From mboxrd@z Thu Jan 1 00:00:00 1970 From: Matthew Garrett Subject: Re: patch for video.c driver Date: Fri, 13 Mar 2009 20:52:56 +0000 Message-ID: <20090313205256.GA5854@srcf.ucam.org> References: <20090312154741.GA1055@hygelac> <1236910092.2807.103.camel@rzhang-dt> <20090313030030.GA23571@srcf.ucam.org> <20090313173604.GB13342@hygelac> <20090313174726.GA2888@srcf.ucam.org> <20090313175812.GE13342@hygelac> <20090313180646.GA3171@srcf.ucam.org> <20090313204544.GR16754@hygelac> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Return-path: Received: from cavan.codon.org.uk ([93.93.128.6]:46137 "EHLO vavatch.codon.org.uk" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750777AbZCMUxG (ORCPT ); Fri, 13 Mar 2009 16:53:06 -0400 Content-Disposition: inline In-Reply-To: <20090313204544.GR16754@hygelac> Sender: linux-acpi-owner@vger.kernel.org List-Id: linux-acpi@vger.kernel.org To: Terence Ripperda Cc: Zhang Rui , "lenb@kernel.org" , linux-acpi On Fri, Mar 13, 2009 at 03:45:44PM -0500, Terence Ripperda wrote: > NVIF methods implement both notification and query commands. so it's required > for generating the event and querying the firmware's preferred set of enabled > outputs. Ok, got you. In that case the best approach would be for you to add an input device to the kernel component of your driver which listens for the notification and then generates KEY_SWITCHVIDEOMODE. The right level of integration for the preferred set of screens would be to extend the video output class to support listing the outputs as well as just performing the switches, then register with that if the platform doesn't support DGS and co. We're trying to deprecate the proc interfaces, and in general platform-specific drivers should hook into generic interfaces rather than overloading the functionality of the existing ACPI drivers. Do you have any examples of machines that have this requirement? -- Matthew Garrett | mjg59@srcf.ucam.org