configuring on AIX

[Robert Latham]

On Mon, Sep 04, 2006 at 12:51:04PM +0200, Joachim Worringen wrote:
> 
> I have a problem to configure pnetcdf 1.0.1 on AIX. configure fails with:
> ./configure
> [snip]
> checking for Fortran "byte"... yes
> checking for Fortran "integer*2"... yes
> checking if Fortran "byte" is C "signed char"... yes
> checking if Fortran "byte" is C "short"... no
> checking if Fortran "byte" is C "int"... no
> checking if Fortran "byte" is C "long"... no
> checking if Fortran "integer*2" is C "short"... yes
> checking if Fortran "integer*2" is C "int"... no
> checking if Fortran "integer*2" is C "long"... no
> checking if Fortran "integer" is C "int"... yes
> checking if Fortran "real" is C "float"... yes
> checking if Fortran "doubleprecision" is C "double"... yes
> checking for Fortran-equivalent to netCDF "byte"... byte
> checking for Fortran-equivalent to netCDF "short"... integer*2
> checking for Fortran "integer*"... no
> configure: error: F77 does not support "integer*"
> 
> The machine is:
> j36:parallel-netcdf-1.0.1 [17] uname -a
> AIX j36 3 5 00223C3B4C00
> 
> Compiler settings:
> MPICC=mpcc
> MPIF77=mpxlf
> CXX=xlC
> CC=xlc
> FC=xlf
> F77=xlf
> 
> Is this a known issue?

Hi Joachim

I believe this is a known issue and also something we have now fixed
in CVS: Christopher Subich <csubich at math.uwaterloo.ca> contributed a
compile-only test for the size of MPI_Offset which helps us a lot in
cross-compile environments.

Looks like we are due for a new pnetcdf release sometime soon...  In
the meantime, you can do 'export ac_cv_sizeof_MPI_Offset=8', and
autoconf will bypass all checks in favor of the vaule suppled in
$ac_cv_sizeof_MPI_Offset

==rob

-- 
Rob Latham
Mathematics and Computer Science Division    A215 0178 EA2D B059 8CDF
Argonne National Lab, IL USA                 B29D F333 664A 4280 315B