kernel preemption....

[Dave Jones]

On Sun, 2004-04-25 at 03:34, Craig Cruden wrote:
> I noticed the latest kernels that have been
> distributed for testing have had the "PREEMPT" flag
> turned off.

It was never turned on in the Fedora kernels.

> The pre-emption was added for things like
> sound -- i.e. desktop applications....  when the
> product is distributed is there going to be a
> "desktop" kernel version with it turned on by default?

Nope. The pain that comes with preempt outweighs the
perceived gain. A lot of folks seem to think it's a silver bullet
that will provide low-latency, but that isn't the case.
Preemption lowers the average latency, not the worse case
latency, which for most things isn't a gain.
If you've got a driver that has a udelay() of some stupid
length of time, preempt won't save you. Drivers that pull
this kind of trick need fixing instead of putting faith
in some magical config option to 'fix everything'.

> Also, a side question -- what are the benefits of the
> "4G/4G" changes (simple explaination) -- right now I
> am unable to use it turned on on my portable -- so I
> was wondering what I am missing out on :p

Executive summary:
It increases the address space usable by each process
at the expense of some performance if you have an application
that makes a *lot* of system calls to the kernel.

	Dave