From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S261875AbULGSVE (ORCPT ); Tue, 7 Dec 2004 13:21:04 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S261877AbULGSVD (ORCPT ); Tue, 7 Dec 2004 13:21:03 -0500 Received: from hirsch.in-berlin.de ([192.109.42.6]:39808 "EHLO hirsch.in-berlin.de") by vger.kernel.org with ESMTP id S261875AbULGSU0 (ORCPT ); Tue, 7 Dec 2004 13:20:26 -0500 X-Envelope-From: kraxel@bytesex.org To: Eyal Lebedinsky , Michael Hunold , Andrew Morton Cc: Linus Torvalds , Kernel Mailing List Subject: Re: Linux 2.6.10-rc3 oops when 'modprobe -r dvb-bt8xx' References: <41B1BD24.4050603@eyal.emu.id.au> From: Gerd Knorr Organization: SUSE Labs, Berlin Date: 07 Dec 2004 19:03:57 +0100 In-Reply-To: <41B1BD24.4050603@eyal.emu.id.au> Message-ID: <87653ex9wy.fsf@bytesex.org> User-Agent: Gnus/5.09 (Gnus v5.9.0) Emacs/21.3 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 Eyal Lebedinsky writes: > Linus Torvalds wrote: > > Please do test this - and don't send me anything but bug-fixes. Let's aim > > for a real 2.6.10 before xmas (or hanukkah, or whatever your favourite > > holiday happens to be). > > In the spirit of festive testing I would like to say that the oops that I > enjoyed throughout rc2-bk* is still present in -rc3. -mm series does not > have this problem. Oh joy. I think thats a different one. And I suspect there are a few more of that kind which simply didn't show up yet because the user base is too small. The current way of doing the initialization of the frontend devices (in Linus' tree) is simply a big mess and you may get all sorts of strange effects depending on the order you load or unload the modules if one of the parties involved doesn't get it right. > modprobe -r dvb-bt8xx > [] mt352_detach_client+0x52/0x54 [mt352] What happens if you "rmmod mt352" first? Guess it works without oopsing then? I somehow feel the best way to deal with that is to merge the redesigned frontend handing pending in -mm at the moment into Linus tree _now_, that should kill that whole class of bugs. That may result in the dvb subsystem not being that stable in 2.6.10. But dvb not being rock solid in 2.6.10 will very likely happen anyway as the code currently in Linus' tree isn't very stable as well. I think the chance that it gets even worse is small enougth that we can take the risk. Additional bonus would be that we don't get bugreports for the old code base which is already obsolete (and nobody wants to work on because of that). Michael? Gerd -- #define printk(args...) fprintf(stderr, ## args)