374312 - crash in XPCJSRuntime::GCCallback

Status: RESOLVED WONTFIX

(Core :: JavaScript Engine, defect)

Description

[Christian Kirbach]

User-Agent:       Opera/9.10 (X11; Linux i686; U; en)
Build Identifier: 

transferred from http://bugzilla.gnome.org/show_bug.cgi?id=419196

#4  <signal handler called>
No symbol table info available.
#5  0x03ec6b70 in XPCJSRuntime::GCCallback (cx=0x8bd9458, status=JSGC_END)
    at xpcjsruntime.cpp:562
        obj = (nsISupports *) 0x8e93d60
        array = (class nsVoidArray *) 0x8bd7d90
        lock = (XPCLock *) 0x0
        dyingWrappedJSArray = <value optimized out>
        self = <value optimized out>
#6  0x01de5b48 in DOMGCCallback (cx=0x8bd9458, status=JSGC_END)
    at nsJSEnvironment.cpp:2234
        result = 1
#7  0x01a2aa83 in js_GC (cx=0x8bd9458, gcflags=0) at jsgc.c:1992
        rt = (JSRuntime *) 0x8bd0300
        iter = (JSContext *) 0x0
        acx = (JSContext *) 0x0
        fp = <value optimized out>
        chain = (JSStackFrame *) 0x0
        i = <value optimized out>
        depth = <value optimized out>
        nslots = <value optimized out>
        type = 4
        sh = <value optimized out>
        tvr = <value optimized out>
        nflags = 10
        a = (JSArena *) 0x0
        ap = (JSArena **) 0x8be7318
        flags = <value optimized out>
        flagp = <value optimized out>
        split = (uint8 *) 0x8be7400 ""
        thing = <value optimized out>
        limit = (JSGCThing *) 0x8be9400
        flp = (JSGCThing **) 0x8be93b0
        oflp = (JSGCThing **) 0x8be93b0
        finalizer = (GCFinalizeOp) 0
        bytesptr = <value optimized out>
        all_clear = 1
        currentThread = <value optimized out>
        requestDebit = 0
#8  0x01a2aadc in js_ForceGC (cx=0x8bd7d90, gcflags=0) at jsgc.c:1516
        i = <value optimized out>
#9  0x01a12432 in js_DestroyContext (cx=0x8bd9458, gcmode=JS_FORCE_GC)
    at jscntxt.c:285
        rt = (JSRuntime *) 0x8bd0300
        last = 0
        map = <value optimized out>
        lrs = <value optimized out>
        lrc = <value optimized out>
#10 0x01a0b81a in JS_DestroyContext (cx=0x8bd9458) at jsapi.c:932
No locals.
#11 0x03eca3b6 in ~XPCJSContextStack (this=0x8bd7c60)
    at xpcthreadcontext.cpp:61
No locals.
#12 0x03ec9e3b in XPCPerThreadData::Cleanup (this=0x8bd7c30)
    at xpcthreadcontext.cpp:385
No locals.
#13 0x03ec9f92 in ~XPCPerThreadData (this=0x8bd7c30)
    at xpcthreadcontext.cpp:394
No locals.
#14 0x03eca038 in xpc_ThreadDataDtorCB (ptr=0x8bd7c30)
    at xpcthreadcontext.cpp:429
No locals.
#15 0x05ba704b in PR_NewRWLock () from /usr/lib/libnspr4.so
No symbol table info available.
#16 0x05bbc2a3 in PR_ProcessExit () from /usr/lib/libnspr4.so
No symbol table info available.
#17 0x05bbc37e in PR_ProcessExit () from /usr/lib/libnspr4.so

Reproducible: Always

Steps to Reproduce:
What were you doing when the application crashed?
Closing Epiphany. This happens consistently.

Comment 1

[Frank Wein [:mcsmurf]]

What Gecko version is that version of Epiphany based on?

Comment 2

[Schlaegel]

I am the original reporter. I am using the current stock Fedora 6 packages.

$ rpm -qa '*epiphany*'
epiphany-2.16.3-3.fc6.i386
epiphany-debuginfo-2.16.3-3.fc6.i386
epiphany-extensions-2.16.1-4.i386
$ rpm -qa '*firefox*'
firefox-devel-1.5.0.10-5.fc6.i386
firefox-debuginfo-1.5.0.10-5.fc6.i386
firefox-1.5.0.10-5.fc6.i386
$ firefox -v
Mozilla Firefox 1.5.0.10, Copyright (c) 1998 - 2007 mozilla.org

Firefox About box:
Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.8.0.10) Gecko/20070313 Fedora/1.5.0.10-5.fc6 Firefox/1.5.0.10

Comment 3

[Bob Clary [:bc] (inactive)]

Mozilla/5.0 (X11; U; Linux i686; en; rv:1.8.0.10) Gecko/20070313 Fedora/1.5.0.10-5.fc6 Epiphany/2.16 Firefox/1.5.0.10

Comment 4

[Martijn Wargers (dead)]

I'm getting this crash sometimes in my testing with trunk builds. Talkback ID: TB32099518W
I'm using accessibility stuff when it happens, so it seems to have somehow have something to do with accessibility, I'm guessing.
Unfortunately, I'm getting this crash rarely, and in a totally unreproducable way, it seems.

Comment 5

[Schlaegel]

I still hit this bug daily, on two different systems. Neither system has any accessibility stuff running.

Since I get this bug on two different systems, I wonder why more people don't suffer from this bug and give this report some more information.

I have noticed that if run epiphany from the command line, it spews messages such as " Message: plugin_get_value 1 (1)". These messages happen while epiphany is running, not while it is closing, so they might not be related to the bug.

Comment 6

[Schlaegel]

I uninstalled both of my plugins and to see if it would make this bug go away. I had no luck. With no plugins I still regularly hit this bug.

Ahhh. Make it stop.

Comment 7

[Bill Gianopoulos [:WG9s]]

Confirming.  Also occurs under windows.

http://crash-stats.mozilla.com/report/index/6acde9b0-b7e6-4045-a841-2876d2090329

Comment 8

[Bill Gianopoulos [:WG9s]]

Oh, meant to mention my crash ins under Firefox happens at random times like clicking refresh.

Comment 9

[Bill Gianopoulos [:WG9s]]

OOPs.  I didn't realize this was such an old bug.  My issue just started recently.  I suspect it is same stack signature, different issue.

Comment 10

[Brendan Eich [:brendan]]

(In reply to comment #9)
> OOPs.  I didn't realize this was such an old bug.  My issue just started
> recently.  I suspect it is same stack signature, different issue.

New bug? Link to this one, but it seems likely to be new in cause and fix. Can you id a rough regression window from when it started happening to you?

/be

Comment 11

[Brendan Eich [:brendan]]

Oops, raced you and only saw bug 485840 too late. Sorry.

/be

Comment 12

[Luke Wagner [:luke]]

Much has changed in this area.  Can anyone still reproduce?