Open Bug 1668853 Opened 4 years ago Updated 1 year ago

Intermittent /html/browsers/history/the-history-interface/005.html | single tracking bug

Categories

(Core :: DOM: Core & HTML, defect, P5)

Product:
Core
Component:
DOM: Core & HTML
Type:
defect

Tracking

()

Status:
REOPENED
Project Flags:
Fission Milestone Future

People

(Reporter: intermittent-bug-filer, Unassigned)

References

Details

(Keywords: intermittent-failure, intermittent-testcase)

Filed by: rmaries [at] mozilla.com
Parsed log: https://treeherder.mozilla.org/logviewer.html#?job_id=317401138&repo=autoland
Full log: https://firefox-ci-tc.services.mozilla.com/api/queue/v1/task/INOsD48_TdGFM1zZhifscw/runs/0/artifacts/public/logs/live_backing.log

[task 2020-10-02T14:48:21.625Z] 14:48:21     INFO - TEST-START | /html/browsers/history/the-history-interface/005.html
[task 2020-10-02T14:48:21.633Z] 14:48:21     INFO - Closing window 6442450948
[task 2020-10-02T14:48:23.523Z] 14:48:23     INFO - 
[task 2020-10-02T14:48:23.523Z] 14:48:23     INFO - TEST-PASS | /html/browsers/history/the-history-interface/005.html | history.pushState support is needed for this testcase 
[task 2020-10-02T14:48:23.524Z] 14:48:23     INFO - TEST-UNEXPECTED-FAIL | /html/browsers/history/the-history-interface/005.html | <body onpopstate="..."> should register a listener for the popstate event - assert_true: expected true got false
[task 2020-10-02T14:48:23.524Z] 14:48:23     INFO - step2/<@http://web-platform.test:8000/html/browsers/history/the-history-interface/005.html:27:16
[task 2020-10-02T14:48:23.524Z] 14:48:23     INFO - Test.prototype.step@http://web-platform.test:8000/resources/testharness.js:2001:25
[task 2020-10-02T14:48:23.525Z] 14:48:23     INFO - test@http://web-platform.test:8000/resources/testharness.js:535:30
[task 2020-10-02T14:48:23.525Z] 14:48:23     INFO - step2@http://web-platform.test:8000/html/browsers/history/the-history-interface/005.html:26:7
[task 2020-10-02T14:48:23.525Z] 14:48:23     INFO - .```

https://wiki.mozilla.org/Bug_Triage#Intermittent_Test_Failure_Cleanup
For more information, please visit auto_nag documentation.

Status: NEW → RESOLVED
Closed: 4 years ago
Resolution: --- → INCOMPLETE

Recent log:
https://treeherder.mozilla.org/logviewer?job_id=335709226&repo=autoland&lineNumber=13645

Status: RESOLVED → REOPENED
Resolution: INCOMPLETE → ---

This WPT failure is an intermittent failure on Linux that happens for both Fission and e10s. Thus it doesn't need to block Fission MVP.

https://searchfox.org/mozilla-central/rev/72951aa826642f048da4c6b71b8b3e36a9606dcd/testing/web-platform/meta/html/browsers/history/the-history-interface/005.html.ini#2-5

  [<body onpopstate="..."> should register a listener for the popstate event]
    expected:
      if (os == "linux") and webrender and fission and debug: ["PASS", "FAIL"]
      if (os == "linux") and not webrender and not debug: ["PASS", "FAIL"]
Whiteboard: [not-a-fission-bug]

https://wiki.mozilla.org/Bug_Triage#Intermittent_Test_Failure_Cleanup
For more information, please visit auto_nag documentation.

Status: REOPENED → RESOLVED
Closed: 4 years ago → 4 years ago
Resolution: --- → INCOMPLETE

Recent failure log: https://treeherder.mozilla.org/logviewer?job_id=341571537&repo=autoland&lineNumber=6449

Status: RESOLVED → REOPENED
Resolution: INCOMPLETE → ---

Over the last two months, this test has only ever failed on Fission. This test only failed on Windows 10 before and today failed on Linux ASan:

https://treeherder.mozilla.org/intermittent-failures/bugdetails?startday=2021-02-02&endday=2021-06-02&tree=trunk&bug=1668853

While this test supposedly fails intermittently on Linux without Fission (according to the WPT expectations in 005.html.ini), Fission seems to make this failure much more likely due to timing changes. Let's monitor this test bug to see if failures are becoming more frequent for some reason.

WPT expectations for this test:

https://searchfox.org/mozilla-central/rev/74f3c420ee54001059e1850bef3be876749ff873/testing/web-platform/meta/html/browsers/history/the-history-interface/005.html.ini#2-9

[<body onpopstate="..."> should register a listener for the popstate event]
  bug:
    if fission and (os == "linux"): https://bugzilla.mozilla.org/show_bug.cgi?id=1668853
    if fission and (os == "win") and debug: https://bugzilla.mozilla.org/show_bug.cgi?id=1668853
  expected:
    if (os == "linux") and webrender and fission and debug: ["PASS", "FAIL"]
    if (os == "win") and fission and debug: ["PASS", "FAIL"]
    if (os == "linux") and not webrender and not debug: ["PASS", "FAIL"]
Blocks: 1696042
Fission Milestone: --- → Future
Summary: Intermittent /html/browsers/history/the-history-interface/005.html | <body onpopstate="..."> should register a listener for the popstate event - assert_true: expected true got false → Fission? Intermittent /html/browsers/history/the-history-interface/005.html | <body onpopstate="..."> should register a listener for the popstate event - assert_true: expected true got false
Whiteboard: [not-a-fission-bug]
Summary: Fission? Intermittent /html/browsers/history/the-history-interface/005.html | <body onpopstate="..."> should register a listener for the popstate event - assert_true: expected true got false → Intermittent [fission] /html/browsers/history/the-history-interface/005.html | <body onpopstate="..."> should register a listener for the popstate event - assert_true: expected true got false

This test is pretty flaky (see also 1692852), it expects timeouts to run after a previous history.go has fired the popstate event, but that's not guaranteed at all (https://searchfox.org/mozilla-central/source/testing/web-platform/tests/html/browsers/history/the-history-interface/005.html#22-23,30-31).

Severity: normal → S3
Keywords: intermittent-testcase
Summary: Intermittent [fission] /html/browsers/history/the-history-interface/005.html | <body onpopstate="..."> should register a listener for the popstate event - assert_true: expected true got false → Intermittent /html/browsers/history/the-history-interface/005.html | single tracking bug
Duplicate of this bug: 1692852
You need to log in before you can comment on or make changes to this bug.