Date: Fri, 18 Apr 1997 18:50:27 +0200 From: Stefan Esser <se@freebsd.org> To: "Daniel O'Callaghan" <danny@panda.hilink.com.au> Cc: dennis <dennis@etinc.com>, hackers@freebsd.org Subject: Re: Probing deflugalty Message-ID: <19970418185027.62509@x14.mi.uni-koeln.de> In-Reply-To: <Pine.BSF.3.91.970418085541.10264R-100000@panda.hilink.com.au>; from Daniel O'Callaghan on Fri, Apr 18, 1997 at 09:00:53AM %2B1000 References: <3.0.32.19970417185259.00b2cdf0@etinc.com> <Pine.BSF.3.91.970418085541.10264R-100000@panda.hilink.com.au>
next in thread | previous in thread | raw e-mail | index | archive | help
On Apr 18, Daniel O'Callaghan <danny@panda.hilink.com.au> wrote: > Like I said, numbering starts *after* the ISA numbers. I tried a kernel > config with > > ed0 > ed1 > ed2 at isa .... > > What happened was the PCI cards started at ed3 and ed4. So just added > > ed0 at isa > ed1 at isa > ed2 at isa > ed3 > > and now I have the flexibility of putting in more ISA cards if necessary, > and the PCI cards start at 3. ISA cards won't automatically number > themselves, so it makes sense to start the number for PCI *after* all of > the defined ISA devices. I have not looked at the code, but I assume > that the PCI attach code finds the end of the ISA devices somehow. Yes, it does. The unit number in the PCI attach code is initialized to "NED +1" ... I had considered several possibilities when I wrote the ED and LNC PCI attach wrappers, and found this to be the cleanest solution. If I get around to make "config" more PCI-aware, I intend to make device ed0 at pci? not reserve unit 0 for use by ISA ... Regards, STefan
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?19970418185027.62509>