[dm-devel] [PATCH v4 6/6] dm rq: Avoid that request processing stalls sporadically

[Mike Snitzer]

On Tue, Apr 11 2017 at  1:51pm -0400,
Bart Van Assche <Bart.VanAssche at sandisk.com> wrote:

> On Tue, 2017-04-11 at 13:47 -0400, Mike Snitzer wrote:
> > Other drivers will very likely be caught about by
> > this blk-mq quirk in the future.
> 
> Hello Mike,
> 
> Are you aware that the requirement that blk-mq drivers rerun the queue after
> having returned BLK_MQ_RQ_QUEUE_BUSY is a requirement that is shared with
> traditional block drivers? From dm_old_request_fn():
> 
> 	if (... || (ti->type->busy && ti->type->busy(ti))) {
> 		blk_delay_queue(q, 10);
> 		return;
> 	}

No, and pointing to DM code that does something with the old .request_fn
case to justify why blk-mq requires the same is pretty specious.

Rather than working so hard to use DM code against me, your argument
should be: "blk-mq drivers X, Y and Z rerun the hw queue; this is a well
established pattern"

I see drivers/nvme/host/fc.c:nvme_fc_start_fcp_op() does.  But that is
only one other driver out of ~20 BLK_MQ_RQ_QUEUE_BUSY returns
tree-wide.

Could be there are some others, but hardly a well-established pattern.