Python 3.12 Performance Profiling with Linux perf - New JIT Compiler and Production Profiling

Discover the groundbreaking features of Python 3.12 in this PyCon US talk by Pablo Galindo Salgado. Explore the integration of one of the world's smallest just-in-time (JIT) compilers and the powerful Linux perf profiler. Learn how this implementation enhances performance analysis capabilities, enabling dynamic production profiling and providing deeper insights into application behavior. Understand the differences between perf and other Python profilers, and gain practical knowledge on effectively utilizing this tool. Delve into the requirements for optimal results, potential limitations, and their impact on metrics. Uncover the importance of cross-correlating performance information with various markers such as branch mispredictions, cache misses, and context switches to solve complex profiling challenges. Gain valuable insights into improving the performance characteristics of Python applications, especially those involving native code written in C, C++, or Rust.