From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S261423AbUBYRHI (ORCPT ); Wed, 25 Feb 2004 12:07:08 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S261421AbUBYRHI (ORCPT ); Wed, 25 Feb 2004 12:07:08 -0500 Received: from mail.acu.edu ([150.252.135.93]:9941 "EHLO nicanor.acu.edu") by vger.kernel.org with ESMTP id S261423AbUBYRHE (ORCPT ); Wed, 25 Feb 2004 12:07:04 -0500 Date: Wed, 25 Feb 2004 11:06:29 -0600 From: Michael Joy Subject: Re: 2.6.3 Boot Failure on Nforce2 Board In-reply-to: <403CCAA2.5070800@pobox.com> To: LKML Message-id: <1077728789.14230.3.camel@physx01.acu.edu> Organization: Abilene Christian University MIME-version: 1.0 X-Mailer: Ximian Evolution 1.4.5-2.norlug Content-type: text/plain Content-transfer-encoding: 7BIT References: <1077723571.9844.22.camel@physx01.acu.edu> <403CCAA2.5070800@pobox.com> Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org This problem is completely independent of nvnet as I'm not using it for 2.6.3.. the dmesg dump is from our WORKING configuration. I can't use forcedeth as the kernel won't even get to module loading to boot! Again, this has nothing to do with binary driver problems. Michael On Wed, 2004-02-25 at 10:17, Jeff Garzik wrote: > Michael Joy wrote: > > Hello, > > We're having an interesting problem with the latest kernel release. On > > an Albatron KM18G, latest bios, 1024MB system with athlon xp proc, 2.6.3 > > refuses to boot. It hangs on initializing the ide devices. > > > > [2.] Full description of the problem/report: The problem is most > > definately related to the ide controller changes made in 2.6.3 as in > > 2.4.22 we did not have this issue. We haven't tried any of the previous > > kernels as this is a production system. > > > > When booting the 2.6.3 kernel, either compiled by Mandrake (cooker) or > > using the straight up source, the kernel hangs without any error on hda: > > max request size : 128KiB. > > > > I don't have a log of this as it won't initialize the HD (wd1200jb on an > > 80pin cable) to log the dmesg dump. Anyways we have two identical > > machines that do this. Both are nforce2 integrated gpu's, using onboard > > networking and sound. They have 2x512 Kingston HyperX memory modules > > which have been thouroughly tested in these machines with memtest and no > > errors are found. > > > > Of note is that these machines exhibit the random freezes (blank screen, > > hard lock, normally associated with heavy disk thrashing) many other > > nforce2 boards seem to be experiencing. To fix this, we boot them with > > the noapic and nolapic option and the problem does not reappear. > > > i2c-nforce2 4392 0 (unused) > > > nvnet 30880 1 (autoclean) > > > You appear to have a binary-only module, nvnet, loaded... we cannot > debug problems with closed source code in your kernel. Try "forcedeth" > NIC driver instead. > > Jeff > -- Michael Joy Abilene Christian University