From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S261411AbVFNX2O (ORCPT ); Tue, 14 Jun 2005 19:28:14 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S261418AbVFNX2N (ORCPT ); Tue, 14 Jun 2005 19:28:13 -0400 Received: from graphe.net ([209.204.138.32]:61087 "EHLO graphe.net") by vger.kernel.org with ESMTP id S261411AbVFNX2C (ORCPT ); Tue, 14 Jun 2005 19:28:02 -0400 Date: Tue, 14 Jun 2005 16:27:55 -0700 (PDT) From: Christoph Lameter X-X-Sender: christoph@graphe.net To: Andi Kleen cc: christoph , akpm@osdl.org, linux-kernel@vger.kernel.org, shai@scalex86.org Subject: Re: [PATCH] Move some variables into the "most_readonly" section?? In-Reply-To: <20050614231911.GV11898@wotan.suse.de> Message-ID: References: <20050608131839.GP23831@wotan.suse.de> <20050614230411.GU11898@wotan.suse.de> <20050614231911.GV11898@wotan.suse.de> MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII X-Spam-Score: -5.9 Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org On Wed, 15 Jun 2005, Andi Kleen wrote: > > If these maps would start in the middle of a cacheline then additional > > cacheline fetches may become necessary to scan an array etc. > > But the CPUs do prefetching anyways for that. Do you have numbers > that this is actually worth it? Its not only for scanning an array. A struct may contains a lot of related information like for example boot_cpu_data. Aligning increases locality and the likelyhood that no additional cachelines have to be fetched.