Subject: Re: processor affinity
To: Erik E. Fair <fair@netbsd.org>
From: Nathan J. Williams <nathanw@wasabisystems.com>
List: tech-smp
Date: 06/02/2003 12:35:42
"Erik E. Fair" <fair@netbsd.org> writes:

> Scheduling on UNIX has been studied nearly to death, and I'll bet
> (without having looked) that there are papers that report on research
> into processor affinity on MP systems, scheduling, and the pros/cons
> of that.
> 
> Do the other readers of this list have any memorable/interesting
> papers to cite?

I'll dig into my bibliography to find some good ones, as you're
correct that it's a subject that's been studied nearly to
death. However, the era in which it was studied to death in published
papers was roughly a decade ago, so many of the machine-specific
results need to be looked at carefully to see whether they still apply
to modern hardware.

        - Nathan