From mboxrd@z Thu Jan 1 00:00:00 1970 From: Thomas Renninger Subject: Re: [Devel] [PATCH 1/2] ACPICA: Introduce acpi_os_phys_table_override function Date: Wed, 31 Aug 2011 11:28:49 +0200 Message-ID: <201108311128.50256.trenn@suse.de> References: <1314179290-31526-1-git-send-email-trenn@suse.de> <1314179290-31526-2-git-send-email-trenn@suse.de> <1314758622.6486.74.camel@minggr.sh.intel.com> Mime-Version: 1.0 Content-Type: Text/Plain; charset="utf-8" Content-Transfer-Encoding: 7bit Return-path: Received: from cantor2.suse.de ([195.135.220.15]:57131 "EHLO mx2.suse.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754812Ab1HaJ2x (ORCPT ); Wed, 31 Aug 2011 05:28:53 -0400 In-Reply-To: <1314758622.6486.74.camel@minggr.sh.intel.com> Sender: linux-acpi-owner@vger.kernel.org List-Id: linux-acpi@vger.kernel.org To: Lin Ming Cc: "lenb@kernel.org" , "eric.piel@tremplin-utc.net" , "jnelson-suse@jamponi.net" , "x86@kernel.org" , "linux-kernel@vger.kernel.org" , "linux-acpi@vger.kernel.org" , "hpa@zytor.com" , "devel@acpica.org" On Wednesday, August 31, 2011 04:43:42 AM Lin Ming wrote: > On Wed, 2011-08-24 at 17:48 +0800, Thomas Renninger wrote: ... > You add a new interface. Yes, is this a bigger problem? > Can we just extend the existing interface: acpi_os_table_override? Not sure how to do that without OS/ACPICA API changes. The virtual address handling is nasty. You have to differ early mappings (early_ioremap) and later mappings (io/memremap). Re-mapping later is not possible because the physical address is lost with the current overriding interface. The physical address usage is transparent and from what I can see the only way to provide proper table overriding. If it's ok to add more paramters to acpi_os_table_override and either pass the virtual (as before) or the physical address, this would work: acpi_os_table_override(struct acpi_table_header *existing_table, struct acpi_table_header **new_table, acpi_physical_address *address, u32 *table_length); This would be an interface change which looked even worse to me, than adding a new function. Thomas