Currently future_into_py() spawns task to rust runtime and do Python::with_gil(... within it. It may block the current executor thread if the python runtime is highly utilized or some long-running python code is occupying the lock. Would this be a room for improvement, or a known non-issue?
Currently
future_into_py()spawns task to rust runtime and doPython::with_gil(...within it. It may block the current executor thread if the python runtime is highly utilized or some long-running python code is occupying the lock. Would this be a room for improvement, or a known non-issue?