Closed Bug 1501138 Opened 2 years ago Closed 2 years ago

Bug 1437064 caused failures on comm-central

Categories

(Core :: DOM: Core & HTML, defect, P2)

defect

Tracking

()

RESOLVED FIXED
mozilla65
Tracking Status
firefox-esr60 --- unaffected
firefox63 --- unaffected
firefox64 --- fixed
firefox65 --- fixed

People

(Reporter: darktrojan, Assigned: farre)

References

Details

(Keywords: regression)

Attachments

(1 file)

I've come to the conclusion that bug 1437064 is what caused a bunch of failures on comm-central. Only on Linux for some reason. All of the failures showed up with perma-fails on our automation when this bug landed, and disappear if it is backed out.

Bug 1499240 and bug 1499286, which both seem to be caused by oncommand events not firing on textboxes (my understanding is that oncommand only fires if the user stops typing, so there's a timer involved somehow), and bug 1499316, which directly involves setTimeout.

I can understand if you're not interested in Thunderbird's problems, but there might be something in this that breaks Firefox too.
Flags: needinfo?(afarre)
Can you point me to some logs for this, or better yet point me to how I can build/debug this issue myself?
Flags: needinfo?(afarre)
If you're prepared to build Thunderbird, here are the instructions:
https://developer.mozilla.org/en-US/docs/Mozilla/Developer_guide/Build_Instructions/Simple_Thunderbird_build

It would have to be on Linux since the tests only fail there. I can let you know how to run those tests.

Or maybe investigate why textboxes don't seem to fire oncommand, see our fix here:
https://hg.mozilla.org/comm-central/rev/379f94cf13ee#l1.21
and proposed fix in attachment 9017461 [details] [diff] [review].

Best to catch us on IRC's #maildev channel.
Priority: -- → P2
Assignee: nobody → afarre
Duplicate of this bug: 1505783
Status: NEW → ASSIGNED
My developer has been debugging this for several hours, and came up with the exact same fix. He wanted to post it here and found that you had been working on it and found the same fix :).
Could you please land this ASAP? This has been causing lots of test failures on our end, and the hot fixes to it have caused further problems down, and the hotfixes to that hotfix have again caused other bugs.
Pushed by afarre@mozilla.com:
https://hg.mozilla.org/integration/autoland/rev/9daebaa1526d
Make sure that timeout iterator advances. r=peterv
r+ from our side.
We found the same fix, independent from you, which gives confidence in the fix. We also verified that this patch fixes the test failure we'd been fighting. The test had been passing before bug 1437064, fails with bug 1437064 and without this fix, and passes with this fix.
Comment on attachment 9023636 [details]
Bug 1501138 - Make sure that timeout iterator advances.

[Beta/Release Uplift Approval Request]

Feature/Bug causing the regression: Bug 1437064

User impact if declined: Breaks setTimeout which is a big deal

Is this code covered by automated tests?: Yes

Has the fix been verified in Nightly?: No

Needs manual test from QE?: No

If yes, steps to reproduce: The tests that managed to expose this issue are actually thunderbird tests, see for example Bug 1499240. So str becomes, for example:

1 Build Thunderbird
2 cd to obj dir
3 Run make SOLO_TEST=subscribe/test-subscribe-news-filter.js mozmill-one

List of other uplifts needed: None

Risk to taking this patch: Low

Why is the change risky/not risky? (and alternatives if risky): The change is small and well contained and doesn't add any new code paths.

String changes made/needed:
Attachment #9023636 - Flags: approval-mozilla-beta?
https://hg.mozilla.org/mozilla-central/rev/9daebaa1526d
Status: ASSIGNED → RESOLVED
Closed: 2 years ago
Resolution: --- → FIXED
Target Milestone: --- → mozilla65
Yay! :-)
Comment on attachment 9023636 [details]
Bug 1501138 - Make sure that timeout iterator advances.

fix a regression with timeouts exposed by tb tests, approved for 64.0b9
Attachment #9023636 - Flags: approval-mozilla-beta? → approval-mozilla-beta+
Component: DOM → DOM: Core & HTML
You need to log in before you can comment on or make changes to this bug.