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: <20200630112123.12076-1-qais.yousef@arm.com> <CAKfTPtCjTCBbGN0gN_=7MJdiY3N-PPy0ApZWH5xd7KDQ9=0yMg@mail.gmail.com>
 <20200706104111.yfopcjm3kiqlxhx2@e107158-lin.cambridge.arm.com>
In-Reply-To: <20200706104111.yfopcjm3kiqlxhx2@e107158-lin.cambridge.arm.com>
From:   Vincent Guittot <vincent.guittot@linaro.org>
Date:   Tue, 7 Jul 2020 14:29:39 +0200
Message-ID: <CAKfTPtDKb0Co5wEcKu3ZE2cJZs87Pma2oQaNQvRq8da26_xjcg@mail.gmail.com>
Subject: Re: [PATCH v6 0/2] sched: Optionally skip uclamp logic in fast path
To:     Qais Yousef <qais.yousef@arm.com>
Cc:     Ingo Molnar <mingo@redhat.com>,
        Peter Zijlstra <peterz@infradead.org>,
        Valentin Schneider <valentin.schneider@arm.com>,
        Juri Lelli <juri.lelli@redhat.com>,
        Dietmar Eggemann <dietmar.eggemann@arm.com>,
        Steven Rostedt <rostedt@goodmis.org>,
        Ben Segall <bsegall@google.com>, Mel Gorman <mgorman@suse.de>,
        Patrick Bellasi <patrick.bellasi@matbug.net>,
        Chris Redpath <chris.redpath@arm.com>,
        Lukasz Luba <lukasz.luba@arm.com>,
        linux-kernel <linux-kernel@vger.kernel.org>
Content-Type: text/plain; charset="UTF-8"
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On Mon, 6 Jul 2020 at 12:41, Qais Yousef <qais.yousef@arm.com> wrote:
>
> On 07/03/20 14:09, Vincent Guittot wrote:
> > I have run the perf bench sched pipe that have have already run
> > previously with this v6 and the results are similar to my previous
> > tests:
> > The impact is -1.61% similarly to v2 which is better compared the
> > original -3.66% without your patch
>
> Thanks Vincent.
>
> Can you afford doing a capture of `perf record` and share the resulting
> perf.dat with vmlinux (with debug symbols)?

Will try to make it by end of the week

>
> Having a before/after capture would be even better.
>
> Not sure if we can do  much about this -1.61% in your case, but it'd be good to
> understand why if possible. perf bench sched pipe is very sensitive to tiniest
> of changes which could be due to binary-to-binary differences.
>
> Thanks
>
> --
> Qais Yousef