---1247997369-209215680-908844973=:24868
Content-Type: TEXT/PLAIN; CHARSET=US-ASCII
Content-ID: <Pine.LNX.3.96.981020095427.20796G@mirkwood.dummy.home>
On Tue, 20 Oct 1998, MOLNAR Ingo wrote:
> On Mon, 19 Oct 1998, Rik van Riel wrote:
>
> > - if you run more SCHED_IDLE processes, only the first
> > one will get any CPU time assigned
>
> nope they will eat up their timeslice and the next one comes, just
> like RT processes do. This makes sense from the cache-usage point of
> view, SCHED_IDLE processes obviosly dont have any inteactivity
> needs, so the longer the timeslice, the better the overall system
> performance. Think RC5 cracker.
Have you actually verified that? I don't see any
constructions to do a move_first/last runqueue
or to return dynamic priorities from goodness()...
With your patch, goodness() will return -999 on
every SCHED_IDLE patch; this means that the first
SCHED_IDLE patch on the runqueue will be selected.
Rik.
+-------------------------------------------------------------------+
| Linux memory management tour guide. H.H.vanRiel@phys.uu.nl |
| Scouting Vries cubscout leader. http://www.phys.uu.nl/~riel/ |
+-------------------------------------------------------------------+
---1247997369-209215680-908844973=:24868--
-
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.tux.org/lkml/