Matthias Drochner wrote:

> simonb@netbsd.org said:
> >  a) Using sys/lib/libsa/nfs.c just plain doesn't work (haven't figured
> >     out why yet).  I had tried both the i386 nfs.c and the libsa one
> > in
> >     earlier attempts to get this working. 
> 
> You'll need the i386 nfs.c. The mi libsa one has a special entry for
> the mount call, which requires to bypass the libsa filesystem selection.
> That's not clean, and it causes larger code. The mi code should be
> changed of course, but a lot of md dev_net.c files depending on it.
> 
> I'm using tftp to load kernels, so I'm not up-to-date wrt nfs
> problems. When I've dealt with some other problems, I'll try it.
> (But alpha and suns use nfs loaded kernels more often, so I'd assume
> that we would have heard if there were big problems.)

Maybe the libsa nfs.c DTRT, whereas the i386 doesn't.  Now that I've
got a working netboot, I'll see where the MI nfs.c differs in path
handling...

> [3c905b]
> > > The buggy chips have a number ending with -001 on them,
> > > newer ones ending in -002 are OK.
> > 
> > Mine's a -004, so I guess that's good :-)
> 
> When you boot NetBSD, is the mii phy chip recognized? I've
> got serious problems with my -002 card's media handling,
> all mii registers are read as 0.
> (Have a 905b-FX too, which doesn't work either.)
> I'll try to find out some more...

This is what I get:

	ex0 at pci0 dev 16 function 0: 3Com 3c905B-TX 10/100 Ethernet
	ex0: interrupting at irq 11
	ex0: MAC address 00:10:5a:63:65:4f
	exphy0 at ex0 phy 24: 3Com internal media interface
	exphy0: 10baseT, 10baseT-FDX, 100baseTX, 100baseTX-FDX, auto

so everything seems fine...

Simon.