Skip to content

Conversation

@Fidget-Spinner
Copy link
Member

@Fidget-Spinner Fidget-Spinner commented Sep 18, 2025

There is no taking up of an additional register on the computed goto build in _PyEval_EvalFrameDefault as the instruction dispatch table is already kept in a register on x86_64.

There is no regression in performance on both linux x86_64 and macOS AArch64 for tailcalling builds https://github.com/facebookexperimental/free-threading-benchmarking/tree/main/results/bm-20250917-3.15.0a0-178b03e-CLANG

Copy link
Member

@markshannon markshannon left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Looks good.

@Fidget-Spinner Fidget-Spinner merged commit a269e69 into python:main Sep 18, 2025
61 checks passed
@Fidget-Spinner Fidget-Spinner deleted the dynamic_opcode_targets branch September 18, 2025 13:12
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants