Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S261883AbUKJFFl (ORCPT ); Wed, 10 Nov 2004 00:05:41 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S261885AbUKJFFk (ORCPT ); Wed, 10 Nov 2004 00:05:40 -0500 Received: from omx2-ext.sgi.com ([192.48.171.19]:20622 "EHLO omx2.sgi.com") by vger.kernel.org with ESMTP id S261883AbUKJFFb (ORCPT ); Wed, 10 Nov 2004 00:05:31 -0500 Date: Wed, 10 Nov 2004 16:05:43 +1100 (EST) From: Mark Goodwin X-X-Sender: markgw@woolami.melbourne.sgi.com To: Matthew Dobson cc: Erich Focht , Jack Steiner , Takayoshi Kochi , linux-ia64@vger.kernel.org, LKML Subject: Re: Externalize SLIT table In-Reply-To: <1100044724.3980.23.camel@arrakis> Message-ID: References: <20041103205655.GA5084@sgi.com> <20041104.105908.18574694.t-kochi@bq.jp.nec.com> <20041104141337.GA18445@sgi.com> <200411041631.42627.efocht@hpce.nec.com> <1100029381.3980.12.camel@arrakis> <1100044724.3980.23.camel@arrakis> MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII; format=flowed Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 1429 Lines: 30 On Tue, 9 Nov 2004, Matthew Dobson wrote: > On Tue, 2004-11-09 at 12:34, Mark Goodwin wrote: >> Once again however, it depends on the definition of distance. For nodes, >> we've established it's the ACPI SLIT (relative distance to memory). For >> cpus, should it be distance to memory? Distance to cache? Registers? Or >> what? >> > That's the real issue. We need to agree upon a meaningful definition of > CPU-to-CPU "distance". As Jesse mentioned in a follow-up, we can all > agree on what Node-to-Node "distance" means, but there doesn't appear to > be much consensus on what CPU "distance" means. How about we define cpu-distance to be "relative distance to the lowest level cache on another CPU". On a system that has nodes with multiple sockets (each supporting multiple cores or HT "CPUs" sharing some level of cache), when the scheduler needs to migrate a task it would first choose a CPU sharing the same cache, then a CPU on the same node, then an off-node CPU (i.e. falling back to node distance). Of course, I have no idea if that's anything like an optimal or desirable task migration policy. Probably depends on cache-trashiness of the task being migrated. -- Mark - 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/