*BSD News Article 25804


Return to BSD News archive

Xref: sserve comp.os.linux.help:15506 comp.os.386bsd.questions:7826
Newsgroups: comp.os.linux.help,comp.os.386bsd.questions
Path: sserve!newshost.anu.edu.au!munnari.oz.au!hp9000.csc.cuhk.hk!cpccspc.cphk.hk!cs221442
From: cs221442@comp.hkp.hk (Hsieh Shing Leung Arthur (BACS3 Class A))
Subject: NE2000 problems in Linux & NetBSD
Message-ID: <CJFBu4.5y8@cpccspc.cphk.hk>
Keywords: NE2000 Ethernet
Sender: usenet@cpccspc.cphk.hk
Nntp-Posting-Host: csun50.comp.hkp.hk
Reply-To: cs221442@comp.hkp.hk
Organization: Department of Computing, Hong Kong Polytechnic
X-Newsreader: TIN [version 1.2 PL2]
Date: Mon, 10 Jan 1994 17:13:16 GMT
Lines: 56

I've a small LAN for connecting a Linux (0.99.14o) box and a NetBSD (0.9)
box together.  Linux box is an NFS server and NetBSD mounts its /usr
partition from Linux box.  Both have been using SMC ethercards (WD8013)
for more than 2 months without any problem.

Recently I replaced SMC cards with NE2000 clones.  The NE2000 is using
Winbond 89C901 chipset (is this similar to the one (83C901) mentioned
in Ethernet-HOWTO part 2 that we should watch out?).  Now I experience
the following problems:

1. Linux hangs frequently when the network load is high like recompiling
   NetBSD kernel across network and occasionally gets the following
   messages:

Jan  9 10:29:01 ahsieh vmunix: eth0: Receiver overrun.
Jan  9 10:29:01 ahsieh vmunix: eth0: transmit timed out, TX status 0x16,
ISR 0x0.
Jan  9 10:29:01 ahsieh vmunix: eth0: Possible network cable problem?

Jan  9 10:34:35 ahsieh vmunix: eth0: Receiver overrun.
Jan  9 10:34:52 ahsieh vmunix: eth0: transmit timed out, TX status 0x40,
ISR 0x40.
Jan  9 10:34:52 ahsieh vmunix: eth0: Possible IRQ conflict on IRQ9?

2. NetBSD always fails to mount /usr partition in time when it's booting
   up.  So it can't boot up properly as all daemons reside in /usr
   partition will be unavailable.  However, /usr partition does get
   mounted later but /etc/rc script has already completed.

3. Both Linux and NetBSD get following output when doing ping.

PING anbsd (200.1.1.2): 56 data bytes
64 bytes from 200.1.1.2: icmp_seq=0 ttl=178 time=89 ms
64 bytes from 200.1.1.2: icmp_seq=0 ttl=178 time=90 ms (DUP!)
64 bytes from 200.1.1.2: icmp_seq=0 ttl=178 time=90 ms (DUP!)
64 bytes from 200.1.1.2: icmp_seq=0 ttl=178 time=91 ms (DUP!)
64 bytes from 200.1.1.2: icmp_seq=1 ttl=179 time=3 ms
64 bytes from 200.1.1.2: icmp_seq=1 ttl=179 time=3 ms (DUP!)
64 bytes from 200.1.1.2: icmp_seq=1 ttl=179 time=4 ms (DUP!)
64 bytes from 200.1.1.2: icmp_seq=2 ttl=180 time=2 ms
64 bytes from 200.1.1.2: icmp_seq=2 ttl=180 time=3 ms (DUP!)

--- anbsd ping statistics ---
3 packets transmitted, 3 packets received, +6 duplicates, 0% packet loss
round-trip min/avg/max = 2/41/91 ms

   Duplicated packets are my worries.  However, dupicated packets
   won't appear if NetBSD box is booted up using Linux bootable disk.

It seems that both OSes don't like the newly bought NE2000 cards.
Could anyone tell me if all problems are related to these
brain damaged NE2000 clones or just the driver problem incidentally
happening in these 2 OSes?
--
Arthur Hsieh
cs221442@comp.hkp.hk