From mboxrd@z Thu Jan 1 00:00:00 1970 From: Daniel Vetter Subject: Re: [PATCH 8/8] drm/i915: move dp clock computations to encoder->compute_config Date: Fri, 5 Apr 2013 20:51:40 +0200 Message-ID: <20130405185140.GY2228@phenom.ffwll.local> References: <1364463723-30252-1-git-send-email-daniel.vetter@ffwll.ch> <1364463723-30252-9-git-send-email-daniel.vetter@ffwll.ch> <20130402142048.4ee9c2d3@jbarnes-desktop> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Return-path: Received: from mail-ee0-f53.google.com (mail-ee0-f53.google.com [74.125.83.53]) by gabe.freedesktop.org (Postfix) with ESMTP id 4E5B1E5BFB for ; Fri, 5 Apr 2013 11:48:43 -0700 (PDT) Received: by mail-ee0-f53.google.com with SMTP id c13so1465673eek.26 for ; Fri, 05 Apr 2013 11:48:42 -0700 (PDT) Content-Disposition: inline In-Reply-To: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: intel-gfx-bounces+gcfxdi-intel-gfx=m.gmane.org@lists.freedesktop.org Errors-To: intel-gfx-bounces+gcfxdi-intel-gfx=m.gmane.org@lists.freedesktop.org To: Jesse Barnes Cc: Intel Graphics Development List-Id: intel-gfx@lists.freedesktop.org On Tue, Apr 02, 2013 at 11:26:06PM +0200, Daniel Vetter wrote: > On Tue, Apr 2, 2013 at 11:20 PM, Jesse Barnes wrote: > > > On Thu, 28 Mar 2013 10:42:03 +0100 > > Daniel Vetter wrote: > > > > > With the exception of hsw, which has dedicated DP clocks which run at > > > the fixed frequency already, and vlv, which doesn't have optmized > > > pre-defined dp clock parameters (yet). > > > > > > > Nice. I think we should do this for common HDMI modes too. We have > > some extra clock manipulation regs we can use to tune things, so having > > fixed dividers for 720p and 1080p along with the tuning params should > > give us better behavior than what we have today. > > > Imo we still have a few lower-hanging fruit before we need to start doing > clock fine-tuning in hdmi-land: Atm we don't really bother with supporting > the 1001/1000 modified clocks in the CEA spec at all ... Once we have the > support code for that, adding fine-tuned clocks for those modes starts to > make sense, so that we really hit them spot-on. I've dropped this patch here again for now, since the previous patch is horribly broken: It completely missed to transform the ironlake dpll functions over to use pipe_config->dpll instead of the intel_clock_t computed. Thanks to Paulo for dissecting this embarassement. /me hides in shame Cheers, Daniel -- Daniel Vetter Software Engineer, Intel Corporation +41 (0) 79 365 57 48 - http://blog.ffwll.ch