The Linux Kernel Mailing List
 help / color / mirror / Atom feed
From: K Prateek Nayak <kprateek.nayak@amd.com>
To: Shrikanth Hegde <sshegde@linux.ibm.com>,
	<linux-kernel@vger.kernel.org>, <mingo@kernel.org>,
	<peterz@infradead.org>, <juri.lelli@redhat.com>,
	<vincent.guittot@linaro.org>, <yury.norov@gmail.com>,
	<iii@linux.ibm.com>
Cc: <tglx@kernel.org>, <gregkh@linuxfoundation.org>,
	<pbonzini@redhat.com>, <seanjc@google.com>, <vschneid@redhat.com>,
	<huschle@linux.ibm.com>, <rostedt@goodmis.org>,
	<dietmar.eggemann@arm.com>, <mgorman@suse.de>,
	<bsegall@google.com>, <maddy@linux.ibm.com>,
	<srikar@linux.ibm.com>, <hdanton@sina.com>, <chleroy@kernel.org>,
	<vineeth@bitbyteword.org>, <frederic@kernel.org>,
	<arighi@nvidia.com>, <pauld@redhat.com>,
	<christian.loehle@arm.com>, <tj@kernel.org>,
	<tommaso.cucinotta@gmail.com>, <maz@kernel.org>,
	<rafael@kernel.org>
Subject: Re: [PATCH v3 08/20] sched/fair: load balance only among preferred CPUs
Date: Thu, 4 Jun 2026 11:39:39 +0530	[thread overview]
Message-ID: <e908bf2d-ee44-4237-8ae6-9efa8d2eeb98@amd.com> (raw)
In-Reply-To: <8281bd2a-31a1-4e18-aea2-830777a454a6@linux.ibm.com>

Hello Shrikanth,

On 6/4/2026 11:15 AM, Shrikanth Hegde wrote:
> Hi Prateek.
> Thanks for reviewing the patchset and tags.
> 
> On 6/4/26 10:50 AM, K Prateek Nayak wrote:
>> Hello Shrikanth,
>>
>> On 5/14/2026 8:51 PM, Shrikanth Hegde wrote:
>>> diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c
>>> index a704285ac55a..0a851d4b0d7e 100644
>>> --- a/kernel/sched/fair.c
>>> +++ b/kernel/sched/fair.c
>>> @@ -12087,6 +12087,9 @@ static int sched_balance_rq(int this_cpu, struct rq *this_rq,
>>>         cpumask_and(cpus, sched_domain_span(sd), cpu_active_mask);
>>>   +    /* Spread load among preferred CPUs */
>>> +    cpumask_and(cpus, cpus, cpu_preferred_mask);
>>
>> nit.
>>
>> Instead of the onine_mask, can we keep the cpu_preferred_mask closer in sync
>> with the active_mask?
> 
> I tried to make it active_mask. When there is high steal time,
> a CPU marked as not-active isn't available for workload which pins
> them. That would break user affinities.

Ack! Sorry for the confusion. More details below ...

> 
>>
>> That way, if we can ensure that a preferred CPU is always a subset of the
>> active_mask, we can save on a couple of cpumask_and() operations by just
>> looking at preferred_mask instead of both the active_mask and the
>> preferred_mask.
>>
> 
> active mask used heavily today. I didn't want to merge the two.

I didn't suggest merging the two - simply clearing the CPU as preferred
before it is transitioned out of the active mask ...

> 
>> Are there any complications in that approach to track it via
>> sched_cpu_activate() / sched_cpu_deactivate()?
> 
> Sorry i didn;t get this part. You mean track preferred mask with it?

... just moving set_cpu_preferred() into sched_cpu_activate() and
sched_cpu_deactivate() instead of doing it at set_cpu_online().

> 
> We end up calling sched domain rebuild too if we go via sched_cpu_activate
>         build_sched_domains+8
>         partition_sched_domains_locked+1316
>         partition_sched_domains+80
>         rebuild_sched_domains_locked+332
>         cpuset_handle_hotplug+1456
>         sched_cpu_deactivate+608
>         cpuhp_invoke_callback+504
>         cpuhp_thread_fun+324
>         smpboot_thread_fn+556
> 
> 
>>
> 
> What is active_mask guaranteed to provide today? Are there any ABI for it
> or we are free to change it?

I thin it is just a transition mask between the CPU being marked
online -> active and inactive -> offline. Simply suggests the CPU is
available for the scheduler to run tasks on it.

-- 
Thanks and Regards,
Prateek


  reply	other threads:[~2026-06-04  6:09 UTC|newest]

Thread overview: 53+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-05-14 15:21 [PATCH v3 00/20] sched: Introduce cpu_preferred_mask and steal-driven vCPU backoff Shrikanth Hegde
2026-05-14 15:21 ` [PATCH v3 01/20] sched/debug: Remove unused schedstats Shrikanth Hegde
2026-06-04  4:01   ` K Prateek Nayak
2026-05-14 15:21 ` [PATCH v3 02/20] sched/docs: Document cpu_preferred_mask and Preferred CPU concept Shrikanth Hegde
2026-05-14 15:21 ` [PATCH v3 03/20] kconfig: Provide PREFERRED_CPU option Shrikanth Hegde
2026-05-14 15:21 ` [PATCH v3 04/20] cpumask: Introduce cpu_preferred_mask Shrikanth Hegde
2026-05-14 15:21 ` [PATCH v3 05/20] sysfs: Add preferred CPU file Shrikanth Hegde
2026-05-14 15:21 ` [PATCH v3 06/20] sched/core: allow only preferred CPUs in is_cpu_allowed Shrikanth Hegde
2026-06-04  5:06   ` K Prateek Nayak
2026-06-04  5:54     ` Shrikanth Hegde
2026-05-14 15:21 ` [PATCH v3 07/20] sched/fair: Select preferred CPU at wakeup when possible Shrikanth Hegde
2026-06-04  5:14   ` K Prateek Nayak
2026-06-04  5:48     ` Shrikanth Hegde
2026-06-08 16:39     ` Ilya Leoshkevich
2026-06-09  4:42       ` Shrikanth Hegde
2026-05-14 15:21 ` [PATCH v3 08/20] sched/fair: load balance only among preferred CPUs Shrikanth Hegde
2026-06-04  5:20   ` K Prateek Nayak
2026-06-04  5:45     ` Shrikanth Hegde
2026-06-04  6:09       ` K Prateek Nayak [this message]
2026-06-04 13:07         ` Shrikanth Hegde
2026-06-09 11:22           ` Peter Zijlstra
2026-06-09 11:33             ` Shrikanth Hegde
2026-05-14 15:21 ` [PATCH v3 09/20] sched/rt: Select a preferred CPU for wakeup and pulling rt task Shrikanth Hegde
2026-06-09 11:25   ` Peter Zijlstra
2026-06-09 11:30     ` Shrikanth Hegde
2026-06-10 17:13       ` Shrikanth Hegde
2026-05-14 15:21 ` [PATCH v3 10/20] sched/core: Keep tick on non-preferred CPUs until tasks are out Shrikanth Hegde
2026-05-14 15:21 ` [PATCH v3 11/20] sched/core: Push current task from non preferred CPU Shrikanth Hegde
2026-06-04  6:58   ` K Prateek Nayak
2026-06-04  9:43     ` Shrikanth Hegde
2026-06-04  9:56       ` K Prateek Nayak
2026-06-04 12:16         ` Shrikanth Hegde
2026-05-14 15:21 ` [PATCH v3 12/20] sched/debug: Add migration stats due to non preferred CPUs Shrikanth Hegde
2026-05-14 15:21 ` [PATCH v3 13/20] sched/debug: Create debugfs folder steal_monitor Shrikanth Hegde
2026-05-14 15:21 ` [PATCH v3 14/20] sched/debug: Provide debugfs to enable/disable steal monitor Shrikanth Hegde
2026-06-08 17:37   ` Ilya Leoshkevich
2026-06-09  4:48     ` Shrikanth Hegde
2026-06-08 18:28   ` Ilya Leoshkevich
2026-06-09  5:04     ` Shrikanth Hegde
2026-05-14 15:21 ` [PATCH v3 15/20] sched/core: Introduce a simple " Shrikanth Hegde
2026-06-08 16:25   ` Ilya Leoshkevich
2026-06-09  5:54     ` Shrikanth Hegde
2026-05-14 15:22 ` [PATCH v3 16/20] sched/core: Compute steal values at regular intervals Shrikanth Hegde
2026-05-14 15:22 ` [PATCH v3 17/20] sched/core: Introduce default arch handling code for inc/dec preferred CPUs Shrikanth Hegde
2026-06-08 16:22   ` Ilya Leoshkevich
2026-06-09  5:14     ` Shrikanth Hegde
2026-06-08 18:40   ` Ilya Leoshkevich
2026-06-09  6:42     ` Shrikanth Hegde
2026-05-14 15:22 ` [PATCH v3 18/20] sched/core: Handle steal values and mark CPUs as preferred Shrikanth Hegde
2026-05-14 15:22 ` [PATCH v3 19/20] sched/core: Mark the direction of steal values to avoid oscillations Shrikanth Hegde
2026-05-14 15:22 ` [PATCH v3 20/20] sched/debug: Add debug knobs for steal monitor Shrikanth Hegde
2026-06-04  3:17 ` [PATCH v3 00/20] sched: Introduce cpu_preferred_mask and steal-driven vCPU backoff Shrikanth Hegde
2026-06-09 19:51 ` Shrikanth Hegde

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=e908bf2d-ee44-4237-8ae6-9efa8d2eeb98@amd.com \
    --to=kprateek.nayak@amd.com \
    --cc=arighi@nvidia.com \
    --cc=bsegall@google.com \
    --cc=chleroy@kernel.org \
    --cc=christian.loehle@arm.com \
    --cc=dietmar.eggemann@arm.com \
    --cc=frederic@kernel.org \
    --cc=gregkh@linuxfoundation.org \
    --cc=hdanton@sina.com \
    --cc=huschle@linux.ibm.com \
    --cc=iii@linux.ibm.com \
    --cc=juri.lelli@redhat.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=maddy@linux.ibm.com \
    --cc=maz@kernel.org \
    --cc=mgorman@suse.de \
    --cc=mingo@kernel.org \
    --cc=pauld@redhat.com \
    --cc=pbonzini@redhat.com \
    --cc=peterz@infradead.org \
    --cc=rafael@kernel.org \
    --cc=rostedt@goodmis.org \
    --cc=seanjc@google.com \
    --cc=srikar@linux.ibm.com \
    --cc=sshegde@linux.ibm.com \
    --cc=tglx@kernel.org \
    --cc=tj@kernel.org \
    --cc=tommaso.cucinotta@gmail.com \
    --cc=vincent.guittot@linaro.org \
    --cc=vineeth@bitbyteword.org \
    --cc=vschneid@redhat.com \
    --cc=yury.norov@gmail.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