From owner-freebsd-stable@FreeBSD.ORG Wed Sep 13 09:40:25 2006 Return-Path: X-Original-To: freebsd-stable@freebsd.org Delivered-To: freebsd-stable@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 4311C16A40F for ; Wed, 13 Sep 2006 09:40:25 +0000 (UTC) (envelope-from robert@mpe.mpg.de) Received: from mpemail.mpe.mpg.de (mpemail.mpe-garching.mpg.de [130.183.137.110]) by mx1.FreeBSD.org (Postfix) with ESMTP id 5948743D82 for ; Wed, 13 Sep 2006 09:40:15 +0000 (GMT) (envelope-from robert@mpe.mpg.de) Received: from robert3.mpe-garching.mpg.de ([130.183.136.84]) by mpemail.mpe.mpg.de with esmtpsa (TLS-1.0:DHE_RSA_AES_256_CBC_SHA:32) (Exim 4.50) id 1GNRE4-0002J2-Sq for freebsd-stable@freebsd.org; Wed, 13 Sep 2006 11:40:13 +0200 Message-ID: <4507D1F8.7040508@mpe.mpg.de> Date: Wed, 13 Sep 2006 11:40:08 +0200 From: Klaus Robert Suetterlin User-Agent: Mozilla/5.0 (X11; U; SunOS sun4v; en-US; rv:1.7) Gecko/20060629 X-Accept-Language: en-us, en MIME-Version: 1.0 To: freebsd-stable@freebsd.org Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit X-Authenticated-Id: robert Subject: PAE-SMP kernel generates NMI (PCI bus errors) X-BeenThere: freebsd-stable@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: Production branch of FreeBSD source code List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Wed, 13 Sep 2006 09:40:25 -0000 Hi all, I have a legacy system --- 4xXeon (PIII, 700MHz, 2MB cache), 16GB ECC RAM and 300GB Harddisk on an ICP Vortex Raid Controller --- which I tried to bring from RedHat6 to something more current and supported. (Un)Fortunately the only system that wanted to install on the machine is FreeBSD6. When I use a non-SMP kernel either with or without PAE everything works ok. When I use a SMP non-PAE kernel everything works ok, too. When I use a SMP+PAE kernel the system starts throwing NMIs (most likely due to PCI system bus errors, which is the only error reporting activated in BIOS) until the system finally freezes (which is most likely when the system event log of the server is full). I can give a lot more diagnostics to anyone interested in solving this, but I need a little handholding. Regards, Robert S