From owner-freebsd-mips@FreeBSD.ORG Tue Nov 11 15:28:25 2003 Return-Path: Delivered-To: freebsd-mips@freebsd.org Received: by hub.freebsd.org (Postfix, from userid 931) id 52B1F16A4CF; Tue, 11 Nov 2003 15:28:25 -0800 (PST) Date: Tue, 11 Nov 2003 17:28:25 -0600 From: Juli Mallett To: Ralf Baechle Message-ID: <20031111172825.A32024@FreeBSD.org> References: <20031111142433.A9560@FreeBSD.org> <20031111210405.GA16834@linux-mips.org> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.2.5.1i In-Reply-To: <20031111210405.GA16834@linux-mips.org>; from ralf@linux-mips.org on Tue, Nov 11, 2003 at 10:04:05PM +0100 X-Title: Code Maven X-Towel: Yes X-Negacore: Yes X-Authentication-Warning: localhost: juli pwned teh intarweb cc: freebsd-mips@FreeBSD.org Subject: Re: A note about status. X-BeenThere: freebsd-mips@freebsd.org X-Mailman-Version: 2.1.1 Precedence: list List-Id: Porting FreeBSD to MIPS List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 11 Nov 2003 23:28:25 -0000 * Ralf Baechle [ Date: 2003-11-11 ] [ w.r.t. Re: A note about status. ] > Juli, > > On Tue, Nov 11, 2003 at 02:24:33PM -0600, Juli Mallett wrote: > > > I'm doing all this on my R4400 Indigo2's. I've seen some weird stuff > > with my Indy, but that may have been related to a dumb mistake in the > > TLB refill code I fixed last night. > > The R4600 only has a 5 stage pipeline unlike the R4400's 8 stages which > means most hazards are shorter or do not exist on the R4600 - but not > all of them. > > Some pre-historic, somewhat over-agressive version of the Linux TLB refill > handlers got shot down by the tiny differences between the R4000 v2.2 > and v3.0 differences, since then I'm behaving like a burned child ;-) Hmmm, fun. I don't know my history as well as I should, I'm going to make a lot of mistakes I shouldn't, I'm sure. In any case, still a few hours of work before I can get along home and see if the TLB refill fix was enough to fix it. I was seeing something odd with badvaddr being a garbage value (0x10, or so), but that could be an artifact of my XTLB refill being broken, in interesting ways :) > > Either than, or the R4600 is just > > unsuitable for running with the KX bit set (extended kernel mode - 64 > > bit address space). I have other hardware on the way, and intend to > > get an O2 to play with, too. I'd like to keep things inclusive from > > an early stage, and not make any assumptions that might bite me badly. > > The other hardware on the way isn't something SGI :) > > I can assure you the R4600 is running with KX just fine. OK, good to know. > > Anyone interested in helping out is welcome to, I'm even willing to put > > in time to make my Perforce tree available to the outside (non-FreeBSD) > > world, if there's enough interest, or help out in other ways. Questions > > welcome :) > > For curiosity sake I'd be interested. Alright, let me see what kind of time I can put into it :) -- juli mallett. email: jmallett@freebsd.org; efnet: juli; aim: bsdflata; the finer things keep shining through the way my soul gets lost in you