From: Con Kolivas <kernel@kolivas.org>
To: Willy Tarreau <w@1wt.eu>, ck list <ck@vds.kolivas.org>
Cc: Ingo Molnar <mingo@elte.hu>,
linux-kernel@vger.kernel.org,
Linus Torvalds <torvalds@linux-foundation.org>,
Andrew Morton <akpm@linux-foundation.org>,
Nick Piggin <npiggin@suse.de>, Mike Galbraith <efault@gmx.de>,
Arjan van de Ven <arjan@infradead.org>,
Peter Williams <pwil3058@bigpond.net.au>,
Thomas Gleixner <tglx@linutronix.de>,
caglar@pardus.org.tr, Gene Heskett <gene.heskett@gmail.com>
Subject: Re: [REPORT] cfs-v4 vs sd-0.44
Date: Sun, 22 Apr 2007 11:51:37 +1000 [thread overview]
Message-ID: <200704221151.38464.kernel@kolivas.org> (raw)
In-Reply-To: <20070421121235.GA2044@1wt.eu>
On Saturday 21 April 2007 22:12, Willy Tarreau wrote:
> 2) SD-0.44
>
> Feels good, but becomes jerky at moderately high loads. I've started
> 64 ocbench with a 250 ms busy loop and 750 ms sleep time. The system
> always responds correctly but under X, mouse jumps quite a bit and
> typing in xterm or even text console feels slightly jerky. The CPU is
> not completely used, and the load varies a lot (see below). However,
> the load is shared equally between all 64 ocbench, and they do not
> deviate even after 4000 iterations. X uses less than 1% CPU during
> those tests.
Found it. I broke SMP balancing again so there is serious scope for
improvement on SMP hardware. That explains the huge load variations. Expect
yet another fix soon, which should improve behaviour further :)
--
-ck
next prev parent reply other threads:[~2007-04-22 1:53 UTC|newest]
Thread overview: 147+ messages / expand[flat|nested] mbox.gz Atom feed top
2007-04-20 14:04 [patch] CFS scheduler, v4 Ingo Molnar
2007-04-20 21:37 ` Gene Heskett
2007-04-21 20:47 ` S.Çağlar Onur
2007-04-22 1:22 ` Gene Heskett
2007-04-20 21:39 ` mdew .
2007-04-21 6:47 ` Ingo Molnar
2007-04-21 7:55 ` [patch] CFS scheduler, v4, for v2.6.20.7 Ingo Molnar
2007-04-21 12:12 ` [REPORT] cfs-v4 vs sd-0.44 Willy Tarreau
2007-04-21 12:40 ` Con Kolivas
2007-04-21 13:02 ` Willy Tarreau
2007-04-21 15:46 ` Ingo Molnar
2007-04-21 16:18 ` Willy Tarreau
2007-04-21 16:34 ` Linus Torvalds
2007-04-21 16:42 ` William Lee Irwin III
2007-04-21 18:55 ` Kyle Moffett
2007-04-21 19:49 ` Ulrich Drepper
2007-04-21 23:17 ` William Lee Irwin III
2007-04-21 23:35 ` Linus Torvalds
2007-04-22 1:46 ` Ulrich Drepper
2007-04-22 7:02 ` William Lee Irwin III
2007-04-22 7:17 ` Ulrich Drepper
2007-04-22 8:48 ` William Lee Irwin III
2007-04-22 16:16 ` Ulrich Drepper
2007-04-23 0:07 ` Rusty Russell
2007-04-21 16:53 ` Willy Tarreau
2007-04-21 16:53 ` Ingo Molnar
2007-04-21 16:57 ` Willy Tarreau
2007-04-21 18:09 ` Ulrich Drepper
2007-04-21 17:03 ` Geert Bosch
2007-04-21 15:55 ` Con Kolivas
2007-04-21 16:00 ` Ingo Molnar
2007-04-21 16:12 ` Willy Tarreau
2007-04-21 16:39 ` William Lee Irwin III
2007-04-21 17:15 ` Jan Engelhardt
2007-04-21 19:00 ` Ingo Molnar
2007-04-22 13:18 ` Mark Lord
2007-04-22 13:27 ` Ingo Molnar
2007-04-22 13:30 ` Mark Lord
2007-04-25 8:16 ` Pavel Machek
2007-04-25 8:22 ` Ingo Molnar
2007-04-25 10:19 ` Alan Cox
2007-04-21 22:54 ` Denis Vlasenko
2007-04-22 0:08 ` Con Kolivas
2007-04-22 4:58 ` Mike Galbraith
2007-04-21 23:59 ` Con Kolivas
2007-04-22 13:04 ` Juliusz Chroboczek
2007-04-22 23:24 ` Linus Torvalds
2007-04-23 1:34 ` Nick Piggin
2007-04-23 15:56 ` Linus Torvalds
2007-04-23 19:11 ` Ingo Molnar
2007-04-23 19:52 ` Linus Torvalds
2007-04-23 20:33 ` Ingo Molnar
2007-04-23 20:44 ` Ingo Molnar
2007-04-23 21:03 ` Ingo Molnar
2007-04-23 21:53 ` Guillaume Chazarain
2007-04-24 7:04 ` Rogan Dawes
2007-04-24 7:31 ` Ingo Molnar
2007-04-24 8:25 ` Rogan Dawes
2007-04-24 15:03 ` Chris Friesen
2007-04-24 15:07 ` Rogan Dawes
2007-04-24 15:15 ` Chris Friesen
2007-04-24 23:55 ` Peter Williams
2007-04-25 9:29 ` Ingo Molnar
2007-04-23 22:48 ` Jeremy Fitzhardinge
2007-04-24 0:59 ` Li, Tong N
2007-04-24 1:57 ` Bill Huey
2007-04-24 18:01 ` Li, Tong N
2007-04-24 21:27 ` William Lee Irwin III
2007-04-24 22:18 ` Bernd Eckenfels
2007-04-25 1:22 ` Li, Tong N
2007-04-25 6:05 ` William Lee Irwin III
2007-04-25 9:44 ` Ingo Molnar
2007-04-25 11:58 ` William Lee Irwin III
2007-04-25 20:13 ` Willy Tarreau
2007-04-26 17:57 ` Li, Tong N
2007-04-26 19:18 ` Willy Tarreau
2007-04-28 15:12 ` Bernd Eckenfels
2007-04-26 23:26 ` William Lee Irwin III
2007-04-24 3:46 ` Peter Williams
2007-04-24 4:52 ` Arjan van de Ven
2007-04-24 6:21 ` Peter Williams
2007-04-24 6:36 ` Ingo Molnar
2007-04-24 7:00 ` Gene Heskett
2007-04-24 7:08 ` Ingo Molnar
2007-04-24 6:45 ` David Lang
2007-04-24 7:24 ` Ingo Molnar
2007-04-24 14:38 ` Gene Heskett
2007-04-24 17:44 ` Willy Tarreau
2007-04-25 0:30 ` Gene Heskett
2007-04-25 0:32 ` Gene Heskett
2007-04-24 7:12 ` Gene Heskett
2007-04-24 7:14 ` Ingo Molnar
2007-04-24 14:36 ` Gene Heskett
2007-04-24 7:25 ` Ingo Molnar
2007-04-24 14:39 ` Gene Heskett
2007-04-24 14:42 ` Gene Heskett
2007-04-24 7:33 ` Ingo Molnar
2007-04-26 0:51 ` SD renice recommendation was: " Con Kolivas
2007-04-24 15:08 ` Ray Lee
2007-04-25 9:32 ` Ingo Molnar
2007-04-23 20:05 ` Willy Tarreau
2007-04-24 21:05 ` 'Scheduler Economy' prototype patch for CFS Ingo Molnar
2007-04-23 2:42 ` [report] renicing X, cfs-v5 vs sd-0.46 Ingo Molnar
2007-04-23 15:09 ` Linus Torvalds
2007-04-23 17:19 ` Gene Heskett
2007-04-23 17:19 ` Gene Heskett
2007-04-23 19:48 ` Ingo Molnar
2007-04-23 20:56 ` Michael K. Edwards
2007-04-22 13:23 ` [REPORT] cfs-v4 vs sd-0.44 Mark Lord
2007-04-21 18:17 ` Gene Heskett
2007-04-22 1:26 ` Con Kolivas
2007-04-22 2:07 ` Gene Heskett
2007-04-22 8:07 ` William Lee Irwin III
2007-04-22 11:11 ` Gene Heskett
2007-04-22 1:51 ` Con Kolivas [this message]
2007-04-21 20:35 ` [patch] CFS scheduler, v4 S.Çağlar Onur
2007-04-22 8:30 ` Michael Gerdau
2007-04-23 22:47 ` Ingo Molnar
2007-04-23 1:12 ` [patch] CFS scheduler, -v5 Ingo Molnar
2007-04-23 1:25 ` Nick Piggin
2007-04-23 2:39 ` Gene Heskett
2007-04-23 3:08 ` Ingo Molnar
2007-04-23 2:55 ` Ingo Molnar
2007-04-23 3:22 ` Nick Piggin
2007-04-23 3:43 ` Ingo Molnar
2007-04-23 4:06 ` Nick Piggin
2007-04-23 7:10 ` Ingo Molnar
2007-04-23 7:25 ` Nick Piggin
2007-04-23 7:35 ` Ingo Molnar
2007-04-23 9:25 ` Ingo Molnar
2007-04-23 3:19 ` [patch] CFS scheduler, -v5 (build problem - make headers_check fails) Zach Carter
2007-04-23 10:03 ` Ingo Molnar
2007-04-23 5:16 ` [patch] CFS scheduler, -v5 Markus Trippelsdorf
2007-04-23 5:27 ` Markus Trippelsdorf
2007-04-23 6:21 ` Ingo Molnar
2007-04-25 11:43 ` Srivatsa Vaddagiri
2007-04-25 12:51 ` Ingo Molnar
2007-04-23 12:20 ` Guillaume Chazarain
2007-04-23 12:36 ` Ingo Molnar
2007-04-24 16:54 ` Christian Hesse
2007-04-25 9:25 ` Ingo Molnar
2007-04-25 10:51 ` Christian Hesse
2007-04-25 10:56 ` Ingo Molnar
2007-04-23 9:28 ` crash with CFS v4 and qemu/kvm (was: [patch] CFS scheduler, v4) Christian Hesse
2007-04-23 10:18 ` Ingo Molnar
2007-04-24 10:54 ` Christian Hesse
-- strict thread matches above, loose matches on Subject: below --
2007-04-22 4:38 [REPORT] cfs-v4 vs sd-0.44 Al Boldi
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=200704221151.38464.kernel@kolivas.org \
--to=kernel@kolivas.org \
--cc=akpm@linux-foundation.org \
--cc=arjan@infradead.org \
--cc=caglar@pardus.org.tr \
--cc=ck@vds.kolivas.org \
--cc=efault@gmx.de \
--cc=gene.heskett@gmail.com \
--cc=linux-kernel@vger.kernel.org \
--cc=mingo@elte.hu \
--cc=npiggin@suse.de \
--cc=pwil3058@bigpond.net.au \
--cc=tglx@linutronix.de \
--cc=torvalds@linux-foundation.org \
--cc=w@1wt.eu \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox