A deep dive into the performance of Python 3.14's tail-call interpreter: How the performance results were confounded by an LLVM regression, the surprising complexity of compiling interpreter loops, and some reflections on performance work, software engineering, and optimizing compilers.
