From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S262036AbULHGBj (ORCPT ); Wed, 8 Dec 2004 01:01:39 -0500 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S262037AbULHGBj (ORCPT ); Wed, 8 Dec 2004 01:01:39 -0500 Received: from willy.net1.nerim.net ([62.212.114.60]:32522 "EHLO willy.net1.nerim.net") by vger.kernel.org with ESMTP id S262036AbULHGB3 (ORCPT ); Wed, 8 Dec 2004 01:01:29 -0500 Date: Wed, 8 Dec 2004 06:48:34 +0100 From: Willy Tarreau To: Simon Byrnand Cc: linux-kernel@vger.kernel.org Subject: Re: 2.4.27 -> 2.4.28 breaks i810-tco watchdog timer Message-ID: <20041208054834.GD17946@alpha.home.local> References: <17545.210.54.153.131.1102449575.squirrel@210.54.153.131> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <17545.210.54.153.131.1102449575.squirrel@210.54.153.131> User-Agent: Mutt/1.4i Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Hi, On Wed, Dec 08, 2004 at 08:59:35AM +1300, Simon Byrnand wrote: (...) > e400-e47f : motherboard > e800-e81f : motherboard > ec00-ec3f : motherboard > f000-f00f : Intel Corp. 82801DB Ultra ATA Storage Controller > f000-f007 : ide0 > f008-f00f : ide1 > > Clearly the IO range the driver is trying to open is already in use by > "motherboard". If I check another almost identical machine still running > 2.4.27 but with the watchdog timer unloaded, /proc/ioports gives: grep -r shows that it's ACPI which declares this "motherboard" name. I seem to remember about a change in the ACPI resource reservation. You might want to check with Len or Marcelo. I believe you cannot disable it if you want to use HT. (...) > For some reason a blanket range of IO addresses are being allocated which > include the range needed by the i810-tco watchdog timer. Why ? I don't see > any changes to the code for the watchdog timer itself so it seems that > something else is allocating that range first. Does anyone know of a patch > to fix this ? Running without a watchdog timer until the next kernel > version doesnt appeal :( (I can't go back to 2.4.27 due to other problems > with that version) Maybe you can hack the driver if you absolutely need to load 2.4.28+TCO. Simply comment out the "return -EIO" at line 402. It will still print the message but should load. Regards, Willy