[snip]

> 
> All the above test results has the impact of oprofile included.  Running
> oprofile also may significantly increase mmap_sem contention.
> 
> I Will run the tests again without oprofile to understand the impact of
> oprofile itself.
> 

Hi,

I have run the same tests without oprofile and also tried to plot the
impact of oprofile.

There are two attached graphs.

rcu-vma-scaling.png contains the plot of Peter's RCU-VMA lookup without
running oprofile.

There are some variations beyond 10 CPUs.  I am trying to wonder if those
were due to errors in test/measurement!

oprofile-scaling.png plots similar tests with and without oprofile enabled.
 We can see more impact beyond 12 CPUs.

--Vaidy