Hi,
strictly speaking not a linux-rt specific questions but I think
that the folks here have the needed experience and knowledge:
what is the relation of
pthread_mutexattr_setprotocol(&attr, PTHREAD_PRIO_INHERIT);
and scheduler class used for a particular thread? Does the priority
inheritance work across all threads in the system or is it limited
to one scheduling class?
Specifically: my application has some SCHED_RR threads with fixed
priorities and realtime needs and a few that are quite low-prio,
actually have the SCHED_OTHER semantics and I don't want them
to count against the time spent in RT (keyword: throttler).
I need to occasionally queue data from from the RT-ones to
the low-prio ones, so for short periods of time I need to lock
a mutex. With it comes the question of priority inversion.
Can I leave the low-prio threads at SCHED_OTHER and rely on
the PTHREAD_PRIO_INHERIT to boost their priority in case they
were preempted with the lock held and the RT-thread needs
to lock it too, or do they have to be SCHED_RR too?
Thanks
--
Stano
--
To unsubscribe from this list: send the line "unsubscribe linux-rt-users" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
