On Sat, 2017-09-16 at 00:44 +0800, Ming Lei wrote:
> +static void save_path_queue_depth(struct pgpath *p)
> +{
> + struct request_queue *q = bdev_get_queue(p->path.dev->bdev);
> +
> + p->old_nr_requests = q->nr_requests;
> + p->queue_depth = q->queue_depth;
> +
> + /* one extra request for making the pipeline full */
> + if (p->queue_depth)
> + blk_update_nr_requests(q, p->queue_depth + 1);
> +}
blk_mq_init_allocated_queue() initializes nr_requests to the tag set queue depth.
Does that mean that the above code increases nr_requests by one? If so, does
that change only result in a performance improvement for the queue depth
mentioned in the path description (3)? Sorry but I doubt that this change will
yield a significant improvement for higher queue depths. Does that mean that
this patch can be left out?
Thanks,
Bart.
