [email protected]

Applied to md-next. Thanks!

Song