From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id ; Wed, 22 May 2002 08:24:39 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id ; Wed, 22 May 2002 08:24:39 -0400 Received: from louise.pinerecords.com ([212.71.160.16]:11026 "EHLO louise.pinerecords.com") by vger.kernel.org with ESMTP id ; Wed, 22 May 2002 08:24:38 -0400 Date: Wed, 22 May 2002 14:24:28 +0200 From: Tomas Szepe To: Joris Braakman Cc: linux-kernel@vger.kernel.org Subject: Re: 2.4.18 SPARC SMP oops Message-ID: <20020522122428.GB26844@louise.pinerecords.com> In-Reply-To: <20020522131202.B6096@lama.euronet.nl> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.3.99i X-OS: Linux/sparc 2.2.21-rc4-ext3-0.0.7a SMP (up 6 days, 6:02) Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org > The following oops looks a lot like Tomas Szepe oops, so I copied the > subject. > > I could still login with ssh, su to root and issue a reboot, which > didn't work completely, but after a halt command I got the bootprompt > on the console and issued a boot. Other commands as ps or top didn't > work. ls did work. It happened under heavy load. The machine was running > for more than one day with this load. > > It is a sparc 5, single processor, 128MB. I assume this is indeed the well-known SRMMU bug. It's very easy to reproduce actually, just stress the machine so that the load reaches 20+. For a temporary and imperfect solution, look up the relevant thread on the aurora sparc mailing list. Someone is supposed to have been working on a fix for some time now. T.