[Intel-gfx] [PATCH] drm/i915: Prioritize SSC quirk table when determining SSC settings

[Michel Alexandre Salim]

On Wed, 2011-11-09 at 14:20 +0000, Chris Wilson wrote:
> On Wed, 09 Nov 2011 15:11:40 +0100, Michel Alexandre Salim <salimma at fedoraproject.org> wrote:
> > From ab11cf4b1b28d0abb1db8bc992d016b257489ab5 Mon Sep 17 00:00:00 2001
> > From: Michel Alexandre Salim <salimma at fedoraproject.org>
> > Date: Wed, 9 Nov 2011 14:18:45 +0100
> > Subject: [PATCH] drm/i915: Prioritize SSC quirk table when determining SSC
> >  settings
> > 
> > Commit 72bbe58cd9568c7766cc219a779ea68a02132797 makes the check against
> > the quirk table unreachable if i915_panel_use_ssc is set. This patch
> > reorders the tests so that the quirk table is checked first, the
> > i915_panel_use_ssc next and the original per-device setting last.
> 
> The whole point of having a module parameter is to override the defaults
> for the hardware, be it quirked or not. You are complaining that we are
> honoring a parameter explicitly set by the user, are you not?
> -Chris
> 
Fixed in v3:
- if i915_panel_use_ssc is set to exactly 1 (so the user is explicitly
setting this value) then quirk is ignored
- else if the hardware is in the quirk table, SSC is always disabled
- else consider the inherited value of i915_panel_use_ssc per Keith's
72bbe58cd commit
- lastly, use the original device SSC setting (no longer checking
against the quirk table here, since that's already checked before).

I've tested it by booting:
- without any kernel parameter override -- SSC is off and my panel works
- with i915.lvds_use_ssc=1 -- SSC is forced on, panel turns black as
soon as KMS is enabled
- with i915.lvds_use_ssc=0 -- SSC is off

Let me know if there's any problem with the v3 patch.

Thanks,

-- 
Michel Alexandre Salim