[MITgcm-support] mpi problem

[Kung H.S.]

Hi all,

I have a problem in using mpi. When I tried to use mpi to run /tutorial_deep_convection, there's a problem in reading the initial data. 
In STDERR.0000, there's an error: "S/R INI_THETA: theta=0 identically. If this is intentional you will need to edit ini_theta.F to avoid this safety check".
However, when I output the data directly from ini_theta.F, theta is not equal to zero. And there's no problem when I ran this experiment with serial. 
Is this related to the mpi compiler I used or something in the configuration ?  Thanks a lot.

Here's my option file for the mpi:
---------------
#!/bin/bash

FC='mpif90'
CC='mpicc'
LINK='mpif90'
INCLUDES='-I/local/software/mpich/include/'
CPP='/usr/bin/cpp -P -tradition'
MPI='true'
DEFINES='-DWORDLENGTH=4'
------------------

Cheers,
Hiusuet


_________________________________________________________________
利用 Windows Live™ Photos 簡易上載照片, 與親友分享歡樂時刻
http://profile.live.com/connect
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mitgcm.org/pipermail/mitgcm-support/attachments/20090610/470b8342/attachment.htm>