[MITgcm-support] mitgcm 52k

Dmitri Leonov dleonov at ocean.washington.edu
Wed Apr 5 23:20:04 EDT 2006


Hi Michael,

What optfile are you using? the -qsuffix=f=for option is used when the 
original .F files are first preprocessed and saved as .for files which 
are then compiled:

cat write_grid.F | ../../../tools/set64bitConst.sh |  /usr/bin/cpp 
-traditional -P -xassembler-with-cpp -DWORDLENGTH=4 -DNML_TERMINATOR=" 
&" -I/sw/netcdf/include -I/sw/openmpi-1.0.1/include > write_grid.for
/sw/openmpi-1.0.1/bin/mpif77 -qfixed=132 -qsuffix=f=for 
-qsuppress=cmpmsg -O3 -qarch=g5 -qtune=g5 -c write_grid.for

I use an optfile based on one by Nicholas with xlf 8.1. Sometimes it's 
helpful to run a clean genmake2 and 'make clean' before running 'make' 
(depending on what you've changed).

Cheers,
Dmitri

Michael Beck wrote:
> I am trying to run the verification exp0 in version 52k_post on a G5 
> machine and running make I come up across the error message:
>
> $ make
> /opt/ibmcmp/xlf/8.1/bin/f77 -qfixed=132 -qsuffix=f=for -O3 -c 
> debug_call.f
> /opt/ibmcmp/xlf/8.1/bin/f77: 1501-218 file debug_call.f contains an 
> incorrect fi
> make: *** [debug_call.o] Error 1
>
> Any ideas? I checked out this version from CVS and have not yet 
> altered it
>
> michael
>
>
> ----------------------------------------------------------------
> This message was sent using IMP, the Internet Messaging Program.
>
> _______________________________________________
> MITgcm-support mailing list
> MITgcm-support at mitgcm.org
> http://mitgcm.org/mailman/listinfo/mitgcm-support




More information about the MITgcm-support mailing list