From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1755880AbXJaA3g (ORCPT ); Tue, 30 Oct 2007 20:29:36 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1753302AbXJaA32 (ORCPT ); Tue, 30 Oct 2007 20:29:28 -0400 Received: from netops-testserver-3-out.sgi.com ([192.48.171.28]:55041 "EHLO relay.sgi.com" rhost-flags-OK-OK-OK-FAIL) by vger.kernel.org with ESMTP id S1752608AbXJaA31 (ORCPT ); Tue, 30 Oct 2007 20:29:27 -0400 Date: Tue, 30 Oct 2007 17:29:23 -0700 From: Paul Jackson To: David Rientjes Cc: clameter@sgi.com, Lee.Schermerhorn@hp.com, akpm@linux-foundation.org, ak@suse.de, linux-kernel@vger.kernel.org Subject: Re: [patch 2/2] cpusets: add interleave_over_allowed option Message-Id: <20071030172923.d5d8dac3.pj@sgi.com> In-Reply-To: References: <1193434278.5032.106.camel@localhost> <20071026180713.aeedfac2.pj@sgi.com> <20071026194144.6042316a.pj@sgi.com> <20071027161955.e9d4d2db.pj@sgi.com> <20071028164637.8d1b45b0.pj@sgi.com> <20071028212745.57a7db67.pj@sgi.com> <20071029001554.c12b6b6c.pj@sgi.com> <20071030164414.b353e41c.pj@sgi.com> Organization: SGI X-Mailer: Sylpheed version 2.2.4 (GTK+ 2.8.3; i686-pc-linux-gnu) Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org David wrote: > Of course they have specific affinity needs, that's why they used > mempolicies. No. Good grief. If they are just looking for some set of memory banks, not to other node-specific hardware, then they might not need a specific node. Consider for example a multi-threaded, compute bound, long running scientific computation that has a substantial and fussy memory layout. Remapping it from one cpuset to another having the same NUMA topology may well work fine, once its memory caches recover. Reverting it to the lowest common denominator MPOL_DEFAULT policy because (Choice C) it no longer has access to its initial nodes might devastate its performance. pj wrote: > I sure wish I knew what real world, actual, not hypothetical, situations > were motivating this. I'm still wishing ... -- I won't rest till it's the best ... Programmer, Linux Scalability Paul Jackson 1.925.600.0401