From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S263622AbTKKRA1 (ORCPT ); Tue, 11 Nov 2003 12:00:27 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S263625AbTKKRA1 (ORCPT ); Tue, 11 Nov 2003 12:00:27 -0500 Received: from fw.osdl.org ([65.172.181.6]:3006 "EHLO mail.osdl.org") by vger.kernel.org with ESMTP id S263622AbTKKRA0 (ORCPT ); Tue, 11 Nov 2003 12:00:26 -0500 Date: Tue, 11 Nov 2003 09:00:23 -0800 (PST) From: Linus Torvalds To: Pascal Schmidt cc: linux-kernel@vger.kernel.org Subject: Re: 2.9test9-mm1 and DAO ATAPI cd-burning corrupt In-Reply-To: Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org On Tue, 11 Nov 2003, Pascal Schmidt wrote: > > Well, that person is me and I tried making it work with ide-cd. Got read > support to work, submitted to Jens, you have it in your kernel. No luck > with write support. I could get it to mount read-write and data actually > made it to disk, but umount lead to a BUG_ON. Hmm.. That looks "impossible", so it's most likely a serious bh corruption issue. But what corrupts it is hard to guess at - the actual trace to that point has nothing to do with the MO drive itself, so you've either hit a generic ext2 bug (hey, it's possible, I guess, but sounds very unlikely), or the ide-scsi driver has corrupted memory. The latter is obviously the more likely schenario, but it does require somebody with the MO device to actually try to figure out how the corruption occurs.. Probably with a big printk buffer and a lot of printk's.. Linus