Bug 86504 - [BYT/BSW] eDP black screen when backlight is set over update_status path
Summary: [BYT/BSW] eDP black screen when backlight is set over update_status path
Status: CLOSED WORKSFORME
Alias: None
Product: DRI
Classification: Unclassified
Component: DRM/Intel (show other bugs)
Version: XOrg git
Hardware: Other All
: high normal
Assignee: Intel GFX Bugs mailing list
QA Contact: Intel GFX Bugs mailing list
URL:
Whiteboard:
Keywords:
Depends on:
Blocks:
 
Reported: 2014-11-21 00:35 UTC by Rodrigo Vivi
Modified: 2017-07-24 22:50 UTC (History)
3 users (show)

See Also:
i915 platform: BSW/CHT, BYT
i915 features: display/eDP


Attachments
dmesg (137.62 KB, text/plain)
2014-11-21 00:35 UTC, Rodrigo Vivi
no flags Details

Description Rodrigo Vivi 2014-11-21 00:35:50 UTC
Created attachment 109785 [details]
dmesg

xset dpms force off

leads to a frozen blank screen without any way to recover from it.
It happens that on update_status path backlight is disabled before the regular path and it causes this issue.

If we avoid "&= ~EDP_BLC_ENABLE;" when coming from backlight_ops.update_status path everything works correctly as expected.

For a long time I believed that it was a PSR issue, but when fighting against it I noticed it was happening 100% of the time even with PSR disabled.
So I just reproduced and collected log of it on today's nightly that doesn't have PSR for VLV merged yet.
Comment 1 Jani Nikula 2014-11-21 07:12:18 UTC
What if you do

# echo 4 > /sys/class/backlight/intel_backlight/bl_power
# echo 0 > /sys/class/backlight/intel_backlight/bl_power

when you're in PSR? Does this reproduce the issue?
Comment 2 Rodrigo Vivi 2014-11-21 18:46:25 UTC
Yes, the issue is the same regardless PSR.

Yes, this sequence works fine.

I also tried
echo 4 > /sys/class/backlight/intel_backlight/bl_power
echo 4 > /sys/class/backlight/intel_backlight/bl_power
echo 0 > /sys/class/backlight/intel_backlight/bl_power

to see if disabling it twice would face the issue but no, everything worked correctly.

Also I noticed that 
echo 0 > /sys/class/backlight/intel_backlight/bl_power
is the only thing that can restore my screen after I hit the issue. So forget the part of "without any way to recover from it".
Comment 3 Ville Syrjala 2014-12-18 11:43:31 UTC
Could this be fixed by

commit 7ecc778691c452285f754743a93a46fa1d3da52f
Author: Chris Wilson <chris@chris-wilson.co.uk>
Date:   Tue Nov 4 10:36:26 2014 +0000

    backlight: bl_power uses 0 for on and 4 for off

in the ddx?
Comment 4 Rodrigo Vivi 2014-12-24 18:37:03 UTC
Actually the issue happens even when intel_backlight is not present.

intel_backlight was actually happening to get display back, but I removed in some compilation and noticed I'm without it and faced blank screen issue again here.

So, no ddx patch doesn't fix the issue.
Comment 5 Jani Nikula 2015-01-30 07:06:56 UTC
(In reply to Rodrigo Vivi from comment #4)
> intel_backlight was actually happening to get display back, but I removed in
> some compilation and noticed I'm without it and faced blank screen issue
> again here.

Do I understand you correctly that this has nothing to do with backlight after all?
Comment 6 Jani Nikula 2015-02-12 12:59:26 UTC
Rodrigo, is this still an issue?

The dmesg does hint that Ville was on the correct track with comment #3, but if that doesn't help, I'm pretty confused.
Comment 7 Jani Nikula 2015-10-07 14:16:59 UTC
Reporter is AWOL, closing. ;)

Rodrigo, please reopen if this is still an issue.


Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.