| 
| template<class F , int N, bool simple, int VecLen = YAKL_DEFAULT_VECTOR_LEN, bool B4B = false>  | 
| void  | yakl::fortran::parallel_for (Bounds< N, simple > const &bounds, F const &f, LaunchConfig< VecLen, B4B > config=LaunchConfig<>()) | 
|   | [ASYNCHRONOUS] Launch the passed functor in parallel.  More...
  | 
|   | 
| template<class F , int N, bool simple, int VecLen = YAKL_DEFAULT_VECTOR_LEN, bool B4B = false>  | 
| void  | yakl::fortran::parallel_for (char const *str, Bounds< N, simple > const &bounds, F const &f, LaunchConfig< VecLen, B4B > config=LaunchConfig<>()) | 
|   | [ASYNCHRONOUS] Launch the passed functor in parallel.  More...
  | 
|   | 
| template<class F , int N, bool simple>  | 
| YAKL_INLINE void  | yakl::fortran::parallel_inner (Bounds< N, simple > const &bounds, F const &f, InnerHandler handler) | 
|   | Launch the passed functor in parallel in the finenst-level parallelism on the device.  More...
  | 
|   | 
| template<class F , int N, bool simple, int VecLen = YAKL_DEFAULT_VECTOR_LEN, bool B4B = false>  | 
| void  | yakl::fortran::parallel_outer (Bounds< N, simple > const &bounds, F const &f, LaunchConfig< VecLen, B4B > config=LaunchConfig<>()) | 
|   | [ASYNCHRONOUS] Launch the passed functor in parallel in the coarsest-level parallelism on the device  More...
  | 
|   | 
| template<class F , int N, bool simple, int VecLen = YAKL_DEFAULT_VECTOR_LEN, bool B4B = false>  | 
| void  | yakl::fortran::parallel_outer (char const *str, Bounds< N, simple > const &bounds, F const &f, LaunchConfig< VecLen, B4B > config=LaunchConfig<>()) | 
|   | [ASYNCHRONOUS] Launch the passed functor in parallel in the coarsest-level parallelism on the device  More...
  | 
|   | 
| template<class F >  | 
| YAKL_INLINE void  | yakl::fortran::single_inner (F const &f, InnerHandler handler) | 
|   | Launch the passed functor to only use one of the inner threads (still parallel over outer threads).  More...
  | 
|   |