[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: NetCDF for platforms with 8-byte floats and doubles

>To: address@hidden
>From: Matt Huddleston <address@hidden>
>Subject: Re: 20021111: NetCDF for platforms with 8-byte floats and doubles
>Organization: Met Office, United Kingdom
>Keywords: 

Hi Matt,

> I picked up your response to a problem with NetCDF on a J90 - about F90
> and  platforms where the real and double precision variables are the
> same length.  I'm trying hard to install netcdf-3.5.1-beta5 on a Cray
> T3e but am having the same troubles.  Has any solution been found yet ?
> 
> I've tried removing the FourByte_ / EightByte definitions from the
> netcdf_overloads.f90 file but that didn't work.  Do you have any other
> suggestions to try ?

Here's what Steve Emmerson suggests:

  The solution is to configure the file "f90/netcdf_overloads.f90" so
  that: 1) it doesn't use any primitive types that aren't supported; and
  2) the signatures of all the routines that the file defines are unique.

  The problem is that there's no standard macro capability in Fortran-90.
  This means that the appropriate configuration of the file must be
  determined by the configure script and the configuring process must be
  implemented using UNIX utilities (e.g. cpp, sed). I haven't had time to
  do either.

  He should have been able to manually modify the file in order to
  successfully compile it.

So, sorry, we haven't gotten to this yet.  (We currently have only 1/4
person available for netCDF support and none for netCDF development,
but we're hoping for more resources soon ...)

--Russ

_____________________________________________________________________

Russ Rew                                         UCAR Unidata Program
address@hidden                     http://www.unidata.ucar.edu