how I make to get model RSM?

Hi,
I try to install T248L28 GSM but get segmentation fault when running mtn.sh (mtn.x). I specified inst gsm_latest hi during installation. Please advise. Thanks.

Sai Ming

Hi,
I have just downloaded and run "gsm_latest lo" but it fails at chgr.x. I was running it with the test data files, hence nothing was modified.
Sai Ming

Hi,
I would like to try nesting of a 15km RSM within a 60km RSM. How could this be done?
Sai Ming

I need to compile the programs in libs/etc so that I can read and manage data output.
After calling 'make clean' and 'configure-libs' in my libs directory, I run 'make'. My fortran compiler is ifort (i do not have pgf), so I have set 'MACHINE=intel' in my configure-libs file. A lot of objects are compiled before I get to the following error:

ifort -r8 -O3 -convert big_endian -c grmean.f
ifort -r8 -O3 -convert big_endian -o grmean.x grmean.o /home/dmansbac/usr/local/bin/libs/lib/w3lib.a /home/dmansbac/usr/local/bin/libs/lib/libncaru.a
grmean.o: In function `MAIN__':
grmean.f:(.text+0x6f5): undefined reference to `numchar_'
grmean.f:(.text+0x113d): undefined reference to `numchar_'
make[2]: * [grmean.x] Error 1
make[2]: Leaving directory `/home/dmansbac/usr/local/bin/libs/etc/util_intel'
make[1]: * [all] Error 8
make[1]: Leaving directory `/home/dmansbac/usr/local/bin/libs/etc'
make: *** [libs] Error 8


I have not yet changed any compiler options, but I am not really sure of what type of error this is and what the best way to proceed is.

After making the libs I tried the composit script from the CEC utilities. I made a simple, three line listfile with three different yyyymmdd dates, and set the listfile, outfile, and outdir terms in the composit file.
I run ./composit but it ends badly.
The last few lines are:

IFTS,IFTE= -999 -999
status= 25 45924452 33204 1 58200 500
0 12125028 1215543176 1125103465 1194425581 4096
0 0 0 0 0 0
0 0 0 0 0 0
0 0 0 0 0 0
0 0 0 0 0 0
0 0 0 0 0 0
0 0 0 0 0 0
0 0 0 0 0 0
0 0 0 0 0 0
0 0 0 0 0 0
0 0 0 0 0 0
0 0 0 0 0 0
0 0 0 0 0 0
0 0 0 0 0 0
0 0 0 0 0 0
0 0 0 0
file size= 12125028
INCREASE MDATA to 32200
abort: Fortran Abort Called

---

No output files were written.

I don't know what I'm missing here.

Hello,

We have been struggling to get G-RSM to like a certain GCM data as input (the model works well with NCEP data well). A former collague with one of the people in our work group suggested to try the model version that was being used during 07/27/2007 -- as he seemed to have gotten G-RSM to like the GCM data for that specific version.

How would I find out versions and able to get specific version of the model? Of course our error suggested strongly that our GCM data being used now somehow was incompatible how latest version G-RSM read input data.

Thank you in advance

I'm trying to install RSM on Red Hat Ent ver 4
with pgf90 -V
pgf90 7.2-3 64-bit target on x86-64 Linux -tp p7-64

./inst rsm_latest hi

After I get libs built, the RSM make fails with:

pgf90 -r8 -O2 -Msave -Mrecursive -Mdalign -DLINUX -byteswapio -c tersub.f
if [ linux = mac ] ; then \
ranlib -s /mnt/disk2/users/xtuser/nmod/libs/lib/modelib.a ; \
fi
pgf90 -r8 -O2 -Msave -Mrecursive -Mdalign -DLINUX -byteswapio -o mtn.x equallat.o gausslat.o legendre.o makemt.o makeoa.o mtn.o nntprt.o sphert.o tersub.o /mnt/disk2/users/xtuser/nmod/libs/lib/modelib.a
/opt/pgi/linux86-64/7.2-3/lib/libpgf90.a(initpar.o)(.text+0x2): In function `hpf_myprocnum':
: relocation truncated to fit: R_X86_64_PC32 hpf_lcpu
/opt/pgi/linux86-64/7.2-3/lib/libpgf90.a(initpar.o)(.text+0x12): In function `hpf_ncpus':
: relocation truncated to fit: R_X86_64_PC32 hpf_tcpus
/opt/pgi/linux86-64/7.2-3/lib/libpgf90.a(initpar.o)(.text+0x22): In function `hpf_getioproc':
: relocation truncated to fit: R_X86_64_PC32 hpf_ioproc
/opt/pgi/linux86-64/7.2-3/lib/libpgf90.a(initpar.o)(.text+0x32): In function `hpf_is_ioproc':
: relocation truncated to fit: R_X86_64_PC32 hpf_ioproc
/opt/pgi/linux86-64/7.2-3/lib/libpgf90.a(initpar.o)(.text+0x38): In function `hpf_is_ioproc':
: relocation truncated to fit: R_X86_64_PC32 hpf_lcpu
/opt/pgi/linux86-64/7.2-3/lib/libpgf90.a(initpar.o)(.text+0x5f): In function `hpf_abort':
: relocation truncated to fit: R_X86_64_PC32 hpf_lcpu
/opt/pgi/linux86-64/7.2-3/lib/libpgf90.a(initpar.o)(.text+0xeb): In function `hpf_abortp':
: relocation truncated to fit: R_X86_64_PC32 hpf_lcpu
/opt/pgi/linux86-64/7.2-3/lib/libpgf90.a(initpar.o)(.text+0x127): In function `hpf_initarg':
: relocation truncated to fit: R_X86_64_PC32 .bss
/opt/pgi/linux86-64/7.2-3/lib/libpgf90.a(initpar.o)(.text+0x151): In function `hpf_initarg':
: relocation truncated to fit: R_X86_64_PC32 .bss
/opt/pgi/linux86-64/7.2-3/lib/libpgf90.a(initpar.o)(.text+0x17b): In function `hpf_initarg':
: relocation truncated to fit: R_X86_64_PC32 .bss
/opt/pgi/linux86-64/7.2-3/lib/libpgf90.a(initpar.o)(.text+0x18b): In function `hpf_initarg':
: additional relocation overflows omitted from the output
make[2]: * [mtn.x] Error 2
make[2]: Leaving directory `/mnt/disk2/users/xtuser/nmod/rsm/src/mtn'
make[1]: * [rsm] Error 8
make[1]: Leaving directory `/mnt/disk2/users/xtuser/nmod/rsm/src'
make: *** [all] Error 8

Is this a compiler version error? Should I check with PGI?
Thanks,
Merrill

Hello,

I am trying to install the latest RSM with high topo on NCAR (Bluefire). (By the way, Bluefire is not recognized. I have to manually enter "ncar" and choose "ibmspbv"). The installation fails when building libraries:

---

if [ ibmspbv = mac ] ; then zcat sigft4890030900.asc.Z >/blhome/juliam/RSM/libs/con/sigft4890030900.asc || exit 8 ; else uncompress -c sigft4890030900.asc.Z >/blhome/juliam/RSM/libs/con/sigft4890030900.asc || exit 8 ; fi
Target "rcon" is up to date.
cd con ; make romcon || exit 8
make: Cannot find a rule to create target hcs_ofes-grid.nc from dependencies.
Stop.
make: The error code from the last command is 8.

---

I checked the Makefile under /con. The file "hcs_ofes-grid.nc.Z" that it is looking for doesn't get downloaded. I suspect it has something to do with ROMS. Can I deactivate this part and what is a clean way to do it?

Thanks,

Julia

When I execute ./rsm.in under the directory of expscr , I got the following error information:
+ RUNNAME=r_000
+ USEERA=0
+ ENDHOUR=48
+ DELTAT_REG=@DELTAT_REG_WINTER@
+ NESTING_HOUR=6
+ SWHR_REG=1
+ LWHR_REG=1
+ INTSFCX=24
+ INCHOUR=6
+ INTPOST=3
+ INTPGB=3
+ INTRESTART=6
+ FCST_RESTART=yes
+ DELETE_SIGGLOB=yes
+ '[' 6 -lt 24 ']'
+ echo 'INCHOUR < INTSFCX'
INCHOUR < INTSFCX
+ echo 'INCHOUR set to INTSFCX to make sure sfc is called.'
INCHOUR set to INTSFCX to make sure sfc is called.
+ EXECDIR=@PWD@/r_000
+ SST_ANL_T0=
+ SST_ANL=
+ SNO_ANL_T0=
+ SNO_ANL=
+ ICE_ANL_T0=
+ ICE_ANL=
+ FCST_NPES=@NPES@
+ GLOBAL_SIG_INITIAL=@CON_DIR@/sigft0090030900.asc
+ GLOBAL_SFC_INITIAL=@CON_DIR@/sfcanl90030900.asc
+ INITIAL_SFC_FILE_TYPE=osu1
+ BASE_SIG='@CON_DIR@/sigft${FH}90030900.asc'
+ YYYYS=1990
+ MMS=03
+ DDS=09
+ HHS=00
+ CPROJ=0.0
+ CTRUTH=0.0
+ CORIENT=0.0
+ CDELX=0.0
+ CDELY=0.0
+ CCENLAT=0.0
+ CCENLON=0.0
+ CLFTGRD=0.0
+ CBTMGRD=0.0
+ . @RUNSCR_DIR@/envset
./rsm.in: line 86: @RUNSCR_DIR@/envset: no such file or directory

What should I do to deal with this problem? Thanks.
By the way, is "setenv path .:$path" the right way to "add ‘.’ To your PATH in .cshrc file" ?
Thanks.

Hi all !!

Currently I am trying to port Global Spectral Model. I had followed the steps specified at
http://ecpc.ucsd.edu/projects/G-RSM/docs/INT/index.html
Changes specific to my cluster are as follows:
1> configure-libs
MACHINE=intel
MARCH=mpi

2> configure-model
LIBS_DIR=$HOME/GSM/libs
MODEL=gsm
MODEL_DEFINE=$HOME/GSM/gsm/def/gsm12628
GTOPO30=no
MARCH=mpi
NPES=16
NCPUS=1
INSTITUTION=ecpc

Other configuration compilation steps are same. I was able to run gsm with T62L28 successfully.
Now I am trying to run with higher resolution like T126L28. When I try to run gsm12628 I am getting error as
ERROR.out => "$HOME/GSM/libs/etc/scnvrt: Execution failed"
hit eof while reading sigma file
abort:

======================================================================

Contents of scnvrt.data
&NAMCNV
GSM0RSM='gsm',
SFCFTYP='osu1',
INFMT='asc',OFMT='bin',
JCAP=126,IDIM=384,JDIM=190,KDIM=28,
NEWYR=-1,NEWMO=-1,NEWDY=-1,NEWHR=-1,
FHNEW=-1.,
&END

======================================================================

Contents of scnvrt.out

modified infmt=asc
ofmt=bin
sigma file
fhour,idate= 0.000000000000000E+000 0 3 9
90
number of tracers,clouds= 1.00000000000000 0.000000000000000E+000
second header rec written
write k= 1 array= 449.880000000000
write k= 2 array= 0.179570000000000
write k= 3 array= -0.415878000000000
write k= 4 array= -0.327469000000000
write k= 5 array= 5.410310000000000E-002
write k= 6 array= 2.294540000000000E-002
write k= 7 array= -1.180630000000000E-002
write k= 8 array= 2.168130000000000E-002
write k= 9 array= 1.846800000000000E-007
write k= 10 array= 4.526570000000000E-008
write k= 11 array= -1.440600000000000E-007
write k= 12 array= 1.021350000000000E-007
write k= 13 array= 7.749780000000000E-008
write k= 14 array= 6.192490000000000E-008
write k= 15 array= 5.654470000000000E-008
write k= 16 array= 1.439460000000000E-007
write k= 17 array= -1.247630000000000E-007
write k= 18 array= -7.994810000000000E-008
write k= 19 array= 1.782930000000000E-007
write k= 20 array= -1.163500000000000E-007
write k= 21 array= 1.488370000000000E-007
write k= 22 array= 8.129720000000000E-008
write k= 23 array= 2.776730000000000E-006
write k= 24 array= -7.297590000000000E-006
write k= 25 array= 1.273390000000000E-005
write k= 26 array= -1.434600000000000E-006
write k= 27 array= -8.933020000000000E-010
write k= 28 array= -6.315520000000000E-010
hit eof while reading sigma file
abort:

=========================================================================

I am not able to figure out the problem. What steps I should follow for running GSM with higher resolution? Do I
need to update sigma files? Do I need to have special flags while compiling GSM/libs/etc/scnvrt.x ?
Please let me know if any other information is needed.



Thanks & Regards,
Sandip.

hello !
I am trying to install RSM on NERSC machine (franklin) and by following the instructions I am getting errors. It happens when running the makefile in
/scratch/scratchdirs/hafner/0SCOAR/RSM/RSM_20100127/rsm

the error type :

/global/homes/h/hafner/scratch/0SCOAR/RSM/RSM_20100127/rsm/bin/gsml.a(gbphys.o): In function `gbphys_':
./gbphys.f:640: undefined reference to `sfcdif_'
./gbphys.f:640: undefined reference to `sfcdrv_'
./gbphys.f:640: undefined reference to `sfcdiag_'
./gbphys.f:803: undefined reference to `kf2cnv_'
./gbphys.f:845: undefined reference to `cnvcldn_'
./gbphys.f:845: undefined reference to `lrgscl_'

apparently some routines do not get compiled and linked into library. I discovered that some of the necessary routines are under subdirectories, e.g.
sfcdif.F is
/scratch/scratchdirs/hafner/0SCOAR/RSM/RSM_20100127/rsm/src/gsml
and under two subdirectories : NOALSM1/sfcdif.F VICLSM1/sfcdif.F

please if you have any hint or idea what I am doing wrong let me know ..
thanks !
Jan Hafner (IPRC)

Hello,

I am trying to install RSM on franklin machine, but the automated make script does not complete.
Once the ./install script starts it quits with the message:

=

make[2]: Circular @OBJS_SFC0@ <- @OBJS_SFC0@ dependency dropped.

==

then the @OBJS_SFC0@ is not defined and the compilation quits too.

the offending make file is :
~/rsm/src/rsfc0/Makefile

which looks like:

=============================

EXECUTABLE = rsfc0.x
OBJS = @OBJS_SFC0@

MACHINE = linux
SHELL = /bin/sh
CP = cp
RM = rm
CD = cd
CHMOD = chmod

CC = cc
CPP = gcc -E
F77 = pgf90
FORT_FLAGS = -r8 -O2 -Msave -Mrecursive -Mdalign -DLINUX -byteswapio
LOAD_FLAGS = -r8 -O2 -lpgmp -Mrecursive -Mdalign -DLINUX -byteswapio
INCLUDE_DIR = -I../.. -I../include -I/opt/mpt/3.5.0/xt/mpich2-pgi/include
LIB_DIR = /global/homes/h/hafner/scratch/0SCOAR/RSM/RSM_20100309/libs/lib
INSTALLDLIB = /global/homes/h/hafner/scratch/0SCOAR/RSM/RSM_20100309/rsm/bin
INSTALL_DIR = /global/homes/h/hafner/scratch/0SCOAR/RSM/RSM_20100309/rsm/bin
EXTRA_LIBS =
EXTRA_LIBS_SFC0 =

LIBS = ${INSTALLDLIB}/sfcl.a ${INSTALLDLIB}/rsml.a ${INSTALLDLIB}/share.a ${LIB_DIR}/w3lib.a ${LIB_DIR}/modelib.a ${EXTRA_LIBS_SFC0}

SRCS = ${OBJS:.o=.F}

.DEFAULT :
ln -fs ../sfc0/$@ $@

${EXECUTABLE} : ${OBJS} ${LIBS}
if [ linux = mac ] ; then \
ranlib -s ${LIBS} ; \
fi
${F77} ${LOAD_FLAGS} -o $@ ${OBJS} ${LIBS} ${EXTRA_LIBS}
${CP} $@ ${INSTALL_DIR}

${OBJS} : ${SRCS}

depend :
${LIB_DIR}/../etc/mkdep

.SUFFIXES : .o .f .F

.F.f :
sed '/<paramodel.h>/{G;s/$$/#undef MP/;}' $*.F >TEMP.F
${CPP} -P ${INCLUDE_DIR} TEMP.F >$*.i
sed '/^ *$$/d;s/\/\*/\!/' $*.i >$*.f ; rm $*.i TEMP.F
.F.o :
sed '/<paramodel.h>/{G;s/$$/#undef MP/;}' $*.F >TEMP.F
${CPP} -P ${INCLUDE_DIR} TEMP.F >$*.i
sed '/^ *$$/d;s/\/\*/\!/' $*.i >$*.f ; rm $*.i TEMP.F
${F77} ${FORT_FLAGS} -c $*.f

.f.o :
${F77} ${FORT_FLAGS} -c $*.f

clean :
- ${RM} -f *.f *.o ${EXECUTABLE} core

dist_clean :
- ${RM} -f *.f *.o *.x core Makefile
- ${RM} -rf rii_files

=======================

please if you have any idea what could be wrong and how to possibly fix it - let me know

thanks !
Jan Hafner (IPRC/SOEST U. of Hawaii)

Hi,

I try to perform installation by following the instructions. However, I have encountered following errors :

1. When checkout from rokka.ucsd.edu, i will get error message : can't create temporary directory /tmp/cvs-serv27779
No space left on device

2. I tried checkout "Install" from 219.142.99.7, it works but when i execute the install script, the following error messages prompted :

cvs co libs_rsm
+ cvs co libs_rsm
cvs server: cannot find module `lib/0READ.ME' - ignored
cvs server: cannot find module `lib/Makefile.in' - ignored
cvs server: cannot find module `lib/modelib' - ignored
cvs server: cannot find module `lib/ncaru' - ignored
cvs server: cannot find module `lib/w3lib' - ignored
cvs server: cannot find module `lib/w3lib_cray' - ignored
cvs server: cannot find module `lib/w3lib_dec' - ignored
cvs server: cannot find module `lib/w3lib_hp' - ignored
cvs server: cannot find module `lib/w3lib_ibmsp' - ignored
cvs server: cannot find module `lib/w3lib_origin' - ignored
cvs server: cannot find module `lib/w3lib_sgi' - ignored
cvs server: cannot find module `lib/w3lib_sun' - ignored
cvs server: cannot find module `lib/w3lib_t90' - ignored
cvs server: cannot find module `lib/w3lib_t3e' - ignored
cvs server: cannot find module `lib/w3lib_linux' - ignored
cvs server: cannot find module `lib/w3lib_nec' - ignored
cvs server: cannot find module `lib/w3lib_es' - ignored
cvs server: cannot find module `lib/w3lib_mac' - ignored
cvs server: cannot find module `lib/w3lib_intel' - ignored
cvs server: cannot find module `lib/w3lib_sx6' - ignored
cvs server: cannot find module `lib/w3lib_ibmspbv' - ignored
cvs server: cannot find module `lib/w3lib_ibmspbs' - ignored
cvs server: cannot find module `lib/w3lib_ibmspbl' - ignored
cvs server: cannot find module `etc' - ignored
cvs [checkout aborted]: cannot expand modules

Thanks a lot for any help.

I given ./install and selected rsm script..
then it shows the following error
.. help me in fixing this error



Executing configure-model in rsm
+ ./configure-model
Key variables and their values used to make sysvars definition file:
BIT64=no
COMPILER=intel
DEBUG=no
INSTITUTION=prl
LINUXVSN=0
MACHINE=linux
MARCH=mpi
MPIVSN=1
NETCDFVSN=3
NICKNAME=midhun
= not defined
Edit and add values to MPICH_DIR
+ exit 8

I want to install gsm and rsm on IBM AIX P6.
Anybody is familiar with the installation of g-rsm on IBM machines, If so please help me .. Please share the def/sysvars.defs file

I have installed gsm on last week of march 2013. I retained the same sysvars.defs file for a new installation on today (19-june), but I got the following error.....

=============================

ifort -r8 -O2 -convert big_endian -L/lib64 -o grmap.x grmap.o /scratch/amit/gsm/GSMfree/libs/lib/w3lib.a /scratch/amit/gsm/GSMfree/libs/lib/libncaru.a /scratch/amit/gsm/GSMfree/libs/lib/modelib.a
/opt/software/intel/composer_xe_2011_sp1.7.256/compiler/lib/intel64/libifcore.a(for_diags_intel.o): In function `forio_return':
for_diags_intel.c:(.text+0xb95): undefined reference to `intel_sse2_strchr'
/opt/software/intel/composer_xe_2011_sp1.7.256/compiler/lib/intel64/libifcore.a(for_diags_intel.o): In function `forissue_diagnostic':
for_diags_intel.c:(.text+0x127a): undefined reference to `intel_sse2_strchr'
for_diags_intel.c:(.text+0x13cf): undefined reference to `intel_sse2_strchr'
/opt/software/intel/composer_xe_2011_sp1.7.256/compiler/lib/intel64/libifcore.a(for_diags_intel.o): In function `forget_msg':
for_diags_intel.c:(.text+0x17a0): undefined reference to `intel_sse2_strchr'
/opt/software/intel/composer_xe_2011_sp1.7.256/compiler/lib/intel64/libifcore.a(for_diags_intel.o): In function `redefine_severity_table':
for_diags_intel.c:(.text+0x1d61): undefined reference to `intel_sse2_strchr'
/opt/software/intel/composer_xe_2011_sp1.7.256/compiler/lib/intel64/libifcore.a(for_diags_intel.o):for_diags_intel.c:(.text+0x1eb6): more undefined references to `__intel_sse2_strchr' follow

---

Is there any change made in the code after 2013 march? Pleasse help me in fixing this problem,
thankfully,
midhun

After the rsm run-script compiled, I ran the rsm_runs test then "echo Thu May 12 15:49:37 CST 2016 - rinpg execution failed" happened.
Anyone had same problem as mine?

rmakemt.f(4): error #5192: Lead underscore not allowed

parameter(mtnres_=_rsmtnres_)

------------------------^

rmakemt.f(4): error #6592: This symbol must be a defined parameter, an enumerator, or an argument of an inquiry function that evaluates to a compile-time constant. [_RSMTNRES_]

parameter(mtnres_=_rsmtnres_)

------------------------^

rmakemt.f(4): error #6404: This name does not have a type, and must have an explicit type. [_RSMTNRES_]

parameter(mtnres_=_rsmtnres_)

------------------------^

compilation aborted for rmakemt.f (code 1)

make[2]: * [rmakemt.o] Error 1

make[2]: Leaving directory `/home/lmwu/rsm/crsm/src/rmtn'

make[1]: * [rsm_single] Error 8

make[1]: Leaving directory `/home/lmwu/rsm/crsm/src'

make: *** [all] Error 8

++ exit 8



As above, what is "Lead underscore not allowed" .............???