Subject: Re: Unable to ifconfig - t1 100
To: None <port-sparc64@netbsd.org>
From: Tom Danielsen <tom@mnemonic.no>
List: port-sparc64
Date: 10/15/2002 22:50:00
On 15.10 20:26, Martin Husemann wrote:
> On Tue, Oct 15, 2002 at 11:38:40AM +0200, Tom Danielsen wrote:
> 
> > The problem is that I am not able to set an IP address on either
> > of the interfaces.
> 
> Try using them - they'll probably work.
> 
> On the miniroot (and install kernels with embeded ram disk) ifconfig is
> lying to you when reading back the inet address (toolchain problems, we worked
> around that for the next release.)
> 
> The standard installation does not have this problem, so once your system is
> up, all should be working.
> 
> Martin

I tried the interfaces, they did not work.

After a little searching I found people had the same problem with 1.5.1,
the solution was to use ifconfig from base.tgz.  I tried this - now I
see the IP address I set on the interface.

But I still can't make it work... :-(  And the symptoms are strange, I
can't quite figure it out.

After havnig configured hme0 I ping the netboot/jumpstart (Solaris)
server.  Using tcpdump on the Solaris machine I see ARP requests and
replies, but nothing else.  On the NetBSD side ping says "No route to
host".  "route -n show" lists route to 0.0.0.0 via the NetBSD MAC address
and another route to 0.0.0.0 via the Solaris MAC address (!).  ping
also says on the first line "ping 10.1.0.50 (0.0.0.0) ....", there
seems to be a problem with resolving.  I don't have a resolv.conf,
and both hosts (and localhost) is listed in hosts.

The hardware is fine, and I'm able to boot Solaris on the T1
over the net.



# ifconfig hme0 inet 10.1.0.54 netmask 0xffffff00
# ifconfig hme0
hme0: flags=8a63<UP,BROADCAST,NOTRAILERS,RUNNING,ALLMULTI,SIMPLEX,MULTICAST> mtu 1500
        address: 08:00:20:c1:dd:XX
        media: Ethernet autoselect (100baseTX full-duplex)
        status: active
        inet 10.1.0.54 netmask 0xffffff00 broadcast 10.1.0.255
        inet6 fe80::a00:20ff:fec1:ddXX%hme0 prefixlen 64 scopeid 0x1
#
# route -n show
Routing tables

Internet:
Destination       Gateway            Flags
0.0.0.0           link#1             U

...IP6

# ping 10.1.0.50
PING jumpqfe3 (0.0.0.0): 48 data bytes
ping: sendto: Host is down
ping: sendto: Host is down
ping: sendto: Host is down
ping: sendto: Host is down
^C
----jumpqfe3 PING Statistics----
10 packets transmitted, 0 packets received, 100.0% packet loss
#

========== It seems I am not able to reproduce the funny routing table

# route -n show
Routing tables

Internet:
Destination       Gateway            Flags
0.0.0.0           link#1             U
0.0.0.0           link#1             UHR




========== As seen on the Solaris machine when pinging:

tcpdump: listening on qfe3
22:42:18.650187 8:0:20:c1:dd:XX ff:ff:ff:ff:ff:ff 0806 60: arp who-has 10.1.0.50 tell 10.1.0.54
22:42:18.650266 8:0:20:f6:95:XX 8:0:20:c1:dd:XX 0806 42: arp reply 10.1.0.50 is-at 8:0:20:f6:95:XX
22:42:19.651190 8:0:20:c1:dd:XX ff:ff:ff:ff:ff:ff 0806 60: arp who-has 10.1.0.50 tell 10.1.0.54
22:42:19.651265 8:0:20:f6:95:XX 8:0:20:c1:dd:XX 0806 42: arp reply 10.1.0.50 is-at 8:0:20:f6:95:XX
22:42:20.661294 8:0:20:c1:dd:XX ff:ff:ff:ff:ff:ff 0806 60: arp who-has 10.1.0.50 tell 10.1.0.54
22:42:20.661371 8:0:20:f6:95:XX 8:0:20:c1:dd:XX 0806 42: arp reply 10.1.0.50 is-at 8:0:20:f6:95:XX