Bug 95281 - [ILK,BDW] Display stop working after resumed from S4 due to LP0 WM fail
Summary: [ILK,BDW] Display stop working after resumed from S4 due to LP0 WM fail
Status: CLOSED FIXED
Alias: None
Product: DRI
Classification: Unclassified
Component: DRM/Intel (show other bugs)
Version: unspecified
Hardware: x86-64 (AMD64) Linux (All)
: high critical
Assignee: Intel GFX Bugs mailing list
QA Contact: Intel GFX Bugs mailing list
URL:
Whiteboard:
Keywords:
Depends on:
Blocks:
 
Reported: 2016-05-05 08:48 UTC by Lu, Han
Modified: 2017-07-24 22:41 UTC (History)
1 user (show)

See Also:
i915 platform: BDW, ILK
i915 features: display/watermark, power/suspend-resume


Attachments
dmesg after resume from S4 (117.64 KB, text/plain)
2016-05-05 08:48 UTC, Lu, Han
no flags Details
dmesg with drm.debug=14 (123.54 KB, text/plain)
2016-05-06 03:03 UTC, Lu, Han
no flags Details

Description Lu, Han 2016-05-05 08:48:28 UTC
Created attachment 123487 [details]
dmesg after resume from S4

System Environment
CRB:Intel NUC5i3RYH
Platform: BroadWell
CPU: Hasewell i5-4250U @ 1.30GHz
Audio card: ALC283
BIOS:BDW-E1R1.86C.0093.R00.1409071941
Kernel:4.5.0 drm-intel-nightly
Tree: git://anongit.freedesktop.org/drm-intel
Branch: drm-intel-nightly
Commit: 3e5ecc8c5ff80cb1fb635ce1cf16b7cd4cfb1979

Bug detailed description
Display stop working after resumed from S4

Reproduce Steps
rtcwake -m disk -s 30

Expected Result
Display work normally after resumed from S4.

Actual Result
Display stops working.
Comment 1 Lu, Han 2016-05-05 08:49:30 UTC
Remove the audio module from kernel config and the issue is still reproduced.
Comment 2 Lu, Han 2016-05-06 03:03:34 UTC
Created attachment 123514 [details]
dmesg with drm.debug=14

add dmesg with drm.debug=14
Comment 3 Ville Syrjala 2016-05-10 15:18:52 UTC
[   63.184167] [drm:ilk_validate_wm_level] Sprite WM0 too large 120 (max 0)
[   63.184168] [drm:ilk_validate_pipe_wm] LP0 watermark invalid
[   63.184168] [drm:intel_crtc_atomic_check] No valid intermediate pipe watermarks are possible

I saw this on my ILK recently. Regression from some watermark work.
Comment 4 Ville Syrjala 2016-05-10 15:21:22 UTC
[   63.184167] [drm:ilk_validate_wm_level] Sprite WM0 too large 120 (max 0)
[   63.184168] [drm:ilk_validate_pipe_wm] LP0 watermark invalid
[   63.184168] [drm:intel_crtc_atomic_check] No valid intermediate pipe watermarks are possible

I saw this on my ILK recently. Regression from some watermark work.
Comment 5 Ville Syrjala 2016-05-16 09:00:16 UTC
Shoukd be fixed by 
commit e3d5457c7caa ("drm/i915: Ignore stale wm register values on resume on ilk-bdw (v2)")


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.