All of lore.kernel.org
 help / color / mirror / Atom feed
From: Peter Zijlstra <peterz@infradead.org>
To: Fabio Checconi <fchecconi@gmail.com>
Cc: Ingo Molnar <mingo@elte.hu>, Thomas Gleixner <tglx@linutronix.de>,
	Paul Turner <pjt@google.com>,
	Dario Faggioli <faggioli@gandalf.sssup.it>,
	Michael Trimarchi <michael@evidence.eu.com>,
	Dhaval Giani <dhaval@retis.sssup.it>,
	Tommaso Cucinotta <t.cucinotta@sssup.it>,
	linux-kernel@vger.kernel.org
Subject: Re: [PATCH 2/3] sched: enforce per-cpu utilization limits on runtime balancing
Date: Tue, 23 Mar 2010 21:33:58 +0100	[thread overview]
Message-ID: <1269376438.5283.8.camel@laptop> (raw)
In-Reply-To: <20100303170000.GO2490@gandalf.sssup.it>

On Wed, 2010-03-03 at 18:00 +0100, Fabio Checconi wrote:
> > From: Peter Zijlstra <peterz@infradead.org>
> > Date: Thu, Feb 25, 2010 09:28:25PM +0100
> >
> > On Tue, 2010-02-23 at 19:56 +0100, Fabio Checconi wrote:
> > > +static u64 from_ratio(unsigned long ratio, u64 period)
> > > +{
> > > +       return (ratio * period) >> 20;
> > > +}
> > > +
> > > +/*
> > > + * Try to move *diff units of runtime from src to dst, checking
> > > + * that the utilization does not exceed the global limits on the
> > > + * destination cpu.  Returns true if the migration succeeded, leaving
> > > + * in *diff the actual amount of runtime moved, false on failure, which
> > > + * means that no more bandwidth can be migrated to rt_rq.
> > > + */
> > > +static int rt_move_bw(struct rt_rq *src, struct rt_rq *dst,
> > > +                      s64 *diff, u64 rt_period)
> > > +{
> > > +       struct rq *rq = rq_of_rt_rq(dst), *src_rq = rq_of_rt_rq(src);
> > > +       struct rt_edf_tree *dtree = &rq->rt.rt_edf_tree;
> > > +       struct rt_edf_tree *stree = &src_rq->rt.rt_edf_tree;
> > > +       unsigned long bw_to_move;
> > > +       int ret = 0;
> > > +
> > > +       double_spin_lock(&dtree->rt_bw_lock, &stree->rt_bw_lock);
> > > +
> > > +       if (dtree->rt_free_bw) {
> > > +               bw_to_move = to_ratio(rt_period, *diff);
> > > +               if (bw_to_move > dtree->rt_free_bw) {
> > > +                       bw_to_move = dtree->rt_free_bw;
> > > +                       *diff = from_ratio(bw_to_move, rt_period);
> > > +               }
> > > +
> > > +               stree->rt_free_bw -= bw_to_move;
> > > +               dtree->rt_free_bw += bw_to_move;
> > > +               ret = 1;
> > > +       }
> > > +
> > > +       double_spin_unlock(&dtree->rt_bw_lock, &stree->rt_bw_lock);
> > > +
> > > +       return ret;
> > > +} 
> > 
> > The from_ratio() stuff smells like numerical instability for
> > ->rt_free_bw, I can't see anything that would, given sufficient balance
> > cycles keep the sum of rt_free_bw over the cpus equal to what it started
> > out with.
> 
> You're right...  What would you think about the following solution?
> It just keep tracks of the bw accounted for every rt_rq when it is
> updated, and that should be enough to avoid accumulating the errors.
> 
> static inline void rt_update_bw(struct rt_rq *rt_rq, struct rt_edf_tree *tree,
>                                 s64 diff, u64 rt_period)
> {       
> 	unsigned long bw;
> 
> 	rt_rq->rt_runtime += diff;
> 	bw = to_ratio(rt_period, rt_rq->rt_runtime);
> 	tree->rt_free_bw += bw - rt_rq->rt_bw;
> 	rt_rq->rt_bw = bw;
> }
>  
> static bool rt_move_bw(struct rt_rq *src, struct rt_rq *dst,
> 		       s64 *diff, u64 rt_period)
> {
> 	struct rq *rq = rq_of_rt_rq(dst), *src_rq = rq_of_rt_rq(src);
> 	struct rt_edf_tree *dtree = &rq->rt.rt_edf_tree;
> 	struct rt_edf_tree *stree = &src_rq->rt.rt_edf_tree;
> 	unsigned long bw_to_move;
> 	bool ret = false;
> 
> 	double_spin_lock(&dtree->rt_bw_lock, &stree->rt_bw_lock);
> 
> 	if (dtree->rt_free_bw) {
> 		bw_to_move = to_ratio(rt_period, *diff);
> 		if (bw_to_move > dtree->rt_free_bw)
> 			*diff = from_ratio(dtree->rt_free_bw, rt_period);
> 
> 		if (*diff) {
> 			rt_update_bw(src, stree, -(*diff), rt_period);
> 			rt_update_bw(dst, dtree, *diff, rt_period);
> 
> 			ret = true;
> 		}
> 	}
> 
> 	double_spin_unlock(&dtree->rt_bw_lock, &stree->rt_bw_lock);
> 
> 	return ret;
> }

OK, I think that should work, add a little comment on why we're doing it
this way and all should be well ;-)


  reply	other threads:[~2010-03-23 20:34 UTC|newest]

Thread overview: 26+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2010-02-23 18:56 [PATCH 0/3] sched: use EDF to throttle RT task groups v2 Fabio Checconi
2010-02-23 18:56 ` [PATCH 1/3] sched: use EDF to schedule groups Fabio Checconi
2010-02-25 20:28   ` Peter Zijlstra
2010-03-03 16:59     ` Fabio Checconi
2010-02-23 18:56 ` [PATCH 2/3] sched: enforce per-cpu utilization limits on runtime balancing Fabio Checconi
2010-02-25 20:28   ` Peter Zijlstra
2010-03-03 16:59     ` Fabio Checconi
2010-02-25 20:28   ` Peter Zijlstra
2010-03-03 17:00     ` Fabio Checconi
2010-03-23 20:32       ` Peter Zijlstra
2010-02-25 20:28   ` Peter Zijlstra
2010-03-03 16:59     ` Fabio Checconi
2010-02-25 20:28   ` Peter Zijlstra
2010-03-03 17:00     ` Fabio Checconi
2010-03-23 20:33       ` Peter Zijlstra [this message]
2010-02-25 20:28   ` Peter Zijlstra
2010-03-03 17:00     ` Fabio Checconi
2010-02-23 18:56 ` [PATCH 3/3] sched: make runtime balancing code more EDF-friendly Fabio Checconi
2010-02-25 20:28   ` Peter Zijlstra
2010-03-03 17:01     ` Fabio Checconi
2010-02-25 20:28 ` [PATCH 0/3] sched: use EDF to throttle RT task groups v2 Peter Zijlstra
2010-02-27 12:33 ` Peter Zijlstra
2010-03-03 17:01   ` Fabio Checconi
2010-03-23 20:30     ` Peter Zijlstra
2010-03-23 20:56       ` Dhaval Giani
2010-03-23 21:51         ` Tommaso Cucinotta

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=1269376438.5283.8.camel@laptop \
    --to=peterz@infradead.org \
    --cc=dhaval@retis.sssup.it \
    --cc=faggioli@gandalf.sssup.it \
    --cc=fchecconi@gmail.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=michael@evidence.eu.com \
    --cc=mingo@elte.hu \
    --cc=pjt@google.com \
    --cc=t.cucinotta@sssup.it \
    --cc=tglx@linutronix.de \
    --subject='Re: [PATCH 2/3] sched: enforce per-cpu utilization limits on runtime balancing' \
    /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

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.