Closed Bug 1461899 Opened 6 years ago Closed 6 years ago

Intermittent devtools/client/inspector/animation/test/browser_animation_keyframes-graph_computed-value-path_easing-hint.js | stroke-opacity of hintEl for hint[0] of margin-left in both-easing should be 0 while mouse is out from the element

Categories

(DevTools :: Inspector, defect, P5)

defect

Tracking

(firefox62 fixed)

RESOLVED FIXED
Tracking Status
firefox62 --- fixed

People

(Reporter: intermittent-bug-filer, Assigned: daisuke)

References

Details

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

Attachments

(2 files)

Filed by: apavel [at] mozilla.com

https://treeherder.mozilla.org/logviewer.html#?job_id=178694881&repo=autoland

https://queue.taskcluster.net/v1/task/SDK4jMZcT4mAmOWY1d9-Tw/runs/0/artifacts/public/logs/live_backing.log

23:16:40     INFO - Checking interaction for hint[0] of opacity in keyframe-easing
23:16:40     INFO - TEST-PASS | devtools/client/inspector/animation/test/browser_animation_keyframes-graph_computed-value-path_easing-hint.js | stroke-opacity of hintEl for hint[0] of opacity in keyframe-easing should be 0 while mouse is out from the element - 
23:16:40     INFO - TEST-PASS | devtools/client/inspector/animation/test/browser_animation_keyframes-graph_computed-value-path_easing-hint.js | stroke-opacity of hintEl for hint[0] of opacity in keyframe-easing should be 1 while mouse is over the element - 
23:16:40     INFO - Checking keyframes graph for both-easing
23:16:40     INFO - Click on animation whose selector of target element is '.both-easing'
23:16:40     INFO - Checking easing hint for margin-left in both-easing
23:16:40     INFO - Checking easing hint existence for margin-left in both-easing
23:16:40     INFO - TEST-PASS | devtools/client/inspector/animation/test/browser_animation_keyframes-graph_computed-value-path_easing-hint.js | Count of easing hint elements of margin-left in both-easing should be 1 - 
23:16:40     INFO - Checking hint[0] of margin-left in both-easing
23:16:40     INFO - Checking <title> in hint[0] of margin-left in both-easing
23:16:40     INFO - TEST-PASS | devtools/client/inspector/animation/test/browser_animation_keyframes-graph_computed-value-path_easing-hint.js | <title> element in hint[0] of margin-left in both-easing should be existence - 
23:16:40     INFO - TEST-PASS | devtools/client/inspector/animation/test/browser_animation_keyframes-graph_computed-value-path_easing-hint.js | Content of <title> in hint[0] of margin-left in both-easing should be steps(1) - 
23:16:40     INFO - Checking <path> in hint[0] of margin-left in both-easing
23:16:40     INFO - TEST-PASS | devtools/client/inspector/animation/test/browser_animation_keyframes-graph_computed-value-path_easing-hint.js | The <path> element  in hint[0] of margin-left in both-easing should be existence - 
23:16:40     INFO - TEST-PASS | devtools/client/inspector/animation/test/browser_animation_keyframes-graph_computed-value-path_easing-hint.js | The tested element should have pathSegList - 
23:16:40     INFO - TEST-PASS | devtools/client/inspector/animation/test/browser_animation_keyframes-graph_computed-value-path_easing-hint.js | The path segment of x 0, y 0 should be passing through - 
23:16:40     INFO - TEST-PASS | devtools/client/inspector/animation/test/browser_animation_keyframes-graph_computed-value-path_easing-hint.js | The path segment of x 999, y 0 should be passing through - 
23:16:40     INFO - TEST-PASS | devtools/client/inspector/animation/test/browser_animation_keyframes-graph_computed-value-path_easing-hint.js | The path segment of x 1000, y 100 should be passing through - 
23:16:40     INFO - Checking interaction for hint[0] of margin-left in both-easing
23:16:40     INFO - Buffered messages finished
23:16:40     INFO - TEST-UNEXPECTED-FAIL | devtools/client/inspector/animation/test/browser_animation_keyframes-graph_computed-value-path_easing-hint.js | stroke-opacity of hintEl for hint[0] of margin-left in both-easing should be 0 while mouse is out from the element - Got 1, expected 0
23:16:40     INFO - Stack trace:
23:16:40     INFO - chrome://mochikit/content/browser-test.js:test_is:1284
23:16:40     INFO - chrome://mochitests/content/browser/devtools/client/inspector/animation/test/browser_animation_keyframes-graph_computed-value-path_easing-hint.js:null:249
23:16:40     INFO - TEST-PASS | devtools/client/inspector/animation/test/browser_animation_keyframes-graph_computed-value-path_easing-hint.js | stroke-opacity of hintEl for hint[0] of margin-left in both-easing should be 1 while mouse is over the element - 
23:16:40     INFO - Checking easing hint for opacity in both-easing
23:16:40     INFO - Checking easing hint existence for opacity in both-easing
23:16:40     INFO - TEST-PASS | devtools/client/inspector/animation/test/browser_animation_keyframes-graph_computed-value-path_easing-hint.js | Count of easing hint elements of opacity in both-easing should be 1 - 
23:16:40     INFO - Checking hint[0] of opacity in both-easing
23:16:40     INFO - Checking <title> in hint[0] of opacity in both-easing
23:16:40     INFO - TEST-PASS | devtools/client/inspector/animation/test/browser_animation_keyframes-graph_computed-value-path_easing-hint.js | <title> element in hint[0] of opacity in both-easing should be existence - 
23:16:40     INFO - TEST-PASS | devtools/client/inspector/animation/test/browser_animation_keyframes-graph_computed-value-path_easing-hint.js | Content of <title> in hint[0] of opacity in both-easing should be steps(2) -
Flags: needinfo?(gl) → needinfo?(dakatsuka)
It seems like this is quite tough to hit on a retrigger. I wasn't sure how my patches could've caused this failure, but it took a bunch of retries on an earlier revision before I hit the error: https://treeherder.mozilla.org/#/jobs?repo=autoland&filter-searchStr=OS%20X%2010.10%20opt%20Mochitests%20with%20e10s%20test-macosx64%2Fopt-mochitest-devtools-chrome-e10s-5%20M-e10s(dt5)&tochange=d5db2927b771045c2521f048cea5cf0034befda6&fromchange=1299ab00899ffe1512b7416bdceb92a1eda8e0f7&group_state=expanded&selectedJob=179336451

So it looks like the regression happened sometime earlier, but it would probably take a lot of retriggers to find exactly where.
Flags: needinfo?(mshal)
This bug has failed 115 times in the last 7 days so created a disable patch if needed.
Assignee: nobody → aiakab
Attachment #8980833 - Flags: review?(jmaher)
There are only a couple of failures on asan and debug for those two platforms. I think we need to skip only on opt/pgo here, and only linux64. I would do something like this:
skip-if = !debug  && (os == "win" || os == "linux" && bits == 64) #Bug 1461899.
Thank you for the patch, Arthur, Cosmin!
And I'm so sorry for delay.

I'll take a look this tomorrow once.
If I could not find the cause by tomorrow, please disable the test.
Comment on attachment 8981052 [details]
Bug 1461899: Use mouseover of another element instead of mouseout.

Let me cancel the reviewing since the try has errors.
Attachment #8981052 - Flags: review?(gl)
Comment on attachment 8980833 [details] [diff] [review]
Disabled on Linux and Windows

Review of attachment 8980833 [details] [diff] [review]:
-----------------------------------------------------------------

this patch looks good, lets see if the real fix gets r+ and landed ok today
Attachment #8980833 - Flags: review?(jmaher) → review+
For now, the try looks no failed of this test.
Whiteboard: [stockwell disable-recommended] → [stockwell disabled]
Although I have found the cause and made the patch...
the patch is not r+ or landed, we have waited on many patches in this state for a week+, when it is r+ and ready to land, then enable the test- it is simple to do.
Pushed by csabou@mozilla.com:
https://hg.mozilla.org/integration/mozilla-inbound/rev/7fd4f748a785
Disable browser_animation_keyframes-graph_computed-value-path_easing-hint.js on Windows and Linux. r=jmaher
Keywords: checkin-needed
Comment on attachment 8981052 [details]
Bug 1461899: Use mouseover of another element instead of mouseout.

https://reviewboard.mozilla.org/r/247172/#review254588

::: commit-message-6b3d9:1
(Diff revision 2)
> +Bug 1461899: Change the way for mouse out. r?gl

This needs a better commit message.
Attachment #8981052 - Flags: review?(gl) → review+
Hello Joel.
I'd like to enable the tests again.
I did tests (20 times respectively) that had failed in the past, but there were no errors.
https://treeherder.mozilla.org/#/jobs?repo=try&revision=03eb1d8d3421092774e905aa09b3d58af2b2a8a5
Comment on attachment 8981052 [details]
Bug 1461899: Use mouseover of another element instead of mouseout.

https://reviewboard.mozilla.org/r/247172/#review256490

thanks for doing the retriggers on try and enabling this test again :)
Attachment #8981052 - Flags: review?(jmaher) → review+
Thanks, will land the patch!
Pushed by dakatsuka@mozilla.com:
https://hg.mozilla.org/integration/autoland/rev/b586283c989d
Use mouseover of another element instead of mouseout. r=gl,jmaher
Product: Firefox → DevTools
Updating assignee to Daisuke and marking as fixed since it appears his patch has fixed this.
Assignee: aiakab → dakatsuka
Status: NEW → RESOLVED
Closed: 6 years ago
Resolution: --- → FIXED
Whiteboard: [stockwell disabled] → [stockwell fixed]
Keywords: leave-open
Whiteboard: [stockwell fixed] → [stockwell fixed:product]
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Created:
Updated:
Size: