Marc, On Tue, Sep 23, 2008 at 3:56 PM, <[EMAIL PROTECTED]> wrote: > Hey Stephane, > > thanks for your _very_ fast response. In the meantime I took a break and had > a coffee. It turned out that coffee is good for realizing one's faults. I > guess the reported amount of 32 retired instructions is totally valid.
Glad to hear it actually works for this code! I'd like to emphasize that using this option on small functions may not be a good idea depending on the event you are measuring. To start/stop on code boundaries pfmon is using debug breakpoints and as you can imagine it incurs a significant overhead which may affect the counts you are observing. Obviously, it should not have impact on instructions_retired because you execute all the instructions no matter what. > > However, I do have 2 more questions: > How do convince pfmon to report the number of retired instructions for each > call (using --trigger-code-repeat)? You mean aggregate over all calls? If so, then yes, use --trigger-code-repeat. > Is there something similar to the --opc-match8 option in 3.x? > This option (opcode matching) is Itanium specific. If you are trying this on x86, you will not get it. ------------------------------------------------------------------------- This SF.Net email is sponsored by the Moblin Your Move Developer's challenge Build the coolest Linux based applications with Moblin SDK & win great prizes Grand prize is a trip for two to an Open Source event anywhere in the world http://moblin-contest.org/redirect.php?banner_id=100&url=/ _______________________________________________ perfmon2-devel mailing list perfmon2-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/perfmon2-devel
