Scheduled Downtime
On Tuesday 24 October 2023 @ 5pm MT the forums will be in read only mode in preparation for the downtime. On Wednesday 25 October 2023 @ 5am MT, this website will be down for maintenance and expected to return online later in the morning.
Normal Operations
The forums are back online with normal operations. If you notice any issues or errors related to the forums, please reach out to help@ucar.edu

clm3 on linux cluster with mpich-gm

W

world100@gmail_com

New Member
Hello,

I am trying to run clm3 offline on linux cluster with mpi. but I cann't compile it.

configure as following:

[yqian@master offline]$ ./jobscript.csh.qian

compiling clm model
The following Makefile environment is used
ROOTDIR is /home3/yqian/clm3
MODEL_EXEDIR is /home3/yqian/clm3/clm3run
SMP is FALSE
SPMD is TRUE
DEBUG is FALSE
INC_NETCDF is /home3/yqian/local/netcdf/include
LIB_NETCDF is /home3/yqian/local/netcdf/lib
USER_FC is pgf90
INC_MPI is /home/mpich-gm/include
LIB_MPI is /home/mpich-gm/lib
Linux machine name is master
Linux mpirun cmnd is /home/mpich-gm/bin/mpirun
GMAKE ERROR failed: see /home3/yqian/clm3/clm3run/compile_log.clm

Could you give any hints

Thanks



compile_log.clm
......
gmpi_regcache.o(.text+0xf08): undefined reference to `gm_lookaside_alloc'
gmpi_regcache.o(.text+0xf6d): undefined reference to `GM_PAGE_LEN'
gmpi_regcache.o(.text+0xf8a): undefined reference to `gm_hash_insert'
/home/mpich-gm/lib/libmpich.a(gmpi_regcache.o): In function `gmpi_use_interval':
gmpi_regcache.o(.text+0x1017): undefined reference to `GM_PAGE_LEN'
gmpi_regcache.o(.text+0x1059): undefined reference to `gm_hash_find'
gmpi_regcache.o(.text+0x1166): undefined reference to `GM_PAGE_LEN'
/home/mpich-gm/lib/libmpich.a(gmpi_regcache.o): In function `gmpi_unuse_interval':
gmpi_regcache.o(.text+0x1219): undefined reference to `GM_PAGE_LEN'
gmpi_regcache.o(.text+0x1247): undefined reference to `gm_hash_find'
gmpi_regcache.o(.text+0x1258): undefined reference to `GM_PAGE_LEN'
/home/mpich-gm/lib/libmpich.a(gmpi_regcache.o): In function `gmpi_clear_interval':
gmpi_regcache.o(.text+0x12ae): undefined reference to `GM_PAGE_LEN'
gmpi_regcache.o(.text+0x12f0): undefined reference to `gm_hash_find'
gmpi_regcache.o(.text+0x130a): undefined reference to `gm_hash_remove'
gmpi_regcache.o(.text+0x1367): undefined reference to `gm_lookaside_free'
gmpi_regcache.o(.text+0x1382): undefined reference to `GM_PAGE_LEN'
/home/mpich-gm/lib/libmpich.a(gmpi_regcache.o): In function `gmpi_clear_all_intervals':
gmpi_regcache.o(.text+0x13e8): undefined reference to `gm_destroy_hash'
gmpi_regcache.o(.text+0x13f6): undefined reference to `gm_destroy_lookaside'
/home/mpich-gm/lib/libmpich.a(adi2init.o): In function `MPID_Init':
adi2init.o(.text+0xf95): undefined reference to `gm_get_host_name'
/home/mpich-gm/lib/libmpich.a(gmpi_conf.o): In function `gmpi_allocate_port':
gmpi_conf.o(.text+0x9d2): undefined reference to `gm_open'
gmpi_conf.o(.text+0xa82): undefined reference to `gm_open'
/home/mpich-gm/lib/libmpich.a(gmpi_conf.o): In function `gmpi_getconf':
gmpi_conf.o(.text+0xf5a): undefined reference to `gm_receive'
gmpi_conf.o(.text+0xf86): undefined reference to `gm_receive'
gmpi_conf.o(.text+0xfaf): undefined reference to `gm_blocking_receive_no_spin'
gmpi_conf.o(.text+0xfd8): undefined reference to `gm_blocking_receive'
gmpi_conf.o(.text+0xfe5): undefined reference to `gm_receive'
gmpi_conf.o(.text+0xff4): undefined reference to `gm_receive'
gmpi_conf.o(.text+0x1142): undefined reference to `gm_get_node_id'
gmpi_conf.o(.text+0x1193): undefined reference to `gm_node_id_to_unique_id'
gmpi_conf.o(.text+0x1251): undefined reference to `gm_bzero'
gmpi_conf.o(.text+0x1271): undefined reference to `gm_get_host_name'
gmpi_conf.o(.text+0x12fa): undefined reference to `gm_bzero'
gmpi_conf.o(.text+0x13af): undefined reference to `gm_bzero'
gmpi_conf.o(.text+0x19fb): undefined reference to `gm_receive'
gmpi_conf.o(.text+0x1a27): undefined reference to `gm_receive'
gmpi_conf.o(.text+0x1a50): undefined reference to `gm_blocking_receive_no_spin'
gmpi_conf.o(.text+0x1a79): undefined reference to `gm_blocking_receive'
gmpi_conf.o(.text+0x1a86): undefined reference to `gm_receive'
gmpi_conf.o(.text+0x1a95): undefined reference to `gm_receive'
gmpi_conf.o(.text+0x1bdc): undefined reference to `gm_get_node_id'
gmpi_conf.o(.text+0x1c2d): undefined reference to `gm_node_id_to_unique_id'
gmpi_conf.o(.text+0x1d18): undefined reference to `gm_bzero'
gmpi_conf.o(.text+0x1d49): undefined reference to `gm_bcopy'
gmpi_conf.o(.text+0x1f12): undefined reference to `gm_bzero'
gmpi_conf.o(.text+0x1f3f): undefined reference to `gm_bcopy'
gmpi_conf.o(.text+0x1f63): undefined reference to `gm_ticks'
gmpi_conf.o(.text+0x1fa9): undefined reference to `gm_ticks'
gmpi_conf.o(.text+0x21f9): undefined reference to `gm_bzero'
gmpi_conf.o(.text+0x273b): undefined reference to `gm_get_host_name'
gmpi_conf.o(.text+0x2781): undefined reference to `gm_unique_id_to_node_id'
gmpi_conf.o(.text+0x2829): undefined reference to `gm_crc'
gmpi_conf.o(.text+0x2850): undefined reference to `gm_crc'
gmpi_conf.o(.text+0x2999): undefined reference to `GM_PAGE_LEN'
gmpi_conf.o(.text+0x29b7): undefined reference to `GM_PAGE_LEN'
/home/mpich-gm/lib/libmpich.a(gmpi_conf.o): In function `MPID_GetConfigInfo':
gmpi_conf.o(.text+0x2b9a): undefined reference to `gm_receive'
/home/mpich-gm/lib/libmpich.a(gmpi_priv.o): In function `gmpi_dma_alloc':
gmpi_priv.o(.text+0xc55): undefined reference to `gm_dma_malloc'
/home/mpich-gm/lib/libmpich.a(gmpi_priv.o): In function `gmpi_dma_free':
gmpi_priv.o(.text+0xc85): undefined reference to `gm_dma_free'
/home/mpich-gm/lib/libmpich.a(gmpi_priv.o): In function `gmpi_drop_send_packet_callback':
gmpi_priv.o(.text+0xff6): undefined reference to `gm_send_with_callback'
gmpi_priv.o(.text+0x1008): undefined reference to `gm_strerror'
/home/mpich-gm/lib/libmpich.a(gmpi_priv.o): In function `gmpi_drop_put_data_callback':
gmpi_priv.o(.text+0x10ea): undefined reference to `gm_directed_send_with_callback'
gmpi_priv.o(.text+0x10fc): undefined reference to `gm_strerror'
/home/mpich-gm/lib/libmpich.a(gmpi_priv.o): In function `gmpi_check_send_status':
gmpi_priv.o(.text+0x11fb): undefined reference to `gm_drop_sends'
gmpi_priv.o(.text+0x127c): undefined reference to `gm_send_with_callback'
gmpi_priv.o(.text+0x12dc): undefined reference to `gm_directed_send_with_callback'
gmpi_priv.o(.text+0x132a): undefined reference to `gm_node_id_to_host_name'
gmpi_priv.o(.text+0x137d): undefined reference to `gm_node_id_to_unique_id'
gmpi_priv.o(.text+0x13d6): undefined reference to `gm_strerror'
/home/mpich-gm/lib/libmpich.a(gmpi_priv.o): In function `gmpi_flush_fifo':
gmpi_priv.o(.text+0x1910): undefined reference to `gm_send_with_callback'
gmpi_priv.o(.text+0x19c1): undefined reference to `gm_directed_send_with_callback'
/home/mpich-gm/lib/libmpich.a(gmpi_priv.o): In function `gmpi_send_packet_fifo':
gmpi_priv.o(.text+0x1bdf): undefined reference to `gm_send_with_callback'
/home/mpich-gm/lib/libmpich.a(gmpi_priv.o): In function `gmpi_put_data':
gmpi_priv.o(.text+0x1e03): undefined reference to `gm_directed_send_with_callback'
/home/mpich-gm/lib/libmpich.a(gmpi_priv.o): In function `gmpi_packet_recv_event':
gmpi_priv.o(.text+0x1f0c): undefined reference to `gm_node_id_to_unique_id'
/home/mpich-gm/lib/libmpich.a(gmpi_priv.o): In function `gmpi_net_lookup':
gmpi_priv.o(.text+0x242c): undefined reference to `gm_receive'
gmpi_priv.o(.text+0x255f): undefined reference to `gm_provide_receive_buffer'
gmpi_priv.o(.text+0x25a0): undefined reference to `gm_unknown'
/home/mpich-gm/lib/libmpich.a(gmpi_priv.o): In function `gmpi_init':
gmpi_priv.o(.text+0x26b8): undefined reference to `gm_allow_remote_memory_access'
gmpi_priv.o(.text+0x26d1): undefined reference to `gm_perror'
gmpi_priv.o(.text+0x26fe): undefined reference to `gm_num_send_tokens'
gmpi_priv.o(.text+0x2719): undefined reference to `gm_num_receive_tokens'
gmpi_priv.o(.text+0x272f): undefined reference to `gm_num_receive_tokens'
gmpi_priv.o(.text+0x27ba): undefined reference to `gm_dma_malloc'
gmpi_priv.o(.text+0x27d5): undefined reference to `gm_dma_malloc'
gmpi_priv.o(.text+0x282a): undefined reference to `gm_set_acceptable_sizes'
gmpi_priv.o(.text+0x2843): undefined reference to `gm_perror'
gmpi_priv.o(.text+0x2887): undefined reference to `gm_set_acceptable_sizes'
gmpi_priv.o(.text+0x28a0): undefined reference to `gm_perror'
gmpi_priv.o(.text+0x2927): undefined reference to `gm_provide_receive_buffer_with_tag'
/home/mpich-gm/lib/libmpich.a(gmpi_priv.o): In function `gmpi_finish':
gmpi_priv.o(.text+0x2a98): undefined reference to `gm_close'
/home/mpich-gm/lib/libmpich.a(gmpi_chself.o): In function `MPID_CH_Self_isend':
gmpi_chself.o(.text+0xd7f): undefined reference to `gm_bcopy'
/home/mpich-gm/lib/libmpich.a(gmpi_chself.o): In function `MPID_CH_Self_send':
gmpi_chself.o(.text+0xf17): undefined reference to `gm_bcopy'
/home/mpich-gm/lib/libmpich.a(gmpi_smppriv.o): In function `smpi_queue_send':
gmpi_smppriv.o(.text+0xeb0): undefined reference to `gm_lookaside_alloc'
/home/mpich-gm/lib/libmpich.a(gmpi_smppriv.o): In function `MPID_SMP_Check_incoming':
gmpi_smppriv.o(.text+0x1fe2): undefined reference to `gm_lookaside_free'
/home/mpich-gm/lib/libmpich.a(gmpi_smppriv.o): In function `smpi_init':
gmpi_smppriv.o(.text+0x20a4): undefined reference to `gm_create_lookaside'
gmpi_smppriv.o(.text+0x2185): undefined reference to `GM_PAGE_LEN'
gmpi_smppriv.o(.text+0x23dd): undefined reference to `GM_PAGE_LEN'
gmpi_smppriv.o(.text+0x24e5): undefined reference to `GM_PAGE_LEN'
gmpi_smppriv.o(.text+0x26b4): undefined reference to `GM_PAGE_LEN'
/home/mpich-gm/lib/libmpich.a(gmpi_smppriv.o): In function `smpi_finish':
gmpi_smppriv.o(.text+0x2774): undefined reference to `GM_PAGE_LEN'
gmpi_smppriv.o(.text+0x27b7): undefined reference to `gm_destroy_lookaside'
gmake: *** [/home3/yqian/clm3/clm3run/clm] Error 2
 
W

world100@gmail_com

New Member
Songroy,

Thank you so much for your reply.

after I change pgf90 to mpif90, I can compile clm, but can not run the example successfully.
it shows :error: logname not defined.
could u give some hints?

Thanks


[yqian@master clm3run]$ more clm.log.050906-153653
running /home3/yqian/clm3/clm3run/clm on 2 LINUX ch_gm processors
Program binary is: /home3/yqian/clm3/clm3run/clm
Machines file is /home/mpich-gm-1.2.6..13/share/machines.ch_gm.LINUX
Shared memory for intra-nodes coms is enabled.
GM receive mode used: polling.
2 processes will be spawned:
Process 0 (/home3/yqian/clm3/clm3run/clm ) on node5
Process 1 (/home3/yqian/clm3/clm3run/clm ) on node5
Open a socket on master...
Got a first socket opened on port 47647.
Shared memory file: /tmp/gmpi_shmem-5984828:[0-9]*.tmp

rsh node5 cd /home3/yqian/clm3/clm3run ; env GMPI_MASTER=master GMPI_PORT=47647 GMPI_SHMEM=1 GMPI_VERBOSE=1 DISPLAY=localho
st:23.0 GMPI_MAGIC=5984828 GMPI_ID=0 GMPI_NP=2 GMPI_BOARD=-1 GMPI_SLAVE=10.0.1.5 /home3/yqian/clm3/clm3run/clm
rsh node5 -n cd /home3/yqian/clm3/clm3run ; env GMPI_MASTER=master GMPI_PORT=47647 GMPI_SHMEM=1 GMPI_VERBOSE=1 DISPLAY=loca
lhost:23.0 GMPI_MAGIC=5984828 GMPI_ID=1 GMPI_NP=2 GMPI_BOARD=-1 GMPI_SLAVE=10.0.1.5 /home3/yqian/clm3/clm3run/clm
t_setoption: option disabled: Usr Sys
t_setoption: option disabled: Usr Sys
2 pes participating in computation

-----------------------------------

NODE# NAME
( 0) node5.cluster
( 1) node5.cluster
(shr_orb_params) Calculate characteristics of the orbit:
(shr_orb_params) CVS revision: $Revision: 1.2.6.1 $
(shr_orb_params) CVS Tag : $Name: clm3_deva_21 $
(shr_orb_params) Calculate orbit for year: 1950
(shr_orb_params) ------ Computed Orbital Parameters ------
(shr_orb_params) Eccentricity = 1.672393E-02
(shr_orb_params) Obliquity (deg) = 2.344627E+01
(shr_orb_params) Obliquity (rad) = 4.092146E-01
(shr_orb_params) Long of perh(deg) = 1.020390E+02
(shr_orb_params) Long of perh(rad) = 4.922510E+00
(shr_orb_params) Long at v.e.(rad) = -3.262537E-02
(shr_orb_params) -----------------------------------------
CLM MODEL version 3.0

Attempting to initialize the land model .....
Preset Fortran unit numbers:
unit 5 = standard input
unit 6 = standard output

Attempting to initialize run control settings .....
error: logname not defined
ENDRUN: called without a message string
[0] MPI Abort by user Aborting program !
[0] Aborting program!
MPI Id 0 is using GM port 5, board 0 (MAC 0060dd4886de).
MPI Id 1 is using GM port 6, board 0 (MAC 0060dd4886de).
Received data from all 2 MPI processes.
Sending mapping to MPI Id 0.
Sending mapping to MPI Id 1.
Data sent to all processes.
Received valid abort message !
Reap remote processes:
rsh node5 -n kill -9 8253 2>/dev/null
rsh node5 -n kill -9 8254 2>/dev/null
Abort in progress...
 
W

world100@gmail_com

New Member
after I add LOGNAME in .bashrc, it can recognize logname.

but i meet new problem:
/bin/sh: msread: command not found

any suggests?

Thanks a lot
 
T

thomas_lavergne@jrc_it

New Member
goldensky said:
but i meet new problem:
/bin/sh: msread: command not found
Click to expand...

Dear all,
I am trying to use CLM3.0 under Linux and have the same issue as goldensky with the msread command: it cannot find it. Does anybody know what this command is supposed to do, how to get it, etc..

Thanks
 
T

thomas_lavergne@jrc_it

New Member
Answering to myself and for future reference...
It seems the msread command (although hardcoded in CLM3.0 :-( ) is not available outside a NCAR environment. It should extract a file from a mass-storage system. As I am working very locally (like many downloading and building the CLM3.0), I created a bash script msread (which should be executable and reachable through your PATH) which is a wrapper for the cp command:


Code: 
#!/bin/bash

if [ ! $# = '2' ]; then
   exit 1;
fi

# invert the arguments order ( is local,  is original)
cp  
exit 0


It seems to solve my problem.
Thomas
 
F

forrest@climate_ornl_gov

New Member
goldensky said:
Hello,

I am trying to run clm3 offline on linux cluster with mpi. but I cann't compile it.

configure as following:

. . .
Click to expand...

It looks to me like you merely need to add your GM library into the link step of the compile. This may happen automatically if you switch USER_FC to something like /home/mpich-gm/bin/mpif90

Good luck!
 
You must log in or register to reply here.
Top