Messages in this thread | | | Date | Thu, 21 Jan 1999 15:02:50 GMT | From | "Stephen C. Tweedie" <> | Subject | Re: Scheduler problems |
| |
Hi,
On Thu, 21 Jan 1999 12:05:11 +0100 (CET), Rik van Riel <riel@nl.linux.org> said:
> The biggest problem is that p->counter is essentially a > random value because of the gross recalculation done in > the schedule() function.
Doesn't matter: it's the only usable indication of recent behaviour that we have, and in any case, the goodness is tuned so that CPU bound process always have goodness lower than interactive processes of the same priority, regardless of the current counter.
Using the last known goodness on that CPU is still good enough.
--Stephen
- 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/
| |