mca_coll_cuda_reduce   48 ompi/mca/coll/cuda/coll_cuda.h int mca_coll_cuda_reduce(const void *sbuf, void *rbuf, int count,
mca_coll_cuda_reduce  104 ompi/mca/coll/cuda/coll_cuda_module.c     cuda_module->super.coll_reduce     = mca_coll_cuda_reduce;