From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S264766AbUEUAvu (ORCPT ); Thu, 20 May 2004 20:51:50 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S265217AbUEUAvu (ORCPT ); Thu, 20 May 2004 20:51:50 -0400 Received: from mail.fastclick.com ([205.180.85.17]:41410 "EHLO mail.fastclick.net") by vger.kernel.org with ESMTP id S264766AbUEUAvt (ORCPT ); Thu, 20 May 2004 20:51:49 -0400 Message-ID: <40AD52A4.3060607@fastclick.com> Date: Thu, 20 May 2004 17:51:48 -0700 From: "Brett E." Reply-To: brettspamacct@fastclick.com User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.4) Gecko/20030624 X-Accept-Language: en-us, en MIME-Version: 1.0 To: linux-kernel mailing list Subject: How can I optimize a process on a NUMA architecture(x86-64 specifically)? Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Say you have a bunch of single-threaded processes on a NUMA machine. Does the kernel make sure to prefer allocations using a certain CPU's memory, preferring to run a given process on the CPU which contains its memory? Or should I use the NUMA API(libnuma) to spell this out to the kernel? Does the kernel do the right thing in this case? Thanks, Brett