[Intel-gfx] Breaking suspend/resume by the Pipe A quirk

[Thomas Richter]

Am 03.06.2014 17:14, schrieb Chris Wilson:
> On Tue, Jun 03, 2014 at 05:04:52PM +0200, Thomas Richter wrote:
>> Am 03.06.2014 16:45, schrieb Daniel Vetter:
>>
>>>
>>> Yeah, both connectors use CRTC 0. Have you tried what happens if you:
>>> - disable DVI1 first (--off)
>>> - then enable it on crtc 1?
>>
>> Same difference, internal screen goes blank with --off, and stays
>> blank after moving it to crtc 1 if I try to re-enable it with --auto
>> or --mode.
>
> The oddity in the config is that the LVDS is reported as disconnected.
> That should not be happening unless there is some sharing going on
> inside the DVO chip.

Please note that on this specific notebook, the internal screen is 
connected via DVI, not via LVDS. I don't know what they did with the 
LVDS output, it's probably - as said - just disconnected.

The R31 has its panel connected via LVDS, and here I can set resolutions 
independently just fine.

Currently compiling the patch, takes a while on a 1GHhz/1GB P-3 machine. (-;

Greetings,
	Thomas