From owner-freebsd-current Thu Dec 7 21:59:03 1995 Return-Path: owner-current Received: (from root@localhost) by freefall.freebsd.org (8.7.3/8.7.3) id VAA01208 for current-outgoing; Thu, 7 Dec 1995 21:59:03 -0800 (PST) Received: from phaeton.artisoft.com (phaeton.Artisoft.COM [198.17.250.211]) by freefall.freebsd.org (8.7.3/8.7.3) with SMTP id VAA01203 for ; Thu, 7 Dec 1995 21:58:58 -0800 (PST) Received: (from terry@localhost) by phaeton.artisoft.com (8.6.11/8.6.9) id WAA00289; Thu, 7 Dec 1995 22:56:14 -0700 From: Terry Lambert Message-Id: <199512080556.WAA00289@phaeton.artisoft.com> Subject: Re: FYI To: jhay@mikom.csir.co.za (John Hay) Date: Thu, 7 Dec 1995 22:56:14 -0700 (MST) Cc: julian@ref.tfs.com, terry@lambert.org, current@freebsd.org In-Reply-To: <199512080532.HAA03443@zibbi.mikom.csir.co.za> from "John Hay" at Dec 8, 95 07:32:26 am X-Mailer: ELM [version 2.4 PL24] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-current@freebsd.org Precedence: bulk > > neat.. so xxx is running on the 2nd processor? [ ... ] > Hey what about the two most important questions? > 1. Are we going to see it in the cvs tree at some stage? > 2. If we are, when? :-))) > > Jokes aside, I am realy interrested. My machine at home have a dual pentium > motherboard. It only need a second processor. I need someone who cares about Intel page tables to look at the locore.s. Trying to move to the most recent -current from stuff that's maybe a month old blow up. Basically, I'm working on somone elses code with a less than perfect knowledge of the SMP memory design (other than there is a dual mapped memory window). If I cheat an use a counter and basically start two idle procs and murder the process from the fork for the second one, then I'm fine, but this is a hell of a kludge with some very obvious races in it. If I start the second process by going into "secondary_main" from finishing the init, I get an integer divide by 0 fault. 8-(. That's why the idle0/idle1 didn't show in the ps I posted. 8-(. The kernel debuggers can't go in as it sits, I'd like to fix that. The code needs a cleanup for generalization for 'n' processors. There are a lot of hacks specific to 2 of them. I can do everything but the page tables. I'm too busy learning about other processors right now to learn enough about Intel to do the work myself right now. So nothing is going in immediately. Anyone want to hack a locore.s? Terry Lambert terry@lambert.org --- Any opinions in this posting are my own and not those of my present or previous employers.