From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S268415AbUHXWHE (ORCPT ); Tue, 24 Aug 2004 18:07:04 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S268435AbUHXWF7 (ORCPT ); Tue, 24 Aug 2004 18:05:59 -0400 Received: from atlrel8.hp.com ([156.153.255.206]:46293 "EHLO atlrel8.hp.com") by vger.kernel.org with ESMTP id S268365AbUHXV7B (ORCPT ); Tue, 24 Aug 2004 17:59:01 -0400 Subject: Re: [BK PATCH] I2C update for 2.6.8-rc1 From: Alex Williamson To: Greg KH Cc: torvalds@osdl.org, akpm@osdl.org, linux-kernel , sensors@Stimpy.netroedge.com In-Reply-To: <20040715000527.GA18923@kroah.com> References: <20040715000527.GA18923@kroah.com> Content-Type: text/plain Organization: LOSL Date: Tue, 24 Aug 2004 15:58:42 -0600 Message-Id: <1093384722.8445.10.camel@tdi> Mime-Version: 1.0 X-Mailer: Evolution 1.5.93 Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org On Wed, 2004-07-14 at 17:05 -0700, Greg KH wrote: > : > o I2C: patch quirks.c - SMBus hidden on hp laptop This particular patch, along w/ the new 20040715 ACPI drop has made my nc6000 laptop unusable. The problem is we're exposing a device that firmware considers hidden. The new motherboard driver in ACPI goes out and tries to claim resources to prevent them from being stepped on. It rightfully considers the hidden SMBus device a motherboard resource. The PCI code then stumbles onto this device, sees that the BAR it's using is unavailable and moves it somewhere else in the address space. At this point, I lose two for the three thermal zones on the laptop because the AML that deals with them assumes they haven't moved. I'm not sure what the point on un-hiding this devices it. ACPI sets up an OpRegion to access this device and should have exclusive access to that region. Letting a sensor driver poke at it may be fun, but I'd rather not fry my laptop. Can we drop the un-hiding of the SMBus for this laptop (probably the nc8000 too), or is there some way to make the ACPI motherboard driver and this quirk live together? Thanks, Alex -- Alex Williamson HP Linux & Open Source Lab