From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S267445AbUH1QmP (ORCPT ); Sat, 28 Aug 2004 12:42:15 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S267401AbUH1QlD (ORCPT ); Sat, 28 Aug 2004 12:41:03 -0400 Received: from parcelfarce.linux.theplanet.co.uk ([195.92.249.252]:9095 "EHLO www.linux.org.uk") by vger.kernel.org with ESMTP id S267438AbUH1QfZ (ORCPT ); Sat, 28 Aug 2004 12:35:25 -0400 Date: Sat, 28 Aug 2004 17:35:21 +0100 From: Matthew Wilcox To: Jon Smirl Cc: Matthew Wilcox , Greg KH , Jesse Barnes , Martin Mares , "Pallipadi, Venkatesh" , linux-pci@atrey.karlin.mff.cuni.cz, Alan Cox , Linux Kernel Mailing List , Petr Vandrovec , Benjamin Herrenschmidt Subject: Re: [PATCH] add PCI ROMs to sysfs Message-ID: <20040828163521.GD16196@parcelfarce.linux.theplanet.co.uk> References: <20040827164303.GW16196@parcelfarce.linux.theplanet.co.uk> <20040827222938.12618.qmail@web14922.mail.yahoo.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20040827222938.12618.qmail@web14922.mail.yahoo.com> User-Agent: Mutt/1.4.1i Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org On Fri, Aug 27, 2004 at 03:29:38PM -0700, Jon Smirl wrote: > I would think that it would return 4MB. If it can't find standard ROM > headers it should return the window size. The patch I just sent changes this ... if we can find PCI headers, we should return nothing. > Are you sure you are getting the correct contents of those ROMs? Would > it be worthwhile to try and get the author of the ROMs to add standard > ROM headers? The content you included doesn't look that useful unless > that is what IA64 instructions look like. It looks like nothing that makes any sense to me. Maybe it's a blanked EEPROM or something. This particular device is on the motherboard of a prototype machine, so there could be almost anything going on with it ;-) > If reading past 1MB for those ROMs causes a reboots, could something be > wrong in the IA64 fault handing code? No, that's normal behaviour on ia64 -- unacknowledged PCI reads cause a machine check rather than reading ffffffff like x86 does. -- "Next the statesmen will invent cheap lies, putting the blame upon the nation that is attacked, and every man will be glad of those conscience-soothing falsities, and will diligently study them, and refuse to examine any refutations of them; and thus he will by and by convince himself that the war is just, and will thank God for the better sleep he enjoys after this process of grotesque self-deception." -- Mark Twain