From: Andi Kleen Subject: Re: [PATCH V2 1/3] lib: introduce some memory copy macros and functions Date: Wed, 8 Sep 2010 15:05:19 +0200 Message-ID: <20100908150519.2f9fe283@basil.nowhere.org> References: <46238.91.60.152.217.1283948365.squirrel@www.firstfloor.org> <4C878823.60709@cn.fujitsu.com> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Cc: Peter Zijlstra , Ingo Molnar , Andrew Morton , "Theodore Ts'o" , Linux Kernel , Linux Ext4 , Linux Btrfs To: miaox@cn.fujitsu.com Return-path: Received: from one.firstfloor.org ([213.235.205.2]:49915 "EHLO one.firstfloor.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752741Ab0IHNFW (ORCPT ); Wed, 8 Sep 2010 09:05:22 -0400 In-Reply-To: <4C878823.60709@cn.fujitsu.com> Sender: linux-ext4-owner@vger.kernel.org List-ID: On Wed, 08 Sep 2010 20:57:07 +0800 Miao Xie wrote: > On Wed, 8 Sep 2010 14:19:25 +0200 (cest), Andi Kleen wrote: > > > >> According to the data, the length of the most copies is>=128. > > > > Thanks for the data. Large is easier to optimize than small, that's > > good. > > > > Could you also measure how many memsets need the backwards copy? > > (should be easy to add) > > I think memset doesn't need the backwards copy. I meant for memmove of course. Obviously memset doesn't need a backwards copy. That was just the only thing the script didn't measure because the original version didn't have memmove support. Your whole thread was about making memmove faster, right? -Andi -- ak@linux.intel.com -- Speaking for myself only.