From mboxrd@z Thu Jan 1 00:00:00 1970 From: Bjorn Helgaas Subject: Re: EC driver registration (was: IPMI/ACPI: Locate the IPMI system interface in ACPI namespace) Date: Tue, 27 Oct 2009 12:14:00 -0600 Message-ID: <200910271214.00839.bjorn.helgaas@hp.com> References: <1256564026-9855-1-git-send-email-yakui.zhao@intel.com> <200910271105.10056.bjorn.helgaas@hp.com> <4AE73599.5050109@gmail.com> Mime-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit Return-path: Received: from g4t0014.houston.hp.com ([15.201.24.17]:43690 "EHLO g4t0014.houston.hp.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1756548AbZJ0SN5 (ORCPT ); Tue, 27 Oct 2009 14:13:57 -0400 In-Reply-To: <4AE73599.5050109@gmail.com> Content-Disposition: inline Sender: linux-acpi-owner@vger.kernel.org List-Id: linux-acpi@vger.kernel.org To: Alexey Starikovskiy Cc: "lenb@kernel.org" , "linux-acpi@vger.kernel.org" On Tuesday 27 October 2009 12:02:01 pm Alexey Starikovskiy wrote: > Currently EC driver work by doing either ECDT parse or acpi_walk_namespace, > Do you think it will be appropriate to change it to use PNP layer? I don't see this. In acpi_ec_init(), there's an acpi_bus_register_driver(). There's also the ECDT parsing in acpi_ec_ecdt_probe(). There's a comment near the acpi_ec_ecdt_probe() call that says it must happen before the EC is found in the namespace. There is an acpi_walk_namespace() inside acpi_ec_add(), but that's only to find _Qx methods under an EC device we've already found. It's conceivable that the EC driver could use PNP registration rather than ACPI registration -- that would let you get rid of the acpi_walk_resources() in ec_parse_device(). But I don't know that it's worth it right now. For IPMI, we're adding a bunch of new code and there are structural problems that EC doesn't have. So IPMI feels more urgent to me. Bjorn