Date: Sun, 10 Jun 2001 12:50:03 -0600 From: Warner Losh <imp@village.org> To: Doug Rabson <dfr@nlsystems.com> Cc: Sergey Babkin <babkin@bellatlantic.net>, j mckitrick <jcm@FreeBSD-uk.eu.org>, "Matthew N. Dodd" <winter@jurai.net>, Mike Smith <msmith@FreeBSD.ORG>, freebsd-hackers@FreeBSD.ORG Subject: Re: newbussifying drivers Message-ID: <200106101850.f5AIo3l21071@billy-club.village.org> In-Reply-To: Your message of "Sun, 10 Jun 2001 10:57:43 BST." <Pine.BSF.4.33.0106101056090.3242-100000@herring.nlsystems.com> References: <Pine.BSF.4.33.0106101056090.3242-100000@herring.nlsystems.com>
next in thread | previous in thread | raw e-mail | index | archive | help
In message <Pine.BSF.4.33.0106101056090.3242-100000@herring.nlsystems.com> Doug Rabson writes: : Actually, the ex driver does have an identify method. The issue is that : the identify method uses 'unallocated' ports to do its search. It ought to : allocate them, do its tests and release them for the real probe but it : gets messy due to the fact that a device instance doesn't exist yet. This is ugly. Obviously you don't want to tresspass into another device's address space. The other problem is that the other device might not exist yet and hasn't snagged its resources. You could use the isa bus itself at the resource holder for these temporary resources. One big area we need to clean up is PNPBIOS devices. They should be probed *FIRST* before anything else (which is hard because the isa bus gets attached late on pci systems). We'd need to update our drivers in the tree, plus there's the "hints vs pnp" issue where some people want to specify hints to wire a device unit (eg, the sio that's at address 0xXXXX should be sio9). This is one rats nest that's still around that is going to be hard to solve. Warner To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-hackers" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200106101850.f5AIo3l21071>