Process input queue - trickle_fast_inputs = false : process all events, turn into flattened input state (e.g. successive down/up/down/up will be lost) - trickle_fast_inputs = true : process as many events as possible (successive down/up/down/up will be trickled over several frames so nothing is lost) (new feature in 1.87)
source not stored for this graph (policy: none)
nothing calls this directly
no test coverage detected