Re: High CPU load when machine is idle (related to PROBLEM: Unusually high load average when idle in 2.6.35, 2.6.35.1 and later)

[Peter Zijlstra <peterz@infradead.org>]

On Tue, 2010-11-09 at 21:37 -0500, tmhikaru@gmail.com wrote:
> On Tue, Nov 09, 2010 at 08:02:28PM +0100, Peter Zijlstra wrote:
> > On Tue, 2010-11-09 at 13:55 -0500, Kyle McMartin wrote:
> > > On Fri, Oct 29, 2010 at 09:42:23PM +0200, Peter Zijlstra wrote:
> > > > 
> > > > not tested other than compile, but how does that look...?
> > > > 
> > > 
> > > Hi Peter,
> > > 
> > > I finally got around to building a test kernel for a user who had been
> > > reporting this against F-14 (2.6.35) but they report there's not much
> > > improvement.
> > > 
> > > Sorry. :/
> > 
> > Right, I tested it myself yesterday and got similar results. Does the
> > 'simple' patch from a few posts back work for people?
> > 
> >  http://lkml.org/lkml/2010/10/26/150
> > 
> > If that does work I fudged the fancy math, if that doesn't work either
> > we're back to square zero.
> 
> I seem to remember someone saying this would only work on x86-64; if this is
> meant to work with a 32bit intel single core processor, please let me know -
> I'd like to try it. It would also help if I knew which kernel version this
> simpler patch is meant to apply to.

Venki's patch was, the one I linked should work on every arch.