Stephane, > Fluctations are not unexpected when the system is doing nothing (or at > least you think it is). > What is more important is to verify that when you have a constant > workload the count for > REF remains fairly constant and close to peak for a nonblocking > workload, i.e., a stupid noploop > program such as (pfmon/tests/noploop).
good point, I was probably running a too non-homogenous task. Pfmon idle loop gets assigned to one core of the two I have in my Core2Duo, which is then at the peak rate (near to the official 2,33 GHz clock rate). By looking at the values output by pfmon in 100ms intervals, the fluctuations are now limited to the lower 4 orders of magnitude. I.e. the total count is cca. 2.3 * 10^8 per each 100 ms quantum +- 10^4 (no statistics yet, just computations by eye!). This makes 0.05 permille error. Tom ------------------------------------------------------------------------- This SF.net email is sponsored by the 2008 JavaOne(SM) Conference Don't miss this year's exciting event. There's still time to save $100. Use priority code J8TL2D2. http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone _______________________________________________ perfmon2-devel mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/perfmon2-devel
