[MITgcm-support] compiler flags to make output human readable

[Gus Correa]

It depends on your compiler, whatever your "ftn" stands for.
The compiler man page ('man ftn' ?) may give a hint
on how you trace back the routine that caused the error.
The syntax varies a bit.
For instance,
Gnu gfortran has -fbacktrace,
Intel ifort has -traceback,
PGI pgfortran has -traceback.

My two cents,
Gus Correa

On 08/05/2014 07:10 AM, Jonny Williams wrote:
> Hi there
>
> With regard to some emails on another thread ("Forcing files in data.dic
> (biogeochemistry)") I am trying to get more output on a failure where I
> am trying to restart a model with the OBCS, EXF, DIC, GCHEM, and
> PTRACERS packages turned on.
>
> Currently I am getting output of the form...
>
> /forrtl: severe (64): input conversion error, unit -5, file Internal
> Formatted Read/
> /Image              PC                Routine            Line
>   Source /
> /mitgcmuv           0000000000BAB3FE  Unknown               Unknown
>   Unknown/
> /mitgcmuv           0000000000BA973A  Unknown               Unknown
>   Unknown/
> /mitgcmuv           000000000051D78C  Unknown               Unknown
>   Unknown/
> /mitgcmuv           00000000005CE598  Unknown               Unknown
>   Unknown/
> /mitgcmuv           00000000007410B0  Unknown               Unknown
>   Unknown/
> /mitgcmuv           000000000070FE8E  Unknown               Unknown
>   Unknown/
> /mitgcmuv           0000000000735AD2  Unknown               Unknown
>   Unknown/
> /mitgcmuv           000000000075BCEF  Unknown               Unknown
>   Unknown/
> /mitgcmuv           000000000075BF19  Unknown               Unknown
>   Unknown/
> /mitgcmuv           00000000006A2B14  Unknown               Unknown
>   Unknown/
> /mitgcmuv           0000000000400F26  Unknown               Unknown
>   Unknown/
> /mitgcmuv           0000000000C5DBA4  Unknown               Unknown
>   Unknown/
> /mitgcmuv           0000000000400D8D  Unknown               Unknown
>   Unknown/
> /
> /
> /... /which clearly isn't of much use. I am trying to implement
> something like the FORTRAN "-g" flag to get human readable output. My
> build script (compilation script) is below and I have highlighted where
> I have added the "-g" options, which do not give me any more output at
> the moment.
>
> Does anyone have any ideas about how to get more useful output?
>
> Many thanks!
>
>
> /#!/bin/bash/
> /
> /
> /FC=ftn/
> /F90C=ftn/
> /CC=cc/
> /CPP='cpp -P -traditional'/
> /
> /
> /MPI='true'/
> /DEFINES='-DALLOW_USE_MPI -DALWAYS_USE_MPI -DWORDLENGTH=4'/
> /
> /
> /FFLAGS=' -g -m64 -fPIC -W0 -WB -convert big_endian -assume byterecl'/
> /FOPTIM=' -g -O3 -align -ip -mp1 -prec-div'/
> /
> /
> /INCLUDES='-I/usr/local/include -I${MPICH_DIR}/include
> -I${NETCDF_DIR}/include'/
> /
> /
> /LIBS=' '/
>
>
> --
> Dr Jonny Williams
> School of Geographical Sciences
> University of Bristol
> University Road
> BS8 1SS
>
> +44 (0)117 3318352
> jonny.williams at bristol.ac.uk <mailto:jonny.williams at bristol.ac.uk>
> bit.ly/jonnywilliams <http://bit.ly/jonnywilliams>
>
>
> _______________________________________________
> MITgcm-support mailing list
> MITgcm-support at mitgcm.org
> http://mitgcm.org/mailman/listinfo/mitgcm-support
>