Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!usc!snorkelwacker.mit.edu!bu.edu!att!cbnewsj!nhc
From: nhc@cbnewsj.att.com (n.h.chandler)
Newsgroups: comp.os.minix
Subject: Re: Problems with Minix Amoeba System
Keywords: amoeba ethernet WD8003E
Message-ID: <1991Feb3.191259.20424@cbnewsj.att.com>
Date: 3 Feb 91 19:12:59 GMT
Organization: AT&T Bell Laboratories
Lines: 53


In Article <XBR4D75G%DDATHD21.BITNET> Hans-Juergen Knobloch, (TH Darmstadt,
Germany ) writes
>Subject: Problems with Minix Amoeba System
>Message-ID: <26461@nigel.udel.EDU>
>Date: 2 Aug 90 15:26:06 GMT

>After upgrading my MINIX-system to version 1.5.10, I finally found some time
>to generate the corresponding AMOEBA-version. The generation of my new

-other stuff

>The included examples server1/client1 and server2/client2 are running
>local on one PC successfully but I could not run them between two
>different PC's.
>Ethernet-cards are installed on both PC's and when booting, the
>hardware-address is shown on the screen by the minix system.

>When I started the examples server1/client1 on one machine, I used
>the other PC to monitor the the ethernet with "NETWATCH", here are the 4
>broadcast messages I observed on the net:

>0040   FFFF FFFF FFFF 0000  C0CD A517 8145 FF00                 E
>0050   0000 0000 1E00 0001  7879 7A00 0000 7624            xyz   v$
>0060   0F0F 0F0F 2F1F 0F0F  4949 4949 4959 4B49        /   IIIIIYKI
>0070   71F0 E0E8 F890 F8E0  DCB7 B6B3 3D00 3C00    q           = <
>
-other stuff
>
>When running only client1 on a machine, I had exspected to see at least
>one broadcast message on the net, but nothing appeared.
>
>Has anyone similar problems or any idea to fix this network problem?
>
>about interrupt level (0x3), io-address (0x280) and memory-address (0xd000),
>have I forgotten something during system generation?
>
>Any help and hints are appreciated.
>

I am using Minix 1.5 (PH 1991) and have the same problem with 
Amoeba (exactly). Can anyone suggest a solution?  
For both machines int-level (0x03), io-address (0x280) and
mem-address (0xd400).
Both systems execute server[n]/client[n] locally but fail
the remote tests.  With client1 running on either machine,
and NETWATCH on the other, traffic on the ethernet line is displayed.
The hardware passes all diagnostic tests under MS-DOS.
Any help or hints will be appreciated.

Neville H. Chandler
 nhc@cbnewsj.att.com
 nhc@mtdcc.att.com