From mboxrd@z Thu Jan 1 00:00:00 1970 From: Imre Deak Subject: Re: [PATCH 00/13] drm/i915: backlight rewrite Date: Tue, 12 Nov 2013 23:22:20 +0200 Message-ID: <1384291340.2462.12.camel@ideak-mobl> References: Reply-To: imre.deak@intel.com Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Return-path: Received: from mga01.intel.com (mga01.intel.com [192.55.52.88]) by gabe.freedesktop.org (Postfix) with ESMTP id 9E097FA406 for ; Tue, 12 Nov 2013 13:22:34 -0800 (PST) In-Reply-To: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: intel-gfx-bounces@lists.freedesktop.org Errors-To: intel-gfx-bounces@lists.freedesktop.org To: Jani Nikula Cc: intel-gfx@lists.freedesktop.org List-Id: intel-gfx@lists.freedesktop.org On Fri, 2013-11-08 at 16:48 +0200, Jani Nikula wrote: > Hi all - > > This series is a successor to [1], but with too many changes to list so > I'll just consider this v1 of the patches. > > This is pretty much a rewrite of our backlight code. Highlights: > > * All the platform specific stuff in platform specific functions. The > code is much easier to follow, and it will be easier to enable future > platforms without breaking old ones all the time. > > * Backlight enable writes all the backlight registers based on > information retrieved at setup stage. We no longer depend on what's in > the registers prior to enable, which should make things more > deterministic, and we no longer need to save/restore backlight > registers. > > * We no longer keep reading max backlight value or combination mode bits > etc. from the registers all over the place. All that is based on the > initial values read in the setup stage. > > TODO: > > * If initial setup in registers is bogus, we should try to read the > values from VBT and/or try to use sane defaults. (This problem we have > already, not introduced by the series.) > > I've tested this so far on ILK and IVB, trying carefully keep it working > commit by commit to keep things bisectable. More testing across > platforms is very much needed. We have a history with backlight... > > Also available at backlight-rework branch at [2]. Looks like a nice improvement to code readability and separation of init vs. run-time HW programming steps. The diff is big and I only managed to review patches 1-9 so far; afaics those could be applied even separately. r-bs and some nitpicks inlined. --Imre