New issue
Advanced search Search tips
Note: Color blocks (like or ) mean that a user may not be available. Tooltip shows the reason.

Issue 804887 link

Starred by 1 user

Issue metadata

Status: WontFix
Owner:
OOO until 2019-01-24
Closed: Sep 11
Components:
EstimatedDays: ----
NextAction: ----
OS: Mac
Pri: 2
Type: Bug



Sign in to add a comment

webgl_conformance_tests flaky on retina mac with AMD GPU

Project Member Reported by ellyjo...@chromium.org, Jan 23 2018

Issue description

[350/459] gpu_tests.webgl_conformance_integration_test.WebGLConformanceIntegrationTest.WebglConformance_conformance_textures_misc_texture_size failed unexpectedly 2.4463s:
  
  Traceback (most recent call last):
    _RunGpuTest at content/test/gpu/gpu_tests/gpu_integration_test.py:132
      self.RunActualGpuTest(url, *args)
    RunActualGpuTest at content/test/gpu/gpu_tests/webgl_conformance_integration_test.py:203
      getattr(self, test_name)(test_path, *args[1:])
    _RunConformanceTest at content/test/gpu/gpu_tests/webgl_conformance_integration_test.py:217
      self._CheckTestCompletion()
    _CheckTestCompletion at content/test/gpu/gpu_tests/webgl_conformance_integration_test.py:213
      self.fail(self._WebGLTestMessages(self.tab))
    fail at /System/Library/Frameworks/Python.framework/Versions/2.7/lib/python2.7/unittest/case.py:412
      raise self.failureException(msg)
  AssertionError: 2D texture of size 2048x2048 with mips should draw with yellow
  at (16, 16) expected: 255,255,0,255 was 253,255,0,255
  FAIL 2D texture of size 2048x2048 with mips should draw with yellow
  at (16, 16) expected: 255,255,0,255 was 253,255,0,255
  
  Locals:
    msg : u'2D texture of size 2048x2048 with mips should draw with yellow\nat (16, 16) expected: 255,255,0,255 was 253,255,0,255\nFAIL 2D texture of size 2048x2048 with mips should draw with yellow\nat (16, 16) expected: 255,255,0,255 was 253,255,0,255\n'
  

Marking for Blink>WebGL triage.
 

Comment 1 by zmo@chromium.org, Jan 25 2018

Labels: OS-Mac
This looks like we just need to relax the threshold a bit to make the test pass again.

That said, I am really curious what triggers the flakiness in rendering.

Comment 2 by kbr@chromium.org, Jan 25 2018

Owner: kbr@chromium.org
Status: Started (was: Untriaged)
Thanks for the report. Please always include a link to the failing build. In this case it was this build:

https://ci.chromium.org/buildbot/chromium.gpu/Mac%20Retina%20Release%20%28AMD%29/22175

and this shard:

https://chromium-swarm.appspot.com/task?id=3b3c424cb886c110&refresh=10&show_raw=1

It turns out this test flaked before in this way and a tolerance was added. https://github.com/KhronosGroup/WebGL/pull/2586 is up increasing the tolerance a bit.

Status: WontFix (was: Started)
This isn't flaky any more according to the flakiness dashboard:

https://test-results.appspot.com/dashboards/flakiness_dashboard.html#testType=webgl_conformance_tests&builder=chromium.gpu%3AMac%20Retina%20Release%20(AMD)

Since there's no CL even associated with the WebGL conformance roll that pulled in https://github.com/KhronosGroup/WebGL/pull/2586 , closing this as WontFix (no longer reproducible).

Sign in to add a comment