Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S964903AbVJZT42 (ORCPT ); Wed, 26 Oct 2005 15:56:28 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S964904AbVJZT41 (ORCPT ); Wed, 26 Oct 2005 15:56:27 -0400 Received: from ms005msg.fastwebnet.it ([213.140.2.50]:55982 "EHLO ms005msg.fastwebnet.it") by vger.kernel.org with ESMTP id S964903AbVJZT41 (ORCPT ); Wed, 26 Oct 2005 15:56:27 -0400 Date: Wed, 26 Oct 2005 21:56:32 +0200 From: Paolo Ornati To: Dave Airlie Cc: Linux Kernel Mailing List Subject: Re: [2.6.14-rc? on x86-64] Total machine freeze Message-ID: <20051026215632.447c96e4@localhost> In-Reply-To: <21d7e9970510260320w5d9e9e6fvcfdf16f29366a03f@mail.gmail.com> References: <20051022124306.36d13c39@localhost> <21d7e9970510260320w5d9e9e6fvcfdf16f29366a03f@mail.gmail.com> X-Mailer: Sylpheed-Claws 1.9.13 (GTK+ 2.6.8; x86_64-pc-linux-gnu) Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 1913 Lines: 55 On Wed, 26 Oct 2005 20:20:44 +1000 Dave Airlie wrote: > > 2.6.14-rc[45] works fine here, usually. But sometimes (~2 times in a > > week) I get an hard-freeze: > > Sys-RQ doesn't work > > machine doesn't reply to ping / ssh > > nothing in the logs > > > > Seems like it loops somewhere with interrupt disabled... but I don't > > know. > > > Does it get into X and crash?, did it do it with 2.6.13? I got every crash under X so far (but I run X all the time so I can't tell if it is related to X or not). The crash doesn't happen when X starts, it happens at RANDOM times, without any particular workload. Now it's some days that I don't see the crash (from when I updated to 2.6.14-rc5-gd475f3f4)... so five days without crash. I'm waiting to see if it happens again. The last kernel that crashed was 2.6.14-rc5-g93918e9a. I don't remember to have seen this crash with 2.6.13... so I suppose 2.6.13 is OK. > > Do you have any "ricer" X options turned on like AGPFastWrite.. No. $ grep "AGP Fast" /var/log/Xorg.0.log (II) RADEON(0): AGP Fast Write disabled by default I have other issues with AGP Fast Write: if it is DISABLED in BIOS I can enable it in Xorg config without problems. If it's enabled in BIOS and I enable it in Xorg config I get this: when X starts the screen goes in SUSPEND mode and X eats all the CPU (I've seen this whith ssh). X is unkillable and I cannot debug it with gdb... only a reboot kills it! But this is another issue. The current AGP Fast Write config is: - enabled in BIOS (don't ask me why...) - disabled in xorg.conf -- Paolo Ornati Linux 2.6.14-rc5-gd475f3f4 on x86_64 - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/