Messages in this thread | | | Date | Fri, 18 Sep 1998 08:15:40 +0200 (CEST) | From | Rik van Riel <> | Subject | Re: Interesting scheduling times |
| |
On Thu, 17 Sep 1998, Linus Torvalds wrote: > On Fri, 18 Sep 1998, Richard Gooch wrote: > > > > Linus: what do you think of this idea? A valid project for 2.3? > > A single run-queue is almost always better than multiple run-queues, and > I'm very unlikely to change that.
Not necessarily. In a QNX-sched style scheduler, we may have to scan multiple queues, but we don't have to calculate process priority, as we need to do now.
It's just a different way of spending your cycles, but it doesn't mean you have to do more. I am currently rewriting the specs for the QNX-style scheduler (on request, for the Alliance OS) and the docs will be ready in about a week. After that, pseudo-code (for alliance) and maybe some real code (when school doesn't take up too much time) for Linux will emerge.
> Even under heavy load, the runqueue is seldom more than a few entries > deep. More than 10 entries on the run-queue is already very rare, and
This is a good point. Maybe the code should be semantically QNX-like but with a different implementation? I'll think it over...
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.tux.org/lkml/
| |