Re: [tlug] 2.6 Kernel vs 2.4 Kernel

[Edward Middleton <edward@example.com>]

Gerald Naughton wrote:
> Between new kernel/new drivers/new X-Server etc , I`m
> wondering if this a general trend  ?
>   
There seemed to be a few cases which people kept hitting when the 2.6
kernels first came out.  Some of the problems were things like.

- SMP kernels being slower then non-SMP kernels on non-SMP systems.
- kernel preemption lowering throughput.
- the new scheduler using more system resources then the 2.4 scheduler
- the higher system clock putting more load on the system.
- various hacks that had been used to reduce poor latency in
applications running on 2.4 kernels, like reniceing X resulted in very
bad system response on 2.6.
- a large number of highly optimized 2.4 forks.

These aren't 2.6 kernel problems but that didn't stop people from
characterizing them as such.

Edward