Re: Strange scheduling behavoir in SMP (kernel 2.2.14)

[Alan Cox]

> What about cpu0: xosview, cpu1: tight loop and they stay where they are? The scheduler
> was written in order to exactly do that, but somehow it doesn't. Why?
> 

Which CPU is running X11, kswapd, kpiod etc

Im not saying the scheduler is right - the speed evidence is good evidence
it is not, just that you want to instrument the kernel to measure the goings
on not use Xosview/X11