Re: [announce] [patch] ultra-scalable O(1) SMP and UP scheduler

[Davide Libenzi <davidel _at_ xmailserver.org>]

On Sat, 5 Jan 2002, Linus Torvalds wrote:

>
> On Sun, 6 Jan 2002, Ingo Molnar wrote:
> >
> > (if Davide's post gave you the impression that my patch doesnt do per-CPU
> > queues then i'd like to point out that it does so. Per-CPU runqueues are a
> > must for good performance on 8-way boxes and bigger. It's just the actual
> > implementation of the 'per CPU queue' that is O(1).)
>
> Ahh, ok. No worries then.
>
> At that point I don't think O(1) matters all that much, but it certainly
> won't hurt. UNLESS it causes bad choices to be made. Which we can only
> guess at right now.

You're the men :)
Ok, this stops all the talks Ingo.
Can you send me a link, there're different things to be fixed IMHO.
The load estimator can easily use the current dyn_prio/time_slice by
simplyfing things a _lot_
I would suggest a lower number of queues, 32 is way more than necessary.
The rt code _must_ be better, it can be easily done by a smartest wakeup.
There's no need to acquire the whole lock set, at least w/out a checkpoint
solution ( look at BMQS ) that prevents multiple failing lookups inside
the RT queue.




- Davide


-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo _at_ vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/