[PyQt] I'm seeing some strange behaviour whenever I quit my PyQt5 application

[Chris Pezley]

I don't have any clue as to why the segfault could be popping up, but it 
would probably help you to generate a core dump and analyze it with gdb. 
You can do that with the following steps (note you'll need to have 
python compiled in debug):

1. Enable core dumps with "ulimit -c unlimited"

2. Get your program to crash

3. Open the core dump with gdb "gdb python core"

4. Use the python gdb extensions to analyze where the segfault is 
happening in the python code (see 
https://wiki.python.org/moin/DebuggingWithGdb)


On 09/06/2016 07:56 AM, Xavion wrote:
> Hi All,
>
> I've just finished converting my PyQt4 application to PyQt5, and I 
> keep seeing the following error message whenever I _exit_ it:
>
>     fish: “./App.pyw” terminated by signal SIGSEGV (Address boundary
>     error)
>
>
> This doesn't happen with the PyQt4 edition of the program, so I'm 
> guessing it's either due to a bug in Qt5 or PyQt5.
>
> When I run the program via GDB, I get the following output.  The 
> backtrace is available in the first attachment.
>
>     Thread 1 "python" received signal SIGSEGV, Segmentation
>     fault.0x00007ffff5cf3643 in ?? () from /usr/lib/libQt5Core.so.5
>
>
> It appeared to be related to the 'Breeze' style, so I installed qt5ct 
> <https://sourceforge.net/projects/qt5ct/> and switched to 'Fusion'.  
> When I then ran my program (again via GDB), the output and backtrace 
> were entirely different.
>
>     Thread 2 "Qt bearer threa" received signal SIGSEGV, Segmentation
>     fault.
>     [Switching to Thread 0x7fffea52b700 (LWP 22844)]
>     0x00007fffe98cbbe5 in QDBusMetaType::typeToSignature(int) () from
>     /usr/lib/libQt5DBus.so.5
>
>
> I'm on Arch Linux, by the way.  I'm using Python v3.5.2-1, Qt 
> v5.7.0-2, and PyQt v5.7-2.  All are packages installed from the 
> [extra] repository.
>
> With all of this in mind, is anyone able to pinpoint the nature of the 
> problem(s)?  If so, is it the sort of thing that'd be easy enough to 
> fix in time for the next (Py)Qt5 release?
>
>
> -- 
> Regards, Xavion.
>
>
>
> _______________________________________________
> PyQt mailing list    PyQt at riverbankcomputing.com
> https://www.riverbankcomputing.com/mailman/listinfo/pyqt

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://www.riverbankcomputing.com/pipermail/pyqt/attachments/20160906/ba086355/attachment.html>