From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S267749AbUHWUYX (ORCPT ); Mon, 23 Aug 2004 16:24:23 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S267509AbUHWUXr (ORCPT ); Mon, 23 Aug 2004 16:23:47 -0400 Received: from ee.oulu.fi ([130.231.61.23]:7413 "EHLO ee.oulu.fi") by vger.kernel.org with ESMTP id S267490AbUHWTeS (ORCPT ); Mon, 23 Aug 2004 15:34:18 -0400 Date: Mon, 23 Aug 2004 22:34:16 +0300 From: Pekka Pietikainen To: linux-kernel@vger.kernel.org Subject: Re: Broadcom 4401 problem Message-ID: <20040823193416.GA21390@ee.oulu.fi> References: <20040822205346.GA17895@em.ca> <20040822221734.GA10372@ee.oulu.fi> <20040823155712.GA21840@em.ca> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Disposition: inline In-Reply-To: <20040823155712.GA21840@em.ca> User-Agent: Mutt/1.4.2i Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org On Mon, Aug 23, 2004 at 09:57:12AM -0600, Bruce Guenter wrote: > On Mon, Aug 23, 2004 at 01:17:34AM +0300, Pekka Pietikainen wrote: > > Could you try the driver from http://www.ee.oulu.fi/~pp/b44-095-2.tgz , > > which has some fixes that have been submitted but not yet merged. > In order to try to get it to compile, I removed the PCI table entry for > the BCM4713 and made every conditional that depends on it false (by > removing the appropriate code). However, when trying to ifconfig the > interface I get: > SIOCSIFFLAGS: Cannot allocate memory > SIOCSIFFLAGS: Cannot allocate memory Oh right... Yea, the code assumes that someone added the 4713 id (0x4713 ;) ) to pci_ids.h. Interesting that it runs out of memory though. The new version does use quite a bit of GFP_DMA (== memory under 16MB since there's no way of allocating memory under 1GB, which would do.). Apart from some legacy ISA stuff I don't think anything else should be sucking it up, so there should be enough, at least after a fresh boot. > > If that > > doesn't help, the broadcom driver > > ( http://www.broadcom.com/drivers/downloaddrivers.php ) might be > > worth a try. > The Broadcom Linux driver version 3.0.7 also locks up on me, but it > doesn't support "ethtool -A", so I have to stop and start the interface > completely to get it working again. Whew, maybe it's not a driver issue. Only thing I can think of is trying a few different kernels (say 2.6.7, 2.4.x if that fails), acpi=off, pci=noacpi on the command line etc. -- Pekka Pietikainen