[MITgcm-support] no mnc output files with -enable=mnc
Martin Losch
mlosch at awi-bremerhaven.de
Wed Jul 5 02:28:53 EDT 2006
Hi Jill,
this time I read your email starting at the end to make sure I don't
miss it (o:
In order to get mnc-output you need to do 3 things:
1. compile with mnc, you have done that (instead of the -enable=mnc
flag you could put a line "mnc" into ../code/packages.conf)
2. enable the package at runtime, that is in your data.pkg you should
have a line " useMNC=.TRUE.,"
3. create a data.mnc file (as for every other package there needs to
be data.pkgname for mnc too), where you can set mnc-specific runtime
flags.
I assume that you haven't done 2. and 3. For an example with mnc that
is similar to verification/exp2 have a look at verification/
global_ocean.90x40x15 (which incidentally is a bit more realistic
than exp2 with seasonal forcing and a nonlinear equation of state).
Martin
On Jul 5, 2006, at 8:13 AM, jschwarz at awi-bremerhaven.de wrote:
> Hullo,
>
> running genmake with -enable=mnc, i don't seem to get any mnc
> output files.. as many specific details as i thought useful are
> pasted below (what i ran, where i've checked..).
> Grateful for any ideas!
>
> Also, is there somewhere you can/should set $ROOTDIR ?
> jill.
>
>
> tasman:/usr/MITgcm/verification/exp2/build # ../../../tools/
> genmake2 -mods=../code -of ../../../tools/build_options/
> linux_ia32_gfortran_jill -enable=mnc
>
> ---------------->works fine and tells me (abridged):
> === Checking system libraries ===
> Do we have the system() command using gfortran... yes
> Do we have the fdate() command using gfortran... yes
> Do we have the etime() command using gfortran... yes
> Can we call simple C routines (here, "cloc()") using gfortran...
> yes
> Can we unlimit the stack size using gfortran... yes
> Can we register a signal handler using gfortran... yes
> Can we use stat() through C calls... yes
> Can we create NetCDF-enabled binaries... yes
> ...
> packages are: cd_code debug generic_advdiff mdsio mnc
> mom_common mom_fluxform mom_vecinv monitor rw
> applying package dependency rules
> packages are: cd_code debug generic_advdiff mdsio mnc
> mom_common mom_fluxform mom_vecinv monitor rw
>
> tasman:/usr/MITgcm/verification/exp2/build # make depend
> tasman:/usr/MITgcm/verification/exp2/build # make CLEAN
> ---------------------> work fine.
>
> tasman:/usr/MITgcm/verification/exp2/build # make
> --------------------> reports a few warnings but no errors, and
> ends on:
>
> make: warning: Clock skew detected. Your build may be incomplete.
>
> tasman:/usr/MITgcm/verification/exp2/build # ./mitgcmuv > output.txt
> -------------> works fine, provides much detail in output.txt and
> completes on:
> STOP NORMAL END
>
> In the build directory, the mnc* files are .h, .o, .f, .F;
> specifically:
> # ls mnc*
> mnc_common.h mnc_cw_init.o mnc_dim.F
> mnc_init.F
> mnc_create_dir.c mnc_cw_model_attr.F mnc_dim.f
> mnc_init.f
> mnc_create_dir.o mnc_cw_model_attr.f mnc_dim.o
> mnc_init.o
> mnc_cw_citer.F mnc_cw_model_attr.o mnc_dump.F
> mnc_readparms.F
> mnc_cw_citer.f mnc_cw_udim.F mnc_dump.f
> mnc_readparms.f
> mnc_cw_citer.o mnc_cw_udim.f mnc_dump.o
> mnc_readparms.o
> mnc_cw_cvars.F mnc_cw_udim.o mnc_file.F
> mnc_utils.F
> mnc_cw_cvars.f mnc_cw_write_grid_info.F mnc_file.f
> mnc_utils.f
> mnc_cw_cvars.o mnc_cw_write_grid_info.f mnc_file.o
> mnc_utils.o
> mnc_cw_fsize.c mnc_cw_write_grid_info.o mnc_grid.F mnc_var.F
> mnc_cw_fsize.o mnc_cwrapper.F mnc_grid.f mnc_var.f
> mnc_cw_init.F mnc_cwrapper.f mnc_grid.o mnc_var.o
> mnc_cw_init.f mnc_cwrapper.o mnc_id_header.h
>
> i've checked all the directories und /exp2 for mnc* and MNC* output
> files, or an mnc_test_$DATE_$SEQ- format directory name. Do i need
> to set another flag somewhere to force creation of netcdf output?
>
>
> _______________________________________________
> MITgcm-support mailing list
> MITgcm-support at mitgcm.org
> http://mitgcm.org/mailman/listinfo/mitgcm-support
More information about the MITgcm-support
mailing list