From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S934188AbXGKVEi (ORCPT ); Wed, 11 Jul 2007 17:04:38 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S933978AbXGKVCk (ORCPT ); Wed, 11 Jul 2007 17:02:40 -0400 Received: from opinicus.com ([24.73.193.242]:46238 "EHLO thing2.opinicus.com" rhost-flags-OK-OK-OK-FAIL) by vger.kernel.org with ESMTP id S933886AbXGKVCj (ORCPT ); Wed, 11 Jul 2007 17:02:39 -0400 X-Greylist: delayed 725 seconds by postgrey-1.27 at vger.kernel.org; Wed, 11 Jul 2007 17:02:39 EDT Message-ID: <46954273.9090207@opinicus.com> Date: Wed, 11 Jul 2007 16:49:55 -0400 From: William Montgomery User-Agent: Mozilla Thunderbird 1.0.6 (X11/20050716) X-Accept-Language: en-us, en MIME-Version: 1.0 To: linux-kernel@vger.kernel.org Subject: e100 driver hard lockups with AEI-P430C 4 port card Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Hello, The AEI-P430C PCI 10/100 Quad Fast Ethernet card causes two computers to lockup hard but runs OK on a third computer. I have spent several days trying to determine the cause. I have used the following OS/driver combos: Slackware 10.1 Linux kernel 2.6.17 / eepro100 / e100 Slackware 10.1 Linux kernel 2.6.13 / eepro100 / e100 Slackware 9.1 Linux kernel 2.4.22 / e100 The following computer hardware configurations were used: PC_1: Rackmount Industrial PC (Advantech 4U chassis) 12 slot PCI backplane (Advantech) Pentium 4 2.80GHz, 512M Ram PCMIG (Advantech) PC_2: Rackmount Industrial PC (Axiomtek 4U chassis) 12 slot PCI backplane (Axiomtek) Pentium 4 3.00GHz, 512M Ram PCMIG (Axiomtek) PC_3: Desktop (ShuttleX) 1 AGP, 1 PCI slot motherboard with Pentium 4 2.80GHz, 512M Ram The hard lockups occur on PC_1 and PC_2 but not on PC_3. I have used all three OS/driver combos on PC_1 and PC_2 to try to eliminate the OS/driver as the problem; in each case the lockup is repeatable within a few minutes of heavy traffic. The test script uses the "iperf" utility which measures bandwidth on an ethernet port by moving as much data as possible in 10 seconds then computing bandwidth. The script runs iperf on 3 ports simultaneously and repeats until killed. I have tested the AEI card by connecting to 3 separate Intel PRO/100 S cards and (in trying a different configuration) by connecting it to a Routerboard RB44 (4 port card). I have tried the obvious methods to get an oops but no success so far. I have enabled the Magic SysRq key and enabled the NMI watchdog. The lockup is so hard that none of these works. The only thing that brings the CPU out of the lockup is the i810 watchdog but that causes a reset/reboot and no oops so I cant see any backtrace info. Any suggestions would be welcome. Regards, William Montgomery