> > CPU's passed the 200 MHz level. The chronic problems of pre-2.0.32
> > kernels with SMP deadlocks were also in many cases related to speed -- a
>
> gee, this sounds like a great reason to run modern (2.1) kernels...
Actually they aren't related to speed at all. Specifically the two bugs
which seem to have fixed most of the problems for SMP are
1. A deadlock for 3 or more way machines related to nothing much but
CPU count/memory load
2. An irq re-entry bug. Nothing to do with CPU speed but interrupt
timings. So 100Mbit tended to hit it more than 10Mbit.
Both are now certified dead and Im hoping .36 will be amazingly more stable
on .36 final
