From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S264528AbTLQTvA (ORCPT ); Wed, 17 Dec 2003 14:51:00 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S264526AbTLQTu7 (ORCPT ); Wed, 17 Dec 2003 14:50:59 -0500 Received: from mail3.bluewin.ch ([195.186.1.75]:31175 "EHLO mail3.bluewin.ch") by vger.kernel.org with ESMTP id S264415AbTLQTu5 (ORCPT ); Wed, 17 Dec 2003 14:50:57 -0500 Date: Wed, 17 Dec 2003 20:49:51 +0100 From: Roger Luethi To: Rik van Riel Cc: Andrew Morton , Andrea Arcangeli , wli@holomorphy.com, kernel@kolivas.org, chris@cvine.freeserve.co.uk, linux-kernel@vger.kernel.org, mbligh@aracnet.com Subject: Re: 2.6.0-test9 - poor swap performance on low end machines Message-ID: <20031217194950.GA9375@k3.hellgate.ch> Mail-Followup-To: Rik van Riel , Andrew Morton , Andrea Arcangeli , wli@holomorphy.com, kernel@kolivas.org, chris@cvine.freeserve.co.uk, linux-kernel@vger.kernel.org, mbligh@aracnet.com References: <20031216112307.GA5041@k3.hellgate.ch> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: X-Operating-System: Linux 2.6.0-test11 on i686 X-GPG-Fingerprint: 92 F4 DC 20 57 46 7B 95 24 4E 9E E7 5A 54 DC 1B X-GPG: 1024/80E744BD wwwkeys.ch.pgp.net User-Agent: Mutt/1.5.4i Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org On Wed, 17 Dec 2003 13:53:28 -0500, Rik van Riel wrote: > On Tue, 16 Dec 2003, Roger Luethi wrote: > > > One potential problem with the benchmarks is that my test box has > > just one bar with 256 MB RAM. The kbuild and efax tests were run with > > mem=64M and mem=32M, respectively. If the difference between mem=32M > > OK, I found another difference with 2.4. > > Try "echo 256 > /proc/sys/vm/min_free_kbytes", I think > that should give the same free watermarks that 2.4 has. I played around with that knob after wli posted his findings in the "mem=16MB laptop testing" thread. IIRC tweaking min_free_kbytes didn't help nearly as much as I had hoped. I'm running the efax benchmark right now just to make sure. It's going to take a couple of hours, I'll follow up with results. FWIW akpm posted a patch to initialize min_free_kbytes depending on available RAM which seemed to make sense but it hasn't made it into mainline yet. > Using 1MB as the min free watermark for lowmem is bound > to result in more free (and less used) memory on systems > with less than 128 MB RAM ... significantly so on smaller > systems. Possibly. If memory pressure is high enough, though, the allocator ignores the watermarks. And on the other end kswapd seems to be pretty busy anyway during the benchmarks. Roger