Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
MIME-Version: 1.0
References: <1634278612-17055-1-git-send-email-huangzhaoyang@gmail.com>
 <YYGV1TxsZXzGXFmx@cmpxchg.org> <YYqMJLXcQ4a+Lh/4@hirez.programming.kicks-ass.net>
 <CAGWkznHGZeAeQzpuwi-ZH50R_fm4Nw5r8TSonk8AQatZaB_xSA@mail.gmail.com> <YYuEoftqXxOhStJ7@hirez.programming.kicks-ass.net>
In-Reply-To: <YYuEoftqXxOhStJ7@hirez.programming.kicks-ass.net>
From:   Zhaoyang Huang <huangzhaoyang@gmail.com>
Date:   Wed, 10 Nov 2021 16:47:46 +0800
Message-ID: <CAGWkznED8Zsy5jxcPWk7u70cmBFER1fY1WDYRo0zD6ptPbyWVw@mail.gmail.com>
Subject: Re: [Resend PATCH] psi : calc cfs task memstall time more precisely
To:     Peter Zijlstra <peterz@infradead.org>
Cc:     Vincent Guittot <vincent.guittot@linaro.org>,
        Dietmar Eggemann <dietmar.eggemann@arm.com>,
        Ke Wang <ke.wang@unisoc.com>, xuewen.yan@unisoc.com,
        Johannes Weiner <hannes@cmpxchg.org>,
        Andrew Morton <akpm@linux-foundation.org>,
        Michal Hocko <mhocko@kernel.org>,
        Vladimir Davydov <vdavydov.dev@gmail.com>,
        Zhaoyang Huang <zhaoyang.huang@unisoc.com>,
        "open list:MEMORY MANAGEMENT" <linux-mm@kvack.org>,
        LKML <linux-kernel@vger.kernel.org>
Content-Type: text/plain; charset="UTF-8"
Precedence: bulk

On Wed, Nov 10, 2021 at 4:36 PM Peter Zijlstra <peterz@infradead.org> wrote:
>
> On Wed, Nov 10, 2021 at 09:37:00AM +0800, Zhaoyang Huang wrote:
> > On Tue, Nov 9, 2021 at 10:56 PM Peter Zijlstra <peterz@infradead.org> wrote:
> > >
> > > On Tue, Nov 02, 2021 at 03:47:33PM -0400, Johannes Weiner wrote:
> > > > CC peterz as well for rt and timekeeping magic
> > > >
> > > > On Fri, Oct 15, 2021 at 02:16:52PM +0800, Huangzhaoyang wrote:
> > > > > From: Zhaoyang Huang <zhaoyang.huang@unisoc.com>
> > > > >
> > > > > In an EAS enabled system, there are two scenarios discordant to current design,
> > > > >
> > > > > 1. workload used to be heavy uneven among cores for sake of scheduler policy.
> > > > > RT task usually preempts CFS task in little core.
> > > > > 2. CFS task's memstall time is counted as simple as exit - entry so far, which
> > > > > ignore the preempted time by RT, DL and Irqs.
> > >
> > > It ignores preemption full-stop. I don't see why RT/IRQ should be
> > > special cased here.
> > As Johannes comments, what we are trying to solve is mainly the
> > preempted time of the CFS task by RT/IRQ, NOT the RT/IRQ themselves.
> > Could you please catch up the recent reply of Dietmar, which maybe
> > provide more information.
>
> In that case NAK.
Would you please explaining if there is any constraint to prevent from
doing so? We do think eliminating the preempted time is reasonable and
doable as it is memory irrelevant but probably related to lack of CPU
etc.