!!****h* ariane/mod_precision !! NAME !! mod_precision (mod_precision.f90 - Fortran90 module) !! !! USAGE !! Include 'USE mod_precision' in the header of your Fortran 90 source !! code. !! !! FUNCTION !! Specify Integer and Real precision. You could change all the variable !! of all the code from here ans only here. !! !! AUTHOR !! * Origin : Nicolas Grima (April-May 2005) !! !! CREATION DATE !! * April-May 2005 !! !! HISTORY !! Date (dd/mm/yyyy/) - Modification(s) !! !! RESULT !! !! !! EXAMPLES !! !! !! NOTES !! ROBODoc header style. !! !! TODO !! !! !! PORTABILITY !! Machine-OS - Fortran90/95 compiler !! * i686-pc-linux-gnu - ifort !! !! SEE ALSO !! !! !! USES !! * No another USE !! !! USED BY !! * All the routines of the Ariane application !! !! SOURCE !!=========================================================================MODULE mod_precision 38 !-------------! ! Declaration ! !-------------! IMPLICIT NONE INTEGER(kind=4), PARAMETER :: & ishort=4 , & ! Here we decided that single precision is 4 ilong =8 ! and double precision is 8 ! ! (IEEE norm, except some cray machines) INTEGER(kind=ishort), PARAMETER :: & iprec=ishort ! We decided that single precsion for ! ! Integer variables is right. !- We use here the F90 intinsinc function KIND to determine the real precision -! INTEGER(kind=ishort), PARAMETER :: & rshort=KIND(1.0) , & ! Real single precision rlong =KIND(1.0d0) ! Real double precision !- By default compilation will be made in double-precision for real variables. -! INTEGER(kind=ishort), PARAMETER :: & rprec=rlong ! If needed, you can change rlong by rshort (not tested). END MODULE mod_precision !!***