Bug in interface of MAPL_BcastShared procedures?

[bena-nasa]

I was looking at something in MAPL_CommsMod.F90 and noticed by chance that there are all these MAPL_BcastShared procedures that have this interface:

  subroutine MAPL_BcastShared_2DI4(VM, Data, N, Root, RootOnly, rc)
    type(ESMF_VM) :: VM
    integer, pointer,  intent(INOUT) :: Data(:,:)
    integer,           intent(IN   ) :: N
    integer, optional, intent(IN   ) :: Root
    logical,           intent(IN   ) :: RootOnly
    integer, optional, intent(  OUT) :: rc

Seems weird that the optional argument argument Root is before RootOnly, figured I would make an issue and ask our Fortran guru. Should we fix this?

[mathomp4]

Ben showed me this and part of me thinks we should just remove the optional from Root. It's never looked at with a present() in the code so it seems non-optional!

[tclune]

Agree with @mathomp4's suggestion. But am also wondering what "RootOnly" means in this context. I mean, it's a bcast!

[mathomp4]

Agree with @mathomp4's suggestion. But am also wondering what "RootOnly" means in this context. I mean, it's a bcast!

I have no idea. It looks like this code (at least the 1d r4) predates Git. Maybe @atrayano remembers what this is for?