[email protected] testing - undefined references to maximize_system_stacksize_limit and mpp_init

[climbfuji]

Describe the bug
On my mac I am hitting the following problem:

Undefined symbols for architecture x86_64:
  "_maximize_system_stacksize_limit", referenced from:
      ___fms_mod_MOD_check_nml_error in libfms_r8.a(fms.F90.o)
      ___fms_mod_MOD_fms_init.part.0 in libfms_r8.a(fms.F90.o)
      ___fms_mod_MOD_write_version_number in libfms_r8.a(fms.F90.o)
      ___fms_mod_MOD_error_mesg in libfms_r8.a(fms.F90.o)
      ___fms_mod_MOD_fms_init in libfms_r8.a(fms.F90.o)
  "_mpp_init_", referenced from:
      ___module_wrt_grid_comp_MOD_wrt_initialize_p1 in libfv3atm.a(module_wrt_grid_comp.F90.o)
ld: symbol(s) not found for architecture x86_64
collect2: error: ld returned 1 exit status

I already had to add use mpp_mod and use fms_mod before the use fms statement in module_wrt_grid_comp.F90, but it seems that the C function maximize_system_stacksize_limit isn't defined anywhere in the installed library:

> grep -Rie maximize_system_stacksize_limit envs/fms-beta-chained-clang/install/apple-clang/13.1.6/fms-2023.02-beta1-q26chym
Binary file envs/fms-beta-chained-clang/install/apple-clang/13.1.6/fms-2023.02-beta1-q26chym/lib/libfms_r4.a matches
Binary file envs/fms-beta-chained-clang/install/apple-clang/13.1.6/fms-2023.02-beta1-q26chym/lib/libfms_r8.a matches

> nm envs/fms-beta-chained-clang/install/apple-clang/13.1.6/fms-2023.02-beta1-q26chym/lib/libfms_r4.a |grep maximize
envs/fms-beta-chained-clang/install/apple-clang/13.1.6/fms-2023.02-beta1-q26chym/lib/libfms_r4.a:yaml_parser_binding.c.o: no symbols
envs/fms-beta-chained-clang/install/apple-clang/13.1.6/fms-2023.02-beta1-q26chym/lib/libfms_r4.a:yaml_output_functions.c.o: no symbols
envs/fms-beta-chained-clang/install/apple-clang/13.1.6/fms-2023.02-beta1-q26chym/lib/libfms_r4.a:constants.F90.o: no symbols
envs/fms-beta-chained-clang/install/apple-clang/13.1.6/fms-2023.02-beta1-q26chym/lib/libfms_r4.a:constantsr4.F90.o: no symbols
envs/fms-beta-chained-clang/install/apple-clang/13.1.6/fms-2023.02-beta1-q26chym/lib/libfms_r4.a:fm_yaml.F90.o: no symbols
                 U _maximize_system_stacksize_limit
envs/fms-beta-chained-clang/install/apple-clang/13.1.6/fms-2023.02-beta1-q26chym/lib/libfms_r4.a:yaml_parser.F90.o: no symbols
envs/fms-beta-chained-clang/install/apple-clang/13.1.6/fms-2023.02-beta1-q26chym/lib/libfms_r4.a:fms_yaml_output.F90.o: no symbols
envs/fms-beta-chained-clang/install/apple-clang/13.1.6/fms-2023.02-beta1-q26chym/lib/libfms_r4.a:platform.F90.o: no symbols

(and same for libfms_r8.a).

To Reproduce
Try to build current ufs-weather-model with [email protected] - note that you need to enable the legacy fms_io bindings.

Expected behavior
No undefined references in final linker step

System Environment

• OS: macOS Mojave
• Compiler(s): [email protected]
• MPI type, and version (e.g. MPICH, Cray MPI, openMPI): [email protected]
• netCDF Version: netCDF 4.9.2, netCDF-Fortran 4.6.0
• Configure options: all from spack-stack

'-DGFS_PHYS:BOOL=ON'
'-DOPENMP:BOOL=ON'
'-DENABLE_QUAD_PRECISION:BOOL=ON'
'-DWITH_YAML:BOOL=OFF'
'-DCONSTANTS:STRING=GFS'
'-DFPIC:BOOL=ON'
'-D32BIT:BOOL=ON'
'-D64BIT:BOOL=ON'
'-DUSE_DEPRECATED_IO:BOOL=ON'
'-DCMAKE_C_COMPILER:STRING=/Users/heinzell/prod/spack-stack-1.4.1/envs/unified-env/install/apple-clang/13.1.6/openmpi-4.1.5-j7pjg6h/bin/mpicc'
'-DCMAKE_CXX_COMPILER:STRING=/Users/heinzell/prod/spack-stack-1.4.1/envs/unified-env/install/apple-clang/13.1.6/openmpi-4.1.5-j7pjg6h/bin/mpic++'
'-DCMAKE_Fortran_COMPILER:STRING=/Users/heinzell/prod/spack-stack-1.4.1/envs/unified-env/install/apple-clang/13.1.6/openmpi-4.1.5-j7pjg6h/bin/mpif90'
'-DCMAKE_Fortran_FLAGS:STRING=-fallow-argument-mismatch'

Additional context
Add any other context about the problem. If applicable, include where any files
that help describe, or reproduce the problem exist.

[bensonr]

@climbfuji - I suspect this is an issue with the newly created fms/fms_stacksize.c file not existing in the CMakeLists.txt fms_c_src_files list. Also, we have not yet tested the functionality on a Mac - still trying to get a working autotools environment for our tests. So if you do encounter issues with any of the logic in this file, please keep us updated via this issue.