From owner-freebsd-hackers Wed Sep 10 23:18:43 1997 Return-Path: Received: (from root@localhost) by hub.freebsd.org (8.8.7/8.8.7) id XAA00368 for hackers-outgoing; Wed, 10 Sep 1997 23:18:43 -0700 (PDT) Received: from word.smith.net.au (word.smith.net.au [202.0.75.3]) by hub.freebsd.org (8.8.7/8.8.7) with ESMTP id XAA00360 for ; Wed, 10 Sep 1997 23:18:38 -0700 (PDT) Received: from word.smith.net.au (localhost.smith.net.au [127.0.0.1]) by word.smith.net.au (8.8.7/8.8.5) with ESMTP id PAA06738; Thu, 11 Sep 1997 15:46:47 +1000 (EST) Message-Id: <199709110546.PAA06738@word.smith.net.au> X-Mailer: exmh version 2.0zeta 7/24/97 To: John-Mark Gurney cc: Mike Smith , Eivind Eklund , freebsd-hackers@freebsd.org Subject: Re: PnP support In-reply-to: Your message of "Wed, 10 Sep 1997 23:10:15 MST." <19970910231015.01243@hydrogen.nike.efn.org> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Thu, 11 Sep 1997 15:46:47 +1000 From: Mike Smith Sender: owner-freebsd-hackers@freebsd.org X-Loop: FreeBSD.org Precedence: bulk > Mike Smith scribbled this message on Sep 11: > > > > > > well... actually.. this won't be neccessary soon.. I plan on adding > > > resource registering to the pnp code soon... once this is done, the > > > isa probe code should automaticly skip over this... > > > > Please, if you are going to do this, go back to the discussion that > > came up last time on this topic, most particularly the allocation and > > attachment strategy that Stefan and I discussed. Doing this job > > half-assed is just going to mean that it will have to be done again. > > Do it right and people will love you forever. > > are you talking about making auto resource assignment?? if you are, > that wasn't what I was refering too... I mean go back and read the discussion that came up last time "how to integrate PnP" was raised. Right now it's arriving via feeping creaturism, which is OK in that it gives us functionality that we didn't have, but it's basically Wrong, and gives us momentum in the wrong direction. If you have the time to do this, *please* take a little more of it to do it right. We've been over all this several times, and have hashed out some pretty good models. Doug Rabson got a long way through completely redesigning the ISA driver model with huge advantages before disappearing. Don't just throw all this work away. Please? > currently the PnP code DOES NOT claim that a device is using port space > (i.e. to prevent conflics) or memory space for that mater... this is > what I was talking about... not automaticly assigned free resources... If you want a decent resource/extent manager, have a look at the NetBSD extent manager. Whilst Jason may disagree, a little work would see it fitting _very_ well into the picture when it comes to carving up space, and we do need it very much. mike