Re: PAE and balloon benchmarks

[Antti Kantee <pooka%cs.hut.fi@localhost>]

On Sun Jul 11 2010 at 01:44:51 +0200, Jean-Yves Migeon wrote:
> Hi lists,
> 
> FWIW, I just uploaded a wiki page for my results regarding PAE
> benchmarks [1].
> 
> In short:
> 
> - expect 10-20% overhead in specific benchmarks for PAE; nothing new
> here, there have been critics on PAE performance loss multiple times.

Hi,

Thanks for taking the time to throughly benchmark this.

A remarks/questions out of pure interest:

* Is the cost expected?  Naiively thinking, 15-20% seems quite high.
  How much do other systems pay for PAE?  Did you attempt to pinpoint
  where extra time is spent?

* How jittery were the results?  I'd be careful to not dismiss performance
  degradation of several % as "noise" without just cause.  Those percent
  might be important to someone!

* Did you do any hot cache build.sh testing?  IMHO that is at least as
  interesting, if not more interesting, than the coldstart one.