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

[via flang-commits]

agozillon wrote:

> > Thank you @razvanlupusoru for the input. I don't believe the offload runtime currently accepts non-contiguous data, but it's good to see that this case has been thought about thoroughly.
> 
> I think https://github.com/llvm/llvm-project/blob/ebb8ffde94476333cb2e95aacfd2e023d7112efb/llvm/include/llvm/Frontend/OpenMP/OMPIRBuilder.h#L1663C37-L1663C37 allows mapping of non-contiguous data.

Ah, I stand corrected then! I do remember getting complaints about non-contiguous array data when trying to create an N-D map example in C++, but that must have been unrelated 

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