1505790 - toolkit/mozapps/update/tests/browser/browser_updatesBackgroundWindow.js | A promise chain failed to handle a rejection: win.gBrowser is undefined; can't access its "tabs" property when Gecko 65 merges to Beta on 2018-12-03

Status: VERIFIED FIXED

(Toolkit :: Application Update, defect)

Description

[Stefan Hindli [:stefan_hindli]]

[Tracking Requested - why for this release]:

This occurs on late beta: https://treeherder.mozilla.org/#/jobs?repo=try&resultStatus=testfailed%2Cbusted%2Cexception&classifiedState=unclassified&tochange=e002ac26fe2fa2e4e079dba4ff71cde2e35d0636&fromchange=667ddffc382ff2864cd02d4bd4d2f25c4e74dd85&selectedJob=210557876

https://treeherder.mozilla.org/logviewer.html#?job_id=210557876&repo=try&lineNumber=5383

Comment 1

[Sebastian Hengst [:aryx] (needinfo me if it's about an intermittent or backout)]

This showed up in the Try run for bug 1458308 and in today's late beta simulation: https://treeherder.mozilla.org/#/jobs?repo=try&revision=6c15f7d241b231cba58708952c67d546167d0ff9&selectedJob=209676942

Comment 2

[Robert Strong (they/them - no direct email)]

It showed up in an earlier version of the patch for bug 1458308, the patch was changed, and it didn't show up in later versions
https://treeherder.mozilla.org/#/jobs?repo=try&revision=48b9aec5261e75a10b3365ab56838f60c58aa56e

I don't think this has anything to do with bug 1458308.

Comment 3

[Robert Strong (they/them - no direct email)]

Also, bug 1458308 didn't land on beta so this doesn't have anything to do with bug 1458308.

Comment 4

[Sebastian Hengst [:aryx] (needinfo me if it's about an intermittent or backout)]

Beta simulations are done with central, not with beta. They take the code from central and run with beta configurations.

Comment 5

[Robert Strong (they/them - no direct email)]

So, this is in regards to the mozilla-central merge to mozilla-beta and not concerned with current beta.

I was suspicious of the actual code that is failing and will look into it.

Comment 6

[Robert Strong (they/them - no direct email)]

irc convo today

rstrong> jaws: not sure who to talk with about this. I have a test that was converted to async and is now intermittently failing in code in nsBrowserGlue.js in unrelated code with A promise chain failed to handle a rejection: win.gBrowser is undefined; can't access its "tabs" property - stack:
<rstrong> BG__setSyncAutoconnectDelay@jar:file:///Z:/task_1541678884/build/application/firefox/browser/omni.ja!/components/nsBrowserGlue.js:637:7
<rstrong> https://dxr.mozilla.org/mozilla-central/source/browser/components/nsBrowserGlue.js#634
<rstrong> jaws: tempted to just add a check for win.gBrowser there
<•jaws> looking at the code...
<•jaws> this code is so arbitrary
<•jaws> like why 1 second delay for each tab?
<•jaws> rstrong: i'd be fine with replacing that with something else, thinking about what else now
<•MattN> you could workaround by setting the services.sync.autoconnectDelay pref in your test
<•jaws> rstrong: for your test can you set services.sync.autoconnectDelay ?
<•jaws> yes, same as what mattn said
<rstrong> jaws: definitely and if you are fine with that I am :)
<•jaws> rstrong: yeah i'm fine with setting the pref

Comment 7

[Robert Strong (they/them - no direct email)]

Attachment: patch rev1

Comment 8

[Robert Strong (they/them - no direct email)]

Attachment: patch rev1

Just fixed a comment

Comment 9

[Pulsebot]

Pushed by [e-mail]:
https://hg.mozilla.org/integration/mozilla-inbound/rev/3b764f11b980
Fix intermittent failure of browser_updatesBackgroundWindow.js. r=bytesize

Comment 10

[Daniel Varga [:dvarga]]

https://hg.mozilla.org/mozilla-central/rev/3b764f11b980

Comment 11

[Sebastian Hengst [:aryx] (needinfo me if it's about an intermittent or backout)]

Unfortunately, this still fails >10% of runs: https://treeherder.mozilla.org/#/jobs?repo=try&revision=8c41a81e674a51e6375d7ce0eca0036684ecbe4d&searchStr=windows%2Cdevedition%2Cbrowser-chrome&selectedJob=210887942

Comment 12

[Robert Strong (they/them - no direct email)]

Attachment: patch rev2

The previous patch didn't set the pref for browser_updateAutoPrefUI.js since it doesn't call runUpdateTest.

Comment 13

[Robert Strong (they/them - no direct email)]

Note: the reason browser_updateAutoPrefUI.js is that browser_updatesBackgroundWindow.js immediately follows browser_updateAutoPrefUI.js in the test run and the call from sync that isn't able to find the tabs property happens in that test.

Comment 14

[Pulsebot]

Pushed by [e-mail]:
https://hg.mozilla.org/integration/mozilla-inbound/rev/999652c2fc2b
Second attempt to fix Bug 1505790 - Fix intermittent failure of browser_updatesBackgroundWindow.js. r=bytesize

Comment 15

[Robert Strong (they/them - no direct email)]

aryx, could you do another push and include this patch? Thanks!

Comment 16

[Robert Strong (they/them - no direct email)]

FYI: there are other tests that intermittently fail like this one did. Bug 1506271

Comment 17

[Stefan Hindli [:stefan_hindli]]

https://hg.mozilla.org/mozilla-central/rev/999652c2fc2b

Comment 18

[Intermittent Failures Robot]

6 failures in 3007 pushes (0.002 failures/push) were associated with this bug in the last 7 days.

Repository breakdown:
* try: 5
* autoland: 1

Platform breakdown:
* windows7-32: 1
* windows7-32-nightly: 2
* windows7-32-devedition: 3

For more details, see:
https://treeherder.mozilla.org/intermittent-failures.html#/bugdetails?bug=1505790&startday=2018-11-05&endday=2018-11-11&tree=all

Comment 19

[Andreea Pavel [:apavel]]

No longer appears on latest beta simulation:

https://treeherder.mozilla.org/#/jobs?repo=try&resultStatus=success%2Ctestfailed%2Cbusted%2Cexception%2Cretry%2Cusercancel%2Crunnable&revision=e9150112ebed08f3a278e28259c4683a44ba8e31&selectedJob=211093728

Comment 20

[Sebastian Hengst [:aryx] (needinfo me if it's about an intermittent or backout)]

It moved from bc2 to bc5 and didn't fail any of 26 runs (it's perma orange due to a different failing test): https://treeherder.mozilla.org/#/jobs?repo=try&revision=e9150112ebed08f3a278e28259c4683a44ba8e31&selectedJob=211201245

Comment 21

[Robert Strong (they/them - no direct email)]

That permaorange test actually runs before this test or any of the update tests so it looks like this bug is fixed.

Comment 22

[Cristina Coroiu [:ccoroiu]]

Checked on the latest beta sim, no failure found.
https://treeherder.mozilla.org/#/jobs?repo=try&resultStatus=testfailed%2Cbusted%2Cexception%2Crunnable&revision=df5ed7277d57f90489105a63575f6257dcecb028