Closed Bug 1302876 Opened 8 years ago Closed 7 years ago

Intermittent layout/style/test/chrome/test_display_mode.html | Test timed out.

Categories

(Core :: CSS Parsing and Computation, defect)

Product:
Core
Component:
CSS Parsing and Computation
Type:
defect
Priority:
Not set
Severity:
normal

Tracking

()

Status:
RESOLVED INCOMPLETE
Tracking Flags:
Tracking Status
firefox50 --- unaffected
firefox51 --- affected
firefox52 --- wontfix
firefox53 --- affected

People

(Reporter: intermittent-bug-filer, Unassigned)

References

Details

(Keywords: intermittent-failure, Whiteboard: [stockwell unknown]) 

Cam, any idea what might have landed around a week ago to make this test start failing?
status-firefox50: --- → unaffected
status-firefox51: --- → affected
status-firefox52: --- → affected
Component: Layout → CSS Parsing and Computation
Flags: needinfo?(cam)
Xidorn, you said there were some full screen changes recently.  Could they be causing the intermittent test timeouts here?
Flags: needinfo?(cam) → needinfo?(xidorn+moz)
As far as I know, there was only one fullscreen-related change landed in the past two weeks, which is bug 1299795. It seems to me that patch is unlikely related to this failure.
Flags: needinfo?(xidorn+moz)
Ryan, I don't recall seeing anything land that would impact this.  I couldn't reproduce the intermittent failure locally.  Can we bisect on the build machines easily?
Flags: needinfo?(ryanvm)
See Also: → 1302878
That changeset doesn't seem to be something related...
(In reply to Xidorn Quan [:xidorn] (UTC+10) from comment #7)
> That changeset doesn't seem to be something related...

Pro tip: assume our automation is a house of cards that could collapse at any time.

There are 2 potential reasons I can think of that would cause things to change:

1) Landing the change caused the Docker images to be rebuilt with newer packages and a newer package changed behavior somehow. See bug 1289812 for some of the horrors this has caused.

2) Before run-task was put in place, stdin for test processes (likely mozharness) was a TTY. run-task changed stdin to be a non-TTY pipe. I /think/ the test harness would ensure Firefox's stdin was a pipe and this wouldn't change behavior, but you never know.
After even more retriggers, I was able to hit this once on the push prior to gps', which corresponds to bug 1300606. That at least seems to make more sense it made changes to events-related code.
Flags: needinfo?(bugs)
It seems all timeouts are from waiting on sizemodechange event after synthesizing an F11 key. Not sure what's wrong there.
this has been reducing in frequency the last 6 weeks.
Whiteboard: [stockwell unknown]
Too late for firefox 52, mass-wontfix.
status-firefox52: affected → wontfix
Seems to have gone away for good on trunk around mid-June. Only hitting on ESR52 at this point.
Status: NEW → RESOLVED
Closed: 7 years ago
Flags: needinfo?(bugs)
Resolution: --- → INCOMPLETE
You need to log in before you can comment on or make changes to this bug.