Compilation of FEAP on a 64 bit machine

[salil19]

Hi, I am using FEAP ver83. I am able to compile it and successfully run an example problem (Ex8, Iplast_2d, finite deformation problem, file attached) on my 32 bit machine (Ubuntu 12.04). When I try to run the same problem on my 64 bit machine (OpenSuse 12.2) I get an error. The last two lines of the output file are:
Computing solution at time  1.0000E-01: Total proportional load  1.0000E-01
*ERROR* Residual norm is NaN or Inf

The corresponding lines when run on a 32 bit machine are:
 Computing solution at time  1.0000E-01: Total proportional load  1.0000E-01
   Residual norm =     1.1423660E+03    1.0000000E+00      t=     0.07     0.01
In this case the code continues and terminates with no error.
 
I have made the necessary changes to the makefile.in before compiling FEAP on the 64 bit machine
and I could successfully run some other example problems (Ex1 (patch test)and Ex 7 (contact problem)) on my 64 bit machine.

Could you please help me fix the problem.

Thanks.
Salil

[FEAP_Admin]

can you look at include/comblk.h?
what size are the arrays?  they should be 1024.  if not, change to 1024, remove your archive and all *.o files and completely rebuild.

[salil19]

Thanks a lot. The array sizes were 1. I changed them to 1024 and now I am able to run the example problem.

Salil