Re: [PATCH] make poll_idle behave more like the other idle methods

[Joe Korty <joe.korty@ccur.com>]

On Thu, Aug 28, 2008 at 05:00:36AM -0400, Ingo Molnar wrote:
> 
> * Joe Korty <joe.korty@ccur.com> wrote:
> 
> > Make poll_idle() behave more like the other idle methods.
> >
> > Currently, poll_idle() returns immediately.  The other
> > idle methods all wait indefinately for some condition
> > to come true before returning.  poll_idle should emulate
> > these other methods and also wait for a return condition,
> > in this case, for need_resched() to become 'true'.
> >
> > Without this delay the idle loop spends all of its time
> > in the outer loop that calls poll_idle.  This outer loop,
> > these days, does real work, some of it under rcu locks.
> > That work should only be done when idle is entered and
> > when idle exits, not continuously while idle is spinning.
> 
> i'm wondering, what's the motivation, have you actually seen
> anything bad/undesired happen due to that?

I saw the outer loop running continuously, from the old
trace patch which I had applied.

Nowdays the outer loop runs the NO_HZ stuff, which touches
quite a few memory locations.  Having say two cpus in idle
and there is potential for cache thrashing to suck up a good
part of the local memory bus bandwidth.

And I suspect as time goes on cpu_idle will end up with
more work to do.

Joe