1915974 - Remove toolkit/components/places/tests/expiration/test_notifications.js

Status: RESOLVED FIXED

(Toolkit :: Places, defect)

Description

[Treeherder Bug Filer]

Filed by: sstanca [at] mozilla.com
Parsed log: https://treeherder.mozilla.org/logviewer?job_id=472381352&repo=autoland
Full log: https://firefox-ci-tc.services.mozilla.com/api/queue/v1/task/BhawfF5bRaixfq4A3Xp6VQ/runs/0/artifacts/public/logs/live_backing.log

---

[task 2024-08-30T17:10:05.410Z] 17:10:05     INFO -  TEST-PASS | browser/components/migration/tests/unit/test_Chrome_corrupt_history.js | took 1808ms
[task 2024-08-30T17:10:05.413Z] 17:10:05     INFO -  TEST-START | toolkit/components/places/tests/expiration/test_notifications.js
[task 2024-08-30T17:10:09.407Z] 17:10:09  WARNING -  TEST-UNEXPECTED-FAIL | toolkit/components/places/tests/expiration/test_notifications.js | xpcshell return code: 0
[task 2024-08-30T17:10:09.407Z] 17:10:09     INFO -  TEST-INFO took 3993ms

Comment 1

[Emilio Cobos Álvarez (:emilio)]

So this is really weird, and my only thing where my patch is different is that we now initialize gfxplatform, because for xpcshell we happened to be overriding srgb here which didn't hit the problematic code-path...

Marco, do you know what might be going on? I've never seen an xpcshell test failing with "return code: 0" and no error, and it's not clear if the right test is even being blamed for this, since I don't see why this test in particular would hit any color management code...

Comment 2

[Serban Stanca [:SerbanS]]

The culprit was backed out due to the appearance of another failures.

Comment 3

[Intermittent Failures Robot]

1 failures in 3879 pushes (0.0 failures/push) were associated with this bug in the last 7 days.

Repository breakdown:

• autoland: 1

Platform and build breakdown:

• linux1804-64-tsan-qr: 1
  • opt: 1

For more details, see:
https://treeherder.mozilla.org/intermittent-failures/bugdetails?bug=1915974&startday=2024-08-26&endday=2024-09-01&tree=all

Comment 4

[Marco Bonardo [:mak]]

So it looks like we're not receiving TOPIC_EXPIRATION_FINISHED at all, instead of once.
The test opening comment is obsolete, there's no category cache involved here.

The only things I could think of is:

1. expiration is broken and it throws somewhere in here
2. the 2 seconds used by the test are no longer sufficient, something became slower or started happening.

Does increasing the timer in the test to 10s solve the problem?
We may want to await for promiseForceExpirationStep(1) in addition to the 2s timer.

Though, I'm not really sure this test is necessary. it was originally checking category cache didn't notify twice, while now we basically call .observe() directly, that is unlikely to be wrong.
Thus I'd suggest to just remove this test and I can review a patch doing so.

Comment 5

[Emilio Cobos Álvarez (:emilio)]

Attachment: Bug 1915974 - Remove toolkit/components/places/tests/expiration/test_notifications.js. r=mak!

Comment 6

[Pulsebot]

Pushed by [e-mail]:
https://hg.mozilla.org/integration/autoland/rev/b1c57d6b9796
Remove toolkit/components/places/tests/expiration/test_notifications.js. r=mak,places-reviewers

Comment 7

[Cristian Tuns]

https://hg.mozilla.org/mozilla-central/rev/b1c57d6b9796