Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
MIME-Version: 1.0
References: <Y4r/iuqyLG4MOK4d@chenyu5-mobl1> <3A5DA66F-4330-4FC4-9229-998CF98F663E@joelfernandes.org>
 <Y42uH5fglluWYOm7@chenyu5-mobl1> <CAKfTPtA61DDk-VbNDvAycp8Et5fTwaOS=Q3559rBF7QDHDpJfQ@mail.gmail.com>
In-Reply-To: <CAKfTPtA61DDk-VbNDvAycp8Et5fTwaOS=Q3559rBF7QDHDpJfQ@mail.gmail.com>
From:   Josh Don <joshdon@google.com>
Date:   Tue, 6 Dec 2022 18:23:47 -0800
Message-ID: <CABk29Nuc+VVOpAwFsbBo+OppayszebT29Hadc8qK-xeyZpY6NQ@mail.gmail.com>
Subject: Re: [PATCH v3 1/2] sched/fair: Introduce short duration task check
To:     Vincent Guittot <vincent.guittot@linaro.org>,
        Chen Yu <yu.c.chen@intel.com>
Cc:     Joel Fernandes <joel@joelfernandes.org>,
        Honglei Wang <wanghonglei@didichuxing.com>,
        Peter Zijlstra <peterz@infradead.org>,
        Tim Chen <tim.c.chen@intel.com>,
        Mel Gorman <mgorman@techsingularity.net>,
        Juri Lelli <juri.lelli@redhat.com>,
        Rik van Riel <riel@surriel.com>,
        Aaron Lu <aaron.lu@intel.com>,
        Abel Wu <wuyun.abel@bytedance.com>,
        K Prateek Nayak <kprateek.nayak@amd.com>,
        Yicong Yang <yangyicong@hisilicon.com>,
        "Gautham R . Shenoy" <gautham.shenoy@amd.com>,
        Ingo Molnar <mingo@redhat.com>,
        Dietmar Eggemann <dietmar.eggemann@arm.com>,
        Steven Rostedt <rostedt@goodmis.org>,
        Ben Segall <bsegall@google.com>,
        Daniel Bristot de Oliveira <bristot@redhat.com>,
        Valentin Schneider <vschneid@redhat.com>,
        Hillf Danton <hdanton@sina.com>,
        Len Brown <len.brown@intel.com>,
        Chen Yu <yu.chen.surf@gmail.com>,
        Tianchen Ding <dtcccc@linux.alibaba.com>,
        linux-kernel@vger.kernel.org
Content-Type: text/plain; charset="UTF-8"
Precedence: bulk

> We don't want to add more dedicated knobs. So using
> sysctl_sched_min_granularity as you are doing in this patch looks ok

If not a knob, then maybe at least a smaller hardcoded value? Several
milliseconds seems like too long for this optimization; the
opportunity cost of not doing the idle search likely doesn't make
sense if the waker is going to be active for several additional
milliseconds. I would have guessed something on the order of 100us.

Chen, what is the run duration of tasks in your ping pong example?