The reason for choosing a limited set of events is that if you enable all
possible available events then they each get a low dutycycle and hence a
high sampling error. On my Intel Core2 box if you enable all of the
possible core PMU events then the dutycycle varies from 10^-2 to 10^-4
depending on the event.
Joe
On 6/26/14, 6:01 PM, "Frank Ch. Eigler" <fche@xxxxxxxxxx> wrote:
>Hi -
>
>> > [...]
>> >- The configuration file appears to hard-code the perfcounters to be
>> > multiplexed/available. Did you consider automatically enumerating
>> > all the available ones on the platform instead?
>>
>> I don't understand your question. The code checks which hardware PMUs
>>are
>> available and then enumerates the available counters. The code then
>> enables the counters that are specified in the relevant hardware PMU
>> section in the configuration file. [...]
>
>To clarify, I wonder what value the configuration file brings. If we
>can enumerate hardware-available counters based on the lower-level
>library, why not represent apprx. all of them to the PMNS?
>
>
>- FChE
|