Jeff Northon <jeffo@sasq.net> writes:
> We are looking into problems with 3Com905 cards that seem to loose the
> ability to autosence the media they are attached to. It looks like a part
> of the chipset exphy0 gets hosed and then the card just "Dumbs Down" 

This sounds odd.


> With exphy0 in place:
> 
> A piece from messages:
> 
> Jun 11 12:55:40 www3 /netbsd: ex0 at pci0 dev 11 function 0: 3Com
> 3c905B-TX 10/100 Ethernet
> Jun 11 12:55:40 www3 /netbsd: ex0: interrupting at irq 10
> Jun 11 12:55:40 www3 /netbsd: ex0: MAC address 00:50:04:71:f0:8c
> Jun 11 12:55:40 www3 /netbsd: exphy0 at ex0 phy 24: 3Com internal media
> interface
> Jun 11 12:55:40 www3 /netbsd: exphy0: 10baseT, 10baseT-FDX, 100baseTX,
> 100baseTX-FDX, auto

OK, card at PCI device 11, MAC 00:50:04:71:f0:8c


> After the "problem"
> 
> Jun 14 01:18:47 www3 /netbsd: ex0 at pci0 dev 9 function 0: 3Com 
> 3c905B-FDX 10/100 Ethernet
> Jun 14 01:18:47 www3 /netbsd: ex0: interrupting at irq 11
> Jun 14 01:18:47 www3 /netbsd: ex0: MAC address 00:50:04:a5:6e:75

Card at PCI device 9, MAC 00:50:04:a5:6e:75

I'd say that these are _different cards_, put into different slots,
etc.  (Maybe one's built on to a motherboard...)

What's the dmesg output 'around' the ex0 card messages in the second
example?

I'll note that the 3c905 on the motherboard of one of my boxes (a
DELL), says:

ex0 at pci1 dev 8 function 0: 3Com 3c905-TX 10/100 Ethernet
ex0: interrupting at irq 11
ex0: MAC address 00:c0:4f:b6:ec:30
nsphy0 at ex0 phy 24: DP83840 10/100 media interface, rev. 1
nsphy0: 10baseT, 10baseT-FDX, 100baseTX, 100baseTX-FDX, auto

i.e. it uses a phy other than the "exphy".



> What now?

Well, first, how about let's figure out what's really going on
w.r.t. the cards that you're trying to use...  8-)



cgd
-- 
Chris Demetriou - cgd@netbsd.org - http://www.netbsd.org/People/Pages/cgd.html
Disclaimer: Not speaking for NetBSD, just expressing my own opinion.