On Wed, 2005-07-13 at 15:17, Eric Dantan Rzewnicki wrote:
> On Wed, Jul 13, 2005 at 02:20:20PM -0400, Lee Revell wrote:
> > On Wed, 2005-07-13 at 13:52 -0400, Lee Revell wrote:
> > > On Wed, 2005-07-13 at 13:42 -0400, Eric Dantan Rzewnicki wrote:
> > > > > > What is driving the kernel-devs to regress on this issue?
> > > > > Saving battery on laptops. The only performance numbers anyone posted
> > > > > indicated HZ=250 sped up a kernel compile on a 16 CPU machine (!) by
> > > > > ~5%, and this was after the fact. Not exactly compelling...
> > > > > But since Linus and Andrew apparently all use laptops, us desktop people
> > > > > are screwed...
> > > > Any chance they would make it a config option?
> > > It is a config option, the available settings are 100, 250, and 1000.
> > > The problem is that the default has changed to 250.
> > Update: Linus has said that this is a done deal. So now we need to
> > figure out how to work around it. I guess we'll have to go back to
> > using the RTC like on 2.4.
>
> I'm confused ... most of us build our own kernels or use kernels built
> by Fernando or Free. Why can't kernels just be built with the config
> option set to 1000?
They can, what is changing is the default. Which means that out of the
box "normal" distros will not be as suitable to reasonable midi timing
as a tuned kernel. That was not the case in 2.6.x till now.
I don't know about Free but I'd rather not be building kernels forever,
it'd be nice that in a not too distant future regular unadulterated
linux kernels would be good for audio/midi work...
-- Fernando
Received on Thu Jul 14 04:15:17 2005
This archive was generated by hypermail 2.1.8 : Thu Jul 14 2005 - 04:15:18 EEST