From mboxrd@z Thu Jan 1 00:00:00 1970 From: Jeff Garzik Subject: Re: ISDN testing Date: Thu, 24 Apr 2008 22:39:07 -0400 Message-ID: <4811444B.1@garzik.org> References: <20080420223959.GA17998@havoc.gtf.org> <200804210205.38891.elendil@planet.nl> <480BF15D.2090100@garzik.org> <200804211536.57655.elendil@planet.nl> <480CE960.4050000@garzik.org> <878wz76lbr.fsf@basil.nowhere.org> Mime-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Cc: Frans Pop , linux-kernel@vger.kernel.org, netdev@vger.kernel.org, isdn4linux@listserv.isdn4linux.de To: Andi Kleen Return-path: Received: from srv5.dvmed.net ([207.36.208.214]:46238 "EHLO mail.dvmed.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753285AbYDYCjN (ORCPT ); Thu, 24 Apr 2008 22:39:13 -0400 In-Reply-To: <878wz76lbr.fsf@basil.nowhere.org> Sender: netdev-owner@vger.kernel.org List-ID: Andi Kleen wrote: > Jeff Garzik writes: >> Yeah, that's what my patchset does -- splits the "hisax" driver into >> multiple smaller drivers. Thus, "modprobe hisax" produces the results >> that you are seeing. > > That will break a lot of user's setup. At present, yes :/ Yet another why it's not upstream :) >> You'll need to determine which newly-created kernel module applies to >> your hardware. Let me know if you need help with that. The new >> drivers should be in >> /lib/modules/2.6.25-isdn/kernel/drivers/isdn/hisax/ provided that you >> enabled them in the kernel configuration. > > One way to avoid that issue would be to continue having a dummy > "hisax" module that just depends on all the other drivers with > a symbol reference. I'm definitely open to any solution like this. Unfortunately (as Frans, I think, pointed out) the hisax module routinely depends on critical setup info being passed as module options, mainly for non-PCI hardware. One solution might be to create a dummy hisax module that calls foo_module_init() and foo_module_exit(), and stitch them together in an unconventional way. Jeff