Help on Win32::Perflib

shonorio has asked for the wisdom of the Perl Monks concerning the following question:

Hi monks,
I'm looking for help on Win32::Perflib to do correct calculation of one counter type PERF_100NSEC_MULTI_TIMER_INV, I got the formula (B-((N1-N0)/(D1-D0))) on Jutta site for this.
I have no doubt about "(N1-N0)/(D1-D0)", but i can't understand how to get "B". His site say "B -> Base count (number of entries)" or "B represents the number of components being monitored" on MSDN site", but "B" still unclear for me.
Thanks for any help.

Solli Moreira Honorio
Sao Paulo - Brazil

Comment on Help on Win32::Perflib

Replies are listed 'Best First'.
Re: Help on Win32::Perflib by BrowserUk (Patriarch) on Jun 06, 2004 at 02:27 UTC
According to this, the PERF_COUNTER_MULTI_BASE is the base required for calculations involving ERF_100NSEC_MULTI_TIMER_INV. Ie. B = the value of PERF_COUNTER_MULTI_BASE. However, the calculation shown here is somewhat different to the formula you have. Examine what is said, not who speaks. "Efficiency is intelligent laziness." -David Dunham "Think for yourself!" - Abigail	[reply]
Re^2: Help on Win32::Perflib by shonorio (Hermit) on Jun 06, 2004 at 14:09 UTC
BrowserUk, Thanks for your attention, I read on MS site about PERF_COUNTER_MULTI_BASE but I didn't understand how they works, how to get it, and I found no examples. Because this I moved to the next counter (PERF_RAW_FRACTION) and I could undestand that this base value are on same Perflib hash, I just have to continue looking for a BASE counter type. For example, I must to get the counter value of a PERF_RAW_FRACTION, then get the counter value of PERF_RAW_BASE (of course of same perf counter), and finally do the calculation. It's should be easier, but it's not !!! discovery that they work like PERF_100NSEC_MULTI_TIMER_INV. It's need a base too and discovery that I must Solli Moreira Honorio Sao Paulo - Brazil	[reply]