Closed Bug 1043721 Opened 9 years ago Closed 9 years ago

Intermittent test_animLengthUnits.xhtml | (3) Unexpected value for animVal after changing font-size whilst frozen - got 50, expected 70

Categories

(Core :: SVG, defect)

Product:
Core
Component:
SVG
Platform:
x86_64
Android
Type:
defect
Priority:
Not set
Severity:
normal

Tracking

()

Status:
RESOLVED FIXED
Milestone:
mozilla35
Tracking Flags:
Tracking Status
firefox33 --- unaffected
firefox34 --- fixed
firefox35 --- fixed
firefox-esr31 --- unaffected

People

(Reporter: KWierso, Assigned: birtles)

Details

(Keywords: intermittent-failure)

Attachments

(1 file)

Make test_animLengthUnits.xhtml use requestAnimationFrame
1.98 KB, patch
dholbert
: review+
Details | Diff | Splinter Review 
https://tbpl.mozilla.org/php/getParsedLog.php?id=44533259&tree=Fx-Team
Android 2.3 Armv6 Emulator fx-team opt test mochitest-5 on 2014-07-24 10:05:21 PDT for push 72cdf2155ca6

slave: tst-linux64-spot-751




11:08:06     INFO -  3050 INFO TEST-START | /tests/content/svg/content/test/test_a_href_02.xhtml
11:08:06     INFO -  3051 INFO TEST-OK | /tests/content/svg/content/test/test_a_href_02.xhtml | took 2658ms
11:08:06     INFO -  3052 INFO TEST-START | /tests/content/svg/content/test/test_animLengthObjectIdentity.xhtml
11:08:06     INFO -  3053 INFO TEST-OK | /tests/content/svg/content/test/test_animLengthObjectIdentity.xhtml | took 3599ms
11:08:06     INFO -  3054 INFO TEST-START | /tests/content/svg/content/test/test_animLengthReadonly.xhtml
11:08:06     INFO -  3055 INFO TEST-OK | /tests/content/svg/content/test/test_animLengthReadonly.xhtml | took 2837ms
11:08:06     INFO -  3056 INFO TEST-START | /tests/content/svg/content/test/test_animLengthUnits.xhtml
11:08:06     INFO -  3057 INFO dumping last 11 message(s)
11:08:06     INFO -  3058 INFO if you need more context, please use SimpleTest.requestCompleteLog() in your test
11:08:06     INFO -  3059 INFO TEST-PASS | /tests/content/svg/content/test/test_animLengthUnits.xhtml | should be paused by <svg> load handler 
11:08:06     INFO -  3060 INFO TEST-PASS | /tests/content/svg/content/test/test_animLengthUnits.xhtml | should be paused at 0 in <svg> load handler 
11:08:06     INFO -  3061 INFO TEST-PASS | /tests/content/svg/content/test/test_animLengthUnits.xhtml | Unexpected initial baseVal 
11:08:06     INFO -  3062 INFO TEST-PASS | /tests/content/svg/content/test/test_animLengthUnits.xhtml | Unexpected initial baseVal units 
11:08:06     INFO -  3063 INFO TEST-PASS | /tests/content/svg/content/test/test_animLengthUnits.xhtml | Unexpected initial animVal 
11:08:06     INFO -  3064 INFO TEST-PASS | /tests/content/svg/content/test/test_animLengthUnits.xhtml | Unexpected initial animVal units 
11:08:06     INFO -  3065 INFO TEST-PASS | /tests/content/svg/content/test/test_animLengthUnits.xhtml | (1) Unexpected value for baseVal during animation 
11:08:06     INFO -  3066 INFO TEST-PASS | /tests/content/svg/content/test/test_animLengthUnits.xhtml | (1) Unexpected value for animVal during animation 
11:08:06     INFO -  3067 INFO TEST-PASS | /tests/content/svg/content/test/test_animLengthUnits.xhtml | (2) Unexpected value for baseVal after changing font-size during animation 
11:08:06     INFO -  3068 INFO TEST-PASS | /tests/content/svg/content/test/test_animLengthUnits.xhtml | (2) Unexpected value for animVal after changing font-size during animation 
11:08:06     INFO -  3069 INFO TEST-PASS | /tests/content/svg/content/test/test_animLengthUnits.xhtml | (3) Unexpected value for baseVal after changing font-size whilst frozen 
11:08:06     INFO -  3070 INFO TEST-UNEXPECTED-FAIL | /tests/content/svg/content/test/test_animLengthUnits.xhtml | (3) Unexpected value for animVal after changing font-size whilst frozen - got 50, expected 70
11:08:06     INFO -  TEST-INFO | expected PASS
11:08:06     INFO -  3071 INFO TEST-OK | /tests/content/svg/content/test/test_animLengthUnits.xhtml | took 2717ms
11:08:06     INFO -  3072 INFO TEST-START | /tests/content/svg/content/test/test_bbox-with-invalid-viewBox.xhtml
11:08:06     INFO -  3073 INFO TEST-OK | /tests/content/svg/content/test/test_bbox-with-invalid-viewBox.xhtml | took 2323ms
11:08:06     INFO -  3074 INFO TEST-START | /tests/content/svg/content/test/test_bbox.xhtml
11:08:06     INFO -  3075 INFO TEST-OK | /tests/content/svg/content/test/test_bbox.xhtml | took 4375ms
11:08:06     INFO -  3076 INFO TEST-START | /tests/content/svg/content/test/test_bounds.html
11:08:06     INFO -  3077 INFO TEST-OK | /tests/content/svg/content/test/test_bounds.html | took 3682ms
Jonathan, do you have cycles to look at this very frequent failure?
Flags: needinfo?(jwatt)
(In reply to Ryan VanderMeulen [:RyanVM UTC-4] from comment #112)
> Jonathan, do you have cycles to look at this very frequent failure?

Well, yuck, the test uses a 100ms timeout with the assumption that an animation sample will occur in that time. Brian, can't we listen for some sort of animation event?
Flags: needinfo?(jwatt) → needinfo?(birtles)
(In reply to Jonathan Watt [:jwatt] from comment #131)
> Well, yuck, the test uses a 100ms timeout with the assumption that an
> animation sample will occur in that time. Brian, can't we listen for some
> sort of animation event?

In both cases where we have the 100ms timeout, animations are paused (and we're testing that the font-size change has an effect, despite the fact that animations are paused).  So there are no animation events firing here.

Maybe we could set up a MozAfterPaint handler, though, and refuse to proceed until that handler fires with the tested thing being at the expected value?  (If it's not at the expected value yet, we could just assume it's some other random paint, and that our change is still eventually going to trigger a repaint.)  Then, if we happened to get this wrong, we'd fail by timing out (which is fine, since we don't expect to fail).
(In reply to Daniel Holbert [:dholbert] from comment #148)
> Maybe we could set up a MozAfterPaint handler, though, and refuse to proceed
> until that handler fires with the tested thing being at the expected value? 
> (If it's not at the expected value yet, we could just assume it's some other
> random paint, and that our change is still eventually going to trigger a
> repaint.)  Then, if we happened to get this wrong, we'd fail by timing out
> (which is fine, since we don't expect to fail).

I wonder if rAF would work? If not, MozAfterPaint is fine. We already have paint_listener.js for that and I recently made it useable in regular mochitests without Chrome privileges.
Flags: needinfo?(birtles)
Assignee: nobody → birtles
Status: NEW → ASSIGNED
I'm hoping this does the trick. Currently running on try:

  https://tbpl.mozilla.org/?tree=Try&rev=ea27e9ab9c98
Attachment #8495713 - Flags: review?(dholbert)
Attachment #8495713 - Flags: review?(dholbert) → review+
https://hg.mozilla.org/mozilla-central/rev/af9aaed61d20
Status: ASSIGNED → RESOLVED
Closed: 9 years ago
Resolution: --- → FIXED
Target Milestone: --- → mozilla35
You need to log in before you can comment on or make changes to this bug.