Bug 86592 – [Chromium] Some tests are flaky on cr-linux ews

• Status: RESOLVED
• Resolution: WONTFIX
• Component: Tools / Tests

[Kenichi Ishibashi]

For instance:
http://queues.webkit.org/results/12694913
http://queues.webkit.org/results/12719088

Unexpected flakiness: text diff mismatch (5)
  fast/canvas/webgl/shader-precision-format.html = TEXT PASS
  fast/loader/javascript-url-in-object.html = TEXT PASS
  http/tests/inspector/resource-har-pages.html = TEXT PASS
  http/tests/security/sandboxed-iframe-origin-add.html = TEXT PASS
  platform/chromium/virtual/gpu/fast/canvas/webgl/shader-precision-format.html = TEXT PASS

Unexpected flakiness: image mismatch (2)
  svg/batik/text/textOnPath.svg = IMAGE PASS
  svg/batik/text/verticalTextOnPath.svg = IMAGE PASS

Unexpected flakiness: tests timed out (3)
  fast/frames/lots-of-objects.html = TIMEOUT PASS
  fast/loader/text-document-wrapping.html = TIMEOUT PASS
  http/tests/xmlhttprequest/xmlhttprequest-unsafe-redirect.html = TIMEOUT PASS


These flakiness make cqbots sad. I'm going to add suppression for now.

[Kenneth Russell]

Would it be possible to get the bot's output for fast/canvas/webgl/shader-precision-format.html ? There's no reason that test should be flaky.

[Kenichi Ishibashi]

(In reply to comment #1)
> Would it be possible to get the bot's output for fast/canvas/webgl/shader-precision-format.html ? There's no reason that test should be flaky.

Maybe anyone who can ssh to the bot can get the output? Cc'ing haraken who should be able to access the bot.

[Kentaro Hara]

Here is a log on the flakiness dashboard: http://test-results.appspot.com/dashboards/flakiness_dashboard.html#tests=platform%2Fchromium%2Fvirtual%2Fgpu%2Ffast%2Fcanvas%2Fwebgl%2Fshader-precision-format.html

[Kenneth Russell]

Thanks. That's strange; it looks like the expected results expect the test not to run to completion. Mo, possible for you to look into this?

[Zhenyao Mo]

OK, I'll have a look.

[Julien Chaffraix]

The last entries explicitely referencing this bug have been removed as part of http://trac.webkit.org/changeset/130969. The tests hadn't failed for 100 runs according to the flakiness dashboard.

Mo, should we close this bug or is there anything left to investigate?

[Stephen Chenney]

Marked LayoutTest bugs, bugs with Chromium IDs, and some others as WontFix. Test failure bugs still are trackable via TestExpectations or disabled unit tests.