Re: Statistical theory behind OProfile
William Cohen <[email protected]>
| Newsgroups | gmane.linux.oprofile |
|---|---|
| Message-ID | <[email protected]> |
On 05/31/2015 10:15 PM, Donald Johnson wrote: > On Sun, May 31, 2015 at 10:34 AM, Andi Kleen <[email protected]> wrote: >> >>> The environment I am working on is an armv5 board, which means I have >>> to use release 0.9.9 for software interrupts. >> >> That's a very poor environment for performance monitoring. Better use >> some system with a real PMU, where you can use event driven sampling, >> which tends to be more accurate than time driven sampling. > > I would if I could. Next project may have a better environment where > I could be given an updated board that can do hardware interrupts, but > for now it's an old system on an old board. > > ~Donald Johnson Hi, I don't know which kernelversion you are using on this project but if it is newer kernel (later than march 2010) you might make things a bit better on your platform by modifying the oprofile driver code in linux/drivers/oprofile/timer_int.c to have a writeable variable that could be changed rather than constant TICK_NSEC. Or change the oprofile driver to use some other value other than TICK_NSEC. -Will ------------------------------------------------------------------------------