Created attachment 68428 [details]
dmesg drm.debug=6

Created attachment 68431 [details] [review]
fixup ch7xxxx get hw state callback

Let's hope this works ...

Created attachment 68433 [details]
After dvo get_state patch.

Doesn't appear to be the solution. :(

Created attachment 68442 [details] [review]
ch7xxx fixup, v2

I've missed that the connector sanitize code doesn't call disable. Also some additional debug printks. Please attach drm.debug=0xe dmesg with this patch applied.

Created attachment 68443 [details]
patched dmesg, v2

Almost too quiet...

Created attachment 68444 [details]
patched dmesg, v2

Retesting patch 1, also silences the critics. I've arranged for myself to be taken outside and shot.

commit e998eae4cf2804260a92a4eed465ff7f17e0d722
Author: Daniel Vetter <daniel.vetter@ffwll.ch>
Date:   Thu Oct 11 18:43:52 2012 +0200

    drm/i915/dvo-ch7xxx: fix get_hw_state
    
    The boot-up state seems to be all-zeros, so it's safer to check for
    the bits that need to be set when the dvo encoder is in the dpms on
    state, than checking the bits we set when it's in the off state.
    
    Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=55047
    Tested-by: Chris Wilson <chris@chris-wilson.co.uk>
    Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>

A patch referencing this bug report has been merged in Linux v3.7-rc2:

commit 4afa0ace429624fde392b29d05a803672a41192e
Author: Daniel Vetter <daniel.vetter@ffwll.ch>
Date:   Thu Oct 11 18:43:52 2012 +0200

    drm/i915/dvo-ch7xxx: fix get_hw_state