From: Vincent Guittot <vincent.guittot@linaro.org>
To: Valentin Schneider <valentin.schneider@arm.com>
Cc: linux-kernel <linux-kernel@vger.kernel.org>,
Ingo Molnar <mingo@redhat.com>,
Peter Zijlstra <peterz@infradead.org>,
Phil Auld <pauld@redhat.com>,
Srikar Dronamraju <srikar@linux.vnet.ibm.com>,
Quentin Perret <quentin.perret@arm.com>,
Dietmar Eggemann <dietmar.eggemann@arm.com>,
Morten Rasmussen <Morten.Rasmussen@arm.com>,
Hillf Danton <hdanton@sina.com>
Subject: Re: [PATCH v3 04/10] sched/fair: rework load_balance
Date: Wed, 2 Oct 2019 10:30:59 +0200 [thread overview]
Message-ID: <CAKfTPtBQNJfNmBqpuaefsLzsTrGxJ=2bTs+tRdbOAa9J3eKuVw@mail.gmail.com> (raw)
In-Reply-To: <c752dd1a-731e-aae3-6a2c-aecf88901ac0@arm.com>
On Tue, 1 Oct 2019 at 19:47, Valentin Schneider
<valentin.schneider@arm.com> wrote:
>
> On 19/09/2019 08:33, Vincent Guittot wrote:
>
> [...]
>
> > @@ -8283,69 +8363,133 @@ static inline void update_sd_lb_stats(struct lb_env *env, struct sd_lb_stats *sd
> > */
> > static inline void calculate_imbalance(struct lb_env *env, struct sd_lb_stats *sds)
> > {
> > - unsigned long max_pull, load_above_capacity = ~0UL;
> > struct sg_lb_stats *local, *busiest;
> >
> > local = &sds->local_stat;
> > busiest = &sds->busiest_stat;
> >
> > - if (busiest->group_asym_packing) {
> > + if (busiest->group_type == group_misfit_task) {
> > + /* Set imbalance to allow misfit task to be balanced. */
> > + env->balance_type = migrate_misfit;
> > + env->imbalance = busiest->group_misfit_task_load;
> > + return;
> > + }
> > +
> > + if (busiest->group_type == group_asym_packing) {
> > + /*
> > + * In case of asym capacity, we will try to migrate all load to
> > + * the preferred CPU.
> > + */
> > + env->balance_type = migrate_load;
> > env->imbalance = busiest->group_load;
> > return;
> > }
> >
> > + if (busiest->group_type == group_imbalanced) {
> > + /*
> > + * In the group_imb case we cannot rely on group-wide averages
> > + * to ensure CPU-load equilibrium, try to move any task to fix
> > + * the imbalance. The next load balance will take care of
> > + * balancing back the system.
> > + */
> > + env->balance_type = migrate_task;
> > + env->imbalance = 1;
> > + return;
> > + }
> > +
> > /*
> > - * Avg load of busiest sg can be less and avg load of local sg can
> > - * be greater than avg load across all sgs of sd because avg load
> > - * factors in sg capacity and sgs with smaller group_type are
> > - * skipped when updating the busiest sg:
> > + * Try to use spare capacity of local group without overloading it or
> > + * emptying busiest
> > */
> > - if (busiest->group_type != group_misfit_task &&
> > - (busiest->avg_load <= sds->avg_load ||
> > - local->avg_load >= sds->avg_load)) {
> > - env->imbalance = 0;
> > + if (local->group_type == group_has_spare) {
> > + if (busiest->group_type > group_fully_busy) {
> > + /*
> > + * If busiest is overloaded, try to fill spare
> > + * capacity. This might end up creating spare capacity
> > + * in busiest or busiest still being overloaded but
> > + * there is no simple way to directly compute the
> > + * amount of load to migrate in order to balance the
> > + * system.
> > + */
> > + env->balance_type = migrate_util;
> > + env->imbalance = max(local->group_capacity, local->group_util) -
> > + local->group_util;
> > + return;
> > + }
> > +
> > + if (busiest->group_weight == 1 || sds->prefer_sibling) {
> > + /*
> > + * When prefer sibling, evenly spread running tasks on
> > + * groups.
> > + */
> > + env->balance_type = migrate_task;
> > + env->imbalance = (busiest->sum_h_nr_running - local->sum_h_nr_running) >> 1;
>
> Isn't that one somewhat risky?
>
> Say both groups are classified group_has_spare and we do prefer_sibling.
> We'd select busiest as the one with the maximum number of busy CPUs, but it
> could be so that busiest.sum_h_nr_running < local.sum_h_nr_running (because
> pinned tasks or wakeup failed to properly spread stuff).
>
> The thing should be unsigned so at least we save ourselves from right
> shifting a negative value, but we still end up with a gygornous imbalance
> (which we then store into env.imbalance which *is* signed... Urgh).
so it's not clear what happen with a right shift on negative signed
value and this seems to be compiler dependent so even
max_t(long, 0, (local->idle_cpus - busiest->idle_cpus) >> 1) might be wrong
I'm going to update it
>
> [...]
next prev parent reply other threads:[~2019-10-02 8:31 UTC|newest]
Thread overview: 57+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-09-19 7:33 [PATCH v3 0/8] sched/fair: rework the CFS load balance Vincent Guittot
2019-09-19 7:33 ` [PATCH v3 01/10] sched/fair: clean up asym packing Vincent Guittot
2019-09-27 23:57 ` Rik van Riel
2019-09-19 7:33 ` [PATCH v3 02/10] sched/fair: rename sum_nr_running to sum_h_nr_running Vincent Guittot
2019-09-27 23:59 ` Rik van Riel
2019-10-01 17:11 ` Valentin Schneider
2019-09-19 7:33 ` [PATCH v3 03/10] sched/fair: remove meaningless imbalance calculation Vincent Guittot
2019-09-28 0:05 ` Rik van Riel
2019-10-01 17:12 ` Valentin Schneider
2019-10-02 6:28 ` Vincent Guittot
2019-09-19 7:33 ` [PATCH v3 04/10] sched/fair: rework load_balance Vincent Guittot
2019-09-30 1:12 ` Rik van Riel
2019-09-30 7:44 ` Vincent Guittot
2019-09-30 16:24 ` Dietmar Eggemann
2019-10-01 8:14 ` Vincent Guittot
2019-10-01 16:52 ` Dietmar Eggemann
2019-10-02 6:44 ` Vincent Guittot
2019-10-02 9:21 ` Dietmar Eggemann
2019-10-08 13:02 ` Peter Zijlstra
2019-10-02 8:23 ` Vincent Guittot
2019-10-02 9:24 ` Dietmar Eggemann
2019-10-01 8:15 ` Dietmar Eggemann
2019-10-01 9:14 ` Vincent Guittot
2019-10-01 16:57 ` Dietmar Eggemann
2019-10-01 17:47 ` Valentin Schneider
2019-10-02 8:30 ` Vincent Guittot [this message]
2019-10-02 10:47 ` Valentin Schneider
2019-10-08 14:16 ` Peter Zijlstra
2019-10-08 14:34 ` Valentin Schneider
2019-10-08 15:30 ` Vincent Guittot
2019-10-08 15:48 ` Valentin Schneider
2019-10-08 17:39 ` Peter Zijlstra
2019-10-08 18:45 ` Vincent Guittot
2019-10-08 16:33 ` Peter Zijlstra
2019-10-08 16:39 ` Valentin Schneider
2019-10-08 17:36 ` Valentin Schneider
2019-10-08 17:55 ` Peter Zijlstra
2019-10-08 18:47 ` Vincent Guittot
2019-10-16 7:21 ` Parth Shah
2019-10-16 11:56 ` Vincent Guittot
2019-10-18 5:34 ` Parth Shah
2019-09-19 7:33 ` [PATCH v3 05/10] sched/fair: use rq->nr_running when balancing load Vincent Guittot
2019-09-19 7:33 ` [PATCH v3 06/10] sched/fair: use load instead of runnable load in load_balance Vincent Guittot
2019-09-19 7:33 ` [PATCH v3 07/10] sched/fair: evenly spread tasks when not overloaded Vincent Guittot
2019-09-19 7:33 ` [PATCH v3 08/10] sched/fair: use utilization to select misfit task Vincent Guittot
2019-10-01 17:12 ` Valentin Schneider
2019-09-19 7:33 ` [PATCH v3 09/10] sched/fair: use load instead of runnable load in wakeup path Vincent Guittot
2019-10-07 15:14 ` Rik van Riel
2019-10-07 15:27 ` Vincent Guittot
2019-10-07 18:06 ` Rik van Riel
2019-09-19 7:33 ` [PATCH v3 10/10] sched/fair: optimize find_idlest_group Vincent Guittot
2019-10-08 14:32 ` [PATCH v3 0/8] sched/fair: rework the CFS load balance Phil Auld
2019-10-08 15:53 ` Vincent Guittot
2019-10-09 19:33 ` Phil Auld
2019-10-10 8:20 ` Vincent Guittot
2019-10-16 7:21 ` Parth Shah
2019-10-16 11:51 ` Vincent Guittot
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='CAKfTPtBQNJfNmBqpuaefsLzsTrGxJ=2bTs+tRdbOAa9J3eKuVw@mail.gmail.com' \
--to=vincent.guittot@linaro.org \
--cc=Morten.Rasmussen@arm.com \
--cc=dietmar.eggemann@arm.com \
--cc=hdanton@sina.com \
--cc=linux-kernel@vger.kernel.org \
--cc=mingo@redhat.com \
--cc=pauld@redhat.com \
--cc=peterz@infradead.org \
--cc=quentin.perret@arm.com \
--cc=srikar@linux.vnet.ibm.com \
--cc=valentin.schneider@arm.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).