Home

YAKL (Yet Another Kernel Launcher)

A Simple C++ Framework for Performance Portability and Fortran Code Porting

Author: Matt Norman (Oak Ridge National Laboratory) - mrnorman.github.io

Contributors

• Matt Norman (Oak Ridge National Laboratory)
• Abhishek Bagusetty (Argonne National Laboratory)
• Isaac Lyngaas (Oak Ridge National Laboratory)
• Sarat Sreepathi (Oak Ridge National Laboratory)
• Mark Berrill (Oak Ridge National Laboratory)
• Pierre Kestener (Institut de Recherche sur les lois Fondamentales de l'Univers)

Full YAKL API documentation

Known issues

CDash: https://my.cdash.org/index.php?project=YAKL

Example YAKL Usage

For a self-contained example of how to use YAKL, please checkout the cpp/ folder of the miniWeather repo

• https://github.com/mrnorman/miniWeather
• Particularly miniWeather/cpp/miniWeather_mpi_parallelfor.cpp

YAKL Documentation:

• Introduction and Overview
• Using and Compiling with YAKL
• Common Issues and Errors
• parallel_for, Bounds, C-style and Fortran-style, YAKL_INLINE, YAKL_LAMBDA, fence()
• Array Multi-Dimensional Dynamically-Allocated Arrays, C-style and Fortran-style, Shallow and Deep Copy, Deallocation, Reshaping, Host / Device Management, Slicing, const Data Types, Index Checking
• SArray and FSArray Statically-sized Arrays, C-style and Fortran-style,
• Parallel Reductions (minval, maxval, sum, and product)
• Atomics (atomicAdd, atomicMin, atomicMax)
• Using Multiple GPUs with YAKL
• Scalar Live-Out
• YAKL's Fortran Intrinsics Library
• YAKL_INLINE and Calling Class Methods from Kernels
• YAKL_SCOPE and lambda captures from non-local scope
• const Correctness and intent(in/out/inout)
• Fortran - C++ Interoperability with YAKL
• Gator: YAKL's pool allocator
• YAKL Timers
• Using Multiple Streams in YAKL
• Thread-Private Data
• YAKL FFTs
• Asynchronicity in YAKL
• Floating point determinism / bitwise reproducibility with YAKL
• Hierarchical Parallelism
• Improving Readability with typedef, using, and auto
• The ArrayIR class and YAKL Interoperation with Kokkos and other portability libraries
• Moving Data Between Host and Device Memory
• Can I Use Runtime-Sized Arrays on the Stack Inside Kernels?
• YAKL Debugging Tips
• How To Investigate a Suspected Device Memory Leak
• More Information About YAKL's Multi-Dimensional Array Classes
• How Do I Hide Host Code from the Device Compiler Pass?

Other Articles

• C++ Portability for OpenACC and OpenMP Folks