Open jeffhammond opened 1 year ago
I think I can do this by calling MPI_Reduce_local one element at a time. The performance will be trash but nobody should use this anyways.
MPI_Reduce_local
The blocking case is feasible. The nonblocking case requires crazy stuff, maybe generalized requests.
I think I can do this by calling
MPI_Reduce_local
one element at a time. The performance will be trash but nobody should use this anyways.