Subject: Re: de0 any??
To: Matt Thomas <matt@lkg.dec.com>
From: James Lever <J.Lever@mailbox.uq.oz.au>
List: port-i386
Date: 12/11/1995 23:02:21
Me> 2.	I'm using one of the new SMC9332 PCI e/net cards and there appears
Me> 		to be no network traffic using it.. I've attempted to set
Me> 		it to use utp/10Mbps using the config disk but it still
Me> 		doesn't like it under netbsd..

Peter Galbavy> I am using >15 or these cards (I assume these are the 21140 based
Peter Galbavy> cards using Matt Thomas' if_de driver), and they have all worked
Peter Galbavy> first time. I don't know whether they autodetect at all (Matt ?)
Peter Galbavy> since I only use them in a 10BaseT environment for now.

Matt Thomas> The driver will autodetect 10 or 100Mb mode upon boot.  If neither 
Matt Thomas> is detected it will use 10Mb/s.  The driver *never* selects the 
Matt Thomas> STP port.  Note that the driver will tell whether it is using 
Matt Thomas> 10Mb/s or 100Mb/s in the boot messages.

Matt Thomas> Can post the output of dmesg to help diagnose your problem?

At the bottom of message.  Everything looks fine, I can play with the card, 
switch between 100 and 10 Mb/s mode using link0 and -link0 and the like.  I
still can't see anything on the network.

I've recompiled the kernel with minimal config for this system only and booted
single user, ifconfig'd the device, and attempted to ping out of it.  no go.
I've tried this with 3 cards under the de0 driver, no go.  a dlink 100/10 Mb/s,
an SMC 100/10 Mb/s, and an SMC 10 Mb/s pci card.  None of them seem to work.
they all do the same thing (nothing) and they all work under dos. 

method.  boot single user.  

	ifconfig de0 130.102.128.66 netmask 255.255.255.0

that works.

	route add default 130.102.128.30

no problems, just like a normal machine.

	ping 130.102.128.30

hmm, sits there for a while and then claims the host is down..
checked the hub, and it thinks there's a link there.  The port works on the
ed2 I have in the box (unconfigured during these tests.  Used when de0 breaks).

This is exactly the same with all 3 cards.. I hope there is something stupid
I've forgotten, but this hasn't worked since we got them.

Peter Galbavy> I guess you have checked that there is a hub powered on at the 
Peter Galbavy> other end etc etc... have you tried another card in case it is 
Peter Galbavy> a hardware fault ?

Hub works, no hw fault in the card, any of them.

Matt Thomas>I'd try another card if possible.

yup, 3 of them.. suggestions?

James

----
pci section of boot messages
----
Dec  7 21:28:30 gumble /netbsd: pci0 (root): configuration mode 2
Dec  7 21:28:30 gumble /netbsd: pci0 bus 0 device 0: unknown vendor/product: 0x8086/0x04a3 (class: bridge, subclass: host, revision: 0x11) not configured
Dec  7 21:28:30 gumble /netbsd: pci0 bus 0 device 4: unknown vendor/product: 0x1002/0x4758 (class: display, subclass: VGA, revision: 0x01) not configured
Dec  7 21:28:30 gumble /netbsd: pci0 bus 0 device 11: unknown vendor/product: 0x8086/0x0482 (class: prehistoric, subclass: miscellaneous, revision: 0x03) not configured
Dec  7 21:28:30 gumble /netbsd: de0 at pci0 bus 0 device 12pci_map_int: pin A mapped to line 15
Dec  7 21:28:30 gumble /netbsd:  irq 15: DC21140 [10-100Mb/s] pass 1.1 Ethernet address 00:80:c8:2a:6c:4a
Dec  7 21:28:30 gumble /netbsd: de0: enabling 10baseT UTP port
Dec  7 21:28:30 gumble /netbsd: ncr0 at pci0 bus 0 device 15
Dec  7 21:28:30 gumble /netbsd: pci_map_mem: mapping memory at virtual fa013f00, physical 3efeff00
Dec  7 21:28:30 gumble /netbsd: pci_map_int: pin A mapped to line 11
Dec  7 21:28:30 gumble /netbsd: ncr0: restart (scsi reset).
Dec  7 21:28:30 gumble /netbsd: scsibus0 at ncr0
Dec  7 21:28:30 gumble /netbsd: ncr0 targ 0 lun 0: <SEAGATE, ST3655N, 0190> SCSI2 0/direct fixed
Dec  7 21:28:31 gumble /netbsd: sd0 at scsibus0sd0(ncr0:0:0): FAST SCSI-2 100ns (10 Mb/sec) offset 8.
Dec  7 21:28:31 gumble /netbsd: : 520MB, 2493 cyl, 5 head, 85 sec, 512 bytes/sec