From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S266113AbTGLQIr (ORCPT ); Sat, 12 Jul 2003 12:08:47 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S266164AbTGLQGu (ORCPT ); Sat, 12 Jul 2003 12:06:50 -0400 Received: from yankee.rb.xcalibre.co.uk ([217.8.240.35]:59523 "EHLO yankee.rb.xcalibre.co.uk") by vger.kernel.org with ESMTP id S266153AbTGLQGX (ORCPT ); Sat, 12 Jul 2003 12:06:23 -0400 Envelope-to: linux-kernel@vger.kernel.org From: Alistair J Strachan To: linux-kernel@vger.kernel.org Subject: APIC & ACPI on EPoX 8RDA+ (nForce 2) Date: Sat, 12 Jul 2003 17:21:05 +0100 User-Agent: KMail/1.5.9 MIME-Version: 1.0 Content-Disposition: inline Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Message-Id: <200307121721.05221.alistair@devzero.co.uk> Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Hi, A recent 2.5 kernel compiled with local and IO APIC support fails to boot on an EPoX 8RDA+ (nForce2) mainboard. It does not make any difference whether ACPI is enabled or disabled. Has anybody else had problems with this board and APIC? I've fiddled in the BIOS, but only disabling the BIOS APIC allows me to boot 2.5.75-mm1 (or, clearly, not compiling in APIC support). The second (unrelated) issue is that if I allow the ACPI in 2.5 to control PCI routing, I observe loads of IRQ dropouts on the USB 1.1 and USB 2 IRQs (5, 10, and 11 respectively). The number of interrupts shown in /proc/interrupts is an unrealistic constant for the IRQs and the devices (USB) are not initialised. irq 5: nobody cared! irq 10: nobody cared! irq 11: nobody cared! etc. Not compiing in ACPI support fixes it, but I also discovered that passing in pci=noacpi allows the USB devices to initialise and everything works just fine. I don't get the "nobody cared" messages if pci=noacpi is added to the cmdline. Are both of these known issues with the EPoX 8RDA+ mainboard? I've got the BIOS from the June 9th 2003, "06/09/2003", which I believe is the latest. TIA, Alistair.