From owner-freebsd-mobile Tue Apr 6 10:53:43 1999 Delivered-To: freebsd-mobile@freebsd.org Received: from boreas.isi.edu (boreas.isi.edu [128.9.160.161]) by hub.freebsd.org (Postfix) with ESMTP id 6B41A14BF5 for ; Tue, 6 Apr 1999 10:53:41 -0700 (PDT) (envelope-from faber@ISI.EDU) Received: from ISI.EDU (vex-e.isi.edu [128.9.160.240]) by boreas.isi.edu (8.8.7/8.8.6) with ESMTP id KAA27663; Tue, 6 Apr 1999 10:51:37 -0700 (PDT) Message-Id: <199904061751.KAA27663@boreas.isi.edu> X-Mailer: exmh version 2.0.2 2/24/98 To: Nate Williams Cc: NAKAGAWA Yoshihisa , Nick Sayer , freebsd-mobile@FreeBSD.ORG Subject: Re: Any success with CirrusLogic 6729/6730??? In-reply-to: Your message of "Tue, 06 Apr 1999 11:32:31 MDT." <199904061732.LAA16641@mt.sri.com> X-Url: http://www.isi.edu/~faber Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Date: Tue, 06 Apr 1999 10:51:37 -0700 From: Ted Faber Sender: owner-freebsd-mobile@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org Nate Williams wrote: >This was my idea about a 'boot' config line. The problems with this >solution is: > >1) It assumes there is one PCIC controller. (The existing code has the > same problem) >2) It not 'boot-time' configurable. It requires you to edit a file and > then reboot the box before the change can take place. There's a kld version of the pcic code, right? Can we use that to our advantage? I.e. boot set sysctl variables load the kld We could then set up 2 variables, current unit number and current io port. The on load the kld reads the variables, sets its unit number and io port appropraitely and starts up. The sequence for multiple pcics then looks like boot set sysctls for 1st controller load 1st controller set sysctls for 2nd controller load 2nd controller etc. I think that allows each driver to continue to think that there's one pcic. I haven't checked the code to be sure that it would work, though. For one thing, I'm not sure how it allocates irqs, and certainly don't want it stepping on itself. I'm not sure how the PCI code interfaces with the kld stuff, but I doubt that it (currently) reprobes the PCI bus when a kld is loaded. We may be able to get around this by leaving one version oc the pcic driver in the kernel, but it's probably neater to create a device that only initializes the PCI pcics, leaving them in an appropriate configuration for the dynamically loaded pcic klds to find them. Of course, that idea may be hacky enough that its better to just bite the bullet, and try to get the pcic code to be aware of multiple pcics directly. ---------------------------------------------------------------------- Ted Faber faber@isi.edu USC/ISI Computer Scientist http://www.isi.edu/~faber (310) 822-1511 x190 PGP Key: http://www.isi.edu/~faber/pubkey.asc To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-mobile" in the body of the message