Re: [PATCH] blk-mq: streamline handling of q->mq_ops->queue_rq result

[Ming Lei <ming.lei@redhat.com>]

On Wed, Jul 01, 2020 at 01:54:09PM +0100, Christoph Hellwig wrote:
> On Wed, Jul 01, 2020 at 06:16:17PM +0800, Ming Lei wrote:
> > Current handling of q->mq_ops->queue_rq result is a bit ugly:
> > 
> > - two branches which needs to 'continue' have to check if the
> > dispatch local list is empty, otherwise one bad request may
> > be retrieved via 'rq = list_first_entry(list, struct request, queuelist);'
> > 
> > - the branch of 'if (unlikely(ret != BLK_STS_OK))' isn't easy
> > to follow, since it is actually one error branch.
> > 
> > Streamline this handling, so the code becomes more readable, meantime
> > potential kernel oops can be avoided in case that the last request in
> > local dispatch list is failed.
> 
> I don't really find that much easier to read.  If we want to clean
> this up for rea we should use a proper switch statement.  Something like
> this:
> 
> diff --git a/block/blk-mq.c b/block/blk-mq.c
> index a9aa6d1e44cf32..f3721f274b800e 100644
> --- a/block/blk-mq.c
> +++ b/block/blk-mq.c
> @@ -1275,30 +1275,28 @@ bool blk_mq_dispatch_rq_list(struct request_queue *q, struct list_head *list,
>  		}
>  
>  		ret = q->mq_ops->queue_rq(hctx, &bd);
> -		if (ret == BLK_STS_RESOURCE || ret == BLK_STS_DEV_RESOURCE) {
> -			blk_mq_handle_dev_resource(rq, list);
> +		switch (ret) {
> +		case BLK_STS_OK:
> +			queued++;
>  			break;
> -		} else if (ret == BLK_STS_ZONE_RESOURCE) {
> +		case BLK_STS_RESOURCE:
> +		case BLK_STS_DEV_RESOURCE:
> +			blk_mq_handle_dev_resource(rq, list);
> +			goto out;
> +		case BLK_STS_ZONE_RESOURCE:
>  			/*
>  			 * Move the request to zone_list and keep going through
>  			 * the dispatch list to find more requests the drive can
>  			 * accept.
>  			 */
>  			blk_mq_handle_zone_resource(rq, &zone_list);
> -			if (list_empty(list))
> -				break;
> -			continue;
> -		}
> -
> -		if (unlikely(ret != BLK_STS_OK)) {
> +			break;
> +		default:
>  			errors++;
>  			blk_mq_end_request(rq, BLK_STS_IOERR);
> -			continue;
>  		}
> -
> -		queued++;
>  	} while (!list_empty(list));
> -
> +out:
>  	if (!list_empty(&zone_list))
>  		list_splice_tail_init(&zone_list, list);

I am fine to switch back to 'switch'. I doesn't take 'switch' because you
changed 'switch' to 'if else' before.


Thanks,
Ming