I have a 31 year old FORTRAN code which I want to compile on a Linux system. Specifically, the system is a CentOS.

I made an initial attempt to compile and it does compile and even runs. However, I suspect there may be some issues because of which the code doesn’t run as I was expecting. The code needs an input based on the CPU/compiler. These are the options that need to be given in a file using the icpu variable:

c-----------------------------------------------------------------------
c    S E L E C T   A P P R O P R I A T E   C O M P I L E R / C P U
c-----------------------------------------------------------------------
c    icpu =  0...SVS Fortran (680x0/80x86)
c         =  1...Lahey Fortran/Microsoft Fortran (8088/80x86)
c         =  2...VAX/VMS
c         =  3...SUN Fortran (68020/SPARC)
c         =  4...Cray/Unicos
c         =  5...Silicon Graphics Iris
c-----------------------------------------------------------------------
      icpu=1

The code is of fixed-format and it’s source files have .FOR extension.

I made an initial attempt at compiling for icpu=3 by using the --std='legacy' option with the gfortran compiler. It complained that it cannot find the included files. So, I changed the include directives from include 'file_name. to include 'FILE_NAME' and it worked. With this done, it did compile but now I am stuck at a location where the program needs to read a file as part of it’s operation. I am not sure and it may or may not be related to the way I have compiled it.

Can someone recommend anything else I can troubleshoot ? Are my compiler options ok?