[Intel-gfx] [PATCH 2/2] drm/i915: clean up and simplify i9xx_crtc_mode_set wrt PLL handling

[Ville Syrjälä]

On Fri, Sep 13, 2013 at 04:07:19PM +0200, Daniel Vetter wrote:
> On Fri, Sep 13, 2013 at 04:38:28PM +0300, Jani Nikula wrote:
> > On Fri, 13 Sep 2013, Daniel Vetter <daniel at ffwll.ch> wrote:
> > > On Fri, Sep 13, 2013 at 11:03:09AM +0300, Jani Nikula wrote:
> > >> Flat out skip anything to do with PLL if we have a DSI encoder (and thus
> > >> DSI PLL). Also skip PLL computation if the encoder has already set
> > >> clocks. This allows for some tidying up of the code, including a
> > >> superfluous call to intel_limit() for LVDS downclock path.
> > >> 
> > >> Signed-off-by: Jani Nikula <jani.nikula at intel.com>
> > >> Reviewed-by: Ville Syrjälä <ville.syrjala at linux.intel.com>
> > >
> > > Tried to merge this but the baseline seems to be off, at least wrt dinq.
> > > Do I miss some patches that I should apply first?
> > 
> > This one depends on "drm/i915: do not update cursor in crtc mode set" in
> > this thread (my patch but with Ville's commit message amendmend). Did
> > you push the assert patch before that? It's a good order.
> 
> I've pushed the assert patch to dinq, the real fixes need to go to -fixes.
> Tbh I'm a bit confused about what to put where, so please scream ;-)
> 
> For Ville's patches I'm waiting a bit for Paulo to take a look and ack
> them.

For stable we need just (assuming it also fixes Paulo's hangs):
 drm/i915: do not update cursor in crtc mode set

On top of that we can add the cursor assert patch (not sure if we want
that to go to stable):
 drm/i915: add asserts for cursor disabled

I get the impression you applied the assert patch alone. That's probably
not a very good idea since it will (or at least should) scream every time
you do a modeset, unless the other patch is applied as well.

-- 
Ville Syrjälä
Intel OTC