Re: [PATCH] Rmap speedup

[Daniel Phillips]

On Sunday 04 August 2002 00:39, Andrew Morton wrote:
> Daniel Phillips wrote:
> > 
> > Wait a second guys, the problem is with the script, look at those CPU
> > numbers:
> > 
> > > ./daniel.sh  39.78s user 71.72s system 368% cpu 30.260 total
> > > quad:/home/akpm> time ./daniel.sh
> > > ./daniel.sh  38.45s user 70.00s system 365% cpu 29.642 total
> > 
> > They should be 399%!!  With my fancy script, the processes themselves are
> > getting serialized somehow.
> > 
> > Lets back up and try this again with this pair of scripts, much closer to
> > the original:
> 
> Still 360%.  I did have a version which achieved 398%, but it
> succumbed to the monthly "why is there so much junk in my home
> dir" disease.
> 
> But it doesn't matter, does it?  We're looking at deltas here.

It matters a whole lot.  If we aren't saturating the CPUs then we're not
testing what we think we're testing.

-- 
Daniel