[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <CAP-5=fXVGfF8LZr23PUsKWkoi9+hQV3x4fG9dVSihK0FNh8Nfw@mail.gmail.com>
Date: Fri, 6 Dec 2019 15:16:54 -0800
From: Ian Rogers <irogers@...gle.com>
To: Peter Zijlstra <peterz@...radead.org>
Cc: Ingo Molnar <mingo@...hat.com>,
Arnaldo Carvalho de Melo <acme@...nel.org>,
Mark Rutland <mark.rutland@....com>,
Alexander Shishkin <alexander.shishkin@...ux.intel.com>,
Jiri Olsa <jolsa@...hat.com>,
Namhyung Kim <namhyung@...nel.org>,
Andrew Morton <akpm@...ux-foundation.org>,
Masahiro Yamada <yamada.masahiro@...ionext.com>,
Kees Cook <keescook@...omium.org>,
Catalin Marinas <catalin.marinas@....com>,
Petr Mladek <pmladek@...e.com>,
Mauro Carvalho Chehab <mchehab+samsung@...nel.org>,
Qian Cai <cai@....pw>, Joe Lawrence <joe.lawrence@...hat.com>,
Tetsuo Handa <penguin-kernel@...ove.sakura.ne.jp>,
"Uladzislau Rezki (Sony)" <urezki@...il.com>,
Andy Shevchenko <andriy.shevchenko@...ux.intel.com>,
Ard Biesheuvel <ardb@...nel.org>,
"David S. Miller" <davem@...emloft.net>,
Kent Overstreet <kent.overstreet@...il.com>,
Gary Hook <Gary.Hook@....com>, Arnd Bergmann <arnd@...db.de>,
Kan Liang <kan.liang@...ux.intel.com>,
LKML <linux-kernel@...r.kernel.org>,
Stephane Eranian <eranian@...gle.com>,
Andi Kleen <ak@...ux.intel.com>
Subject: Re: [PATCH v3 00/10] Optimize cgroup context switch
On Thu, Nov 14, 2019 at 10:17 AM Ian Rogers <irogers@...gle.com> wrote:
>
> On Thu, Nov 14, 2019 at 2:45 AM Peter Zijlstra <peterz@...radead.org> wrote:
> >
> > On Wed, Nov 13, 2019 at 04:30:32PM -0800, Ian Rogers wrote:
> > > Avoid iterating over all per-CPU events during cgroup changing context
> > > switches by organizing events by cgroup.
> >
> > When last we spoke (Plumbers in Lisbon) you mentioned that this
> > optimization was yielding far less than expected. You had graphs showing
> > how the use of cgroups impacted event scheduling time and how this patch
> > set only reduced that a little.
> >
> > Any update on all that? There seems to be a conspicuous lack of such
> > data here.
>
> I'm working on giving an update on the numbers but I suspect they are
> better than I'd measured ahead of LPC due to a bug in a script.
>
> Thanks,
> Ian
Apologies for the delay, I'm sending v5 that addresses review
comments. I'm still working on performance numbers.
Thanks,
Ian
Powered by blists - more mailing lists