double precision g77



I have a program that has been working fine, written in fortran 77 and
compiled using g77 under fedora core 3. I needed to increase the precision
so changed all REAL declarations to DOUBLE PRECISION and changed FUNCTION
to DOUBLE PRECISION FUNCTION. Now my variables are returning NAN values!
Any ideas why this may be?

Many thanks
Rich
.



Relevant Pages

  • Re: double precision g77
    ... compiled using g77 under fedora core 3. ... so changed all REAL declarations to DOUBLE PRECISION and changed FUNCTION ...
    (comp.lang.fortran)
  • Re: Numerical accuracy of C++ and Fortran programs on 32 bit machines
    ... the "binary representation" is simply IEEE fp and has precisely ... to do w/ Fortran in general or g77 in particular. ... > precision and just have garbage in the rest of the field. ...
    (comp.programming)
  • Re: G77 double precision
    ... > I have a fortran code which is written entirely in double precision ... > g77 handling double precision real numbers. ... > run on a pentium-IV (with cygwin libraries installed). ... The architecture options for g77 are documented under 'info gcc' It's ...
    (comp.lang.fortran)
  • Re: Interesting problem porting a g77 code to gfortran
    ... different values in the two programmes. ... so that the g77 version got the double ... precision value while the gfortran programme interpreted them as single ...
    (comp.lang.fortran)
  • Re: Interesting problem porting a g77 code to gfortran
    ... so that the g77 version got the double ... precision value while the gfortran programme interpreted them as single ... Programs were put in separate files to prevent g95 and gfortran from ...
    (comp.lang.fortran)