From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S265999AbUBCNUj (ORCPT ); Tue, 3 Feb 2004 08:20:39 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S266000AbUBCNUj (ORCPT ); Tue, 3 Feb 2004 08:20:39 -0500 Received: from fmr04.intel.com ([143.183.121.6]:35032 "EHLO caduceus.sc.intel.com") by vger.kernel.org with ESMTP id S265999AbUBCNUh (ORCPT ); Tue, 3 Feb 2004 08:20:37 -0500 Subject: Re: IRQ 9: nobody cared ;_; From: Len Brown To: DaMouse Networks Cc: linux-kernel@vger.kernel.org In-Reply-To: <20040203064920.087d5546@EozVul.WORKGROUP> References: <20040203064920.087d5546@EozVul.WORKGROUP> Content-Type: text/plain Organization: Message-Id: <1075814415.13728.75.camel@dhcppc4> Mime-Version: 1.0 X-Mailer: Ximian Evolution 1.2.3 Date: 03 Feb 2004 08:20:15 -0500 Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org > ACPI: INT_SRC_OVR (bus 0 bus_irq 9 global_irq 9 dfl dfl) > NR Log Phy Mask Trig IRR Pol Stat Dest Deli Vect: > 09 003 03 0 1 0 1 0 1 1 71 Looks like we're setting up IRQ9 correctly -- level/low -- which is what default/default is. Other systems break if dfl/dfl is not interpreted as level/low. > 9: 100000 0 IO-APIC-level acpi But ACPI (alone on the IRQ) is getting an interrupt storm -- IRQ9 is low and ACPI doesn't know why. Please verify that you've got the latest BIOS for this box, and send along 1. /proc/interrupts and demsg from booting same kernel with "acpi=off" 2. output from dmidecode available in /usr/sbin/, or here: http://ftp.kernel.org/pub/linux/kernel/people/lenb/acpi/utils/ i'll send a patch later to force IRQ9 to level/high -- for if the ACPI interrupt works properly with that setting, then we know we've mis-interpreted what we think should be dfl/dfl on this box. thanks, -Len