Bug 98625

Summary:	[SNB BAT] HDMI: EDID checksum is invalid
Product:	DRI	Reporter:	Imre Deak <imre.deak>
Component:	DRM/Intel	Assignee:	Intel GFX Bugs mailing list <intel-gfx-bugs>
Status:	CLOSED FIXED	QA Contact:	Intel GFX Bugs mailing list <intel-gfx-bugs>
Severity:	major
Priority:	medium	CC:	intel-gfx-bugs
Version:	DRI git
Hardware:	x86-64 (AMD64)
OS:	Linux (All)
Whiteboard:
i915 platform:	SNB	i915 features:	display/HDMI

Description Imre Deak 2016-11-07 12:42:53 UTC

During igt/kms_flip on an SNB-2520m machine, monitor ELD G246HYL:

[  384.622585] [drm:intel_hdmi_detect [i915]] [CONNECTOR:46:HDMI-A-1]
[  384.650610] [drm:gmbus_xfer [i915]] GMBUS [i915 gmbus dpb] NAK for addr: 0040 w(1)
[  384.650659] [drm:gmbus_xfer [i915]] GMBUS [i915 gmbus dpb] NAK on first message, retry
[  384.650956] [drm:gmbus_xfer [i915]] GMBUS [i915 gmbus dpb] NAK for addr: 0040 w(1)
[  384.650968] [drm:drm_dp_dual_mode_detect] DP dual mode HDMI ID:  (err -6)
[  384.650979] [drm:drm_detect_monitor_audio] Monitor has basic audio support
[  384.651016] [drm:drm_edid_block_valid] *ERROR* EDID checksum is invalid, remainder is 162

https://intel-gfx-ci.01.org/CI/Patchwork_2916/fi-snb-2520m/dmesg-during.log

Comment 1 Jari Tahvanainen 2016-11-08 03:01:34 UTC

This sporadic bug has been triggered on fi-snb-2520n by various tests on CI_DRM testing:
2016-10-28 CI_DRM_1778/fi-snb-2520m/igt@kms_force_connector_basic@force-edid:
[  418.279645] [drm:drm_edid_block_valid] *ERROR* EDID checksum is invalid, remainder is 147
2016-10-31 CI_DRM_1782/fi-snb-2520m/igt@kms_force_connector_basic@force-edid
[  413.919373] [drm:drm_edid_block_valid] *ERROR* EDID checksum is invalid, remainder is 158
2016-11-02 CI_DRM_1792/fi-snb-2520m/igt@kms_force_connector_basic@force-connector-state
[  402.714977] [drm:drm_edid_block_valid] *ERROR* EDID checksum is invalid, remainder is 224
2016-11-03 CI_DRM_1798/fi-snb-2520m/igt@kms_force_connector_basic@force-load-detect:
[  417.697562] [drm:drm_edid_block_valid] *ERROR* EDID checksum is invalid, remainder is 161
2016-11-07 CI_DRM_1806/fi-snb-2520m/igt@kms_pipe_crc_basic@bad-source
[  432.017861] [drm:drm_edid_block_valid] *ERROR* EDID checksum is invalid, remainder is 117

See also already resolved bug 98228.

Comment 2 Sagar Kamble 2016-11-14 10:09:32 UTC

Seeing this issue with kms_force_connector_basic@force-edid on SNB.
https://intel-gfx-ci.01.org/CI/Trybot_279/fi-snb-2520m/igt@kms_force_connector_basic@force-edid.html

[  405.905695] Setting dangerous option load_detect_test - tainting kernel
[  405.906442] Setting dangerous option load_detect_test - tainting kernel
[  406.074360] [drm:drm_edid_block_valid] *ERROR* EDID checksum is invalid, remainder is 33

Comment 3 Daniel Vetter 2016-11-17 08:20:35 UTC

Do we have a full list of all testcases affected by this? If this only happens with kms_force_connector then it would very likely be a pure software bug (but don't ask me where ...).

Comment 4 Jani Saarinen 2017-02-13 08:48:00 UTC

EDID issues not seen lately on CI due to patch https://patchwork.freedesktop.org/series/16556/
merged.

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.