[RFC,0/7] NumPy Interface for KernelShark

Message

Yordan Karadzhov March 27, 2019, 4:03 p.m. UTC

NumPy is an efficient multi-dimensional container of generic data.
It uses strong typing in order to provide fast data processing in
Python. The NumPy interface will allow sophisticated analysis of
tracing data via scripts, but it also opens the door for
exposing the kernel tracing data to the instruments provided by
the scientific toolkit of Python (matplotlib, scikit-learn) or
maybe even PyTorch and TensorFlow in the future.

Disclaimer: I am not very good in Python. Please check as carefully
as possible :-)

Yordan Karadzhov (7):
  kernel-shark: kshark_string_config_alloc() must take no arguments
  kernel-shark: Add new dataloading method to be used by the NumPu
    interface
  kernel-shark: Prepare for building the NumPy interface
  kernel-shark: Add the core components of the NumPy API
  kernel-shark: Add Numpy Interface for processing of tracing data
  kernel-shark: Add automatic building of the NumPy interface
  kernel-shark: Add basic example demonstrating the NumPy interface

 kernel-shark/CMakeLists.txt                 |   3 +
 kernel-shark/README                         |  12 +-
 kernel-shark/bin/sched_wakeup.py            |  96 +++++++
 kernel-shark/build/py/libkshark_wrapper.pyx | 264 ++++++++++++++++++++
 kernel-shark/build/py/np_setup.py           |  87 +++++++
 kernel-shark/build/py/pybuild.sh            |  26 ++
 kernel-shark/src/CMakeLists.txt             |  39 +++
 kernel-shark/src/libkshark-configio.c       |   2 +-
 kernel-shark/src/libkshark-py.c             | 176 +++++++++++++
 kernel-shark/src/libkshark.c                | 128 ++++++++++
 kernel-shark/src/libkshark.h                |   9 +-
 11 files changed, 838 insertions(+), 4 deletions(-)
 create mode 100755 kernel-shark/bin/sched_wakeup.py
 create mode 100644 kernel-shark/build/py/libkshark_wrapper.pyx
 create mode 100644 kernel-shark/build/py/np_setup.py
 create mode 100755 kernel-shark/build/py/pybuild.sh
 create mode 100644 kernel-shark/src/libkshark-py.c