From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <linux-kernel-owner+w=401wt.eu-S1752449AbXCaISY@vger.kernel.org>
Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand
	id S1752449AbXCaISY (ORCPT <rfc822;w@1wt.eu>);
	Sat, 31 Mar 2007 04:18:24 -0400
Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1752450AbXCaISY
	(ORCPT <rfc822;linux-kernel-outgoing>);
	Sat, 31 Mar 2007 04:18:24 -0400
Received: from fgwmail6.fujitsu.co.jp ([192.51.44.36]:55710 "EHLO
	fgwmail6.fujitsu.co.jp" rhost-flags-OK-OK-OK-OK) by vger.kernel.org
	with ESMTP id S1752449AbXCaISX (ORCPT
	<rfc822;linux-kernel@vger.kernel.org>);
	Sat, 31 Mar 2007 04:18:23 -0400
Date: Sat, 31 Mar 2007 17:16:24 +0900
Message-ID: <87y7ldhlpj.wl%takeuchi_satoru@jp.fujitsu.com>
From: Satoru Takeuchi <takeuchi_satoru@jp.fujitsu.com>
To: Ingo Molnar <mingo@elte.hu>
Cc: Satoru Takeuchi <takeuchi_satoru@jp.fujitsu.com>,
       Linux Kernel <linux-kernel@vger.kernel.org>
Subject: Re: [BUG] scheduler: strange behavor with massive interactive processes
In-Reply-To: <874po6tbez.wl%takeuchi_satoru@jp.fujitsu.com>
References: <87r6rb1nbm.wl%takeuchi_satoru@jp.fujitsu.com>
	<20070327191420.GB29929@elte.hu>
	<874po6tbez.wl%takeuchi_satoru@jp.fujitsu.com>
User-Agent: Wanderlust/2.14.0 (Africa) SEMI/1.14.6 (Maruoka) FLIM/1.14.8
 (=?ISO-8859-4?Q?Shij=F2?=) APEL/10.6 Emacs/21.4 (i486-pc-linux-gnu)
 MULE/5.0 (SAKAKI)
MIME-Version: 1.0 (generated by SEMI 1.14.6 - "Maruoka")
Content-Type: text/plain; charset=US-ASCII
Sender: linux-kernel-owner@vger.kernel.org
X-Mailing-List: linux-kernel@vger.kernel.org

Hi Ingo,

> > > Hi Ingo and all,
> > > 
> > > When I was executing massive interactive processes, I found that some 
> > > of them occupy CPU time and the others hardly run.
> > 
> > yeah.
> > 
> > > I also attach the test program which easily recreates this problem.
> > 
> > thanks, this is really helpful - does the patch below improve the 
> > situation?

I tested your patch and it seems to work well.

Test environment
================

 - kernel: 2.6.21-rc5 with or without Ingo's patch
 - others: same as my initial mail except for omitting nice 19 cases

Result (without Ingo's patch)
=============================

  +---------+-----------+------+------+------+--------+
  |   # of  |   # of    | avg  | max  | min  |  stdev |
  |   CPUs  | processes | (*1) | (*2) | (*3) |  (*4)  |
  +---------+-----------+------+------+------+--------+
  | 1(i386) |       200 |  162 | 8258 |    1 |   1113 |
  +---------+-----------+------+------+------+--------+
  |         |           |  378 | 9314 |    2 |   1421 |
  | 2(ia64) |       400 +------+------+------+--------+
  |         |           |  189 |12544 |    1 |   1443 |
  +---------+-----------+------+------+------+--------+

  *1) average number of loops among all processes
  *2) maximum number of loops among all processes
  *3) minimum number of loops among all processes
  *4) standard deviation

Result (with Ingo's patch)
==========================

  +---------+-----------+------+------+------+--------+
  |   # of  |   # of    | avg  | max  | min  |  stdev |
  |   CPUs  | processes |      |      |      |        |
  +---------+-----------+------+------+------+--------+
  | 1(i386) |       200 |  153 |  232 |  128 |   7.67 |
  +---------+-----------+------+------+------+--------+
  |         |           |  376 |  451 |  291 |   17.6 |
  | 2(ia64) |       400 +------+------+------+--------+
  |         |           |  188 |  236 |  137 |   14.5 |
  +---------+-----------+------+------+------+--------+

Although it is not perfectly fair, it is certain that this patch really improve
the situation in dramatic form. Thank you very much!

Satoru