linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Tim Chen <tim.c.chen@linux.intel.com>
To: Vineeth Remanan Pillai <vpillai@digitalocean.com>,
	Aubrey Li <aubrey.intel@gmail.com>
Cc: "Aaron Lu" <aaron.lwe@gmail.com>,
	"Julien Desfossez" <jdesfossez@digitalocean.com>,
	"Nishanth Aravamudan" <naravamudan@digitalocean.com>,
	"Peter Zijlstra" <peterz@infradead.org>,
	"Ingo Molnar" <mingo@kernel.org>,
	"Thomas Gleixner" <tglx@linutronix.de>,
	"Paul Turner" <pjt@google.com>,
	"Linus Torvalds" <torvalds@linux-foundation.org>,
	"Linux List Kernel Mailing" <linux-kernel@vger.kernel.org>,
	"Dario Faggioli" <dfaggioli@suse.com>,
	"Frédéric Weisbecker" <fweisbec@gmail.com>,
	"Kees Cook" <keescook@chromium.org>,
	"Greg Kerr" <kerrnel@google.com>, "Phil Auld" <pauld@redhat.com>,
	"Valentin Schneider" <valentin.schneider@arm.com>,
	"Mel Gorman" <mgorman@techsingularity.net>,
	"Pawan Gupta" <pawan.kumar.gupta@linux.intel.com>,
	"Paolo Bonzini" <pbonzini@redhat.com>
Subject: Re: [RFC PATCH v4 00/19] Core scheduling v4
Date: Fri, 28 Feb 2020 15:55:24 -0800	[thread overview]
Message-ID: <bcd601e7-3f15-e340-bebe-a6ca3635dacb@linux.intel.com> (raw)
In-Reply-To: <CANaguZDQZg-Z6aNpeLcjQ-cGm3X8CQOkZ_hnJNUyqDRM=yVDFQ@mail.gmail.com>

On 2/26/20 1:54 PM, Vineeth Remanan Pillai wrote:

> rq->curr being NULL can mean that the sibling is idle or forced idle.
> In both the cases, I think it makes sense to migrate a task so that it can
> compete with the other sibling for a chance to run. This function
> can_migrate_task actually only says if this task is eligible and
> later part of the code decides whether it is okay to migrate it
> based on factors like load and util and capacity. So I think its
> fine to declare the task as eligible if the dest core is running
> idle. Does this thinking make sense?
> 
> On our testing, it did not show much degradation in performance with
> this change. I am reworking the fix by removing the check for
> task_est_util. It doesn't seem to be valid to check for util to migrate
> the task.
> 

In Aaron's test case, there is a great imbalance in the load on one core
where all the grp A tasks are vs the other cores where the grp B tasks are
spread around.  Normally, load balancer will move the tasks for grp A.

Aubrey's can_migrate_task patch prevented the load balancer to migrate tasks if the core
cookie on the target queue don't match.  The thought was it will induce
force idle and reduces cpu utilization if we migrate task to it.
That kept all the grp A tasks from getting migrated and kept the imbalance
indefinitely in Aaron's test case.

Perhaps we should also look at the load imbalance between the src rq and
target rq.  If the imbalance is big (say two full cpu bound tasks worth
of load), we should migrate anyway despite the cookie mismatch.  We are willing
to pay a bit for the force idle by balancing the load out more.
I think Aubrey's patch on can_migrate_task should be more friendly to
Aaron's test scenario if such logic is incorporated.

In Vinnet's fix, we only look at the currently running task's weight in
src and dst rq.  Perhaps the load on the src and dst rq needs to be considered
to prevent too great an imbalance between the run queues?

Tim

  parent reply	other threads:[~2020-02-28 23:55 UTC|newest]

Thread overview: 84+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-10-30 18:33 [RFC PATCH v4 00/19] Core scheduling v4 Vineeth Remanan Pillai
2019-10-30 18:33 ` [RFC PATCH v4 01/19] stop_machine: Fix stop_cpus_in_progress ordering Vineeth Remanan Pillai
2019-10-30 18:33 ` [RFC PATCH v4 02/19] sched: Fix kerneldoc comment for ia64_set_curr_task Vineeth Remanan Pillai
2019-10-30 18:33 ` [RFC PATCH v4 03/19] sched: Wrap rq::lock access Vineeth Remanan Pillai
2019-10-30 18:33 ` [RFC PATCH v4 04/19] sched/{rt,deadline}: Fix set_next_task vs pick_next_task Vineeth Remanan Pillai
2019-10-30 18:33 ` [RFC PATCH v4 05/19] sched: Add task_struct pointer to sched_class::set_curr_task Vineeth Remanan Pillai
2019-10-30 18:33 ` [RFC PATCH v4 06/19] sched/fair: Export newidle_balance() Vineeth Remanan Pillai
2019-10-30 18:33 ` [RFC PATCH v4 07/19] sched: Allow put_prev_task() to drop rq->lock Vineeth Remanan Pillai
2019-10-30 18:33 ` [RFC PATCH v4 08/19] sched: Rework pick_next_task() slow-path Vineeth Remanan Pillai
2019-10-30 18:33 ` [RFC PATCH v4 09/19] sched: Introduce sched_class::pick_task() Vineeth Remanan Pillai
2019-10-30 18:33 ` [RFC PATCH v4 10/19] sched: Core-wide rq->lock Vineeth Remanan Pillai
2019-10-30 18:33 ` [RFC PATCH v4 11/19] sched: Basic tracking of matching tasks Vineeth Remanan Pillai
2019-10-30 18:33 ` [RFC PATCH v4 12/19] sched: A quick and dirty cgroup tagging interface Vineeth Remanan Pillai
2019-10-30 18:33 ` [RFC PATCH v4 13/19] sched: Add core wide task selection and scheduling Vineeth Remanan Pillai
2019-10-30 18:33 ` [RFC PATCH v4 14/19] sched/fair: Add a few assertions Vineeth Remanan Pillai
2019-10-30 18:33 ` [RFC PATCH v4 15/19] sched: Trivial forced-newidle balancer Vineeth Remanan Pillai
2019-10-30 18:33 ` [RFC PATCH v4 16/19] sched: Debug bits Vineeth Remanan Pillai
2019-10-30 18:33 ` [RFC PATCH v4 17/19] sched/fair: wrapper for cfs_rq->min_vruntime Vineeth Remanan Pillai
2019-10-30 18:33 ` [RFC PATCH v4 18/19] sched/fair: core wide vruntime comparison Vineeth Remanan Pillai
2019-10-30 18:33 ` [RFC PATCH v4 19/19] sched/fair : Wake up forced idle siblings if needed Vineeth Remanan Pillai
2019-10-31 11:42 ` [RFC PATCH v4 00/19] Core scheduling v4 Li, Aubrey
2019-11-01 11:33   ` Li, Aubrey
2019-11-08  3:20     ` Li, Aubrey
2019-10-31 18:42 ` Phil Auld
2019-11-01 14:03   ` Vineeth Remanan Pillai
2019-11-01 16:35     ` Greg Kerr
2019-11-01 18:07       ` Dario Faggioli
2019-11-12  1:45     ` Dario Faggioli
2019-11-13 17:16       ` Tim Chen
2020-01-02  2:28       ` Aubrey Li
2020-01-10 23:19         ` Tim Chen
2019-11-11 19:10 ` Tim Chen
2020-01-14  1:12   ` Tim Chen
2020-01-14 15:40     ` Vineeth Remanan Pillai
2020-01-15  3:43       ` Li, Aubrey
2020-01-15 19:33         ` Tim Chen
2020-01-16  1:45           ` Aubrey Li
2020-01-17 16:00             ` Vineeth Remanan Pillai
2020-01-22 18:04               ` Gruza, Agata
2020-01-28  2:40       ` Dario Faggioli
     [not found]         ` <CANaguZDDpzrzdTmvjXvCmV2c+wBt6mXWSz4Vn-LJ-onc_Oj=yw@mail.gmail.com>
2020-02-01 15:31           ` Dario Faggioli
2020-02-06  0:28       ` Tim Chen
2020-02-06 22:37         ` Julien Desfossez
2020-02-12 23:07         ` Julien Desfossez
2020-02-13 18:37           ` Tim Chen
2020-02-14  6:10             ` Aubrey Li
     [not found]               ` <CANaguZC40mDHfL1H_9AA7H8cyd028t9PQVRqQ3kB4ha8R7hhqg@mail.gmail.com>
2020-02-15  6:01                 ` Aubrey Li
     [not found]                   ` <CANaguZBj_x_2+9KwbHCQScsmraC_mHdQB6uRqMTYMmvhBYfv2Q@mail.gmail.com>
2020-02-21 23:20                     ` Julien Desfossez
2020-03-17  0:55                       ` Joel Fernandes
2020-03-17 19:07                         ` Tim Chen
2020-03-17 20:18                           ` Tim Chen
2020-03-18  1:10                             ` Joel Fernandes
2020-03-17 21:17                           ` Thomas Gleixner
2020-03-17 21:58                             ` Tim Chen
2020-03-18  1:03                             ` Joel Fernandes
2020-03-18  2:30                               ` Joel Fernandes
2020-03-18  0:52                           ` Joel Fernandes
2020-03-18 11:53                             ` Thomas Gleixner
2020-03-19  1:54                               ` Joel Fernandes
2020-02-25  3:44               ` Aaron Lu
2020-02-25  5:32                 ` Aubrey Li
2020-02-25  7:34                   ` Aaron Lu
2020-02-25 10:40                     ` Aubrey Li
2020-02-25 11:21                       ` Aaron Lu
2020-02-25 13:41                         ` Aubrey Li
     [not found]                 ` <CANaguZD205ccu1V_2W-QuMRrJA9SjJ5ng1do4NCdLy8NDKKrbA@mail.gmail.com>
2020-02-26  3:13                   ` Aaron Lu
2020-02-26  7:21                   ` Aubrey Li
     [not found]                     ` <CANaguZDQZg-Z6aNpeLcjQ-cGm3X8CQOkZ_hnJNUyqDRM=yVDFQ@mail.gmail.com>
2020-02-27  4:45                       ` Aubrey Li
2020-02-28 23:55                       ` Tim Chen [this message]
2020-03-03 14:59                         ` Li, Aubrey
2020-03-03 23:54                           ` Li, Aubrey
2020-03-05  4:33                             ` Aaron Lu
2020-03-05  6:10                               ` Li, Aubrey
2020-03-05  8:52                                 ` Aaron Lu
2020-02-27  2:04                   ` Aaron Lu
2020-02-27 14:10                     ` Phil Auld
2020-02-27 14:37                       ` Aubrey Li
2020-02-28  2:54                       ` Aaron Lu
2020-03-05 13:45                         ` Aubrey Li
2020-03-06  2:41                           ` Aaron Lu
2020-03-06 18:06                             ` Tim Chen
2020-03-06 18:33                               ` Phil Auld
2020-03-06 21:44                                 ` Tim Chen
2020-03-07  3:13                                   ` Aaron Lu

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=bcd601e7-3f15-e340-bebe-a6ca3635dacb@linux.intel.com \
    --to=tim.c.chen@linux.intel.com \
    --cc=aaron.lwe@gmail.com \
    --cc=aubrey.intel@gmail.com \
    --cc=dfaggioli@suse.com \
    --cc=fweisbec@gmail.com \
    --cc=jdesfossez@digitalocean.com \
    --cc=keescook@chromium.org \
    --cc=kerrnel@google.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mgorman@techsingularity.net \
    --cc=mingo@kernel.org \
    --cc=naravamudan@digitalocean.com \
    --cc=pauld@redhat.com \
    --cc=pawan.kumar.gupta@linux.intel.com \
    --cc=pbonzini@redhat.com \
    --cc=peterz@infradead.org \
    --cc=pjt@google.com \
    --cc=tglx@linutronix.de \
    --cc=torvalds@linux-foundation.org \
    --cc=valentin.schneider@arm.com \
    --cc=vpillai@digitalocean.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).