To: Paul McKenney/Beaverton/IBM@IBMUS
Cc: andi@suse.de, andrea@suse.de, aono@ed2.com1.fc.nec.co.jp, beckman@turbolabs.com, bjorn_helgaas@hp.com, Hubertus Franke/Watson/IBM@IBMUS, Jerry.Harrow@Compaq.com, jwright@engr.sgi.com, kanoj@engr.sgi.com, Kenneth Rozendal/Austin/IBM@IBMUS, kumon@flab.fujitsu.co.jp, mikek@sequent.com, norton@mclinux.com, suganuma@hpc.bs1.fc.nec.co.jp, sunil.saxena@intel.com, tbutler@hi.com, woodman@missioncriticallinux.com
From: kanoj@google.engr.sgi.com
Date: 04/01/01 11:06 PM
Subject: Re: your mail (NUMA-on-Linux roadmap)
>
> Kanoj, is there a pointer to public info on hwgfs that would given people
> rough idea of what you have in mind? (I will search www.sgi.com when I
> get back to a higher bandwidth connection, but thought you might have a
> good pointer off the top of your head.)
>
> Thanx, Paul
>
>
I added some quick hwgfs manpages at
http://oss.sgi.com/projects/numa/download/irix.
Btw, I am not proposing exact hwgfs, because as you can see, it is
very platform dependent ... unless we decide we absolutely need a
platform dependent graph to extract 100% performance.
IRIX hwgfs does not deal with exposing shared multilevel caches to
the user.
Kanoj
PS - I will respond to the rest of the mail later.