[Intel-gfx] [PATCH 0/2] drm/i915/opregion: proper handling of DIDL and CADL

[Peter Wu]

Hi Jani,

The method is somehow not reliable. At boot I was still able to capture
brightness hotkeys (via acpi_listen). After starting SSDM and logging
in, it still worked. Then I played a bit with PRIME output slaves, Xorg
crashed at some points and hotkeys stopped working.

Inspection of CADL (via acpidbg) showed that it contained zeroes instead
of 0x400. By poking more, suddenly the expected 0x400 value returned
into CADL and hotkeys functioned again. At later moments it reverted to
0 again and hotkeys broke again.

Tracking down setters of active_crcts, I noticed that
intel_modeset_readout_hw_state somehow does not detect eDP1 as active.
Full dmesg log (9MiB) with drm.debug=0x1e (via sysfs) is available at

https://lekensteyn.nl/files/logs/journal-v4.8-rc3-183-g5e608a0-plus-2-cadl-patches.txt

Kind regards,
Peter

On Thu, Aug 25, 2016 at 03:53:02PM +0300, Jani Nikula wrote:
> This is the next iteration of [1] and [2]. Please review and/or test,
> according to your abilities.
> 
> Thanks,
> Jani.
> 
> Cc: Peter Wu <peter at lekensteyn.nl>
> Cc: Rainer Koenig <Rainer.Koenig at ts.fujitsu.com>
> Cc: Jan-Marek Glogowski <glogow at fbihome.de>
> Cc: Maarten Lankhorst <maarten.lankhorst at linux.intel.com>
> Cc: Marcos Paulo de Souza <marcos.souza.org at gmail.com>
> Cc: Paolo Stivanin <paolostivanin at fastmail.fm>
> 
> [1] http://mid.mail-archive.com/cover.1467214151.git.jani.nikula@intel.com
> [2] http://mid.mail-archive.com/1471315782-925-1-git-send-email-marcos.souza.org@gmail.com
> 
> Jani Nikula (2):
>   drm/i915: make i915 the source of acpi device ids for _DOD
>   drm/i915/opregion: update cadl based on actually active outputs
> 
>  drivers/gpu/drm/i915/i915_drv.h       |   4 +
>  drivers/gpu/drm/i915/intel_display.c  |   6 ++
>  drivers/gpu/drm/i915/intel_drv.h      |   3 +
>  drivers/gpu/drm/i915/intel_opregion.c | 157 +++++++++++++---------------------
>  4 files changed, 74 insertions(+), 96 deletions(-)
> 
> -- 
> 2.1.4
> 

-- 
Kind regards,
Peter Wu
https://lekensteyn.nl