Closed Bug 390225 Opened 17 years ago Closed 17 years ago

Mac-only trunk topcrash at [@ JS_GetGlobalObject]

Categories

(Core Graveyard :: Java: Live Connect, defect)

x86
macOS
defect
Not set
critical

Tracking

(Not tracked)

RESOLVED FIXED
mozilla1.9alpha8

People

(Reporter: benjamin, Unassigned)

Details

(Keywords: regression, topcrash+)

There was a trunk-only topcrash regression introduced around 7/16 at JS_GetGlobalObject. It seems probably related to liveconnect and maybe JEP: http://crash-stats.mozilla.com/report/list?range_unit=weeks&branch=1.9&range_value=2&signature=JS_GetGlobalObject
Flags: blocking1.9+
I'm pretty sure this is a dupe of bug 387975, which is supposedly fixed by the new JEP, which is pending review/approval (bug 386918).
> I'm pretty sure this is a dupe of bug 387975, It's more likely that other factors (as yet unknown) explain both that bug and this one. From the timing, clearly _some_ factor external to the JEP is involved. Did a new version of Adblock or Adblock Plus appear around that time? If so, this would confirm a connection to bug 387975. But why was the change so sudden, and why did it only happen on the trunk? I think the "external factor" is much more likely to be some change in the browser. Over the next week or so I'll try to find out what this might have been. (By the way, JEP 0.9.6.3 _does_ fix a crash bug that might have caused (or partially caused) the bad interaction with Adblock Plus, discussed at 387975. It's #1 in the JEP change log that I posted at bug 386918 (attachment 271050 [details]). It's possible the same JEP bug is (partially) implicated here. We'll only be able to tell after JEP 0.9.6.3 gets landed on the trunk.)
Another possible factor is bug 383783, which broke JavaScript-to-Java LiveConnect -- which seems to be involved in many of these crashes. But once again the timing is wrong -- bug 383783 was landed on 2007-06-18. (JEP 0.9.6.3 also works around this problem, which is #2 in my change log at attachment 271050 [details].) Side note: I think some of these crash logs are corrupt, and have nothing at all to do with Java. For example: http://crash-stats.mozilla.com/report/index/a6637bd9-3dbd-11dc-9b27-001a4bd46e84 http://crash-stats.mozilla.com/report/index/218170f0-3dbd-11dc-aff0-001a4bd43e5c
Severity: normal → critical
Summary: Mac-only trunk topcrash at JS_GetGlobalObject → Mac-only trunk topcrash at [@ JS_GetGlobalObject]
JEP 0.9.6.3 was landed on the trunk yesterday, and is in today's Minefield nightly. It'll be interesting to see what effect this has (over the next week or two) on this crash's frequency.
Reports of this problem at comment #0's URL stop dead with the Aug 14th trunk nightly. This seems to indicate that my JEP 0.9.6.3 upgrade fixed (or worked around) this problem ... whatever it was.
Status: NEW → RESOLVED
Closed: 17 years ago
Resolution: --- → FIXED
Product: Core → Core Graveyard
You need to log in before you can comment on or make changes to this bug.