Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > linux.kernel > #1356939 > unrolled thread

[PATCH 0/8] schedutil enhancements

Back to article view | Back to linux.kernel

Contents

  [PATCH 0/8] schedutil enhancements Michael Turquette <mturquette@baylibre.com> - 2016-03-14 06:30 +0100
    [PATCH 4/8] cpufreq/schedutil: sysfs capacity margin tunable Michael Turquette <mturquette@baylibre.com> - 2016-03-14 06:30 +0100
      Re: [PATCH 4/8] cpufreq/schedutil: sysfs capacity margin tunable Peter Zijlstra <peterz@infradead.org> - 2016-03-15 22:30 +0100
        Re: [PATCH 4/8] cpufreq/schedutil: sysfs capacity margin tunable Peter Zijlstra <peterz@infradead.org> - 2016-03-15 22:50 +0100
          Re: [PATCH 4/8] cpufreq/schedutil: sysfs capacity margin tunable Steve Muckle <steve.muckle@linaro.org> - 2016-03-16 04:40 +0100
            Re: [PATCH 4/8] cpufreq/schedutil: sysfs capacity margin tunable Peter Zijlstra <peterz@infradead.org> - 2016-03-16 09:10 +0100
              Re: [PATCH 4/8] cpufreq/schedutil: sysfs capacity margin tunable Juri Lelli <Juri.Lelli@arm.com> - 2016-03-16 11:10 +0100
                Re: [PATCH 4/8] cpufreq/schedutil: sysfs capacity margin tunable Steve Muckle <steve.muckle@linaro.org> - 2016-03-16 19:00 +0100
                  Re: [PATCH 4/8] cpufreq/schedutil: sysfs capacity margin tunable Michael Turquette <mturquette@baylibre.com> - 2016-03-16 23:10 +0100
                  Re: [PATCH 4/8] cpufreq/schedutil: sysfs capacity margin tunable Juri Lelli <Juri.Lelli@arm.com> - 2016-03-17 10:40 +0100
                    Re: [PATCH 4/8] cpufreq/schedutil: sysfs capacity margin tunable Steve Muckle <steve.muckle@linaro.org> - 2016-03-17 15:00 +0100
                      Re: [PATCH 4/8] cpufreq/schedutil: sysfs capacity margin tunable Patrick Bellasi <patrick.bellasi@arm.com> - 2016-03-17 17:00 +0100
              Re: [PATCH 4/8] cpufreq/schedutil: sysfs capacity margin tunable "Rafael J. Wysocki" <rafael@kernel.org> - 2016-03-16 13:50 +0100
            Re: [PATCH 4/8] cpufreq/schedutil: sysfs capacity margin tunable Michael Turquette <mturquette@baylibre.com> - 2016-03-16 23:10 +0100
    [PATCH 1/8] sched/cpufreq: remove cpufreq_trigger_update() Michael Turquette <mturquette@baylibre.com> - 2016-03-14 06:30 +0100
      Re: [PATCH 1/8] sched/cpufreq: remove cpufreq_trigger_update() Peter Zijlstra <peterz@infradead.org> - 2016-03-15 22:20 +0100
        Re: [PATCH 1/8] sched/cpufreq: remove cpufreq_trigger_update() Peter Zijlstra <peterz@infradead.org> - 2016-03-15 23:00 +0100
      Re: [PATCH 1/8] sched/cpufreq: remove cpufreq_trigger_update() Peter Zijlstra <peterz@infradead.org> - 2016-03-16 09:10 +0100
    [PATCH 5/8] sched/cpufreq: pass sched class into cpufreq_update_util Michael Turquette <mturquette@baylibre.com> - 2016-03-14 06:30 +0100
      Re: [PATCH 5/8] sched/cpufreq: pass sched class into  cpufreq_update_util Peter Zijlstra <peterz@infradead.org> - 2016-03-15 22:30 +0100
        Re: [PATCH 5/8] sched/cpufreq: pass sched class into  cpufreq_update_util Steve Muckle <steve.muckle@linaro.org> - 2016-03-16 05:00 +0100
        Re: [PATCH 5/8] sched/cpufreq: pass sched class into  cpufreq_update_util Peter Zijlstra <peterz@infradead.org> - 2016-03-16 08:50 +0100
          Re: [PATCH 5/8] sched/cpufreq: pass sched class into cpufreq_update_util Vincent Guittot <vincent.guittot@linaro.org> - 2016-03-16 09:40 +0100
            Re: [PATCH 5/8] sched/cpufreq: pass sched class into  cpufreq_update_util Peter Zijlstra <peterz@infradead.org> - 2016-03-16 10:00 +0100
              Re: [PATCH 5/8] sched/cpufreq: pass sched class into cpufreq_update_util Vincent Guittot <vincent.guittot@linaro.org> - 2016-03-16 10:20 +0100
              Re: [PATCH 5/8] sched/cpufreq: pass sched class into cpufreq_update_util "Rafael J. Wysocki" <rafael@kernel.org> - 2016-03-16 13:40 +0100
                Re: [PATCH 5/8] sched/cpufreq: pass sched class into  cpufreq_update_util Peter Zijlstra <peterz@infradead.org> - 2016-03-16 14:20 +0100
                  Re: [PATCH 5/8] sched/cpufreq: pass sched class into cpufreq_update_util "Rafael J. Wysocki" <rafael@kernel.org> - 2016-03-16 14:30 +0100
                    Re: [PATCH 5/8] sched/cpufreq: pass sched class into  cpufreq_update_util Peter Zijlstra <peterz@infradead.org> - 2016-03-16 14:50 +0100
    [PATCH 8/8] sched: prefer cpufreq_scale_freq_capacity Michael Turquette <mturquette@baylibre.com> - 2016-03-14 06:30 +0100
      Re: [PATCH 8/8] sched: prefer cpufreq_scale_freq_capacity Dietmar Eggemann <dietmar.eggemann@arm.com> - 2016-03-15 20:20 +0100
        Re: [PATCH 8/8] sched: prefer cpufreq_scale_freq_capacity Michael Turquette <mturquette@baylibre.com> - 2016-03-15 21:50 +0100
          Re: [PATCH 8/8] sched: prefer cpufreq_scale_freq_capacity Dietmar Eggemann <dietmar.eggemann@arm.com> - 2016-03-16 20:50 +0100
            Re: [PATCH 8/8] sched: prefer cpufreq_scale_freq_capacity Peter Zijlstra <peterz@infradead.org> - 2016-03-16 21:10 +0100
              Re: [PATCH 8/8] sched: prefer cpufreq_scale_freq_capacity "Rafael J. Wysocki" <rjw@rjwysocki.net> - 2016-03-16 22:40 +0100
      Re: [PATCH 8/8] sched: prefer cpufreq_scale_freq_capacity Peter Zijlstra <peterz@infradead.org> - 2016-03-15 22:50 +0100
        Re: [PATCH 8/8] sched: prefer cpufreq_scale_freq_capacity Peter Zijlstra <peterz@infradead.org> - 2016-03-16 08:50 +0100
          Re: [PATCH 8/8] sched: prefer cpufreq_scale_freq_capacity Peter Zijlstra <peterz@infradead.org> - 2016-03-16 13:50 +0100
    [PATCH 2/8] sched/fair: add margin to utilization update Michael Turquette <mturquette@baylibre.com> - 2016-03-14 06:30 +0100
      Re: [PATCH 2/8] sched/fair: add margin to utilization update Peter Zijlstra <peterz@infradead.org> - 2016-03-15 22:20 +0100
        Re: [PATCH 2/8] sched/fair: add margin to utilization update Peter Zijlstra <peterz@infradead.org> - 2016-03-15 22:50 +0100
      Re: [PATCH 2/8] sched/fair: add margin to utilization update Steve Muckle <steve.muckle@linaro.org> - 2016-03-16 04:00 +0100
        Re: [PATCH 2/8] sched/fair: add margin to utilization update Michael Turquette <mturquette@baylibre.com> - 2016-03-16 23:20 +0100
    [PATCH 3/8] sched/cpufreq: new cfs capacity margin helpers Michael Turquette <mturquette@baylibre.com> - 2016-03-14 06:30 +0100
      Re: [PATCH 3/8] sched/cpufreq: new cfs capacity margin helpers Peter Zijlstra <peterz@infradead.org> - 2016-03-15 22:20 +0100
    Re: [PATCH 0/8] schedutil enhancements "Rafael J. Wysocki" <rjw@rjwysocki.net> - 2016-03-16 01:10 +0100

Page 3 of 3 — ← Prev page 1 2 [3]

#1358279 — Re: [PATCH 2/8] sched/fair: add margin to utilization update

On Tue, Mar 15, 2016 at 02:28:48PM -0700, Michael Turquette wrote:
> Quoting Peter Zijlstra (2016-03-15 14:16:14)
> > On Sun, Mar 13, 2016 at 10:22:06PM -0700, Michael Turquette wrote:
> > > @@ -2840,6 +2853,8 @@ static inline void update_load_avg(struct sched_entity *se, int update_tg)
> > >  
> > >       if (cpu == smp_processor_id() && &rq->cfs == cfs_rq) {
> > >               unsigned long max = rq->cpu_capacity_orig;
> > > +             unsigned long cap = cfs_rq->avg.util_avg *
> > > +                     cfs_capacity_margin / max;
> > >  
> > >               /*
> > >                * There are a few boundary cases this might miss but it should
> > > @@ -2852,8 +2867,7 @@ static inline void update_load_avg(struct sched_entity *se, int update_tg)
> > >                * thread is a different class (!fair), nor will the utilization
> > >                * number include things like RT tasks.
> > >                */
> > > -             cpufreq_update_util(rq_clock(rq),
> > > -                                 min(cfs_rq->avg.util_avg, max), max);
> > > +             cpufreq_update_util(rq_clock(rq), min(cap, max), max);
> > >       }
> > >  }
> > 
> > I really don't see why that is here, and not inside whatever uses
> > cpufreq_update_util().
> 
> Because I want schedutil to be dumb and not implement any policy of it's
> own. The idea is for the scheduler to select frequency after all.
> 
> I want to avoid a weird hybrid solution where we try to be smart about
> selecting the right capacity/frequency in fair.c (e.g. Steve and Juri's
> patches to fair.c from the sched-freq-v7 series), but then have an
> additional layer of "smarts" massaging those values further in the
> cpufreq governor.

So the problem here is that you add an unconditional division, even if
cpufreq_update_util() then decides to not do anything with it.

Please, these are scheduler paths, do not add (fancy) instructions if
you don't absolutely have to.

[toc] | [prev] | [next] | [standalone]

#1358478 — Re: [PATCH 2/8] sched/fair: add margin to utilization update

On 03/13/2016 10:22 PM, Michael Turquette wrote:
> +unsigned long cfs_capacity_margin = CAPACITY_MARGIN_DEFAULT;
> +
>  #ifdef CONFIG_CFS_BANDWIDTH
>  /*
>   * Amount of runtime to allocate from global (tg) to local (per-cfs_rq) pool
> @@ -2840,6 +2853,8 @@ static inline void update_load_avg(struct sched_entity *se, int update_tg)
>  
>  	if (cpu == smp_processor_id() && &rq->cfs == cfs_rq) {
>  		unsigned long max = rq->cpu_capacity_orig;
> +		unsigned long cap = cfs_rq->avg.util_avg *
> +			cfs_capacity_margin / max;

Doesn't rq->cpu_capacity_orig get scaled per the microarch invariance?
This would mean that the margin we're applying here would differ based
on that.

I'd expect that the margin would be * (cfs_capacity_margin /
SCHED_CAPACITY_SCALE) which would then reduce the division into a shift.

[toc] | [prev] | [next] | [standalone]

#1359397 — Re: [PATCH 2/8] sched/fair: add margin to utilization update

Quoting Steve Muckle (2016-03-15 19:52:59)
> On 03/13/2016 10:22 PM, Michael Turquette wrote:
> > +unsigned long cfs_capacity_margin = CAPACITY_MARGIN_DEFAULT;
> > +
> >  #ifdef CONFIG_CFS_BANDWIDTH
> >  /*
> >   * Amount of runtime to allocate from global (tg) to local (per-cfs_rq) pool
> > @@ -2840,6 +2853,8 @@ static inline void update_load_avg(struct sched_entity *se, int update_tg)
> >  
> >       if (cpu == smp_processor_id() && &rq->cfs == cfs_rq) {
> >               unsigned long max = rq->cpu_capacity_orig;
> > +             unsigned long cap = cfs_rq->avg.util_avg *
> > +                     cfs_capacity_margin / max;
> 
> Doesn't rq->cpu_capacity_orig get scaled per the microarch invariance?
> This would mean that the margin we're applying here would differ based
> on that.
> 
> I'd expect that the margin would be * (cfs_capacity_margin /
> SCHED_CAPACITY_SCALE) which would then reduce the division into a shift.

Will fix.

Thanks,
Mike

[toc] | [prev] | [next] | [standalone]

#1356948 — [PATCH 3/8] sched/cpufreq: new cfs capacity margin helpers

Introduce helper functions that allow cpufreq governors to change the
value of the capacity margin applied to the cfs_rq->avg.util_avg signal.
This allows for run-time tuning of the margin.

A follow-up patch will update the schedutil governor to use these
helpers.

Signed-off-by: Michael Turquette <mturquette+renesas@baylibre.com>
---
 include/linux/sched.h  |  3 +++
 kernel/sched/cpufreq.c | 53 ++++++++++++++++++++++++++++++++++++++++++++++++++
 2 files changed, 56 insertions(+)

diff --git a/include/linux/sched.h b/include/linux/sched.h
index 1fa9b52..f18a99b 100644
--- a/include/linux/sched.h
+++ b/include/linux/sched.h
@@ -2372,6 +2372,9 @@ void cpufreq_set_freq_update_hook(int cpu, struct freq_update_hook *hook,
 			void (*func)(struct freq_update_hook *hook, u64 time,
 				     unsigned long util, unsigned long max));
 void cpufreq_clear_freq_update_hook(int cpu);
+unsigned long cpufreq_get_cfs_capacity_margin(void);
+void cpufreq_set_cfs_capacity_margin(unsigned long margin);
+void cpufreq_reset_cfs_capacity_margin(void);
 #endif
 
 #ifdef CONFIG_SCHED_AUTOGROUP
diff --git a/kernel/sched/cpufreq.c b/kernel/sched/cpufreq.c
index bd012c2..a126b58 100644
--- a/kernel/sched/cpufreq.c
+++ b/kernel/sched/cpufreq.c
@@ -61,6 +61,59 @@ void cpufreq_clear_freq_update_hook(int cpu)
 EXPORT_SYMBOL_GPL(cpufreq_clear_freq_update_hook);
 
 /**
+ * cpufreq_get_cfs_capacity_margin - Get global cfs enqueue capacity margin
+ *
+ * margin is a percentage of capacity that is applied to the current
+ * utilization when selecting a new capacity state or cpu frequency. The value
+ * should be normalized to the range of [0..SCHED_CAPACITY_SCALE], where
+ * SCHED_CAPACITY_SCALE is 100% of the normalized capacity, or equivalent to
+ * multiplying the utilization by one.
+ *
+ * This function returns the current global cfs enqueue capacity margin
+ */
+unsigned long cpufreq_get_cfs_capacity_margin(void)
+{
+	return cfs_capacity_margin;
+}
+EXPORT_SYMBOL_GPL(cpufreq_get_cfs_capacity_margin);
+
+/**
+ * cpufreq_set_cfs_capacity_margin - Set global cfs enqueue capacity margin
+ * @margin: new capacity margin
+ *
+ * margin is a percentage of capacity that is applied to the current
+ * utilization when selecting a new capacity state or cpu frequency. The value
+ * should be normalized to the range of [0..SCHED_CAPACITY_SCALE], where
+ * SCHED_CAPACITY_SCALE is 100% of the normalized capacity, or equivalent to
+ * multiplying the utilization by one.
+ *
+ * For instance, to add a 25% margin to a utilization, margin should be 1280,
+ * which is 1.25x 1024, the default for SCHED_CAPACITY_SCALE.
+ */
+void cpufreq_set_cfs_capacity_margin(unsigned long margin)
+{
+	cfs_capacity_margin = margin;
+}
+EXPORT_SYMBOL_GPL(cpufreq_set_cfs_capacity_margin);
+
+/**
+ * cpufreq_reset_cfs_capacity_margin - Reset global cfs enqueue cap margin
+ *
+ * margin is a percentage of capacity that is applied to the current
+ * utilization when selecting a new capacity state or cpu frequency. The value
+ * should be normalized to the range of [0..SCHED_CAPACITY_SCALE], where
+ * SCHED_CAPACITY_SCALE is 100% of the normalized capacity, or equivalent to
+ * multiplying the utilization by one.
+ *
+ * This function resets the global margin to its default value.
+ */
+void cpufreq_reset_cfs_capacity_margin(void)
+{
+	cfs_capacity_margin = CAPACITY_MARGIN_DEFAULT;
+}
+EXPORT_SYMBOL_GPL(cpufreq_reset_cfs_capacity_margin);
+
+/**
  * cpufreq_update_util - Take a note about CPU utilization changes.
  * @time: Current time.
  * @util: CPU utilization.
-- 
2.1.4

[toc] | [prev] | [next] | [standalone]

#1358249 — Re: [PATCH 3/8] sched/cpufreq: new cfs capacity margin helpers

On Sun, Mar 13, 2016 at 10:22:07PM -0700, Michael Turquette wrote:
> +/**
> + * cpufreq_set_cfs_capacity_margin - Set global cfs enqueue capacity margin
> + * @margin: new capacity margin
> + *
> + * margin is a percentage of capacity that is applied to the current
> + * utilization when selecting a new capacity state or cpu frequency. The value
> + * should be normalized to the range of [0..SCHED_CAPACITY_SCALE], where
> + * SCHED_CAPACITY_SCALE is 100% of the normalized capacity, or equivalent to
> + * multiplying the utilization by one.
> + *
> + * For instance, to add a 25% margin to a utilization, margin should be 1280,
> + * which is 1.25x 1024, the default for SCHED_CAPACITY_SCALE.
> + */
> +void cpufreq_set_cfs_capacity_margin(unsigned long margin)
> +{
> +	cfs_capacity_margin = margin;
> +}
> +EXPORT_SYMBOL_GPL(cpufreq_set_cfs_capacity_margin);

I don't like this as an interface; what's wrong with using percentiles
as per the discussion I had with Rafael last week?

Also, why is this exported?

[toc] | [prev] | [next] | [standalone]

#1358445

Hi Mike,

On Sunday, March 13, 2016 10:22:04 PM Michael Turquette wrote:
> I'm happy that scheduler-driven cpu frequency selection is getting some
> attention. Rafael's recent schedutil governor is a step in the right direction.

Thanks!

> This series builds on top of Rafael's schedutil governor, bringing it to parity
> with some of the features in the schedfreq series posted by Steve[0], as well
> as adding a couple of new things.
> 
> Patch 1 removes cpufreq_trigger_update()
> 
> Patches 2-4 move the cfs capacity margin out of the governor and into
> cfs. This value is made tunable by a sysfs control in schedutil.
> 
> Patches 5-6 make cpufreq_update_util() aware of multiple scheduler
> classes (cfs, rt & dl), and add storage & summation of these per-class
> utilization values into schedutil.
> 
> Patches 7-8 introduces Dietmar's generic cpufreq implementation[1] of the
> frequency invariance hook and changes the preprocessor magic in sched.h to
> favor the cpufreq implementation over arch- or platform-specific ones.

After the discussion mentioned by Peter in one of his responses (the relevant
e-mail thread is here: http://marc.info/?t=145688568600003&r=1&w=4) I have
changed the schedutil series to address some points made during it.

In particular, I've modified it to use the formula from 
http://marc.info/?l=linux-acpi&m=145756618321500&w=4 with the twist that
if the utilization is frequency-invariant, max_freq will be used instead
of current_freq.

The way it recognizes whether or not that is the case is based on the
Peter's suggestion from http://marc.info/?l=linux-kernel&m=145760739700716&w=4.

For this reason, the governor goes into kernel/sched/ as I don't want
arch_scale_freq_invariant() to be exposed to cpufreq at large, because
schedutil will be the only governor using it in foreseeable future.

The fast switching support patch is slightly different too, but that's
not relevant here.

The new series is in the pm-cpufreq-experimental branch of my tree:

git://git.kernel.org/pub/scm/linux/kernel/git/rafael/linux-pm.git \
pm-cpufreq-experimental

I haven't had the time to post it over the last few days (sorry about
that), but I'll do that tomorrow.

Thanks,
Rafael

[toc] | [prev] | [standalone]

Page 3 of 3 — ← Prev page 1 2 [3]

Back to top | Article view | linux.kernel

csiph-web