[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <07518d3f-cfa8-d6a5-2a13-4d8a7457ee15@arm.com>
Date: Wed, 4 Sep 2019 14:21:48 +0100
From: Valentin Schneider <valentin.schneider@....com>
To: Peter Zijlstra <peterz@...radead.org>,
Radim Krčmář <rkrcmar@...hat.com>
Cc: linux-kernel@...r.kernel.org, Ingo Molnar <mingo@...hat.com>,
Thomas Gleixner <tglx@...utronix.de>,
Borislav Petkov <bp@...en8.de>,
Dave Hansen <dave.hansen@...ux.intel.com>,
Steven Rostedt <rostedt@...dmis.org>,
"H. Peter Anvin" <hpa@...or.com>,
Andy Lutomirski <luto@...nel.org>,
Jirka Hladký <jhladky@...hat.com>,
Jiří Vozár <jvozar@...hat.com>,
x86@...nel.org
Subject: Re: [PATCH 2/2] sched/debug: add sched_update_nr_running tracepoint
On 04/09/2019 14:13, Peter Zijlstra wrote:
> On Tue, Sep 03, 2019 at 05:43:40PM +0200, Radim Krčmář wrote:
>
>> Red Hat adapted one of them for the tracepoint framework and created a
>> tool to plot a heatmap of nr_running, where the sched_update_nr_running
>> tracepoint is being used for fine grained monitoring of scheduling
>> imbalance.
>
> You should be able to reconstruct this from wakeup and switch
> tracepoints.
>
>
I never tried to do it myself but know some folks around me gave it a go
(Morten, Qais and maybe even Chris). I can't remember the exact details but
there was something about not getting the right input - duplicated wakeups,
or no migration information (I think sched_migrate doesn't cover all of the
cases) so the information you're building diverges as time progresses.
Powered by blists - more mailing lists