[Wien] zfft3d.f and c3fft.f

[Peter Blaha]

> I am compiling the Wien2k for parallel use with Nag f95. There are a lot of
> error countered. I had manged to solve most of these by changing some
> statements in each subroutine. But, what I found in zff3d.f and c3fft.f
> subroutines is different,where I have to change the ch, Ifac, cc,c1 from real
> to complex. So , do such change affects the physical results.

If you are very good in FORTRAN, than you can change the code. But the
internal "memory management" of these FFT-routines is quite trickey and
it will not be so simple!

PS: I would expect much better performance witht the free Intel ifc instead
of Nag f95

                                      P.Blaha
--------------------------------------------------------------------------
Peter BLAHA, Inst.f. Materials Chemistry, TU Vienna, A-1060 Vienna
Phone: +43-1-58801-15671             FAX: +43-1-58801-15698
Email: blaha at theochem.tuwien.ac.at    WWW: http://info.tuwien.ac.at/theochem/
--------------------------------------------------------------------------