592998 - 4.0b5,6pre Crash Report [@ nsXPCWrappedJS::`scalar deleting destructor''(unsigned int) ]

Status: RESOLVED FIXED

(Core :: XPConnect, defect)

Description

[Carsten Book [:Tomcat]]

New Crash in the Nightly's Windows only so far -> http://crash-stats.mozilla.com/report/list?range_value=2&range_unit=weeks&date=2010-09-02%2001%3A00%3A00&signature=nsXPCWrappedJS%3A%3A%60scalar%20deleting%20destructor%27%27%28unsigned%20int%29&version=Firefox%3A4.0b6pre

Crashing Thread
Frame 	Module 	Signature [Expand] 	Source
0 	xul.dll 	nsXPCWrappedJS::`scalar deleting destructor' 	
1 	xul.dll 	nsXPCWrappedJSClass::CheckForException 	js/src/xpconnect/src/xpcwrappedjsclass.cpp:1094
2 	xul.dll 	nsXPCWrappedJSClass::CallMethod 	js/src/xpconnect/src/xpcwrappedjsclass.cpp:1726
3 	xul.dll 	nsXPCWrappedJS::CallMethod 	js/src/xpconnect/src/xpcwrappedjs.cpp:571
4 	xul.dll 	PrepareAndDispatch 	xpcom/reflect/xptcall/src/md/win32/xptcstubs.cpp:114
5 	xul.dll 	SharedStub 	xpcom/reflect/xptcall/src/md/win32/xptcstubs.cpp:141
6 	xul.dll 	nsNSSComponent::ShowAlert 	security/manager/ssl/src/nsNSSComponent.cpp:2339

Show/hide other threads

Comment 1

[Scoobidiver (away)]

For the Windows 7 crashes : module DropBox 1.0.0.14 is always present.

Only one crash for Windows XP with a different crashing thread :
6 	xul.dll 	NS_InvokeByIndex_P 	xpcom/reflect/xptcall/src/md/win32/xptcinvoke.cpp:102

Comment 2

[chris hofmann]

this appears to have been around in low volume on 3.5.x and 3.6.x with 0-1 crashes per day, but turned into #30 topcrash on beta5 yesterday.  need to keep an eye on continuing volume.

date     tl crashes at, count build, count build, ...
         nsXPCWrappedJS::.scalar.deleting.destructor...unsigned.int.
20100830 1 ,1 4.0b42010081813
20100831
20100901 11 ,10 4.0b6pre2010090104 ,1 4.0b52010083108
20100902 2 ,2 4.0b6pre2010090104
20100903
20100904 1 ,1 4.0b42010081813
20100905
20100906 1 ,1 3.6.82010072215
20100907 13 ,1 4.0b6pre2010090604 ,12 4.0b52010083108
20100908 30 ,30 4.0b52010083108

Correlation to startup or time of session
30 total crashes for nsXPCWrappedJS::.scalar.deleting.destructor...unsigned.int. on 20100908-crashdata.csv
20 startup crashes inside 30 sec.
27 startup crashes inside 3 min.
17 repeated crashes inside 3 min. of last crash

os breakdown
nsXPCWrappedJS::.scalar.deleting.destructor...unsigned.int.Total 30
Win5.1  0.47
Win6.0  0.03
Win6.1  0.50

some changes going on in the source files on the stack but not sure anything directly related.

more reports at
http://crash-stats.mozilla.com/report/list?signature=%20%20%20%20nsXPCWrappedJS::%60scalar%20deleting%20destructor%27%27%28unsigned%20int%29

Comment 3

[Johnny Stenback (:jst)]

Looking at the crash reports here it seems like this dropped to only 10 crashes in the last 2 weeks, and looking at the installed extensions they're all probably from the same user. Not blocking on this unless this takes a turn for the worse.

Comment 4

[chris hofmann]

yeah, that 30 crash day looks like an anomaly.  It does look like this regressed a bit in 4.0 but volume is still low.

date     tl crashes at, count build, count build, ...
         nsXPCWrappedJS::.scalar.deleting.destructor...unsigned.int.
20100910 2 ,, 1 4.0b6pre2010091004, 1 3.6.92010082415, 
20100911 11 4.0b52010083108, 11 ,, 
20100912 7 4.0b52010083108, 7 ,, 
20100913 2 ,, 1 3.6.92010082415, 1 3.6.32010040108, 
20100914 2 4.0b52010083108, 2 ,, 
20100915  ,, 
20100916 4 4.0b62010091408, 4 ,, 
20100917 8 ,, 4 4.0b62010091408, 4 4.0b52010083108, 
20100918 3 ,, 2 4.0b62010091408, 1 4.0b52010083108, 
20100919 4 ,, 3 4.0b62010091408, 1 3.6.92010082415, 
20100920 5 4.0b62010091408, 5 ,, 
20100921 10 ,, 6 4.0b7pre2010091904, 4 4.0b62010091408, 
20100922 5 ,, 3 4.0b62010091408, 2 4.0b7pre2010091904,

Comment 5

[Marcia Knous [:marcia]]

According to http://people.mozilla.com/~chofmann/crash-stats/20101114/compare-rank-40b6-40b7-40b8pre.txt, this is no longer seen in Beta 7. Removing Crashkill keyboard and resolving as WFM.

Comment 6

[Henrik Gemal]

just crashes with this signature. should I reopen or file new bug?
https://crash-stats.mozilla.com/report/index/bp-c7c6acbe-9921-45b1-bbc9-419592110225

Comment 7

[Johnny Stenback (:jst)]

Reopening. Unclear whether this is the same exact issue or something slightly different, but the stacks are close enough that a new bug doesn't really help us here.

Comment 8

[Marcia Knous [:marcia]]

Bug 641378 references this stack, although it may not be the same issue since it does not sound like a very common operation.

Comment 9

[Blake Kaplan (:mrbkap) (inactive)]

Bug 641378 has landed. Optimistically marking as FIXED. Please reopen if I'm mistaken.

(As a note: the exact operation in bug 641378 that caused the crash was causing a certain type of exception to be thrown from JS -> C++, so while the steps to reproduce might not have been common, we do throw exceptions fairly regularly so it isn't too difficult to imagine that both bugs describe the same problem.)