From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1030201AbXBLWIJ (ORCPT ); Mon, 12 Feb 2007 17:08:09 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1030202AbXBLWIJ (ORCPT ); Mon, 12 Feb 2007 17:08:09 -0500 Received: from smtp.osdl.org ([65.172.181.24]:43380 "EHLO smtp.osdl.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1030201AbXBLWIH (ORCPT ); Mon, 12 Feb 2007 17:08:07 -0500 Date: Mon, 12 Feb 2007 14:08:00 -0800 From: Andrew Morton To: Larry Finger Cc: linux-kernel@vger.kernel.org Subject: Re: Regress for 8139too in 2.6.20 Message-Id: <20070212140800.4d9378ad.akpm@linux-foundation.org> In-Reply-To: <45CFAED5.2090808@lwfinger.net> References: <45CFAED5.2090808@lwfinger.net> X-Mailer: Sylpheed version 2.2.4 (GTK+ 2.8.19; i686-pc-linux-gnu) Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org > On Sun, 11 Feb 2007 18:03:33 -0600 Larry Finger wrote: > In kernel version 2.6.20, my network interface, a Uniden PCN 300 PCMCIA card that uses the 8139too > driver fails to initialize with Yenta and 8139too as modules. It worked correctly in 2.6.19. Using > git bisect, the bad commit is shown below: > > finger@sonylap:~/linux-2.6> git bisect good > 8d610dd52dd1da696e199e4b4545f33a2a5de5c6 is first bad commit > commit 8d610dd52dd1da696e199e4b4545f33a2a5de5c6 > Author: Linus Torvalds > Date: Mon Dec 11 12:12:04 2006 -0800 > > Make sure we populate the initroot filesystem late enough > > We should not initialize rootfs before all the core initializers have > run. So do it as a separate stage just before starting the regular > driver initializers. > > Signed-off-by: Linus Torvalds > > :040000 040000 12bc13def90d15921d41d2b285854b3e157a970f f936e01f6b1b2b75478484861d34e81830c73860 M > include > :040000 040000 5673719c3f6b47b329cfc9554c112077634a9b19 9c2e768964af1500d62849836b9f8e801fe7f29e M > init > > > If Yenta is built into the kernel, rather than a module, it works again with 2.6.20. > That's a bit surprising - the initcall levels don't affect modules. Presumably something went wrong in core kernel which later caused yenta and/or 8139too to fail. Have you tried diffing the before- and after-dmesgs to see if that particular commit has caused any interesting differences? And are you able to test any other cardbus devices to see whether it's specific to 8139too?