cyPAPI

A Python API which provides a convenient interface to access PAPI profiling features with minimum overhead. Currently cyPAPI supports PAPI version >= 7.0.0.

What is PAPI?

The Performance Application Programming Interface (PAPI) supplies a consistent interface and methodology for collecting performance counter information from various hardware and software components, including most major CPUs, GPUs, accelerators, interconnects, I/O systems, and power interfaces, as well as virtual cloud environments.

Installation

To use cyPAPI:

Clone the cyPAPI repository with the following command:

git clone https://github.com/icl-utk-edu/cyPAPI.git

This will create a complete copy of the cyPAPI git respoitory in a folder titled cyPAPI.

Once cyPAPI has been successfully cloned, execute the commands below to install from source:

cd cyPAPI
make install

Installation Requirements

• PAPI library

  • cyPAPI takes advantage of existing PAPI functionality such as timers, eventsets, and software defined events. If PAPI is not installed see the following link.

  Note: cyPAPI will check the following locations for PAPI:

  • First checks if the user has set the PAPI_PATH environment variable which points to the location of your PAPI installation
  • Then it checks if pkg-config utility can locate the PAPI installation
  • Then it relies on C_INCLUDE_PATHS & LIBRARY_PATHS to be used by the compiler
  • It also looks for the PAPI library in LD_LIBRARY_PATH
  • Lastly it expects that PAPI is installed in default OS installation paths

• Cython

  • Allows access to PAPI profiling features by the support of calling C functions and gives the ease of Python with the speed of native code.

• C compiler, e.g. gcc

  • Cython requires a C compiler to be present on the system such that .pyx files can be compiled into C/C++ files respectively. Afterwards the C/C++ files will then be compiled into an extension module which is then directly importable from Python.

• Numpy

  • Internal functions utilize numpy.ndarrays. Due to this the Numpy library is a requirement such that installation will be successful.