Re: hdgeant crash

[Beni Zihlmann <zihlmann@jlab.org>]

I forgot one more important fact.
right at the beginning of the subroutine  hitForwardDC(...) the function
getlayer_() returns zero as well as the function getpackage_()
which causes the chamber number to be -10. needless to say that this should
never happen.

cheers,
Beni

> Hi All,
> I finally managed to get the hdgeant running in debug mode again. 
> Unfortunately only on
> my local machine not on the ifarms.
> I run hdgeant with pythia generated events as input. and I have 
> NOSECONDARIES 0
> because I want to have showers in the BCAL.
> the code crashes unpredictably at an arbitrary event but always with 
> the same error message:
>         400  events simulated
> **** GTCKOV: unable to reflect at NTMULT =    29347 step No.       37 
> photon abandoned!
> **** GTCKOV: unable to reflect at NTMULT =     8492 step No.       50 
> photon abandoned!
> hdgeant_d: bintree.c:33: getTwig: Assertion `node->mark >= 0' failed.
>
> Program received signal SIGABRT, Aborted.
> 0x00b5b402 in __kernel_vsyscall ()
> (gdb)
> attached is the output from the debugger while I go through the 
> different levels.
> the code crashes in hitFDC.c at line 544 when calling getTwig() 
> because as it turns out
> the "chamber" number is determined to be -10! (see attached file)
> I have not looked into detail why this can happen but at this point I 
> through
> the ball out in the field maybe Simon or Richard will see immediately 
> where the
> root of the problem is.
>
> cheers,
> Beni
>