From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S263695AbUAGXtV (ORCPT ); Wed, 7 Jan 2004 18:49:21 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S263666AbUAGXtV (ORCPT ); Wed, 7 Jan 2004 18:49:21 -0500 Received: from smtp-out7.blueyonder.co.uk ([195.188.213.10]:17452 "EHLO smtp-out7.blueyonder.co.uk") by vger.kernel.org with ESMTP id S265508AbUAGXsR (ORCPT ); Wed, 7 Jan 2004 18:48:17 -0500 From: James Cort To: linux-kernel@vger.kernel.org Subject: Stability problem on a Gigabyte motherboard Date: Wed, 7 Jan 2004 23:52:38 +0000 User-Agent: KMail/1.5.4 MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Content-Disposition: inline Message-Id: <200401072352.38578.jim@whitepost.org.uk> X-OriginalArrivalTime: 07 Jan 2004 23:48:32.0980 (UTC) FILETIME=[C2575140:01C3D578] Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org I've been using Linux for ages now, and I think the time has finally come to ask a question in this forum. Doubtless I'm totally wrong and will be flamed into oblivion, but hey... I have a Gigabyte GA-7VAXP motherboard, and I'm using kernel 2.4.23. This board is based on a VIA-KT400 chipset, with UDMA-133 and USB 2 support. Versions of the BIOS prior to F11 are stable but USB 2 support is broken. Later BIOS versions have fixed USB 2 support, but render my system unstable - va-ctcs dies in under 10 seconds, locking the entire machine such that there are no logs, no visible kernel panic to console and the magic sys-rq key doesn't work. Similar symptoms can be observed doing heavy disk writes to a USB 2 external hard disk or decompressing a large archive (such as the kernel source) to the internal IDE disk. I've searched the mailing list archives, asked Gigabyte, Googled, asked my local LUG and looked at the VIA web forums, all to no avail. So now I ask the good people here - are there any known issues with VIA chipset support which might cause this? TBH, the lack of logs I'm getting means I'm not sure if the problem is because of lousy BIOS writing on the part of Gigabyte or a genuine bug in the kernel. However, any helpful suggestions will be most gratefully received. James Cort -- Conserve wildlife - pickle a squirrel