Bill.Holler at Sun.COM wrote:

> Li, Aubrey wrote:
>> Bill.Holler at Sun.COM wrote:
>> 
>> 
>>> Li, Aubrey wrote:
>>> 
>>>> Bill Holler wrote:
>>>> 
>>>> 
>>>> 
>>>>>>> You can probably take it from another CPU that's not in deep-c
>>>>>>> state (if there is one?) 
>>>>>>> 
>>>>>>> 
>>>>>>> 
>>>>>>> 
>>>>>>> 
>>>>>> Yes.  Our initial thoughts are to sync up with a running CPU
>>>>>> via cross call.  We need to run performance experiments with
>>>>>> different implementations to see what the costs are.
>>>>>> 
>>>>>> 
>>>>>> 
>>>>> Synching TSC/APIC with a running cpu is potentially much more
>>>>> accurate than using the HPET.  However this will induce drift.
>>>>> 
>>>>> 
>>>>> 
>>>> if all the cpus are in the deeper c-state,...?
>>>> 
>>>> 
>>> An external source such as the HPET or platform pm timer needs to
>>> be used if all cpus are in deeper c-state.  :-)
>>> 
>>> 
>> 
>> Are you working on the ACPI timer driver as well?
>> 
> 
> Yes.
> 
> Do you have an estimate for how many systems have an ACPI timer
> but do not have an HPET timer?
> 
> Bill
> 
I have no idea, ;-)

-Aubrey


Reply via email to