On Mon, Jun 13, 2022 at 10:23 PM Ming Lei <ming.lei@xxxxxxxxxx> wrote:
>
> On Mon, Jun 13, 2022 at 09:08:11PM +0800, Yi Zhang wrote:
> > Hi Ming
> >
> > The kmemleak also can be reproduced on 5.19.0-rc2, pls try to enable
> > nvme_core multipath and retest.
> >
> > # cat /sys/module/nvme_core/parameters/multipath
> > Y
> >
>
> OK, I can understand the reason now since rqos is only removed for blk-mq queue,
> then rqos allocated for bio queue is leaked, see disk_release_mq().
>
> The following patch should fix it:
Hi Ming
The kmemleak was fixed by this change, feel free to add
Tested-by: Yi Zhang <yi.zhang@xxxxxxxxxx>
>
> diff --git a/block/genhd.c b/block/genhd.c
> index 556d6e4b38d9..6e7ca8c302aa 100644
> --- a/block/genhd.c
> +++ b/block/genhd.c
> @@ -1120,9 +1120,10 @@ static const struct attribute_group *disk_attr_groups[] = {
> NULL
> };
>
> -static void disk_release_mq(struct request_queue *q)
> +static void disk_release_queue(struct request_queue *q)
> {
> - blk_mq_cancel_work_sync(q);
> + if (queue_is_mq(q))
> + blk_mq_cancel_work_sync(q);
>
> /*
> * There can't be any non non-passthrough bios in flight here, but
> @@ -1166,8 +1167,7 @@ static void disk_release(struct device *dev)
> might_sleep();
> WARN_ON_ONCE(disk_live(disk));
>
> - if (queue_is_mq(disk->queue))
> - disk_release_mq(disk->queue);
> + disk_release_queue(disk->queue);
>
> blkcg_exit_queue(disk->queue);
>
>
> Thanks,
> Ming
>
--
Best Regards,
Yi Zhang
