Re: VolanoMark performance improvements (was: Re: volanoMark 12% regression with 2.6.25-rc6)

["Zhang, Yanmin" <yanmin_zhang@linux.intel.com>]

On Fri, 2008-03-21 at 15:14 +0800, Zhang, Yanmin wrote:
> On Thu, 2008-03-20 at 09:35 +0800, Zhang, Yanmin wrote:
> > On Wed, 2008-03-19 at 14:48 +0100, Ingo Molnar wrote:
> > > i just tried a handful of kernels with VolanoMark on an Intel quad-core 
> > > testbox,
> > How many physical processors does it have?
> > > these numbers are totally reproducible when running them over a long 
> > > time, i only included 6 runs for brevity. (A sidenote: the 
> > > /proc/sys/kernel/compat_sched_yield switch of 0 or 1 has no effect on 
> > > the .25-rc7 results, and it degraded the .24 results by about 35% so i 
> > > kept it off there.)
> I used java-openjdk to test it with 2.6.25-rc5 by switching compat_sched_yield
> to 0/1. The result difference is about 30%.
> 
> vmstat data when compat_sched_yield=0:
> procs -----------memory---------- ---swap-- -----io---- --system-- -----cpu------
>  r  b   swpd   free   buff  cache   si   so    bi    bo   in   cs us sy id wa st
> 85  0      0 7584984  17128  79264    0    0     0     0    2 716494 37 60  3  0  0
> 94  0      0 7585168  17136  79256    0    0     0    64    7 723299 39 57  4  0  0
> 
> 
> vmstat data when compat_sched_yield=0:
Sorry. below data is when compat_sched_yield=1:

> procs -----------memory---------- ---swap-- -----io---- --system-- -----cpu------
>  r  b   swpd   free   buff  cache   si   so    bi    bo   in   cs us sy id wa st
> 369  0      0 7580656  14364  75852    0    0     0     0    4 1013260 42 57  1  0  0
> 335  0      0 7580836  14364  75852    0    0     0     0    1 971081 41 58  1  0  0
> 414  0      0 7580932  14372  75852    0    0     0    64    7 955333 40 58  1  0  0