On 22/11/2023 14.01, Frieder Schrempf wrote: > On 22.11.23 09:03, Rasmus Villemoes wrote: >> On 21/11/2023 21.49, Eberhard Stoll wrote: >>> But IMHO there is one next issue with this situation: When the port >>> operates with WAIT_AFTER_RTS and WAIT_AFTER_SEND then some timers >>> for callback functions might be active. I did not discover where they >>> are stopped for the case when the serial port is closed. Maybe stopping >>> is not required ... >> >> Indeed, that's an extra complication. Adding two hrtimer_try_to_cancel() >> in shutdown would probably not hurt, along with setting tx_state OFF. >> >> I wonder if at least mode switching should simply be disallowed (-EBUSY) >> if tx_state is anything but OFF. > > Is there a valid use-case for switching the mode while the device is > transmitting? Is this something we need to support for whatever reason? > It sounds rather an obscure thing to do. No, I don't think there is, and that's not at all what I'm trying to do. I was just thinking out loud that maybe we should explicitly disallow it [further, in fact, if possible, it should be disallowed in the serial core]. But of course that requires that ->tx_state can actually be relied upon, which is what my patch attempts to do for the rs232 case - though it might not be an entirely complete fix, as described by Eberhard, since we can get to ->shutdown and thus disable the TC interrupt before it has a chance to fire. Rasmus
