perf_cnt: add the tracing of window-based average calculation

The average number of cpu cycles/ticks spent could be provided per 1024
scheduled tasks by the performance counter. It would be more accurate for
MCPS measurement than the peak number. Even so, the tracing macro will log
both of them to provide more insights.

This is enabled by CONFIG_PERFORMANCE_COUNTERS_RUN_AVERAGE, which depends
on CONFIG_PERFORMANCE_COUNTERS.

Signed-off-by: Pin-chih Lin <johnylin@google.com>
6 files changed
tree: d071974bd1ff94e4d3a81b578fc8033aac7b6164
  1. .github/
  2. doc/
  3. installer/
  4. keys/
  5. overlays/
  6. scripts/
  7. smex/
  8. src/
  9. test/
  10. third_party/
  11. tools/
  12. zephyr/
  13. .gitignore
  14. .gitmodules
  15. .travis.yml
  16. CMakeLists.txt
  17. CODEOWNERS
  18. Kconfig
  19. Kconfig.sof
  20. Kconfig.xtos-build
  21. Kconfig.xtos-dbg
  22. Kconfig.zephyr-log
  23. LICENCE
  24. README.md
README.md

Sound Open Firmware

Status

GitHub Actions Build Status Gitter chat

Documentation

See docs

Running the tests

See unit testing documentation

Deployment

TODO: Add additional notes about how to deploy this on a live system

Contributing

See Contributing to the Project

License

This project is licensed under the BSD Clause 3 - see the LICENCE file for details