From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S261312AbUBYNGI (ORCPT ); Wed, 25 Feb 2004 08:06:08 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S261316AbUBYNGI (ORCPT ); Wed, 25 Feb 2004 08:06:08 -0500 Received: from intra.cyclades.com ([64.186.161.6]:26837 "EHLO intra.cyclades.com") by vger.kernel.org with ESMTP id S261312AbUBYNF7 convert rfc822-to-8bit (ORCPT ); Wed, 25 Feb 2004 08:05:59 -0500 Date: Wed, 25 Feb 2004 11:01:17 -0300 (BRT) From: Marcelo Tosatti X-X-Sender: marcelo@logos.cnet To: Mika =?ISO-8859-1?Q?=20Bostr=F6m?= Cc: linux-kernel@vger.kernel.org Subject: Re: Another case of lockup using combination of devices (sound+net) In-Reply-To: <20040224172310.GA6083@odin.lnet.lut.fi> Message-ID: References: <20040224172310.GA6083@odin.lnet.lut.fi> MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=ISO-8859-1 Content-Transfer-Encoding: 8BIT X-Cyclades-MailScanner-Information: Please contact the ISP for more information X-Cyclades-MailScanner: Found to be clean Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org On Tue, 24 Feb 2004, Mika Boström wrote: > I hope this bug report has all the required bits in it, I haven't done > this before. I tried to follow the instructions and included some > additional information usually seen here (.config, dmesg). On with the > bug report... > > > [1.] Complete hang when using sound and network simultaneously (2.6.x) > > [2.] I noticed that another person has had similar problems: > http://marc.theaimsgroup.com/?t=107699689400003&r=1&w=2 > While listening to music (xmms, ALSA output) I have had complete > lockups. There is no indication of a problem, and nothing ever gets > written to logs. When the system locks up, even keyboard goes > unresponsive: capslock, numlock, scroll-lock don't flash lights, magic > key does not register. Before the posting above I couldn't find any > reason, but my latest lockup happened while listening to music pulling > data in at a rate of about 6.5 MB/s. The transfer had lasted for perhaps > 10 seconds before system hung. > > This has occurred with all the 2.6 kernels I have tried: 2.6.0-test8, > 2.6.1, and now 2.6.3. Looking at the thread above I took notice that I > had previously had an enormous amount of interrupt errors (~120k/day). > Disabling IO-APIC from kernel configuration apparently solved that, now > they occur approximately 1 error/day. No difference there, the lockup > still takes place. ACPI has never been enabled. Sound card is C-Media > 8738 based Trust, NIC is a 3com 905B. ATI's proprietary kernel module > has not been loaded at any of these instances. (It broke all kinds of > things on its own.) Radeon and the 3com are sharing interrupt 11. That can cause problems. Try changing the interrupt of one of the cards.