Bug 71653 – [Chromium] Layout Test platform/chromium-cg-mac/editing/input/ime-candidate-window-position.html is flaky

• Component: Tools / Tests
• Status: RESOLVED WONTFIX

[Yuta Kitamura]

The following layout test is flaky (pass / timeout) on all platforms.

platform/chromium-cg-mac/editing/input/ime-candidate-window-position.html

See http://test-results.appspot.com/dashboards/flakiness_dashboard.html#tests=platform%2Fchromium-cg-mac%2Fediting%2Finput%2Fime-candidate-window-position.html

Probable cause:

Unknown, as it's not clear when this flakiness started. This test is less flaky on DEPS bots.

[Kenichi Ishibashi]

Umm, I don't have any idea. Maybe recent editing changes trigger this flaky?

[Ryosuke Niwa]

My suspicion is that doTest() is called before iframe is loaded, and doTest is encountering some javascript error and end up not calling finishJSTest.

We should probably move finishJSTest above if with setTimeout(~, 0) so that we can see the diff.

[Kenichi Ishibashi]

Created attachment 113983 [details]
Patch

[Kenichi Ishibashi]

(In reply to comment #2)
> My suspicion is that doTest() is called before iframe is loaded, and doTest is encountering some javascript error and end up not calling finishJSTest.
> 
> We should probably move finishJSTest above if with setTimeout(~, 0) so that we can see the diff.

It seems you are right. I changed the test to call doTest() after iframe is loaded and iterated layout test 10 times, but no timeout occurred.

[Kent Tamura]

Comment on attachment 113983 [details]
Patch

ok, let's land this and watch the dashboard.

[Ryosuke Niwa]

(In reply to comment #4)
> It seems you are right. I changed the test to call doTest() after iframe is loaded and iterated layout test 10 times, but no timeout occurred.

Could you reproduce the issue (intermittent failure) without your patch? It's still odd that load event is fired before the iframe is loaded though. HTML5 spec is explicit about delaying load event until the iframe's content is loaded.

Adam, do you know anything about this?

[WebKit Review Bot]

Comment on attachment 113983 [details]
Patch

Clearing flags on attachment: 113983

Committed r99524: <http://trac.webkit.org/changeset/99524>

[WebKit Review Bot]

All reviewed patches have been landed.  Closing bug.

[Kent Tamura]

(In reply to comment #7)
> Committed r99524: <http://trac.webkit.org/changeset/99524>

It seems this didn't fix the issue.

[Ryosuke Niwa]

Hm...
http://trac.webkit.org/browser/trunk/LayoutTests/platform/chromium-cg-mac/editing/resources/ime-candidate-window-position-iframe.html
and
http://trac.webkit.org/browser/trunk/LayoutTests/platform/chromium-cg-mac/editing/input/ime-candidate-window-position.html#L27

both try to focus the same node. So maybe there's some race condition there?

Also, can we do setTimeout(finishJSTest, 0) before if statement so that we can at least see the diff if the test were to fail?

[Stephen Chenney]

Marking test failures as WontFix. Bug is still accessible and recording in TestExpectations.