On Thu, Feb 17, 2022 at 06:34:02AM -0700, Jens Axboe wrote:
> On 2/17/22 12:48 AM, Christoph Hellwig wrote:
> > On Wed, Feb 16, 2022 at 07:32:12PM +0800, Wang Jianchao (Kuaishou) wrote:
> >> From: Wang Jianchao <wangjianchao@xxxxxxxxxxxx>
> >>
> >> When __alloc_disk_node() failed, there will not not del_gendisk()
> >> any more, then resource in rqos policies is leaked. Add rq_qos_exit()
> >> into blk_cleanup_queue(). rqos is removed from the list, so needn't
> >> to worry .exit is called twice.
> >>
> >> Fixes: commit 8e141f9eb803 ("block: drain file system I/O on del_gendisk")
> >> Suggested-by: Bart Van Assche <bart.vanassche@xxxxxxx>
> >> Signed-off-by: Wang Jianchao <wangjianchao@xxxxxxxxxxxx>
> >
> > Ming had a pending patch to move it into disk_release instead, which
> > I think is the right place.
>
> I missed that patch and can't seem to find it, do you have a link?
[PATCH V2 12/13] block: move rq_qos_exit() into disk_release()
from Jan 22. Although it would need a rebase so it can be applied
without the preceding patches.
