Skip to content

Is this warning about command_argument_count() still relevant #263

@tclune

Description

@tclune
CMake Warning at @cmake/@ecbuild/cmake/ecbuild_log.cmake:158 (message):
  WARN - Found Intel oneAPI on macOS.

   MAPL developers have found an issue with Intel oneAPI on macOS
   where GEOSgcm.x would not work. Debugging found the issue was
   that command_argument_count() would return -1 which should *NEVER*
   happen per Fortran Standard and then this broke FLAP.
   A workaround was found that if the MAPL_cfio_r8 library was compiled
   as TYPE STATIC, the model would work. So we are setting MAPL_cfio_r8 as
   a TYPE STATIC library. Note: This might interfere with coupled model.

Metadata

Metadata

Assignees

Labels

questionFurther information is requested

Type

No type

Projects

No projects

Milestone

No milestone

Relationships

None yet

Development

No branches or pull requests

Issue actions