Bug 107119

Summary:	[CI] [BAT] A lot of crc mismatches on fi-byt-squawks
Product:	DRI	Reporter:	Maarten Lankhorst <bugs>
Component:	DRM/Intel	Assignee:	Intel GFX Bugs mailing list <intel-gfx-bugs>
Status:	RESOLVED MOVED	QA Contact:	Intel GFX Bugs mailing list <intel-gfx-bugs>
Severity:	normal
Priority:	low	CC:	intel-gfx-bugs
Version:	unspecified
Hardware:	x86-64 (AMD64)
OS:	Linux (All)
Whiteboard:	Triaged, ReadyForDev
i915 platform:	BYT	i915 features:	display/Other

Description Maarten Lankhorst 2018-07-05 08:41:18 UTC

A lot of tests on fi-byt-squawks fail with a CRC mismatch. Manual verification doesn't show any major visual glitches, and retrieving a crc multiple times may give different values.

PSR, DRRS and FBC are all disabled on that machine, so it must be something else.
No FIFO underruns show up either.

Comment 1 Francesco Balestrieri 2018-07-06 07:57:10 UTC

Can you point to one of the CI failures for reference?

Comment 2 Maarten Lankhorst 2018-07-06 08:07:21 UTC

https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_4441/fi-byt-squawks/igt@kms_frontbuffer_tracking@basic.html

Comment 3 Lakshmi 2018-09-11 07:03:52 UTC

Rising the priority to high as issue occurs in every round with BAT tests.

Comment 4 Don Hiatt 2019-08-28 17:33:11 UTC

These tests have been failing for a long time, even up to CI_DRM_6795 (44 minutes old) but only on fi-byt-squawks. The filter even says "Ignore crc mismatches on fi-byt-squawks". Since the last update to this fdo was over a year ago, can we just close this bug and not stop updating the filter to remove the noise as this doesn't seem to be a real user issue.

Even though there is 100% reproducibility on fi-byt-squawks, I think the severity should be set to low giving the above. As this appears to be a display/visual issue, I'm changing this from power to dispaly/Other.

Lastly, perhaps we should pass the '--skip-crc-compare' to igt_assert_crc_equal() for fi-byt-squawks given the following comment?

 * Passing --skip-crc-compare on the command line will force this function
 * to always pass, which can be useful in interactive debugging where you
 * might know the test will fail, but still want the test to keep going as if
 * it had succeeded so that you can see the on-screen behavior.

Comment 5 Martin Peres 2019-11-29 17:48:31 UTC

-- GitLab Migration Automatic Message --

This bug has been migrated to freedesktop.org's GitLab instance and has been closed from further activity.

You can subscribe and participate further through the new bug through this link to our GitLab instance: https://gitlab.freedesktop.org/drm/intel/issues/127.

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.