1041017 - TEST-UNEXPECTED-FAIL | /tests/layout/style/test/test_animations_omta_start.html | uncaught exception - TypeError: a is null on b2g-desktop in taskcluster docker image only

Status: RESOLVED INVALID

(Core :: Graphics: Layers, defect)

Description

[Geoff Brown [:gbrown]]

test_animations_omta_start.html is generally trouble-free and passes consistently on all tbpl test platforms, including b2g-desktop. 

In bug 1005907, I am trying to run b2g-desktop mochitests in a new environment (in a docker image). All tests are passing except for 3 tests -- test_animations_omta_start.html is one of the failing tests:

23:21:58     INFO -  21704 INFO TEST-START | /tests/layout/style/test/test_animations_omta_start.html
23:21:58     INFO -  21705 INFO dumping last 2 message(s)
23:21:58     INFO -  21706 INFO if you need more context, please use SimpleTest.requestCompleteLog() in your test
23:21:58     INFO -  21707 INFO TEST-PASS | /tests/layout/style/test/test_animations_omta_start.html | opacity is set on compositor thread after delayed start
23:21:58     INFO -  21708 INFO TEST-PASS | /tests/layout/style/test/test_animations_omta_start.html | transform is set on compositor thread after delayed start
23:21:58     INFO -  21709 INFO TEST-UNEXPECTED-FAIL | /tests/layout/style/test/test_animations_omta_start.html | uncaught exception - TypeError: a is null at http://mochi.test:8888/tests/layout/style/test/animation_utils.js:399
23:21:58     INFO -  TEST-INFO | expected PASS
23:21:58     INFO -  21710 INFO JavaScript error: http://mochi.test:8888/tests/layout/style/test/animation_utils.js, line 399: a is null
23:21:58     INFO -  21711 INFO TEST-UNEXPECTED-FAIL | /tests/layout/style/test/test_animations_omta_start.html | test left refresh driver under test control


As far as I can tell, tests are being run in the new, experimental environment the same way as on tbpl: the same arguments to the same mozharness script. Of course, I expect there are timing differences between tbpl and the new, experimental environment. There may also be missing system packages, differences in environment variables, or other unknown differences.

Comment 1

[Geoff Brown [:gbrown]]

I debugged a little. At http://hg.mozilla.org/mozilla-central/annotate/50a8424cf1b8/layout/style/test/test_animations_omta_start.html#l118:

var transform = gUtils.getOMTAStyle(target, "transform");

transform is set to null. That seems to be because at http://hg.mozilla.org/mozilla-central/annotate/50a8424cf1b8/layout/base/FrameLayerBuilder.cpp#l3546:

nsTArray<DisplayItemData*> *array =
   reinterpret_cast<nsTArray<DisplayItemData*>*>(aFrame->Properties().Get(LayerManagerDataProperty())); 

array is set to null and FrameLayerBuilder::GetDedicatedLayer() returns null.

Comment 2

[Geoff Brown [:gbrown]]

Disabled on b2g desktop so we can proceed with the taskcluster work: 

https://hg.mozilla.org/integration/mozilla-inbound/rev/6e1ee059db88

Comment 3

[Ed Morley [:emorley]]

https://hg.mozilla.org/mozilla-central/rev/6e1ee059db88

Comment 4

[Geoff Brown [:gbrown]]

No longer testing b2g.

Comment 5

[Firefox Bug Husbandry Bot]

Removing leave-open keyword from resolved bugs, per :sylvestre.