445204 - crash [@ js_GC]

Status: RESOLVED WORKSFORME

(Core :: XPConnect, defect)

Description

[Wolfgang Rosenauer [:wolfiR]]

http://crash-stats.mozilla.com/report/index/99369fee-4f8c-11dd-bf25-001a4bd43ed6

I don't have much more information. This report was created when a user tried to reproduce a certain crash reported to Novell's bugzilla.
https://bugzilla.novell.com/show_bug.cgi?id=406288

Comment 1

[Jesse Ruderman]

Brendan, is there enough information in the stack trace and novell bug report?  If not, can you suggest what information Wolfgang can add, or mark it as incomplete?

Comment 2

[Brendan Eich [:brendan]]

If my line numbering is right, the faulting line is

        a->u.hasMarkedDoubles = JS_FALSE;

Please confirm, anyone.

Cc'ing Igor, to see what he can make of it.

/be

Comment 3

[Wayne Mery (:wsmwk)]

(In reply to comment #0)
> http://crash-stats.mozilla.com/report/index/99369fee-4f8c-11dd-bf25-001a4bd43ed6

0	libmozjs.so	js_GC	 mozilla/js/src/jsgc.c:3225
1	libmozjs.so	JS_GC	mozilla/js/src/jsapi.c:2469
2	libxul.so	nsXPConnect::Collect	mozilla/js/src/xpconnect/src/nsXPConnect.cpp:529
3	libxul.so	nsCycleCollector::Collect	mozilla/xpcom/base/nsCycleCollector.cpp:2250
4	libxul.so	nsCycleCollector_collect	mozilla/xpcom/base/nsCycleCollector.cpp:2898
5	libxul.so	nsJSContext::CC	mozilla/dom/src/base/nsJSEnvironment.cpp:3346
6	libxul.so	nsJSContext::MaybeCC	mozilla/dom/src/base/nsJSEnvironment.cpp:3397
7	libxul.so	nsJSContext::CCIfUserInactive	mozilla/dom/src/base/nsJSEnvironment.cpp:3414 

bp-dbf70fef-c118-4e18-97e0-b0eeb2100323 v3.6.2 cites a personas url and has a reporter address
0	js3250.dll	js_GC	 js/src/jsgc.cpp:3622
1	js3250.dll	JS_GC	js/src/jsapi.cpp:2439
2	xul.dll	nsXPConnect::Collect	js/src/xpconnect/src/nsXPConnect.cpp:477
3	xul.dll	nsCycleCollector::Collect	xpcom/base/nsCycleCollector.cpp:2421

Comment 4

[Kai Engert (:KaiE:)]

crasher with same stack has been reported in fedora with seamonkey 2.0.3
https://bugzilla.redhat.com/show_bug.cgi?id=579152

the stack over there has variable contents, maybe it helps you
https://bugzilla.redhat.com/attachment.cgi?id=404240

Comment 5

[Jonathan Watt [:jwatt]]

I just got a crash in js_GC on Windows with FF 3.6.6. The stack looks a bit wrong though?

http://crash-stats.mozilla.com/report/index/2de68051-0301-4b4f-81ba-d189d2100628

Comment 6

[Wayne Mery (:wsmwk)]

FWIW, crash rate for FF 4.0 is near zero 
  https://crash-stats.mozilla.com/query/query?product=Firefox&version=ALL%3AALL&branch=1.9.3&branch=2.0&branch=2.2&range_value=2&range_unit=weeks&date=04%2F28%2F2011+06%3A16%3A36&query_search=signature&query_type=exact&query=js_GC&reason=&build_id=&process_type=any&hang_type=any&do_query=1

far, far less than for 3.x
  https://crash-stats.mozilla.com/query/query?product=Firefox&version=ALL%3AALL&branch=1.9.2&range_value=2&range_unit=weeks&date=04%2F28%2F2011+06%3A16%3A36&query_search=signature&query_type=exact&query=js_GC&reason=&build_id=&process_type=any&hang_type=any&do_query=1

Comment 7

[Wayne Mery (:wsmwk)]

all crashes are version 3.x

Comment 8

[Wayne Mery (:wsmwk)]

(In reply to Wayne Mery (:wsmwk) from comment #7)
> all crashes are version 3.x

all crashes on crash-stats that is