Subject: More on SONIC driver
To: None <port-mac68k@NetBSD.ORG>
From: Mark Andres <mark@giganet.net>
List: port-mac68k
Date: 04/16/1997 18:06:26
Hi,

An update as I have been fighting with this SONIC driver all
afternoon.  The part that makes me want to cry is that for a
brief few minutes, I had it working perfectly.  Pings were
flying around at 2~3 milliseconds.  I could FTP from my IIcx
to my C650 at about 300K/sec.  Even FTP-ing to my Mac running
OT and Fetch, I was getting 70K/sec.  But I rebooted (to 
make sure it would stay fixed) and I am back to problems.

Here is what I did so far:

I installed EtherTalk and MacTCP on the MacOS side.  It worked
fine.  I removed it.  I used TechTool to zap the PRAM.  I 
booted into NetBSD.  The strange thing was that when I did 
ifconfig -a, the SONIC was not configured at all.  So I entered

  ifconfig sn0 inet 10.0.0.1 netmask 255.255.255.0

After that, the ifconfig -a looked OK.  That is when I was able
to ping and ftp to my hearts delight.  It was great to get 
300K/sec instead of 300 bytes/sec.  Maybe this is where I went
wrong.  So next, I put the above line in /etc/netstart.  I
then rebooted.  The ifconfig -a looked OK after bootup, but
I have the same problems as before.  I get a lot of 

  sn0: receive buffer area exhausted

and it says host is down.  I can ping 10.0.0.1, localhost, 
127.0.0.1, or ratbert all without any problems.  I did notice
a new line in /var/log/messages:

  routed[107]: lo0 has no address
  routed[107]: ifinit: out of sync

Since my network is isolated (not connected to "the Net"), I 
turned off routed (RIP is pretty braindead anyway).    

I rebooted without the ifconfig line in /etc/netstart and 
sn0 does not get configured.  I can configure it manually, and
it will work for a little while.  But after a sending a few 
ping packets, I start to get messages like:

  sn0: receive buffer area exhausted
  
and

  sn0: receive buffers exhausted

These messages fill up my screen and eventually take down the 
system.  The system halts and I cannot even get to the debugger.

I put the line back into /etc/netstart and restarted.  For now,
it is basically working.  Here is a ping:

ratbert% ping bullwinkle
PING bullwinkle (10.0.0.3): 56 data bytes
64 bytes from 10.0.0.3: icmp_seq=0 ttl=255 time=1.773 ms
64 bytes from 10.0.0.3: icmp_seq=1 ttl=255 time=2.056 ms
64 bytes from 10.0.0.3: icmp_seq=2 ttl=255 time=2.033 ms
64 bytes from 10.0.0.3: icmp_seq=3 ttl=255 time=2.038 ms
64 bytes from 10.0.0.3: icmp_seq=4 ttl=255 time=2.035 ms
64 bytes from 10.0.0.3: icmp_seq=5 ttl=255 time=2.041 ms
^C
----bullwinkle PING Statistics----
6 packets transmitted, 6 packets received, 0% packet loss
round-trip min/avg/max = 1.773/1.996/2.056 ms

But I am afraid to try a heavy FTP between machines because 
I don't want to get those "receive buffers exhausted" 
messages again.

My question now is, Where is the right place to put the 
ifconfig line and in what file?  Is there a problem with my
ifconfig line?

Thanks in advance for any advice.

Mark

Mark Andres             E-mail: mark@giganet.net
    Running NetBSD and 100% Microsoft Free!
URL: http://www2.giganet.net/private/users/mark/