Messages in this thread | | | Date | Wed, 22 Jul 2009 18:02:48 +0200 | From | Martin Schwidefsky <> | Subject | Re: [RFC][PATCH] cache __next_timer_interrupt result |
| |
On Wed, 22 Jul 2009 16:38:18 +0200 (CEST) Thomas Gleixner <tglx@linutronix.de> wrote:
> On Tue, 21 Jul 2009, Martin Schwidefsky wrote: > > > From: Martin Schwidefsky <schwidefsky@de.ibm.com> > > > > Each time a cpu goes to sleep on a NOHZ=y system the timer wheel is > > searched for the next timer interrupt. It can take quite a few cycles > > to find the next pending timer. This patch adds a field to tvec_base > > that caches the result of __next_timer_interrupt. The hit ratio is > > around 80% on my thinkpad under normal use, on a server I've seen > > Nice, I like it.
Thanks :-)
> > hit ratios from 5% to 95% dependent on the workload. > > Which workloads result in lower hit ratios ? Heavy networking ?
5% ping-pong packet over loopback between two cpus. So yes, networking.
-- blue skies, Martin.
"Reality continues to ruin my life." - Calvin.
| |