On Wed, 10 Aug 2005 23:28:07 +0200
Morten Nilsen <[EMAIL PROTECTED]> wrote:

> Brian Wilson wrote:
> > I guess the reason for two kernels would be to allow
> > a pre-empt version for us daredevils and a non preempt
> > for stodgy people who want their systems to have high
> > uptime.
> 
> afaik, preempt won't affect stability at all, merely changes in response 
> times for certain applications (like an X-server)
Yes and no :)

On an UP system without in-kernel preemption all kernel-spinlocks are removed.
(see 
http://www.kernel.org/pub/linux/kernel/people/rusty/kernel-locking/x109.html)

So with preemption and if the code is not 100% ok, there is the possibility of 
locking bugs (same as you have an smp-box).

But from where did you get this thing, that preempt is good[tm] for
routers? iirc (and as morton said) it's good for low-latency like audio
and desktop stuff, but not for servers / networking.

Olaf

_______________________________________________
tsl-discuss mailing list
[email protected]
http://lists.trustix.org/mailman/listinfo/tsl-discuss

Reply via email to