Re: [PATCH 6/8] loop: don't freeze the queue in lo_release

[Jan Kara <jack@suse.cz>]

On Wed 26-01-22 16:50:38, Christoph Hellwig wrote:
> By the time the final ->release is called there can't be outstanding I/O.
> For non-final ->release there is no need for driver action at all.
> 
> Thus remove the useless queue freeze.
> 
> Signed-off-by: Christoph Hellwig <hch@lst.de>

Looks good, just one nit below. Feel free to add:

Reviewed-by: Jan Kara <jack@suse.cz>

> diff --git a/drivers/block/loop.c b/drivers/block/loop.c
> index 4b0058a67c48e..d9a0e2205762f 100644
> --- a/drivers/block/loop.c
> +++ b/drivers/block/loop.c
> @@ -1758,13 +1758,6 @@ static void lo_release(struct gendisk *disk, fmode_t mode)
>  		 */
>  		loop_schedule_rundown(lo);
>  		return;
> -	} else if (lo->lo_state == Lo_bound) {
> -		/*
> -		 * Otherwise keep thread (if running) and config,
> -		 * but flush possible ongoing bios in thread.
> -		 */
> -		blk_mq_freeze_queue(lo->lo_queue);
> -		blk_mq_unfreeze_queue(lo->lo_queue);
>  	}

Maybe worth a comment like:

	/*
	 * No IO can be running at this point since there are no openers
	 * (covers filesystems, stacked devices, AIO) and the page cache is
	 * evicted.
	 */

								Honza
-- 
Jan Kara <jack@suse.com>
SUSE Labs, CR