COMPILATION ERROR [make -j]

[gopikrishnangs44]

I am trying to compile GCHP. Successfully done until 'Configure your build (Cmake)'.

Compilation (make -j) gives the following error:

[ 0%] Built target registry
[ 2%] Built target generate-template-incs
[ 6%] Built target link_target
[ 10%] Built target generate-type-incs
[ 10%] Building Fortran object src/fms_r8/CMakeFiles/fms_r8.dir/mpp/mpp_parameter.F90.o
/home/21cl91p01/Code.GCHP/src/FMS/mpp/mpp_parameter.F90:100:42:

100 | integer(LONG_KIND), parameter :: DOMAIN_ID_BASE=Z'0000000100000000' ! Workaround for 64bit init problem
| 1
Error: BOZ literal constant at (1) is neither a data-stmt-constant nor an actual argument to INT, REAL, DBLE, or CMPLX intrinsic function [see ‘-fno-allow-invalid-boz’]
/home/21cl91p01/Code.GCHP/src/FMS/mpp/mpp_parameter.F90:34:49:

34 | public :: AGRID, GLOBAL, CYCLIC, DOMAIN_ID_BASE, CENTER, CORNER
| 1
Error: Symbol ‘domain_id_base’ at (1) has no IMPLICIT type
make[2]: *** [src/fms_r8/CMakeFiles/fms_r8.dir/mpp/mpp_parameter.F90.o] Error 1
make[2]: *** Waiting for unfinished jobs....
[ 10%] Building Fortran object src/fms_r8/CMakeFiles/fms_r8.dir/constants/constants.F90.o
make[1]: *** [src/fms_r8/CMakeFiles/fms_r8.dir/all] Error 2
make[1]: *** Waiting for unfinished jobs....
[ 18%] Built target gftl-shared
make: *** [all] Error 2

Please help in the regard.

Thanks

[gopikrishnangs44]

I removed -j from Make and made a try. The result is (for make):

[ 8%] Built target generate-type-incs
[ 10%] Built target generate-template-incs
[ 18%] Built target gftl-shared
[ 22%] Built target yafyaml
[ 22%] Building Fortran object src/pFlogger/src/CMakeFiles/pflogger.dir/MpiCommConfig.F90.o
f951: Fatal Error: Reading module ‘/home/opt_ohpc_pub/apps/intel/2019/compilers_and_libraries_2019.5.281/linux/mpi/intel64/include/mpi.mod’ at line 1 column 2: Unexpected EOF
compilation terminated.
make[2]: *** [src/pFlogger/src/CMakeFiles/pflogger.dir/MpiCommConfig.F90.o] Error 1
make[1]: *** [src/pFlogger/src/CMakeFiles/pflogger.dir/all] Error 2
make: *** [all] Error 2

[gopikrishnangs44]

CMAKE OUTPUT

-- ecbuild 3.0.0 /home/21cl91p01/GCHP/ESMA_cmake/ecbuild/cmake
-- cmake 3.15.4 /home/opt_ohpc_pub/utils/cmake/3.15.4/bin/cmake

---

-- Tests have been disabled

---

CMake Warning at ESMA_cmake/ecbuild/cmake/ecbuild_log.cmake:158 (message):
WARN - Argument mismatches will be treated as warnings and not
errors. Per the gfortran 10 man page:

Some code contains calls to external procedures which
mismatches between the calls and the procedure definition,
or with mismatches between different calls. Such code is
non-conforming, and will usually be flagged wi1th an error.
This options degrades the error to a warning, which can
only be disabled by disabling all warnings vial -w. Only a
single occurrence per argument is flagged by this warning.
-fallow-argument-mismatch is implied by -std=legacy.
Using this option is strongly discouraged. It is possible to
provide standard-conforming code which allows different types
of arguments by using an explicit interface and TYPE(*).
Call Stack (most recent call first):
ESMA_cmake/GNU.cmake:57 (ecbuild_warn)
ESMA_cmake/esma.cmake:29 (include)
CMakeLists.txt:90 (include)

CMake Warning at ESMA_cmake/ecbuild/cmake/ecbuild_log.cmake:158 (message):
WARN - Invalid use of BOZ literal constants will be treated as
warnings and not as errors. Per the GCC 10 release notes:

The handling of a BOZ literal constant has been reworked
to provide better conformance to the Fortran 2008 and 2018
standards. In these Fortran standards, a BOZ literal constant is a
typeless and kindless entity. As a part of the rework, documented
and undocumented extensions to the Fortran standard now emit
errors during compilation. Some of these extensions are permitted
with the -fallow-invalid-boz, where the error is degraded to a
warning and the code is compiled as with older gfortran.
Call Stack (most recent call first):
ESMA_cmake/GNU.cmake:77 (ecbuild_warn)
ESMA_cmake/esma.cmake:29 (include)
CMakeLists.txt:90 (include)

-- Performing Test FORTRAN_COMPILER_SUPPORTS_ASSUMED_TYPE
-- Performing Test FORTRAN_COMPILER_SUPPORTS_ASSUMED_TYPE: SUCCESS
-- Performing Test FORTRAN_COMPILER_SUPPORTS_FINDLOC
-- Performing Test FORTRAN_COMPILER_SUPPORTS_FINDLOC: SUCCESS
-- Could NOT find ImageMagick (missing: ImageMagick_mogrify_EXECUTABLE)
-- NOTE: ImageMagick was not found. This will prevent using LaTeX and some postprocessing utilities from running, but does not affect the build
-- Found MPI: TRUE (found version "3.1")
-- BASEDIR: IGNORE
-- HDF5: Using hdf5 compiler wrapper to determine C configuration
-- Found NetCDF4: 1
-- Found NetCDF4: 1
-- Found MPI: TRUE (found version "3.1") found components: C CXX Fortran
-- Performing Test DETECT_F2PY_SUFFIX
-- Setting F2PY_SUFFIX to .so
-- Performing Test DETECT_F2PY_SUFFIX: SUCCESS

---

-- [gchp_ctm] (13.0.0) [a82afe3]
-- Performing Test _LOGICAL_DEFAULT_KIND: SUCCESS (value=4)
-- Performing Test _INT_DEFAULT_KIND: SUCCESS (value=4)
-- Performing Test _ISO_INT8: SUCCESS (value=1)
-- Performing Test _ISO_INT16: SUCCESS (value=2)
-- Performing Test _ISO_INT32: SUCCESS (value=4)
-- Performing Test _ISO_INT64: SUCCESS (value=8)
-- Performing Test _REAL_DEFAULT_KIND: SUCCESS (value=4)
-- Performing Test _DOUBLE_DEFAULT_KIND: SUCCESS (value=8)
-- Performing Test _ISO_REAL32: SUCCESS (value=4)
-- Performing Test _ISO_REAL64: SUCCESS (value=8)
-- Performing Test _ISO_REAL128: SUCCESS (value=16)
CMake Warning at ESMA_cmake/ecbuild/cmake/ecbuild_log.cmake:158 (message):
WARN - Argument mismatches will be treated as warnings and not
errors. Per the gfortran 10 man page:

Some code contains calls to external procedures which
mismatches between the calls and the procedure definition,
or with mismatches between different calls. Such code is
non-conforming, and will usually be flagged wi1th an error.
This options degrades the error to a warning, which can
only be disabled by disabling all warnings vial -w. Only a
single occurrence per argument is flagged by this warning.
-fallow-argument-mismatch is implied by -std=legacy.
Using this option is strongly discouraged. It is possible to
provide standard-conforming code which allows different types
of arguments by using an explicit interface and TYPE(*).
Call Stack (most recent call first):
ESMA_cmake/GNU.cmake:57 (ecbuild_warn)
src/yaFyaml/CMakeLists.txt:16 (include)

CMake Warning at ESMA_cmake/ecbuild/cmake/ecbuild_log.cmake:158 (message):
WARN - Invalid use of BOZ literal constants will be treated as
warnings and not as errors. Per the GCC 10 release notes:

The handling of a BOZ literal constant has been reworked
to provide better conformance to the Fortran 2008 and 2018
standards. In these Fortran standards, a BOZ literal constant is a
typeless and kindless entity. As a part of the rework, documented
and undocumented extensions to the Fortran standard now emit
errors during compilation. Some of these extensions are permitted
with the -fallow-invalid-boz, where the error is degraded to a
warning and the code is compiled as with older gfortran.
Call Stack (most recent call first):
ESMA_cmake/GNU.cmake:77 (ecbuild_warn)
src/yaFyaml/CMakeLists.txt:16 (include)

-- Performing Test SUPPORT_FOR_ASSUMED_TYPE
-- Performing Test SUPPORT_FOR_ASSUMED_TYPE: SUCCESSS
-- Performing Test SUPPORT_FOR_C_LOC_ASSUMED_SIZE
-- Performing Test SUPPORT_FOR_C_LOC_ASSUMED_SIZE: FAILURE
-- Performing Test SUPPORT_FOR_MPI_ALLOC_MEM_CPTR
-- Performing Test SUPPORT_FOR_MPI_ALLOC_MEM_CPTR: SUCCESSS
-- Could NOT find Doxygen (missing: DOXYGEN_EXECUTABLE)

---

-- [MAPL] (2.2.7)

HEMCO 3.0.0
Current status: GEOS_12.8-102-g133e60e

=================================================================
GEOS-Chem 13.0.0 (science codebase)
Current status: gcc_13.0.0-alpha.13-74-gcdd4217

-- Settings:

• OMP: ON OFF
• USE_REAL8: ON OFF
• APM: ON OFF
• RRTMG: ON OFF
• GTMM: ON OFF
• LUO_WETDEP: ON OFF
  -- Configuring done
  -- Generating done
  -- Build files have been written to: /home/21cl91p01/GCHP/build

There were some warnings. I am not sure whether this caused the problem or not.

[LiamBindle]

Hi @gopikrishnangs44, Could you in share CMakeCache.txt (it's in your build directory) as well as the build log as a text file?

I have a feeling the issue is a misconfigured environment (e.g., incompatible compiler or a mismatch of compiler + libraries). The CMakeCache.txt has the details we need to take a closer look.

Could you also tell me your compiler version and MPI version?

[gopikrishnangs44]

@LiamBindle

Attaching my build log and Cmakecache.txt
ecbuild.log
CMakeCache.txt

Libraries used:
netcdf_f/4.5.2
netcdf_c/4.7.4
openmpi/4.0.2
cmake/3.15.4
gcc/10.2.0

[LiamBindle]

Thanks @gopikrishnangs44. For the build log, I actually meant the console output from when you run make. You can do something like

make > build_log.txt

and upload the resulting build_log.txt file.

One thing that jumps out right away is the version number. It looks like the code you're using is 13.0.0. At the top of this thread you mention 13.3.4 though. Did you do

cd GCHP
git checkout 13.3.4
git submodule update --init --recursive

to checkout GEOS-Chem 13.3.4?

[gopikrishnangs44]

@LiamBindle
Sorry for mentioning the wrong version number.

I have done this.
cd GCHP
git checkout 13.3.4
git submodule update --init --recursive

updated to the latest version.

Still there is a make error. I am uploading the log file here.

CMakeCache.txt
make_log.log

[LiamBindle]

@gopikrishnangs44 Sorry I had a typo in my snippet, could you do

make &> build_log.txt

that way the errors will show up in the build log.

[gopikrishnangs44]

@LiamBindle

I am attaching the build log
build_log.txt

[LiamBindle]

It looks like there is a problem with your NetCDF-Fortran installation. Specifically, the netcdf Fortran module at /home/apps/netcdf_c_with_parallel_support_4.7.4/include/netcdf.mod. The compiler says that there is an EOF (end of file) at line 1 column 2 (the file is empty?).

Can you try the following command?

cat /home/apps/netcdf_c_with_parallel_support_4.7.4/include/netcdf.mod | gunzip | head -1

This should tell you what Fortran module version the file is. If it doesn't say something like

GFORTRAN module version '15' created from netcdf4.f90

then it seems like there is something wrong with your netcdf installation.

[gopikrishnangs44]

Thanks @LiamBindle.

It was netcdf error. So I have built netcdf from source and tried re-compiling the model.

Libraries used:
module load compiler/openmpi/4.0.2
module load compiler/cmake/3.15.4
module load compiler/gcc/8.3.0

netcdf-c-4.7.3
netcdf-cxx-4.2
netcdf-fortran-4.5.1
hdf5-1.10.5
zlib-1.2.11

Now the CMake command in build directory gives the following error.

-- The Fortran compiler identification is GNU 4.8.5
-- The CXX compiler identification is GNU 4.8.5
-- The C compiler identification is GNU 4.8.5
-- Check for working Fortran compiler: /usr/bin/f95
-- Check for working Fortran compiler: /usr/bin/f95 -- works
-- Detecting Fortran compiler ABI info
-- Detecting Fortran compiler ABI info - done
-- Checking whether /usr/bin/f95 supports Fortran 90
-- Checking whether /usr/bin/f95 supports Fortran 90 -- yes
-- Check for working CXX compiler: /usr/bin/c++
-- Check for working CXX compiler: /usr/bin/c++ -- works
-- Detecting CXX compiler ABI info
-- Detecting CXX compiler ABI info - done
-- Detecting CXX compile features
-- Detecting CXX compile features - done
-- Check for working C compiler: /usr/bin/cc
-- Check for working C compiler: /usr/bin/cc -- works
-- Detecting C compiler ABI info
-- Detecting C compiler ABI info - done
-- Detecting C compile features
-- Detecting C compile features - done
-- ecbuild 3.0.0 /home/21cl91p01/GCHP/ESMA_cmake/ecbuild/cmake
-- cmake 3.15.4 /home/opt_ohpc_pub/utils/cmake/3.15.4/bin/cmake

---

-- Found Git: /usr/bin/git (found version "1.8.3.1")
-- Performing Test C_FLAG_TEST_1
-- Performing Test C_FLAG_TEST_1 - Success
-- Performing Test CXX_FLAG_TEST_1
-- Performing Test CXX_FLAG_TEST_1 - Success
-- Looking for sys/types.h
-- Looking for sys/types.h - found
-- Looking for stdint.h
-- Looking for stdint.h - found
-- Looking for stddef.h
-- Looking for stddef.h - found
-- Check size of void*
-- Check size of void* - done
-- Check if the system is big endian
-- Searching 16 bit integer
-- Check size of unsigned short
-- Check size of unsigned short - done
-- Using unsigned short
-- Check if the system is big endian - little endian
-- Check size of off_t
-- Check size of off_t - done
-- Tests have been disabled

CMake Error at ESMA_cmake/GNU.cmake:2 (message):
GNU version must be at least 8.3!
Call Stack (most recent call first):
ESMA_cmake/esma.cmake:29 (include)
CMakeLists.txt:84 (include)

I have loaded gcc/8.3.0 but still GNU recognises as 4.3.

netcdf's were built using gcc 8.3.0.

Regards,
Gopikrishnan

[LiamBindle]

The compilers are selected according to the CC, CXX, and FC variables if they are set (if not, it tries to find compilers automatically).

You can find your compiler version with

$ gfortran --version

and the path to your compiler with

$ which gfortran

First, check that gfortran --version is 8.3.0. Assuming it is, try recreating your build directory (delete your current build directory and create a new one). The build directory saves settings like paths to compilers, so you need to completely recreate your build directory to change the selected compiler.

You can do

$ export CC=gcc
$ export CXX=g++
$ export FC=gfortran

to set the CC, CXX, and FC variables (you can also specify absolute paths to the compilers). If you do this before your initial cmake path/to/gchp call, CMake will use the compilers you specified.

[gopikrishnangs44]

Thank you @LiamBindle

CMake is built successfully now.

But make still have some error. I am attaching the make log and cmake file for your reference.

build_log.txt
cmake_log.txt

ESMF
export ESMF_MOD_DIR='/home/21cl91p01/ESMF/mod/modO/Linux.gfortran.64.mpiuni.default/'
export ESMF_LIBRARY='/home/21cl91p01/ESMF/lib/libO/Linux.gfortran.64.mpiuni.default/'
export ESMF_HEADERS_DIR='/home/21cl91p01/ESMF/src/include/'
export CMAKE_PREFIX_PATH=$ESMF_HEADERS_DIR:$ESMF_MOD_DIR:$ESMF_LIBRARY:'/home/apps/gcc/8.3.0'

[gopikrishnangs44]

I had used gcc 10 to build ESMF before. But now I rebuild it using 8.3.0. And I repeated the cmake and make commands and the executables were build successfully.

But now the error is:

[21cl91p01@login02 build]$ ./gchp

WARNING: There was an error initializing an OpenFabrics device.

Local host: login02
Local device: mlx5_0

At line 32 of file /home/21cl91p01/GCHP/src/yaFyaml/src/FileStream.F90
Fortran runtime error: Cannot open file 'logging.yml': No such file or directory

Error termination. Backtrace:
#0 0x7f59f437d314 in already_open
at ../../../libgfortran/io/open.c:714
#1 0x2629181 in __fy_filestream_MOD_new_filestream
at /home/21cl91p01/GCHP/src/yaFyaml/src/FileStream.F90:32
#2 0x25cb7d2 in __pfl_loggermanager_MOD_load_file
at /home/21cl91p01/GCHP/src/pFlogger/src/LoggerManager.F90:340
#3 0x16de40e in initialize_pflogger
at /home/21cl91p01/GCHP/src/MAPL/base/ApplicationSupport.F90:108
#4 0x16e2316 in __mapl_applicationsupport_MOD_mapl_initialize
at /home/21cl91p01/GCHP/src/MAPL/base/ApplicationSupport.F90:39
#5 0xd8b0bb in __mapl_capmod_MOD_new_mapl_cap
at /home/21cl91p01/GCHP/src/MAPL/gridcomps/Cap/MAPL_Cap.F90:108
#6 0x419b8d in gchpctm_main
at /home/21cl91p01/GCHP/src/GCHPctm.F90:30
#7 0x419508 in main
at /home/21cl91p01/GCHP/src/GCHPctm.F90:14

Please check it out.

make log:
build_log.txt

Make also had a warning :

[ 90%] Built target History
[ 92%] Built target KPP
[ 98%] Built target GeosCore
[ 98%] Built target GCHPctmEnv_GridComp
[100%] Built target GCHP
[100%] Built target GCHP_GridComp
/usr/bin/ld: warning: libhdf5.so.8, needed by /home/21cl91p01/libs/lib/libnetcdf.so, may conflict with libhdf5.so.103
/usr/bin/ld: warning: libhdf5.so.8, needed by /home/21cl91p01/libs/lib/libnetcdf.so, may conflict with libhdf5.so.103

[LiamBindle]

@gopikrishnangs44 Could you clarify what commands you're running and what error message you're getting? If you use GitHub's code formatting, and upload all the necessary log files, it makes it easier for me to understand where you have gotten to.

If I understand correct, you tried

$ ./gchp

which gave the error

WARNING: There was an error initializing an OpenFabrics device.

#Local host: login02
#Local device: mlx5_0
At line 32 of file /home/21cl91p01/GCHP/src/yaFyaml/src/FileStream.F90
Fortran runtime error: Cannot open file 'logging.yml': No such file or directory

Error termination. Backtrace:
#0 0x7f59f437d314 in already_open
at ../../../libgfortran/io/open.c:714
https://github.com/geoschem/GCHP/issues/1 0x2629181 in __fy_filestream_MOD_new_filestream
at /home/21cl91p01/GCHP/src/yaFyaml/src/FileStream.F90:32
https://github.com/geoschem/GCHP/issues/2 0x25cb7d2 in __pfl_loggermanager_MOD_load_file
at /home/21cl91p01/GCHP/src/pFlogger/src/LoggerManager.F90:340
https://github.com/geoschem/GCHP/issues/3 0x16de40e in initialize_pflogger
at /home/21cl91p01/GCHP/src/MAPL/base/ApplicationSupport.F90:108
https://github.com/geoschem/GCHP/issues/4 0x16e2316 in __mapl_applicationsupport_MOD_mapl_initialize
at /home/21cl91p01/GCHP/src/MAPL/base/ApplicationSupport.F90:39
https://github.com/geoschem/GCHP/pull/5 0xd8b0bb in __mapl_capmod_MOD_new_mapl_cap
at /home/21cl91p01/GCHP/src/MAPL/gridcomps/Cap/MAPL_Cap.F90:108
https://github.com/geoschem/GCHP/pull/6 0x419b8d in gchpctm_main
at /home/21cl91p01/GCHP/src/GCHPctm.F90:30
https://github.com/geoschem/GCHP/pull/7 0x419508 in main
at /home/21cl91p01/GCHP/src/GCHPctm.F90:14

Is that right?

---

If so, you need to run the GCHP executable in a run directory. A run directory has the necessary configuration files, including logging.yml. Please see Creating a Run Directory and Running GCHP.

[gopikrishnangs44]

Dear @LiamBindle

Our HPC is down for maintenance at the moment. Will make sure to follow up in 2 days.

I have done exactly what you wrote in the previous comment. But I ran the executable in the build directory. I will copy the executable to a run directory and will let you know asap. It will be of great help if you let me know what all log files should I upload if there is any more error.

Thanks for the timely response

[gopikrishnangs44]

Dear @LiamBindle

I have made the run directory. I have done

cmake . -DRUNDIR=/home/21cl91p01/gchp_merra2_fullchem
make install
mpirun -np 6 ./gchp &> run_log.txt

Please find the run log. There is some MPI issues.
run_log.txt

Thanks.

[LiamBindle]

@gopikrishnangs44 It looks like this is a configuration error. Did you run ./runConfig.sh? Please see Creating a Run Directory and Running GCHP and our YouTube tutorial on running GCHP.

[gopikrishnangs44]

Dear @LiamBindle,

Thank you for the suggestion.
I had followed your youtube tutorial. But still there is error in running the model. runConfig.sh was OK.

Attaching the run log.
log.run1.txt

[lizziel]

Hi @gopikrishnangs44, the error in the run log is FATAL: mpp_domains_define.inc: not all the pe_end are in the pelist which indicates a mismatch between number of cores available and number of cores configured to be used in GCHP.

The top of your log file has this:

          SHMEM: INFO: NumCores per Node = 1
          SHMEM: INFO: NumNodes in use   = 1
          SHMEM: INFO: Total PEs         = 1
          SHMEM: INFO: NumNodes in use  = 1

This indicates you are using 1 core. You need to use at least 6 cores to run GCHP. Did you remember -np 6 in your run command? Your previous run log correctly used 6 cores per node.

[gopikrishnangs44]

Dear @lizziel ,
Thanks for the help. Still I am facing issues. Please look into these files.

This is the SBATCH script
jobsub.sh.txt

Output log
slurm-731105.txt

[stale]

This issue has been automatically marked as stale because it has not had recent activity. If there are no updates within 7 days it will be closed. You can add the "never stale" tag to prevent the Stale bot from closing this issue.

[stale]

Closing due to inactivity