From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S932282AbVJCPql (ORCPT ); Mon, 3 Oct 2005 11:46:41 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S932304AbVJCPql (ORCPT ); Mon, 3 Oct 2005 11:46:41 -0400 Received: from geryon.wsm.com ([12.163.128.2]:11454 "EHLO geryon.wsm.com") by vger.kernel.org with ESMTP id S932282AbVJCPqk (ORCPT ); Mon, 3 Oct 2005 11:46:40 -0400 Message-ID: <43415265.7000908@wsm.com> Date: Mon, 03 Oct 2005 08:46:45 -0700 From: Jeff Johnson Organization: Western Scientific, Inc User-Agent: Mozilla Thunderbird 1.0.6-1.1.fc4 (X11/20050720) X-Accept-Language: en-us, en MIME-Version: 1.0 To: linux-kernel@vger.kernel.org Subject: "Kernel panic: VRKADINT" under heavy I/O. What is the source or location of VRKADINT Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit X-Scan-Signature: 3c1be6617017a38cff1032ecd9727baf Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Greetings, I have several systems that will hang under heavy i/o. This usually occurs during a Pallas AllToAll test over Infiniband. Is anyone able to point me to where I can find VRKADINT or what general area of the kernel this is coming from? Everything runs fine otherwise. When I start the AllToAll across 32 machines and get 128 cores all talking to each other at once random machines will panic with the VRKADINT. I cannot seem to find this message anywhere in kernel source. details: arch x86_64 kernel 2.6.5-7.193-smp (SuSE Ent9) proc 2x dual core Opteron 275 ram 16GB --Jeff