Re: Gfortran now compiles ROMS ocean model

[Steve Kargl <sgk at troutmask dot apl dot washington dot edu>]

On Thu, Oct 27, 2005 at 06:49:55PM -0400, Andrew Pinski wrote:
> > 
> > Hello folks
> > 
> > If this is off-topic I apologise.
> 
> No that really, this is good new.
> > 
> > I am happy to report that Gnu Fortran 95 can now be used to build the 
> > ROMS ocean model, see:
> > 
> >    http://marine.rutgers.edu/po/index.php?model=roms
> > 
> > ROMS is written in Fortran 90 and includes some module-declaration code 
> > that used to trigger ICEs in gfortran. (I considered filing a bug report 
> > but decided not to because: I despaired of reducing the problem to a 
> > simple test case; the problem resembled some existing, unresolved PRs). 
> > But it appears that the problem has now been fixed, some time between 
> > 2005-09-12 and 2005-10-15.
> > 
> > BTW is there a "Gfortran works with the following codes" page? I haven't 
> > seen one.
> 
> You can make one on the gcc wiki.
> 

The gfortran wiki, which is probably assessible from the gcc wiki,
has two links you can follow:

http://gcc.gnu.org/wiki/GFortranResults
  LAPACK test program results
  NIST test results
  Polyhedron benchmarks
  Lawrence Livermore Fortran kernel

http://gcc.gnu.org/wiki/GfortranBuild
  Details on building BLAS
  Details on building LAPACK
  Details on LAM/MPI.

There isn't a simple list of program that are known to compile
and run with gfortran.  I personally known that blas, lapack,
arpack, minpack, mpfun90, bom, nswc math lib, nist, moware, mpich2, 
Matran, and few other programs compile.  Of course, all of my
codes compile and run fine with gfortran.

If you want to add to the gfortran wiki, I'm sure no one would
object.

-- 
Steve