Interesting analysis of linux kernel threading by IBM

[Sean Hunter]

On Sun, Jan 23, 2000 at 03:29:32PM +0100, Davide Libenzi wrote:
> On Sat, 22 Jan 2000, Sean Hunter wrote:
> > > You can state that I don't have benfits in uniprocessor systems.
> > > But I have in SMP, that is, IMHO, the future of computing technology.
> > > You can find the needs of parallelism starting from CPUs executions units up
> > > to
> > > complex software systems up to daily work organization.
> > > And if the OS is the bottleneck of parallelism we must try to improve it,
> > > not to avoid multithreading.
> > 
> > No, yet again the _app_ design you propose is the bottleneck.
> 
> I'll respect Your opinion but the mine is different.

This is not _opinion_.  That's the great thing about engineering.  You
test.  You measure.  Then you release your results.  

Your app design sucks.  If you did it in a different way I am 100%
flat-out positive it could be made considerably faster than it is now.
If you tried alternative (ie non-threaded) algorithms and benched them
you may feel the same.  Larry, and others have done this for years,
and they back up this.  All of the published data backs this up, and
all of the rendering folks feel this way too.

Did you read the simple cache-miss calculation that Larry posted?
Have you thought about what that means?  You talk about "only two
cache" misses in your design, but as Larry showed, that could lead to
50% decrease in performance.  Easily.

Sean
 

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