From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1750704AbWKBBvo (ORCPT ); Wed, 1 Nov 2006 20:51:44 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1752468AbWKBBvo (ORCPT ); Wed, 1 Nov 2006 20:51:44 -0500 Received: from rgminet01.oracle.com ([148.87.113.118]:6959 "EHLO rgminet01.oracle.com") by vger.kernel.org with ESMTP id S1750704AbWKBBvn (ORCPT ); Wed, 1 Nov 2006 20:51:43 -0500 Message-ID: <45494F18.4000406@oracle.com> Date: Wed, 01 Nov 2006 17:51:20 -0800 From: Zach Brown User-Agent: Thunderbird 1.5.0.7 (X11/20060913) MIME-Version: 1.0 To: Jeff Moyer CC: Andrew Morton , linux-kernel@vger.kernel.org Subject: Re: [PATCH] dio: lock refcount operations References: <20061027181735.18631.43565.sendpatchset@tetsuo.zabbo.net> <45426D3F.3040304@oracle.com> In-Reply-To: Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Brightmail-Tracker: AAAAAQAAAAI= X-Brightmail-Tracker: AAAAAQAAAAI= X-Whitelist: TRUE X-Whitelist: TRUE Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org > zach.brown> that path before the recent dio completion patch set. We > zach.brown> shouldn't expect significant performance regression from > zach.brown> returning to the behaviour that existed before the completion > zach.brown> clean up work. > > Are you going to quantify this at all? I think we should. I spotted some free time on an old dual athlon and got an initial look at the cpu cost of the dio cleanup patches currently in -mm. I ran two aio-stress instances doing O_DIRECT 64k sequential reads and writes to an existing 1gig file on ext3 on an old pata drive. I ran a pair of cycle soakers measuring cpu load every second. After trimming out the 8 highest and lowest samples the remaining samples were averaged. I did this three times against 2.6.19-rc4-mm1 before and after applying the dio-* patches. before: 5.02% 5.23% 5.27% after: 5.27% 5.33% 5.32% So I'm not *gravely* concerned that we've regressed, but it'd be nice to measure the impact of the dio-* patches on more capable hardware. - z