427934 - new crash in fx3 b5 [@0x20202020]

Status: RESOLVED DUPLICATE of bug 425596

(Core :: JavaScript Engine, defect)

Description

[chris hofmann]

Didn't see anything like this signature in beta4 so it looks like it might be new.

Its cross plaform

total 1605 - 
windows 575 - 
mac 936 - 	 
linux 94 - 

On linux stack sig looks like:

0  	 	@0x20202020  	
1 	libmozjs.so 	js_Invoke 	mozilla/js/src/jsinvoke.c:1303
2 	libxul.so 	nsXPCWrappedJSClass::CallMethod 	mozilla/js/src/xpconnect/src/xpcwrappedjsclass.cpp:1475
3 	libxul.so 	nsXPCWrappedJS::CallMethod 	mozilla/js/src/xpconnect/src/xpcwrappedjs.cpp:559
4 	libxul.so 	PrepareAndDispatch 	mozilla/xpcom/reflect/xptcall/src/md/unix/xptcstubs_gcc_x86_unix.cpp:95
5 	libxul.so 	NS_GetXPTCallStub_P 	
6 	libxul.so 	XPCWrappedNative::CallMethod 	mozilla/js/src/xpconnect/src/xpcwrappednative.cpp:2369
7 	libxul.so 	XPC_WN_CallMethod 	mozilla/js/src/xpconnect/src/xpcwrappednativejsops.cpp:1470
8 	libmozjs.so 	js_Invoke 	mozilla/js/src/jsinvoke.c:1287
9 	libmozjs.so 	js_Interpret 	mozilla/js/src/jsinterp.c:4841
10 	libmozjs.so 	js_Invoke 	mozilla/js/src/jsinvoke.c:1303


On Mac it's:

0  	 	@0x20202020  	
1 	libmozjs.dylib 	js_Invoke 	mozilla/js/src/jsinvoke.c:1303
2 	libmozjs.dylib 	fun_apply 	mozilla/js/src/jsfun.c:1650
3 	libmozjs.dylib 	js_Interpret 	mozilla/js/src/jsinterp.c:4824
4 	libmozjs.dylib 	js_Invoke 	mozilla/js/src/jsinvoke.c:1303
5 	XUL 	nsXPCWrappedJSClass::CallMethod 	mozilla/js/src/xpconnect/src/xpcwrappedjsclass.cpp:1475
6 	XUL 	nsXPCWrappedJS::CallMethod 	mozilla/js/src/xpconnect/src/xpcwrappedjs.cpp:559
7 	XUL 	PrepareAndDispatch 	mozilla/xpcom/reflect/xptcall/src/md/unix/xptcstubs_unixish_x86.cpp:93
8 	XUL 	nsXPTCStubBase::Stub3 	mozilla/xpcom/reflect/xptcall/src/md/unix/xptcstubs_unixish_x86.cpp:1
9 	XUL 	jsds_ExecutionHookProc 	mozilla/js/jsd/jsd_xpc.cpp:690
10 	XUL 	jsd_CallExecutionHook 	mozilla/js/jsd/jsd_hook.c:177


and On windows it's:

0  	 	@0x20202020  	
1 	js3250.dll 	js_Invoke 	mozilla/js/src/jsinterp.c:1303
2 	xul.dll 	nsXPCWrappedJSClass::CallMethod 	mozilla/js/src/xpconnect/src/xpcwrappedjsclass.cpp:1475
3 	xul.dll 	nsXPCWrappedJS::CallMethod 	mozilla/js/src/xpconnect/src/xpcwrappedjs.cpp:559
4 	xul.dll 	PrepareAndDispatch 	mozilla/xpcom/reflect/xptcall/src/md/win32/xptcstubs.cpp:114
5 	xul.dll 	SharedStub 	mozilla/xpcom/reflect/xptcall/src/md/win32/xptcstubs.cpp:141
6 	xul.dll 	NS_InvokeByIndex_P 	mozilla/xpcom/reflect/xptcall/src/md/win32/xptcinvoke.cpp:101
7 	xul.dll 	XPCWrappedNative::CallMethod 	mozilla/js/src/xpconnect/src/xpcwrappednative.cpp:2369

May need to split these out in to separate bugs.

Mac comments look like:

> trying to select a person to mail to in gmail's contacts

> Trying to access a draft in gmail with a link

> my only comment is Beta 5 is FAR worse than beta 4. Beta 5 is crashing on my hourly if not more... Beta 4 crashed maybe 1-2 times during the weeks that I used it.

> master password is newly enabled. firefox appears to be crashing at the point where the panel for entering the master password should appear.

> mail.yahoo.com crashes firefox

> i was working with google maps when it crashed

Comment 1

[chris hofmann]

the reports for this are at 
http://crash-stats.mozilla.com/report/list?product=Firefox&query_search=signature&query_type=contains&range_unit=weeks&version=Firefox%3A3.0b5&signature=%400x20202020&range_value=1

current ranking is #11 top crash for beta 5

Comment 2

[chris hofmann]

getting on the radar until we know more.

Comment 3

[chris hofmann]

on that comment

> master password is newly enabled. firefox appears to be crashing at the point where the panel for entering the master password should appear.

it looks like about 10% or more of the ~1600 the crash url's might be related to login/sign in pages.   anything change in password manager for b5?

Comment 4

[(not currently active) Ted Mielczarek]

I would hazard a guess that the differences in stack signatures between mac and win/lin are just due to compiler optimization. Note that [@ js_Invoke ] is also the #20 topcrash for b5.

Comment 5

[Jesse Ruderman]

0x20202020 means we're looking at something that got GCed, right?

For 3.0pre, these reports pretty much stopped after the 2008032805 build, so I'm tempted to mark this bug as WFM.

http://crash-stats.mozilla.com/report/list?range_unit=weeks&version=Firefox%3A3.0pre&range_value=2&signature=%400x20202020

Comment 6

[Igor Bukanov]

This could be a regression from bug 424376, the patch for which was backed out on 2008-03-28. I suspect the bug is dup of bug 425596, but that has to be checked

Comment 7

[Brian Crowder]

Based on Igor and Jesse's comments, I'd say we can confidently say that this bug went away with the backout of 424376, and is, as Igor suggests, a dup of bug 425596.  We definitely shouldn't keep it open without a better testcase, given the disappearance of signatures like it following the backout.