Jordan Crouse wrote:
On 22/03/07 18:26 +0100, Jens Axboe wrote:
The tickless kernel just _enables_ low number of interrupts per second,
it'll still wakeup everytime something is scheduled to run. So you
probably want to look into _why_ th timer interrupts happen so often -
who is the user of them?
I see ~70 interrupts/sec here with an idle machine in X, FWIW.
I've seen it as low as 10 on a very very quiet machine. Enable the timer
statistics and run a sample - I'll bet you see some offenders like the
console cursor, the libertas, and other such greedy consumers of our time.
Jordan
Jordan,
Thanks for the pointer. Looking through /proc/timer_list give a list of
possible offenders. I also tried running with and without systemtap script
running. Without the systemtap script the mfgpt-timers irq dropped to about
40/sec. So it appears that systemtap probe is causing the timer to fire more often.
-Will
_______________________________________________
Devel mailing list
[email protected]
http://mailman.laptop.org/mailman/listinfo/devel
