Re: [ANNOUNCE] ktimers subsystem

[Roman Zippel <zippel@linux-m68k.org>]

Hi,

On Tue, 27 Sep 2005, Tim Bird wrote:

> > That still means it is used and if an application
> > actually depends on it, it would be penalized by
> > your implementation. These timers may open up new
> > application (in kernel or user space), where
> > this conversion may be needed, so _only_ looking
> > at the current numbers is a bit misleading.
> 
> Oh good heavens!  One can always point to real or
> hypothetical cases where a change like this
> will result in worse performance.  Will you only
> be satisfied if there is provably NO performance
> degradation for ANY app on ANY platform?

I want to get the focus at the complete picture, as this is a rather 
critical area and I will be satisfied, as soon as I can see all 
consequences and possibilities have been considered.

>  Even
> if the code is easier to maintain, and allows
> for improvements in functionality and equal or
> better performance for the majority of apps.
> and platforms?

If that's case, you're hopefully not afraid of a few questions? Why do I 
have to take the code as is and just believe the claims about it?
I like improvements as everyone, but I also want to verify them and look 
at the alternatives and I can't see anything wrong with it.

> Unless I missed something, ktimers has not been
> recommended for mainlining yet.  I suspect (without
> having measured it myself yet) that the
> core abstraction that it proposes (timers
> vs. timeouts) is an important one for improving
> the kernel timing system.

I'm not saying that the idea is wrong, the general direction is fine, but 
some course correction should be possible?

bye, Roman