Open Bug 1335582 Opened 7 years ago Updated 2 years ago

Intermittent browser/components/extensions/test/browser/browser_ext_tabs_cookieStoreId.js | application crashed [@ js::GCMarker::eagerlyMarkChildren]

Categories

(Core :: JavaScript: GC, defect, P3)

defect

Tracking

()

People

(Reporter: intermittent-bug-filer, Unassigned)

References

(Blocks 1 open bug)

Details

(Keywords: intermittent-failure, triage-deferred)

:zombie - Could this failure be fallout from your change in bug 1330167? (I'm just noting that you modified the test a couple of weeks before this failure apparently started.)
Flags: needinfo?(tomica)
I'm not going to claim it is impossible, but IMO very unlikely it is related to that.  My changes are very minimal, basically boil down to reading a property from a JSON object and calling a JS array method.  Nothing even remotely tricky/low level, that could kill the process.

From my limited understanding of reading error stacks, this seems like crash in GC, so maybe ping some of the spidermonkey folks to take a look?
Flags: needinfo?(tomica)
Ah, didn't notice, it's already in the GC component.
I'm not sure why this bug is consistently showing up for this test, and in the same place, but so far it looks like what's happening is that we're trying to trace a function's name atom, and the base pointer for that atom's string is an invalid upcast of a null pointer. Or something along those lines, anyway.

It's hard to say more without being able to produce this locally, and I'm probably not the best person to look into it further, anyway.

[1]: http://searchfox.org/mozilla-central/rev/f5077ad52f8b90183e73038869f6140f0afbf427/browser/components/extensions/test/browser/browser_ext_tabs_cookieStoreId.js#85-89
Given comment 5, this is probably a regression from bug 1324002.
Blocks: 1324002
See Also: → 1346455
Keywords: triage-deferred
Priority: -- → P3
Severity: normal → S3
You need to log in before you can comment on or make changes to this bug.