[PyKDE] Python/pystate.c:441: failed assertion `autoInterpreterState'

[Patrick Stinson]

ok, It must be another extension (probably one of mine...) then. thanks.

On 4/10/06, Phil Thompson <phil at riverbankcomputing.co.uk> wrote:
> On Tuesday 11 April 2006 1:04 am, Patrick Stinson wrote:
> > I assume this is fixed, considering the age of the following post, but
> > I got it again. What is the likelyhood that this is cause by some
> > other c++ extension than PyQt4. (latest snapshots)
> >
> > On Monday 05 July 2004 16.54, Phil Thompson wrote:
> > > On Monday 05 July 2004 3:09 pm, Fredrik Juhlin wrote:
> > > > The only outstanding issue we have right now is a segfault when exiting
> > > > the application. Running the app with a debug-compiled python reveals a
> > > > "python: Python/pystate.c:423: PyGILState_Ensure: Assertion
> > > > `autoInterpreterState' failed.". I'm not sure what could be causing it,
> > > > but I'm compiling a PyQt with debugging turned on as I speak, hoping
> > > > gdb will give me a clue (despite my feeble C/C++ skills.)
> > >
> > > This usually happens when C++ is still "running", ie. C++ virtuals and
> > > dtors are still being executed after the Python interpreter has been
> > > finalised. The sip module is supposed to guard against it, but there may
> > > be one place where it's not happening. Can you try the attached patch to
> > > see if it makes a difference?
> >
> > That fixed it! That took care of our last immediate issue.
> >
> > Again, thanks a lot!
> >
> > //Fredrik
>
> It shouldn't happen - SIP is aware of the issue and should deal with it
> accordingly. There haven't been changes in this area. But I could be wrong,
> so you need to isolate the problem further.
>
> Phil
>


--
Patrick Kidd Stinson
http://www.patrickkidd.com/
http://pkaudio.sourceforge.net/
http://pksampler.sourceforge.net/