> Hmm, strange. Actually 7c64ade53a6f977d73f16243865c42ceae999aea fixed
> a similiar bug, see here:
> http://bugzilla.kernel.org/show_bug.cgi?id=11908
> 
> Your patch with 2.6.28, does:
> 
>  grep NMI /proc/interrupts
> 
> returns exactly 1 NMI per core or some more?

that bug is similar, but it reports different numbers of NMIs for the
cores.
i get exactly one NMI per core, for each time, i start oprofile:

NMI: 0
opcontrol -s
(wait)
opcontrol --deinit
NMI: 1
opcontrol -s
(wait)
opcontrol --deinit
NMI: 2

hth, tim

--
tim@klingt.org
http://tim.klingt.org

Which is more musical, a truck passing by a factory or a truck passing
by a music school?
  John Cage