[flang-commits] [mlir] [openmp] [flang] [Flang][OpenMP] Initial mapping of Fortran pointers and allocatables for target devices (PR #71766)

[via flang-commits]

agozillon wrote:

> > > If anyone can recommend a good test that'd show-case/test the addendum mapping I'd be much appreciated, I'm not sure of the Fortran cases where addendums would be generated and then used within a target region.
> > 
> > 
> > I guess unlimited polymorphic types will be one case. But not sure whether the OpenMP spec fully supports it for the map clause.
> 
> From what I can see it doesn't seem to be expressly covered or disallowed, I can ask internally however.

It is in-fact unspecified at the moment, so up to us what we do with it for now. Perhaps there's other addendum cases? I'm not against adding the unlimited polymorphic case though in a future patch if there's need/want for it 

https://github.com/llvm/llvm-project/pull/71766