> -----Original Message-----
> From: [EMAIL PROTECTED] 
> [mailto:[EMAIL PROTECTED] On Behalf Of 
> Vaidyanathan Srinivasan
> Sent: 2008年3月31日 14:35
> To: Arjan van de Ven
> Cc: [email protected]
> Subject: Re: [Discuss] How does powertop get the wake-up rate?
> 
> * Arjan van de Ven <[EMAIL PROTECTED]> [2008-03-28 12:56:13]:
> 
> > Vaidyanathan Srinivasan wrote:
> >> Hi Arjan,
> >>
> >> I have been using powertop to find wakeup events and try 
> to correlate
> >> with total interrupts received by the idle system.  I want 
> to see the
> >> rate of local timer interrupts go down as we optimise the 
> drivers and
> >> user-space applications to avoid polling.
> >>
> >> I observed that powertop shows very low wake-from-idle rate on SMP
> >> system and the wake-from-idle rate is very high when all 
> other CPUs are
> >> offlined and only CPU0 is operational. I have routed all 
> irqs to CPUs
> >> before the experiment.
> >
> > powertop shows wakeups *per cpu* ...
> 
> Ok, so the wakeups are averaged out? Total wake-ups divided by current
> number of CPUs?
> 

Yes. If you read the code, you will see that 'wakeups_per_second' is divided by 
sysconf(_SC_NPROCESSORS_ONLN).

> Another related question is how is wake *from idle* computed, how is
> the wakeup of non-idle CPUs and idle CPUs distinguished?
> 

'Wakeups-from-idle per second' is computed by taking advantage of the 
timer_stats facility in kernel. 

_______________________________________________
Discuss mailing list
[email protected]
http://mail.lesswatts.org/mailman/listinfo/discuss

Reply via email to