From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S2993080AbXDRT2j (ORCPT ); Wed, 18 Apr 2007 15:28:39 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S2993081AbXDRT2j (ORCPT ); Wed, 18 Apr 2007 15:28:39 -0400 Received: from zrtps0kp.nortel.com ([47.140.192.56]:41702 "EHLO zrtps0kp.nortel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S2993080AbXDRT2i (ORCPT ); Wed, 18 Apr 2007 15:28:38 -0400 Message-ID: <4626713B.1040206@nortel.com> Date: Wed, 18 Apr 2007 13:27:55 -0600 From: "Chris Friesen" User-Agent: Mozilla Thunderbird 1.0.2-6 (X11/20050513) X-Accept-Language: en-us, en MIME-Version: 1.0 To: Mark Glines CC: Linus Torvalds , Matt Mackall , Nick Piggin , Bill Huey , Mike Galbraith , Peter Williams , William Lee Irwin III , linux-kernel@vger.kernel.org, ck list , Thomas Gleixner , Andrew Morton , Arjan van de Ven Subject: Re: [ck] Re: [Announce] [patch] Modular Scheduler Core and Completely Fair Scheduler [CFS] References: <20070417060955.GO8915@holomorphy.com> <20070417061503.GC1057@wotan.suse.de> <20070417062621.GL2986@holomorphy.com> <20070417070155.GF1057@wotan.suse.de> <20070417213954.GE11166@waste.org> <20070418031511.GA18452@wotan.suse.de> <20070418043831.GR11115@waste.org> <20070418050024.GF18452@wotan.suse.de> <20070418055525.GS11115@waste.org> <20070418152355.GU11115@waste.org> <20070418104814.2c6fdda7@chirp> In-Reply-To: <20070418104814.2c6fdda7@chirp> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit X-OriginalArrivalTime: 18 Apr 2007 19:28:00.0258 (UTC) FILETIME=[ACF9FE20:01C781EF] Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Mark Glines wrote: > One minor question: is it even possible to be completely fair on SMP? > For instance, if you have a 2-way SMP box running 3 applications, one of > which has 2 threads, will the threaded app have an advantage here? (The > current system seems to try to keep each thread on a specific CPU, to > reduce cache thrashing, which means threads and processes alike each > get 50% of the CPU.) I think the ideal in this case would be to have both threads on one cpu, with the other app on the other cpu. This gives inter-process fairness while minimizing the amount of task migration required. More interesting is the case of three processes on a 2-cpu system. Do we constantly migrate one of them back and forth to ensure that each of them gets 66% of a cpu? Chris