I would like to develop a G4 application for pion production
at low energy. The scope is essentially to see pions produced from a proton beam
of few GeV impinging light and heavy materials.
At the time being I am still in a learning process, namely I took the
G4 test number 30 to try to understand how the G4 hadronic models work
(Geant4 5.0 version).
My problem is that once I try to use the Bertini model I have suddenly
a Segmentation fault from the application which I don't have if
I use the LEparametrisation model, or the CHIPS or the Kinetic model.

I am calling Bertini in the physics list with:

else if( gen_name == "bertini" )
  {
    G4CascadeInterface* hkm=new G4CascadeInterface();

    sg = new Sh01SecondaryGenerator(hkm, mat);
    theProcess->SetSecondaryGenerator(sg);

    man->AddDiscreteProcess(theProcess);
  }

An there is the only place where the application distinguish between
the different models.

My geometry and beam configuration is a proton beam of 1.8 GeV on lead target 2
mm thick.
First I thought the error I have could come from a wrong compilation of the G4
libraries. I am using the gcc 3.2 on a linux machine but I get the same error
running on w2000 with CYGWIN and gcc 3.2.
Then I am wondering if my Physic list needs something else.