[Intel-gfx] [C220] Display goes blank with PCH PWM1 enabled warning in intel_display.c:7376/6900

[Catalin Hritcu]

On Mon, Nov 10, 2014 at 12:58 PM, Jani Nikula
<jani.nikula at linux.intel.com> wrote:
> On Mon, 10 Nov 2014, Catalin Hritcu <catalin.hritcu at gmail.com> wrote:
>> On Mon, Nov 10, 2014 at 11:47 AM, Jani Nikula
>> <jani.nikula at linux.intel.com> wrote:
>>> On Mon, 10 Nov 2014, Catalin Hritcu <catalin.hritcu at gmail.com> wrote:
>>>> Dear Intel graphics driver community,
>>>>
>>>> I'm a user experiencing display problems and joined this mailing list
>>>> to get some help with filing a proper bug report. In particular,
>>>> I'm unsure against which component I should file the following problem.
>>>
>>> Please file a new bug against DRM/Intel at [1].
>>>
>>> Reference your mail on the mailing list.
>>>
>>> Smells like pm refcounting problems.
>>
>> I've filed a new bug as instructed:
>> https://bugs.freedesktop.org/show_bug.cgi?id=86105
>>
>>> Attach dmesg with
>>> drm.debug=14 module parameter set, from boot to the problem.
>>
>> I'll try to produce a dmesg log with drm.debug=14 set from the boot.
>> This will take a while since restarting causes the problem to go away
>> and it takes days or even weeks for the problem to reappear. The
>> problem is happening right now though, so if someone can suggest a way
>> to pass drm.debug=14 without rebooting the machine this I could
>> provide some preliminary information right away. I've tried this:
>
> # echo 14 > /sys/module/drm/parameters/debug
>
> Unfortunately setting that now likely does not help in figuring out how
> we got into that situation to begin with. My guess is still the same;
> there's some sequence that leads to one pm put too many.
>
Understood, I'll reboot now and try my best to produce a full log with
drm.debug=14 as soon as this happens again. Until then the end of
https://bugs.freedesktop.org/attachment.cgi?id=109217 has drm.debug=14
set.

Catalin