Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
Subject: Re: [RFC V2 2/2] sched/fair: Fallback to sched-idle CPU if idle CPU
 isn't found
To:     Peter Zijlstra <peterz@infradead.org>,
        Viresh Kumar <viresh.kumar@linaro.org>,
        Subhra Mazumdar <subhra.mazumdar@oracle.com>
Cc:     songliubraving@fb.com, Ingo Molnar <mingo@redhat.com>,
        Vincent Guittot <vincent.guittot@linaro.org>, tkjos@google.com,
        Daniel Lezcano <daniel.lezcano@linaro.org>,
        quentin.perret@linaro.org, chris.redpath@arm.com,
        Dietmar.Eggemann@arm.com, linux-kernel@vger.kernel.org
References: <cover.1556182964.git.viresh.kumar@linaro.org>
 <59b37c56b8fcb834f7d3234e776eaeff74ad117f.1556182965.git.viresh.kumar@linaro.org>
 <20190510072125.GG2623@hirez.programming.kicks-ass.net>
 <20190513093418.altqhlhu4zsu75t4@vireshk-i7>
 <20190513113518.GQ2623@hirez.programming.kicks-ass.net>
From:   Steven Sistare <steven.sistare@oracle.com>
Organization: Oracle Corporation
Message-ID: <05ffe5f4-4324-2977-e46e-155e1aef57fa@oracle.com>
Date:   Tue, 14 May 2019 12:03:21 -0400
User-Agent: Mozilla/5.0 (Windows NT 10.0; WOW64; rv:60.0) Gecko/20100101
 Thunderbird/60.6.1
MIME-Version: 1.0
In-Reply-To: <20190513113518.GQ2623@hirez.programming.kicks-ass.net>
Content-Type: text/plain; charset=utf-8
Content-Language: en-US
Content-Transfer-Encoding: 7bit
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On 5/13/2019 7:35 AM, Peter Zijlstra wrote:
> On Mon, May 13, 2019 at 03:04:18PM +0530, Viresh Kumar wrote:
>> On 10-05-19, 09:21, Peter Zijlstra wrote:
> 
>>> I don't hate his per se; but the whole select_idle_sibling() thing is
>>> something that needs looking at.
>>>
>>> There was the task stealing thing from Steve that looked interesting and
>>> that would render your apporach unfeasible.
>>
>> I am surely missing something as I don't see how that patchset will
>> make this patchset perform badly, than what it already does.
> 
> Nah; I just misremembered. I know Oracle has a patch set poking at
> select_idle_siblings() _somewhere_ (as do I), and I just found the wrong
> one.
> 
> Basically everybody is complaining select_idle_sibling() is too
> expensive for checking the entire LLC domain, except for FB (and thus
> likely some other workloads too) that depend on it to kill their tail
> latency.
> 
> But I suppose we could still do this, even if we scan only a subset of
> the LLC, just keep track of the last !idle CPU running only SCHED_IDLE
> tasks and pick that if you do not (in your limited scan) find a better
> candidate.

Subhra posted a patch that incrementally searches for an idle CPU in the LLC,
remembering the last CPU examined, and searching a fixed number of CPUs from there.
That technique is compatible with the one that Viresh suggests; the incremental
search would stop if a SCHED_IDLE cpu was found.

I also fiddled with select_idle_sibling, maintaining a per-LLC bitmap of idle CPUs,
updated with atomic operations.  Performance was basically unchanged for the
workloads I tested, and I inserted timers around the idle search showing it was
a very small fraction of time both before and after my changes.  That led me to
ignore the push side and optimize the pull side with task stealing.

I would be very interested in hearing from folks that have workloads that demonstrate
that select_idle_sibling is too expensive.

- Steve