Towards a python.util.concurrent?

Maybe it’s the wrong place to ask, but will _Py_TRY_INCREF be public? It’s very crucial to implement correct concurrent reference counting.