From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S264696AbTD0RRW (ORCPT ); Sun, 27 Apr 2003 13:17:22 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S264698AbTD0RRW (ORCPT ); Sun, 27 Apr 2003 13:17:22 -0400 Received: from franka.aracnet.com ([216.99.193.44]:24248 "EHLO franka.aracnet.com") by vger.kernel.org with ESMTP id S264696AbTD0RRV (ORCPT ); Sun, 27 Apr 2003 13:17:21 -0400 Date: Sun, 27 Apr 2003 10:29:31 -0700 From: "Martin J. Bligh" To: Mike Galbraith cc: linux-kernel@vger.kernel.org Subject: Re: Houston, I think we have a problem Message-ID: <32170000.1051464570@[10.10.2.4]> In-Reply-To: <5.2.0.9.2.20030427191459.00caed60@pop.gmx.net> References: <5.2.0.9.2.20030427090009.01f89870@pop.gmx.net> <5.2.0.9.2.20030427090009.01f89870@pop.gmx.net> <5.2.0.9.2.20030427191459.00caed60@pop.gmx.net> X-Mailer: Mulberry/2.2.1 (Linux/x86) MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Content-Disposition: inline Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org >> > To reproduce this 100% of the time, simply compile virgin 2.5.68 >> > up/preempt, reduce your ram to 128mb, and using gcc-2.95.3 as to not >> > overload the vm, run a make -j30 bzImage in an ext3 partition on a >> > P3/500 single ide disk box. No, you don't really need to meet all of >> > those restrictions... you'll see the problem on a big hairy chested >> > box as well, just not as bad as I see it on my little box. The first >> > symptom of the problem you will notice is a complete lack of swap >> > activity along with highly improbable quantities of unused ram were >> > all those hungry cc1's getting regular CPU feedings. >> >> Yes, that's why I don't use ext3 ;-) It's known broken, akpm is fixing >> it. > > I'm not at all convinced (must say I wouldn't mind at _all_ being > convinced) that it's ext3... that just _seems_ to be worst easily > reproducible case for some un-(expletive deleted)-known reason. Well, that's easy to test. Mount the fs as ext2, and see if it goes away. If it's your rootfs, it's too stupid to sniff your fstab at r/w remount time (or give you the correct info from "mount", so be careful), so you have to append rootfstype=ext2 to the kernel command line. M.