linux-block.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Omar Sandoval <osandov@osandov.com>
To: Ming Lei <ming.lei@redhat.com>
Cc: Jens Axboe <axboe@fb.com>,
	linux-block@vger.kernel.org,
	Christoph Hellwig <hch@infradead.org>,
	Omar Sandoval <osandov@fb.com>
Subject: Re: [PATCH 3/4] blk-mq: use hw tag for scheduling if hw tag space is big enough
Date: Wed, 3 May 2017 10:00:04 -0700	[thread overview]
Message-ID: <20170503170004.GA10971@vader> (raw)
In-Reply-To: <20170503165529.GC9706@ming.t460p>

On Thu, May 04, 2017 at 12:55:30AM +0800, Ming Lei wrote:
> On Wed, May 03, 2017 at 09:29:36AM -0700, Omar Sandoval wrote:
> > On Fri, Apr 28, 2017 at 11:15:38PM +0800, Ming Lei wrote:
> > > When tag space of one device is big enough, we use hw tag
> > > directly for I/O scheduling.
> > > 
> > > Now the decision is made if hw queue depth is not less than
> > > q->nr_requests and the tag set isn't shared.
> > > 
> > > Signed-off-by: Ming Lei <ming.lei@redhat.com>
> > > ---
> > >  block/blk-mq-sched.c |  8 ++++++++
> > >  block/blk-mq-sched.h | 15 +++++++++++++++
> > >  block/blk-mq.c       | 18 +++++++++++++++++-
> > >  3 files changed, 40 insertions(+), 1 deletion(-)
> > > 
> > > diff --git a/block/blk-mq-sched.c b/block/blk-mq-sched.c
> > > index 45a675f07b8b..4681e27c127e 100644
> > > --- a/block/blk-mq-sched.c
> > > +++ b/block/blk-mq-sched.c
> > > @@ -507,6 +507,7 @@ int blk_mq_init_sched(struct request_queue *q, struct elevator_type *e)
> > >  	struct elevator_queue *eq;
> > >  	unsigned int i;
> > >  	int ret;
> > > +	bool auto_hw_tag;
> > >  
> > >  	if (!e) {
> > >  		q->elevator = NULL;
> > > @@ -519,7 +520,14 @@ int blk_mq_init_sched(struct request_queue *q, struct elevator_type *e)
> > >  	 */
> > >  	q->nr_requests = 2 * BLKDEV_MAX_RQ;
> > >  
> > > +	auto_hw_tag = blk_mq_sched_may_use_hw_tag(q);
> > > +
> > >  	queue_for_each_hw_ctx(q, hctx, i) {
> > > +		if (auto_hw_tag)
> > > +			hctx->flags |= BLK_MQ_F_SCHED_USE_HW_TAG;
> > > +		else
> > > +			hctx->flags &= ~BLK_MQ_F_SCHED_USE_HW_TAG;
> > > +
> > >  		ret = blk_mq_sched_alloc_tags(q, hctx, i);
> > >  		if (ret)
> > >  			goto err;
> > 
> > I think you should also clear the BLK_MQ_F_SCHED_USE_HW_TAG flag in
> > blk_mq_exit_sched()?
> 
> Looks not necessary since the flag is always evaluated in
> blk_mq_init_sched().

What if we're setting the scheduler to "none"? Then blk_mq_init_sched()
will go in here:

if (!e) {
	q->elevator = NULL;
	return 0;
}

  reply	other threads:[~2017-05-03 17:00 UTC|newest]

Thread overview: 55+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2017-04-28 15:15 [PATCH 0/4] blk-mq: support to use hw tag for scheduling Ming Lei
2017-04-28 15:15 ` [PATCH 1/4] blk-mq: introduce BLK_MQ_F_SCHED_USE_HW_TAG Ming Lei
2017-05-03 16:21   ` Omar Sandoval
2017-05-03 16:46   ` Omar Sandoval
2017-05-03 20:13     ` Ming Lei
2017-05-03 21:40       ` Omar Sandoval
2017-05-04  2:01         ` Ming Lei
2017-05-04  2:13           ` Jens Axboe
2017-05-04  2:51             ` Ming Lei
2017-05-04 14:06               ` Jens Axboe
2017-05-05 22:54                 ` Ming Lei
2017-05-05 23:33                   ` Ming Lei
2017-05-10  7:25                 ` Ming Lei
2017-04-28 15:15 ` [PATCH 2/4] blk-mq: introduce blk_mq_get_queue_depth() Ming Lei
2017-04-28 18:23   ` Jens Axboe
2017-04-29  9:55     ` Ming Lei
2017-05-03 16:55   ` Omar Sandoval
2017-05-04  2:10     ` Ming Lei
2017-04-28 15:15 ` [PATCH 3/4] blk-mq: use hw tag for scheduling if hw tag space is big enough Ming Lei
2017-04-28 18:09   ` Bart Van Assche
2017-04-29 10:35     ` Ming Lei
2017-05-01 15:06       ` Bart Van Assche
2017-05-02  3:49         ` Omar Sandoval
2017-05-02  8:46         ` Ming Lei
2017-04-28 18:22   ` Jens Axboe
2017-04-28 20:11     ` Bart Van Assche
2017-04-29 10:59     ` Ming Lei
2017-05-03 16:29   ` Omar Sandoval
2017-05-03 16:55     ` Ming Lei
2017-05-03 17:00       ` Omar Sandoval [this message]
2017-05-03 17:33         ` Ming Lei
2017-04-28 15:15 ` [PATCH 4/4] blk-mq: dump new introduced flag of BLK_MQ_F_SCHED_USE_HW_TAG Ming Lei
2017-04-28 18:10   ` Bart Van Assche
2017-04-29 11:00     ` Ming Lei
2017-04-28 20:29 ` [PATCH 0/4] blk-mq: support to use hw tag for scheduling Jens Axboe
2017-05-03  4:03   ` Ming Lei
2017-05-03 14:08     ` Jens Axboe
2017-05-03 14:10       ` Jens Axboe
2017-05-03 15:03         ` Ming Lei
2017-05-03 15:08           ` Jens Axboe
2017-05-03 15:38             ` Ming Lei
2017-05-03 16:06               ` Omar Sandoval
2017-05-03 16:21                 ` Ming Lei
2017-05-03 16:52               ` Ming Lei
2017-05-03 17:03                 ` Ming Lei
2017-05-03 17:07                   ` Jens Axboe
2017-05-03 17:15                     ` Bart Van Assche
2017-05-03 17:24                       ` Jens Axboe
2017-05-03 17:35                         ` Bart Van Assche
2017-05-03 17:40                           ` Jens Axboe
2017-05-03 17:43                             ` Bart Van Assche
2017-05-03 17:08                 ` Bart Van Assche
2017-05-03 17:11                   ` Jens Axboe
2017-05-03 17:19                   ` Ming Lei
2017-05-03 17:41                     ` Bart Van Assche

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20170503170004.GA10971@vader \
    --to=osandov@osandov.com \
    --cc=axboe@fb.com \
    --cc=hch@infradead.org \
    --cc=linux-block@vger.kernel.org \
    --cc=ming.lei@redhat.com \
    --cc=osandov@fb.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).