While PEP 684 is not accepted yet (and could get rejected, of course), it would be helpful to have an issue against which I can start creating draft PRs for the implementation.

This issue isn't for all the prerequisite work for a per-interpreter GIL, rather just the implementation of what's actually specified in the PEP. The rest is happening separately.

(Also see my checklist: https://github.com/ericsnowcurrently/multi-core-python/wiki/0-The-Plan)

• PR: gh-99113: A Per-Interpreter GIL! #99114

• PR: gh-99113: Share the GIL via PyInterpreterState.ceval.gil #104203

• PR: gh-99113: Add PyInterpreterConfig.own_gil #104204

• PR: gh-99113: Add Py_MOD_PER_INTERPRETER_GIL_SUPPORTED #104205

• PR: gh-99113: Add a check for Py_MOD_PER_INTERPRETER_GIL_SUPPORTED #104206

• PR: gh-99113: Make Sure the GIL is Acquired at the Right Places #104208

• PR: gh-99113: A Per-Interpreter GIL! #104209

• PR: gh-99113: A Per-Interpreter GIL! #104210

• PR: gh-99113: Add a What's New Entry for PEP 684 #107520

• PR: [3.12] gh-99113: Add a What's New Entry for PEP 684 (gh-107520) #107523

• PR: [3.12] gh-99113: restore behaviour of PyThreadState_Swap #123079