On 2019-08-29 13:55:27 [-0700], Austin Schuh wrote: > > This is still RT, but is pretty inefficient. Any suggestions on what > might be going wrong before I start digging in? there is a pi-boost sched_switch with state R->D and then de-boost switch back. This looks like there is a wakeup while a lock is held which is required by the task that is woken up. > Thanks, > Austin Sebastian
