UNIX for Advanced & Expert Users

Hi, I'm trying to get a grip of the SCHED_SOPRADIC (ss) algorithm. I use the OpenSolaris operating system. The way I understand it is(specs: http://www.unix.org/single_unix_specification/):

1. Is it like:

The ss is a thread. That is, the ss is "in charge" of a certain running queue with a certain priority. The ss contains the sched_param struct. The ss is in the ready queue it is "in charge" of. When the ss gets to run it schedules all threads in its queue, according to the rules of the algorithm, then yield the cpu.

2. Or like:

The ss is not a thread. The ss is and extension to the dispatcher / real time class / libc that apply its rules at given times, for ex one can set of timer interrupts and let the scheduler update budget and such after every interrupt on threads in the class. All threads contains the sched_param struct?


If anyone could get this straight for me it would be appreciated.
Thanks.

SS is certainly not an additional thread. It is one of several scheduling policies that affect the behavior of the scheduler. The scheduler is part of the kernel and you may be using the term dispatcher instead. Which term you use is not important as long as you understand that the scheduler adjusts the priorities of the threads as well as finding one to run in an available cpu. See this link: The QNX Neutrino Microkernel which has the best description I have seen on sporadic scheduling.

I have never used it so i can't add too much. All threads contain the sched_param struct, but the elements pertaining to sporadic scheduling are ignored if sporadic scheduling is not in use.