Closed Bug 1262130 Opened 10 years ago Closed 9 years ago

Intermittent test_videocontrols_standalone.html | The media element should eventually be resized to match the height of the audio controls.

Categories

(Core :: Audio/Video: Playback, defect)

Product:
Core
Component:
Audio/Video: Playback
Type:
defect
Priority:
Not set
Severity:
normal

Tracking

()

Status:
RESOLVED WORKSFORME
Tracking Flags:
Tracking Status
firefox-esr45 --- disabled
firefox51 --- disabled
firefox52 --- disabled
firefox53 --- unaffected

People

(Reporter: cbook, Unassigned)

References

(Blocks 1 open bug,
URL
)

Details

(Keywords: intermittent-failure)

02:34:57 INFO - 4233 INFO TEST-UNEXPECTED-FAIL | toolkit/content/tests/widgets/test_videocontrols_standalone.html | The media element should eventually be resized to match the height of the audio controls. https://treeherder.mozilla.org/logviewer.html#?job_id=25175184&repo=mozilla-inbound
Not entirely e10s-only, but definitely more frequent there. Looking at highly, bug 1261536 looks highly-likely.
Blocks: e10s-tests
tracking-e10s: --- → ?
Component: XUL Widgets → Audio/Video: Playback
Flags: needinfo?(gsquelart)
Product: Toolkit → Core
(In reply to Ryan VanderMeulen [:RyanVM] from comment #2) > Not entirely e10s-only, but definitely more frequent there. Looking at > highly, bug 1261536 looks highly-likely. Ugh, MOAR COFFEE. Looking at *Treeherder*
We should change this test to log what the actual bounding rect looks like when waitForCondition times out...
(In reply to Boris Zbarsky [:bz] from comment #4) > We should change this test to log what the actual bounding rect looks like > when waitForCondition times out... In this audio case, only the height is checked, and it's actually logged after the first failure: > The media element should eventually be resized to match the height of the audio controls. > Height of audio element should be 28, which is equal to the controls bar. - got +0, expected 28 A sample of a few logs from different platforms always shows the exact same error (i.e., audio element height staying at 0; never a failure in the video test). I'd be happy to try and debug this, but I haven't been able to even reproduce it locally yet, and I don't have much knowledge in this area. I'll keep trying, but Boris: Any hints/suggestions? And/or could you recommend someone who could help there.
Flags: needinfo?(bzbarsky)
Backed out bug 1261536. This should be fixed.
Status: NEW → RESOLVED
Closed: 10 years ago
Resolution: --- → FIXED
Thank you Wes. I've rework bug 1261536 and verify that it doesn't trigger these failures anymore before landing again.
Flags: needinfo?(gsquelart)
Flags: needinfo?(bzbarsky)
I don't have any good ideas here so far, unfortunately. :( In terms of who might be able to help... maybe cpearce?
(In reply to Gerald Squelart [:gerald] from comment #8) > I've rework bug 1261536 [...] I meant "I'll rework bug 1261536". I'm still trying to find the issue. (In reply to Boris Zbarsky [:bz] from comment #10) > I don't have any good ideas here so far, unfortunately. :( No worries. I'm trying to debug by printfs through Try! Much fun. > In terms of who might be able to help... maybe cpearce? Excellent choice, Sir. I'll check with him...
Bug 1329403 appears to have made this return with a vengeance on ESR45/51/52 (but not on trunk, interestingly). JW, can you please look at this soon? It's pretty frequent, especially on Windows. https://treeherder.mozilla.org/logviewer.html#?job_id=70192405&repo=mozilla-aurora
Status: RESOLVED → REOPENED
status-firefox51: --- → affected
status-firefox52: --- → affected
status-firefox53: --- → unaffected
status-firefox-esr45: --- → affected
tracking-e10s: ? → ---
Flags: needinfo?(jwwang)
Resolution: FIXED → ---
I guess it's plausible that ralin's work that landed on 53 is why the test is OK on trunk.
(In reply to Ryan VanderMeulen [:RyanVM] from comment #14) > I guess it's plausible that ralin's work that landed on 53 is why the test > is OK on trunk. Which bug?
Flags: needinfo?(jwwang)
Bug 1271765 was the main one. And some follow-up fixes.
That's a huge one for uplifting. I would suggest to just disable the test on Aurora and Beta.
Flags: needinfo?(ryanvm)
Status: REOPENED → RESOLVED
Closed: 10 years ago → 9 years ago
status-firefox51: fixed → disabled
status-firefox52: fixed → disabled
status-firefox-esr45: fixed → disabled
Flags: needinfo?(ryanvm)
Resolution: --- → WORKSFORME
You need to log in before you can comment on or make changes to this bug.