Joseph Winston wrote:

> I'm guessing that you are seeing the problem because of a combination
> of g++, mt_allocator, threads, and dlclose (http://                           
>                                                     
> bugs.donarmstrong.com/cgi-bin/bugreport.cgi?bug=293466)  Try setting
> the environment variable GLIBCXX_FORCE_NEW.  For example,

>From the Thread you posted:

 "We believe that the bug you reported is fixed in the latest version of
  gcc-4.0, which is due to be installed in the Debian FTP archive"

I have used both 4.0.4 and 4.1.2 to compile osgPython in addition to the 
specified environment variable, still same problem. 
Do you mean that I should rebuild OSG with the new GLIBCXX_FORCE_NEW flag 
also?

>From my debugging sessions it would seem that the problem is with the 
function-creation in Boost.
When loadWrapper is called Boost will call make_function and then later on 
Python itself croaks when calling Tuple_Pack.

The last portion of gdb tells me this:
  #0  0x080961af in PyTuple_Pack (n=1) at ../Objects/tupleobject.c:149

The line in question does this (Python 2.4.4 sources):
  Py_INCREF(o);

What could possibly make Python's Py_INCREF macro croak like that?
What versions of Python have you managed to make osgPython work with?

/Peter
_______________________________________________
osg-users mailing list
[email protected]
http://openscenegraph.net/mailman/listinfo/osg-users
http://www.openscenegraph.org/

Reply via email to