ERROR: BUILD FAIL: buildexe failed

[sunweihao]

hello,I create a new case ./create_newcase --case test4 --compset FCHIST --res f09_g17 --mach sun --compiler gnu --run-unsupported

and then I ./case.setup and ./case.build. But it shows error:

it seems there is some error when create exe

I cat the log and it:

ERROR: Command make exec_se -j 4 EXEC_SE=/home/sun/cesm/scratch/test4/bld/cesm.exe MODEL=driver LIBROOT=/home/sun/cesm/scratch/test4/bld/lib -f /home/sun/cesm/cime/scripts/test4/Tools/Makefile failed rc=2
out=/home/sun/cesm/cime/scripts/test4/Tools/mkSrcfiles
cp -f Filepath Deppath
/home/sun/cesm/cime/scripts/test4/Tools/mkDepends Deppath Srcfiles > Depends
/home/sun/cesm/cime/scripts/test4/Tools/mkSrcfiles
mpif90 -o /home/sun/cesm/scratch/test4/bld/cesm.exe cime_comp_mod.o cime_driver.o component_mod.o component_type_mod.o cplcomp_exchange_mod.o map_glc2lnd_mod.o map_lnd2glc_mod.o map_lnd2rof_irrig_mod.o mrg_mod.o prep_aoflux_mod.o prep_atm_mod.o prep_glc_mod.o prep_ice_mod.o prep_lnd_mod.o prep_ocn_mod.o prep_rof_mod.o prep_wav_mod.o seq_diag_mct.o seq_domain_mct.o seq_flux_mct.o seq_frac_mct.o seq_hist_mod.o seq_io_mod.o seq_map_mod.o seq_map_type_mod.o seq_rest_mod.o t_driver_timers_mod.o -L/home/sun/cesm/scratch/test4/bld/lib/ -latm -L/home/sun/cesm/scratch/test4/bld/lib/ -lice -L../../gnu/mpich/nodebug/nothreads/mct/noesmf/lib/ -lclm -L/home/sun/cesm/scratch/test4/bld/lib/ -locn -L/home/sun/cesm/scratch/test4/bld/lib/ -lrof -L/home/sun/cesm/scratch/test4/bld/lib/ -lglc -L/home/sun/cesm/scratch/test4/bld/lib/ -lwav -L/home/sun/cesm/scratch/test4/bld/lib/ -lesp -L../../glc/lib/ -lglimmercismfortran -L../../gnu/mpich/nodebug/nothreads/mct/noesmf/c1a1l1i1o1r1g1w1e1/lib -lcsm_share -L../../gnu/mpich/nodebug/nothreads/lib -lpio -lgptl -lmct -lmpeu -L/home/sun/netcdf4/lib -lnetcdf -lnetcdff -L/home/sun/netcdf4/lib -lhdf5 -lhdf5_hl -L/home/sun/netcdf4/lib -lz -L/home/sun/software/lapack-3.10.0 -llapack -L/home/sun/software/BLAS-3.10.0 -lblas -L/home/sun/app/mpich/lib -lmpi -L/home/sun/netcdf4/lib
/usr/bin/ld: ../../glc/lib//libglimmercismfortran.a(dcgs.f.o): in function `isdcgs_':
dcgs.f:(.text+0x105): undefined reference to `dnrm2_'
/usr/bin/ld: dcgs.f:(.text+0x1f9): undefined reference to `dnrm2_'
/usr/bin/ld: dcgs.f:(.text+0x260): undefined reference to `dnrm2_'
/usr/bin/ld: dcgs.f:(.text+0x283): undefined reference to `dnrm2_'
/usr/bin/ld: dcgs.f:(.text+0x2cd): undefined reference to `dnrm2_'
/usr/bin/ld: ../../glc/lib//libglimmercismfortran.a(dcgs.f.o):dcgs.f:(.text+0x2f7): more undefined references to `dnrm2_' follow
collect2: error: ld returned 1 exit status
make: *** [/home/sun/cesm/cime/scripts/test4/Tools/Makefile:874：/home/sun/cesm/scratch/test4/bld/cesm.exe]

HELP,how can I solve it??

 

[jedwards]

these functions are provided by the lapack library, a free version is available here. Many compiler vendors provide their own.

 

[sunweihao]

these functions are provided by the lapack library, a free version is available here. Many compiler vendors provide their own.

sorry，which functions？

 

[jedwards]

The error message is: undefined reference to `dnrm2_'
this means that your code is calling a function that the linker couldn't find. The solution is to link the lapack library. Your system may or may not already have a version of this library installed. If you are unsure, you should contact your system administration staff.

 

[Qinlu]

Hello, Sun. I'm working on ./case.build and encountered the same problem as you. And after I reinstalled lapack, it is still the same, do you have a solution to this problem? How is it solved?

 

[luojx]

Hello, Sun. I'm working on ./case.build and encountered the same problem as you. And after I reinstalled lapack, it is still the same, do you have a solution to this problem? How is it solved?

Sorry to interrupt you. I got the same problem. Did you solve it?

 

[Qinlu]

Sorry to interrupt you. I got the same problem. Did you solve it?

Moring. After searching on google, I saw that this problem was caused by improper installation of the math library lapack, so I tried to reinstall blas, cblas and lapack. Fortunately, when I reinstalled successfully, the problem was solved. You can try to reinstall these three packages to see if it can be resolved. Wish lucky!

 

[luojx]

Moring. After searching on google, I saw that this problem was caused by improper installation of the math library lapack, so I tried to reinstall blas, cblas and lapack. Fortunately, when I reinstalled successfully, the problem was solved. You can try to reinstall these three packages to see if it can be resolved. Wish lucky!

Thank you for the reply. I reinstalled lapack, and this problem was resolved, but a new error occurred:

Have you ever encountered this? Expect any suggestions.
Full information is in: case.submit error

 

[Qinlu]

Thank you for the reply. I reinstalled lapack, and this problem was resolved, but a new error occurred:
View attachment 1899
Have you ever encountered this? Expect any suggestions.
Full information is in: case.submit error

Hi, Luo. I can not get any useful information from the print-out, you may can check the 'cesm.log' at /home/jxluo/moshi/clm5.0/output/run/.

 

[luojx]

Evenning, Li. Thank you again for your kindly reply. This run did not generate cesm.log file. The content in atm.log, cpl.log, and lnd.log etc. is nearly identical, shown in Fig.2.


I submit the case with "mpirun -np 20 ./cesm.exe" directly, and got the error below, which I did not resolve yet.

Hi, Luo. I can not get any useful information from the print-out, you may can check the 'cesm.log' at /home/jxluo/moshi/clm5.0/output/run/.

 

[jedwards]

You are not getting a cesm log because you are not using the case.submit mechanism to start the job. The error you have suggests that the model was built for a different task count than you are running with, another issue that would be avoided by using the case.submit mechanism.

 

[luojx]

You are not getting a cesm log because you are not using the case.submit mechanism to start the job. The error you have suggests that the model was built for a different task count than you are running with, another issue that would be avoided by using the case.submit mechanism.

When I used the case.submit, this error occurred "ERROR: No result from jobs [('case.run', None), ('case.st_archive', 'case.run or case.test')]", and cesm.log file did not generate either. The attachment is the full submit information.

 

[jedwards]

This shows that the model run is completed and logs are written to /home/jxluo/moshi/clm5.0/output/maduo/logs/ but then something is
going wrong in the case.st_archive step. Could you run

case.st_archive --test-case

from the command line and see if that reproduces the error?

 

[Qinlu]

This shows that the model run is completed and logs are written to /home/jxluo/moshi/clm5.0/output/maduo/logs/ but then something is
going wrong in the case.st_archive step. Could you run

case.st_archive --test-case

from the command line and see if that reproduces the error?

Hello, dear respectly Jedwards. There are some difficulties need your help, that when I use the './case.submits' to run the model, it always return an error that "Sumitted job case.run with id None" and "Sumitted job case.st_archive with id None"(Details can be found in the attachment). And I try to run the cesm.exe with "mpiexec -np 64 cesm.exe" at command line, the program begins but end after one time, with the output is the results of 0000-1-1. Follow your reply, I have run the './case.st_archive --test-case' at the command line, but I can not find any obvious error or warning(Forgive my ignorance and novice, and details also in the attachment, and you can change the '.txt' to '.md' with a formatted text view). Is there something wrong with me, and what caused this situation? What should I do, or can respectly you tell me where I can find the relevant output documentation so that I can solve the problem by myself.

 

[luojx]

This shows that the model run is completed and logs are written to /home/jxluo/moshi/clm5.0/output/maduo/logs/ but then something is
going wrong in the case.st_archive step. Could you run

case.st_archive --test-case

from the command line and see if that reproduces the error?

Thank you for your kindly reply. I run "case.st_archive --test-case", and the print-out is in the txt-file. I resubmit the case, this error (ERROR: No result from jobs [('case.run', None), ('case.st_archive', 'case.run or case.test')]) occurred again. You mentioned that the model run is completed, so could I ignore this error directly?

 

[MAHE1319]

Moring. After searching on google, I saw that this problem was caused by improper installation of the math library lapack, so I tried to reinstall blas, cblas and lapack. Fortunately, when I reinstalled successfully, the problem was solved. You can try to reinstall these three packages to see if it can be resolved. Wish lucky!

I am getting the same error. I have installed lapack but still get the error. I am attaching the lib too. Any suggestions, please help.
Thanks.