Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S933493AbXHOQdT (ORCPT ); Wed, 15 Aug 2007 12:33:19 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S933450AbXHOQbx (ORCPT ); Wed, 15 Aug 2007 12:31:53 -0400 Received: from netops-testserver-3-out.sgi.com ([192.48.171.28]:45734 "EHLO relay.sgi.com" rhost-flags-OK-OK-OK-FAIL) by vger.kernel.org with ESMTP id S933473AbXHOQbu (ORCPT ); Wed, 15 Aug 2007 12:31:50 -0400 Date: Wed, 15 Aug 2007 09:31:45 -0700 From: Paul Jackson To: "Serge E. Hallyn" Cc: Lee.Schermerhorn@hp.com, clameter@sgi.com, serge@hallyn.com, dhaval@linux.vnet.ibm.com, bob.picco@hp.com, nacc@us.ibm.com, kamezawa.hiroyu@jp.fujitsu.com, mel@skynet.ie, akpm@linux-foundation.org, balbir@in.ibm.com, vatsa@linux.vnet.ibm.com, linux-kernel@vger.kernel.org, ckrm-tech@lists.sourceforge.net Subject: Re: Regression in 2.6.23-rc2-mm2, mounting cpusets causes a hang Message-Id: <20070815093145.0c0480bb.pj@sgi.com> In-Reply-To: <20070815143125.GA11582@vino.hallyn.com> References: <20070813201215.GA16908@vino.hallyn.com> <1187103831.6281.24.camel@localhost> <20070814180339.GA32553@vino.hallyn.com> <1187115224.6281.40.camel@localhost> <20070814192306.GB32553@vino.hallyn.com> <20070814204951.GA2065@vino.hallyn.com> <1187127685.6281.139.camel@localhost> <1187185392.5422.13.camel@localhost> <20070815143125.GA11582@vino.hallyn.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 Content-Length: 1393 Lines: 31 Serge wrote: > Paul, is the mems stuff in cpusets only really useful for NUMA cases? I haven't been following this closely, but offhand, my take is that: 1) CONFIG_CPUSETS is primarily useful on SMP or NUMA systems. 2) It can be configured on non-NUMA systems (it still seems to need CONFIG_SMP ... not sure why.) 3) Kernels configured for CONFIG_CPUSETS can be (and often are) run on systems with just one CPU, or just one memory node. 4) To the user, running a CONFIG_CPUSETS enabled kernel w/o CONFIG_NUMA, or on a system with just one memory node should allow all the usual CPUSET operations, as if one just had one memory node, number '0'. If all systems in our universe only had one memory node, then cpusets wouldn't do memory -- it would not be useful. But for consistency of interface between various configured systems, it is useful to have cpusets working the same across both NUMA and non-NUMA systems. Some systems just happen to have only one memory node. -- I won't rest till it's the best ... Programmer, Linux Scalability Paul Jackson 1.925.600.0401 - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/