Re: gettimeofday order of magnitude slower with pmtimer, which is default

[Con Kolivas <kernel@kolivas.org>]

On Tuesday 21 March 2006 23:04, Arjan van de Ven wrote:
> On Tue, 2006-03-21 at 22:58 +1100, Con Kolivas wrote:
> > On Tuesday 21 March 2006 19:53, Andreas Mohr wrote:
> > > (and the fact that invoking a function pointer should be similarly
> > > expensive to a conditional) I don't think it's useful.
> >
> > Is
> >
> > *blah();
> >
> > as expensive as
> >
> > if (conditional)
> > 	blah();
> >
> > I don't know the answer. I just know cmp is expensive. Comments?
>
> function pointer is usually MORE expensive.
> for if() the processor has a change to predict the branch right, while
> call <register> (which is what function pointer calls end up being) are
> basically always mispredicted unless you have a really really fancy
> branch predictor...

Thanks! That's something I've been trying to find good info on.

Cheers,
Con
-
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/