I hate to follow up to my own mail, but there are some other
ideas that occurred to me:
- we want to change schedule_idle() to immediately replace a
SCHED_IDLE process when something else is runnable
- we want _long_ timeslices to maximize speed and to minimize
impact on cache/memory/etc. (say 1 minute)
- [for 2.3.x] we want a 2-level scheduling mechanism with CTSS-like
timeslice adaptation, a limited number of SCHED_IDLE tasks on the
runqueue and some other advanced features like a (small) secondary
idle task which only gets run when the first one takes a fault, we
will then remove the SCHED_RR feature too.
I have some ideas on how to do this in under 1 kB of code, so
I guess it's best that I volunteer for the job :-)
Don't expect it the first weeks however, since I'm super-busy
with my summer-job and I'm also working (in my rare free time)
on the out-of-memory program killer (so that the system continues
to run and kills the major suspect instead of a random process).
It should be ready in early September however...
Rik.
+-------------------------------------------------------------------+
| Linux memory management tour guide. H.H.vanRiel@phys.uu.nl |
| Scouting Vries cubscout leader. http://www.phys.uu.nl/~riel/ |
+-------------------------------------------------------------------+
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.altern.org/andrebalsa/doc/lkml-faq.html