[Qt-interest] QtScript segfault on C++ signal (dis)connection/firing

[Kent Hansen]

You could also try with the 4.5 beta or a recent snapshot just to check
if the problem is present there too. It's hard to tell much from the
backtrace, unfortunately.

Regards,
Kent


Kent Hansen wrote:
> Hi Sean,
> Any chance you can share the application and/or reduce it to a test case
> that reproduces the problem? That would be extremely helpful. Feel free
> to reply to me in private.
>
> Regards,
> Kent
>
>
> Sean M. Pappalardo wrote:
>   
>> Hello there.
>>
>> I'm developing a fairly large QtScript file (35K at the moment) which
>> necessarily involves many C++ object signals being connected and
>> disconnected at various points depending on user action.
>>
>> The problem is the application will randomly cause a segmentation fault
>> when a signal fires (it could be any of them.) It seems to have gotten
>> worse the more signals I add to the script.
>>
>> The signals are (dis)connected via a Q_INVOKABLE C++ function that calls
>> qScriptConnect and qScriptDisconnect. This way, scripts need only one
>> line of code to connect/disconnect a signal.
>>
>> This may be related to another strange problem I found where I have to
>> connect all signals once at the beginning of the script or they won't
>> have any effect if connected for the first time later on. (I've found
>> one exception to this too.)
>>
>> (I also originally tried to do this as shown in the QtScript docs by
>> exposing the object whose signals I wanted to connect, but the script
>> slot functions weren't being called once connected.)
>>
>> If it's helpful, here's a link to a backtrace: http://pastebin.com/m5189a6f1
>>
>> I also ran valgrind to check for leaks and it reported:
>>
>> ==21380== Process terminating with default action of signal 11 (SIGSEGV)
>> ==21380==  Access not within mapped region at address 0x4
>> ==21380==    at 0x529203A: (within /usr/lib/libQtScript.so.4.4.3)
>> ==21380==    by 0x5292E1A: (within /usr/lib/libQtScript.so.4.4.3)
>> ==21380==    by 0x432DD6A: QMetaCallEvent::placeMetaCall(QObject*) (in
>> /usr/lib/libQtCore.so.4.4.3)
>> ==21380==    by 0x432F8E0: QObject::event(QEvent*) (in
>> /usr/lib/libQtCore.so.4.4.3)
>> ==21380==    by 0x454079B: QApplicationPrivate::notify_helper(QObject*,
>> QEvent*) (in /usr/lib/libQtGui.so.4.4.3)
>> ==21380==    by 0x454861D: QApplication::notify(QObject*, QEvent*) (in
>> /usr/lib/libQtGui.so.4.4.3)
>> ==21380==    by 0x43200D0: QCoreApplication::notifyInternal(QObject*,
>> QEvent*) (in /usr/lib/libQtCore.so.4.4.3)
>> ==21380==    by 0x4320D54:
>> QCoreApplicationPrivate::sendPostedEvents(QObject*, int, QThreadData*)
>> (in /usr/lib/libQtCore.so.4.4.3)
>> ==21380==    by 0x4320F4C: QCoreApplication::sendPostedEvents(QObject*,
>> int) (in /usr/lib/libQtCore.so.4.4.3)
>> ==21380==    by 0x434A98E: (within /usr/lib/libQtCore.so.4.4.3)
>> ==21380==    by 0x57001B7: g_main_context_dispatch (in
>> /usr/lib/libglib-2.0.so.0.1600.6)
>> ==21380==    by 0x5703852: (within /usr/lib/libglib-2.0.so.0.1600.6)
>>
>> Thank you very much for your time and attention. I would appreciate any
>> help or direction you can give.
>>
>> Sincerely,
>> Sean M. Pappalardo
>>
>> <<--------------------------------------------------------------------------------->>
>> This E-Mail message has been scanned for viruses
>> and cleared by >>SmartMail<< from Smarter Technology, Inc.
>> <<--------------------------------------------------------------------------------->>
>> _______________________________________________
>> Qt-interest mailing list
>> Qt-interest at trolltech.com
>> http://lists.trolltech.com/mailman/listinfo/qt-interest
>>   
>>     
>
> _______________________________________________
> Qt-interest mailing list
> Qt-interest at trolltech.com
> http://lists.trolltech.com/mailman/listinfo/qt-interest
>