[Intel-gfx] [PATCH 08/10] drm/i915: don't rely on previous values set on DDI_BUF_CTL

[Daniel Vetter]

On Wed, Oct 10, 2012 at 03:27:59PM +0100, Lespiau, Damien wrote:
> On Fri, Oct 5, 2012 at 4:05 PM, Paulo Zanoni <przanoni at gmail.com> wrote:
> > From: Paulo Zanoni <paulo.r.zanoni at intel.com>
> >
> > Just set the only bit we need, everything else is either ignored on
> > HDMI or should be set to zero.
> >
> > Signed-off-by: Paulo Zanoni <paulo.r.zanoni at intel.com>
> 
> Is the assumption that intel_enable_ddi() will only ever be used by
> the HDMI encore likely to hold true for ever? looks like a general
> function that could be used by others to me. Also intel_disable_ddi()
> read the register back, so it'd be unfair to only touch
> intel_enable_ddi() if you want to do that.
> 
> I'd just stay on the safe side and retain the programming here.

I'd prefer to go safe and do our own programming ;-) Imo it's better to
fully program a register than to rely on random garbage left behind by the
bios - usually that allows us to bring up hw a bit quicker, but in the end
results in some nasty bug reports once hw starts shipping. For the disable
side things are usually not that important, since we need to assume that
our own code (or our own hw state readout) set things up correctly.

Wrt the general usefullness of, I guess the dp patches will changed that.
I'll just merge this one here.
-Daniel
-- 
Daniel Vetter
Software Engineer, Intel Corporation
+41 (0) 79 365 57 48 - http://blog.ffwll.ch