Re: [PATCH v2] cpuidle: Fix the menu governor to boost IO performance

[Arjan van de Ven <arjan@infradead.org>]

On Sun, 8 Nov 2009 22:59:43 +0100
Corrado Zoccolo <czoccolo@gmail.com> wrote:
> >
> > the exit latency is +/- 1 us, the entry latency is similar, and then
> > you're pretty close to 5 already (esp if you keep in mind that to
> > break even on energy you also need to be in the C state for a
> > little bit)...
> 
> There are also performance considerations for using C1 (HLT).
> Quoting from http://www.intel.com/Assets/PDF/manual/248966.pdf (8-19):
> On processors supporting HT Technology, operating systems should use
> the HLT instruction if one logical processor is active and the other
> is not. HLT will allow an idle
> logical processor to transition to a halted state; this allows the
> active logical
> processor to use all the hardware resources in the physical package.

I think we all agree that long term polling is bad ;-)
(even though we use rep nop in the polling loop which is also a HT
yield).

There's just the very short sleeps (where short is "single digit usecs")
where the rules are slightly different.

> > this check is supposed to catch the known timer cases; those
> > are rather accurate in prediction
> 
> Unfortunately, I have seen polling residency times > 1ms, so it must
> not be so accurate.

well the question is... is this a measurement error or an error in
when polling is chosen.
We obviously need to fix it whatever it is, but... first need to chase
down really which it is.


-- 
Arjan van de Ven 	Intel Open Source Technology Centre
For development, discussion and tips for power savings, 
visit http://www.lesswatts.org