segfault with gfortran > 4.0.3

[John Pye]

Hi Jerry,

Thanks for the message. The problem is that I'm not a fortran 
programmer: I do C and C++ and Python. I only use fortran to compile 
libraries that I need. So here I am reporting this bug, but I'm not 
really in a position to isolate it.

Arne Drud may be of some assistance. Perhaps if he has a chance he'll be 
able to give you a reduced test case.

Arne mentioned to me that this file is basically a big if..else..else, 
and that some versions of some other compilers have given stack 
overflows or something like that. Perhaps there's something relating to 
stack or memory size (or recursion depth etc) that's been changed in 
recent versions of gfortran?

Thanks again for taking the time to look at this,

Cheers
JP

Jerry DeLisle wrote:
> John Pye wrote:
>> Hi again,
>
> The ICE is at line 3313 of conmsg.f90.
>
> John, to be really helpful, we need a reduced test case.  Most of us 
> just don't have time right now to go through over 3000 lines of code 
> to isolate this.
>
> If you could come up with a 10 or less lines, that may make this 
> workable.
>
> Regards,
>
> Jerry
>

-- 
John Pye
Department of Mechanical and Manufacturing Engineering
University of New South Wales, Sydney, Australia
http://pye.dyndns.org/