From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1753877AbYAWL0Y (ORCPT ); Wed, 23 Jan 2008 06:26:24 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1751159AbYAWL0O (ORCPT ); Wed, 23 Jan 2008 06:26:14 -0500 Received: from gir.skynet.ie ([193.1.99.77]:40528 "EHLO gir.skynet.ie" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751849AbYAWL0K (ORCPT ); Wed, 23 Jan 2008 06:26:10 -0500 Date: Wed, 23 Jan 2008 11:24:36 +0000 From: Mel Gorman To: Andi Kleen Cc: mingo@elte.hu, linux-mm@kvack.org, linux-kernel@vger.kernel.org, apw@shadowen.org Subject: Re: [PATCH 0/2] Relax restrictions on setting CONFIG_NUMA on x86 II Message-ID: <20080123112436.GF21455@csn.ul.ie> References: <20080118153529.12646.5260.sendpatchset@skynet.skynet.ie> <200801231145.14915.andi@firstfloor.org> <20080123105757.GE21455@csn.ul.ie> <200801231215.56741.andi@firstfloor.org> MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-15 Content-Disposition: inline In-Reply-To: <200801231215.56741.andi@firstfloor.org> User-Agent: Mutt/1.5.13 (2006-08-11) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On (23/01/08 12:15), Andi Kleen didst pronounce: > > Anyways from your earlier comments it sounds like you're trying to add SRAT > parsing to CONFIG_NUMAQ. Since that's redundant with the old implementation > it doesn't sound like a very useful thing to do. > No, that would not be useful at all as it's redundant as you point out. The only reason to add it is if the Opteron box can figure out the CPU-to-node affinity. Right now everything gets dumped into node 0 where as x86_64 can figure it out properly. > But the patch is applied already i think. Well I'm sure it passed > checkpatch.pl at least. > :| The patches applied so far are about increasing test coverage, not SRAT messing. While there are still breakages for some boxen, more configurations should work on more machines than previously. Those using non-NUMA .configs should not notice the difference. -- Mel Gorman Part-time Phd Student Linux Technology Center University of Limerick IBM Dublin Software Lab