We use -qrealsize=8 with -L /usr/local/lib32/r4i4 -lnetcdf
Then the call "istat = nf_get_vara_double(ncid,.." reads
8-byte reals.
--Ben
-----------------------------------------------------------------------
Ben Foster High Altitude Observatory (HAO)
foster@xxxxxxxxx phone: 303-497-1595 fax: 303-497-1589
Nat. Center for Atmos. Res. P.O. Box 3000 Boulder CO 80307 USA
-----------------------------------------------------------------------
>
>Sir,
>I recently installed the netcdf version 3.5 and I used
>it with fortran 77 interface.I actually run a atm.
>model on a IBM AIX machine(fortran source) with a xlf
>compiler option -qqutodbl=dbl4(used for promoting real
>objects to double). As the fortran source in turn
>calls some of the netcdf routines I have been
>experiencing problems for use of option -qautodbl as
>the netcdf package is of single precision I suppose.So
>I kindly request you to please tell me where could I
>find a double precision version of netcdf package or
>else what option should I use to compile netcdf for
>use in double precion version with xlc and xlf as
>compilers.
>
>
>________________________________________________________________________
>Looking for a job? Visit Yahoo! India Careers
> Visit http://in.careers.yahoo.com
