[Intel-gfx] Prepare for 2.3.2

[Zhenyu Wang]

On 2008.06.13 20:08:20 +0200, Lukas Hejtmanek wrote:
> > These changes are to avoid returning a 0 backlight brightness when the LCD is 
> > off?  Or did you find cases where the value wasn't 0 even though the display 
> > was supposed to be off?

Another point of Lukas's patch is to read back current level before
setting during some operations, which should be able to make backlight
behavior more consistent.

> 
> Well, it happened that my Xserver started with the maximum backlight value.
> I wondered why this happens. I found that it is because the backlight value is
> set to 0 somewhere in to code path during Xserver startup. Thereafter,
> SetDPMSOn kicks the maximum backlight value.
> 
> So I investigated why I got backlight value to 0. And it was because there is
> called DPMS Off during startup and while being off, randr gets initialized and
> it calls lvds_set or lvds_get property (not sure which one) and it reads the
> current backlight value and save it. However, the value is 0 as DPMS if OFF
> now. I wonder why this happens now and not before. But it definitely happens.
> 

Current we still guard 0 backlight in panel setting, and I think we
might be better to ignore it for 'kernel' backlight method, aka acpi
method. Platform will normally define level 0 to be appropriate minimal
value, but not some insane close-to-black brightness. So in this case
we won't need to turn it into max by always.