From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S265256AbUBIQnc (ORCPT ); Mon, 9 Feb 2004 11:43:32 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S265263AbUBIQnb (ORCPT ); Mon, 9 Feb 2004 11:43:31 -0500 Received: from obsidian.spiritone.com ([216.99.193.137]:16570 "EHLO obsidian.spiritone.com") by vger.kernel.org with ESMTP id S265256AbUBIQna (ORCPT ); Mon, 9 Feb 2004 11:43:30 -0500 Date: Mon, 09 Feb 2004 08:43:18 -0800 From: "Martin J. Bligh" To: Timothy Miller , Nick Piggin cc: Rick Lindsley , Anton Blanchard , akpm@osdl.org, linux-kernel@vger.kernel.org, dvhltc@us.ibm.com Subject: Re: [PATCH] Load balancing problem in 2.6.2-mm1 Message-ID: <144560000.1076344998@[10.10.2.4]> In-Reply-To: <4027B758.8060908@techsource.com> References: <200402062311.i16NBdF14365@owlet.beaverton.ibm.com> <40242152.5030606@cyberone.com.au> <231480000.1076110387@flay> <4024261E.5070702@cyberone.com.au> <232690000.1076111266@flay> <40242D14.6070908@cyberone.com.au> <4027B758.8060908@techsource.com> X-Mailer: Mulberry/2.2.1 (Linux/x86) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Content-Disposition: inline Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org >> Can we keep current behaviour default, and if arches want to >> override it they can? And if someone one day does testing to >> show it really isn't a good idea, then we can change the default. >> >> I like to try stick to the fairness first approach. >> >> We got quite a few complaints about unfairness when the >> scheduler used to keep 2 on one cpu and 1 on another, even in >> development kernels. I suspect that most wouldn't have known >> one way or the other if only top showed 66% each, but still. > > Stupid question: Does the balancing consider process priority? Is it > unfair to have two lower pri tasks always on one cpu while the highest > pri of the three is always by itself? No, it doesn't take account of that. We've discussed that before at some point, and yes, I think it's wrong, but on the other hand, we want to be reasonably fast at deciding what to take, and there are a whole bunch of other criteria that we ought to be taking account as well - it's difficult. M.