Re: Context switch times

Mika Liljeberg (Mika.Liljeberg@welho.com)
Mon, 08 Oct 2001 01:33:33 +0300


Davide Libenzi wrote:
>
> On Mon, 8 Oct 2001, Mika Liljeberg wrote:
>
> > How the balance is determined is another issue, though. I basically
> > proposed summing the time slices consumed by tasks executing on a single
> > CPU and using that as the comparison value. Davide Libenzi, on the other
> > hand, simply proposed using the number of tasks.
>
> CPU based number of __running__ tasks.

That's what I meant, sorry for not making it clear.

I suppose in practise this would amount to using the number of _CPU
bound_ running tasks per CPU, since the non-CPU bound tasks would likely
be waiting rather than running? This would then be essentially
equivalent to (and simpler than) my proposal, i.e. using the sum of the
time slices.

> - Davide

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