Closed xylar closed 8 months ago
So far, I have tested this on Omega in conjunction with #163 on Perlmutter-CPU. I ran:
./configure_polaris_envs.py --conda ~/miniforge3 --env_name polaris_test --update_spack --verbose --recreate
to create the conda and spack environments along with the load script. Then, I ran:
cd e3sm_submodules
git submodule update --init Omega
cd Omega
git submodule update --init --recursive externals/YAKL externals/ekat externals/scorpio cime
cd components/omega
mkdir build_${POLARIS_MACHINE}_${POLARIS_COMPILER}
cd build_${POLARIS_MACHINE}_${POLARIS_COMPILER}
cmake \
-DOMEGA_CIME_COMPILER=${POLARIS_COMPILER} \
-DOMEGA_CIME_MACHINE=${POLARIS_MACHINE} \
-DPARMETIS_ROOT=${PARMETIS_ROOT} \
-DMETIS_ROOT=${METIS_ROOT} \
-DOMEGA_BUILD_TEST=ON \
-S .. -B .
./omega_build.sh
ln -s /global/cfs/cdirs/e3sm/inputdata/ocn/mpas-o/oQU240/ocean.QU.240km.151209.nc test/OmegaMesh.nc
salloc --nodes 1 --qos interactive --time 01:00:00 --constraint cpu --account=e3sm
./omega_ctest.sh
The result were successful tests:
Test project /global/homes/x/xylar/e3sm_work/polaris/update-omega/e3sm_submodules/Omega/components/omega/build_pm-cpu_gnu
Start 1: DATA_TYPES_TEST
1/6 Test #1: DATA_TYPES_TEST .................. Passed 2.84 sec
Start 2: MACHINE_ENV_TEST
2/6 Test #2: MACHINE_ENV_TEST ................. Passed 33.45 sec
Start 3: BROADCAST_TEST
3/6 Test #3: BROADCAST_TEST ................... Passed 66.38 sec
Start 4: LOGGING_TEST
4/6 Test #4: LOGGING_TEST ..................... Passed 0.04 sec
Start 5: DECOMP_TEST
5/6 Test #5: DECOMP_TEST ...................... Passed 74.82 sec
Start 6: YAKL_TEST
6/6 Test #6: YAKL_TEST ........................ Passed 13.33 sec
100% tests passed, 0 tests failed out of 6
Total Test time (real) = 190.94 sec
I'm going to hold off on making this update until I add some more supported machines and compilers.
Omega needs parmetis built with system compilers.
Checklist
Testing
comment in the PR documents testing used to verify the changes