On 04.09.23 15:52, Wu, Wentong wrote:
From: Oliver Neukum <oneukum@xxxxxxxx>
And we execute this. rx_urb is killed and does nothing.
I see nothing that terminates the waiting if you hit the wrong window.
I think the auxiliary_device_delete will trigger the remove function of ljca
client in his own sub system, and the delete() or remove() of every subsystem
will not return until the ongoing transfer(probably blocked by above
cmd_completion) complete. And that also makes sure no more transfers
comes to there.
Sure, you will not free used memory. But what allows you to be sure that you
make any progress at all? That is that you will hang arbitrarily long in
disconnect?
Regards
Oliver
