> > I'm not sure either, since qsort doesn't really have much
> > locality of reference but just walks all over the place.
>
> qsort can be made to perform reasonably well providing you try to cache
> colour the objects you sort and try to use prefetches a bit.
That won't quite work when the qsort in question is 150%
the size of your RAM ;)
> > One thing which could make 2.4.10 faster for this single case
> > is the fact that it doesn't keep any page aging info, so IO
> > clustering won't be confused by the process accessing its
> > pages ;)
>
> I don't think that is too unusual a case. If the smarter vm is making
> poorer I/O clustering decisions it wants investigating
Absolutely, this is something we really want to know ...
I guess I'll play with Lorenzo's program a bit to see how
the system behaves and how it can be improved.
regards,
Rik
-- IA64: a worthy successor to i860.http://www.surriel.com/ http://distro.conectiva.com/
Send all your spam to aardvark@nl.linux.org (spam digging piggy)
- 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/