Re: [PATCH 1/5] cfq-iosched: Don't group_idle if cfqq has big thinktime

From	Jan Kara <jack@suse.cz>
Newsgroups	linux.kernel
Subject	Re: [PATCH 1/5] cfq-iosched: Don't group_idle if cfqq has big thinktime
Date	2016-01-11 16:50 +0100
Message-ID	<qPMTf-4l0-5@gated-at.bofh.it> (permalink)
References	<qOkFI-1wu-11@gated-at.bofh.it> <qOkFJ-1wu-39@gated-at.bofh.it> <qOIyn-YW-27@gated-at.bofh.it>
Organization	linux.* mail to news gateway

Show all headers | View raw

On Fri 08-01-16 11:58:31, Tejun Heo wrote:
> Hello, Jan.
> 
> On Thu, Jan 07, 2016 at 04:28:12PM +0100, Jan Kara wrote:
> > @@ -2947,8 +2948,13 @@ static void cfq_arm_slice_timer(struct cfq_data *cfqd)
> >  		return;
> >  	}
> >  
> > -	/* There are other queues in the group, don't do group idle */
> > -	if (group_idle && cfqq->cfqg->nr_cfqq > 1)
> > +	/*
> > +	 * There are other queues in the group or this is the only group and
> > +	 * it has too big thinktime, don't do group idle.
> > +	 */
> > +	if (group_idle &&
> > +	    (cfqq->cfqg->nr_cfqq > 1 ||
> > +	     !cfq_io_thinktime_big(cfqd, &st->ttime, true)))
> >  		return;
> 
> Is the negation in front of the cfq_io_thinktime_big() right?  That
> doesn't seem to match the comment or description.  What am I missing?

Ah, you are right! I wanted to make arming of group_idle timer consistent
with what we test in cfq_should_idle() for normal idle timer but I
accidentally reverted the condition. I'll fix this.

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

Back to linux.kernel | Previous | Next — Previous in thread | Find similar | Unroll thread

Thread

[PATCH 1/5] cfq-iosched: Don't group_idle if cfqq has big thinktime Jan Kara <jack@suse.cz> - 2016-01-07 16:30 +0100
  Re: [PATCH 1/5] cfq-iosched: Don't group_idle if cfqq has big  thinktime Tejun Heo <tj@kernel.org> - 2016-01-08 18:00 +0100
    Re: [PATCH 1/5] cfq-iosched: Don't group_idle if cfqq has big  thinktime Jan Kara <jack@suse.cz> - 2016-01-11 16:50 +0100

csiph-web