sunhme ethernet (receiver?) locking up w/ ultra1 and 2.4.18-.20

[Peter Parkkali]

[My apologies if this is a FAQ; I couldn't find any info on this..]
Hello,

Has anyone else experienced strange lockups with the HME ethernet  
interface
on an Ultra 1? I've had this with 2.4.18 and .20 kernels.

At some random moment, sometimes after moving a "large" amount of
data (from an nfs mount, etc), the machine stops receiving packets.
Data is still sent; I can see the ARP queries with tcpdump on another
machine. tcpdump running on the buggy machine itself can only see the
sent packets.

It seems the receiving part of the i/f gets jammed and never recovers.

This is what the kernel complains when the lockup happens:
  --------><----10M net  
--------><-------------------><-------------------------
  NETDEV WATCHDOG: eth0: transmit timed out
  eth0: transmit timed out, resetting
  eth0: Happy Status 03030000 TX[000003ff:00000101]
  eth0: Auto-Negotiation unsuccessful, trying force link mode
  eth0: Link has been forced up using internal transceiver at 10Mb/s,  
Half Duplex.
  --------><----100M  
net--------><-------------------><-------------------------
  NETDEV WATCHDOG: eth0: transmit timed out
  eth0: transmit timed out, resetting
  eth0: Happy Status 03010000 TX[000003ff:00000301]
  eth0: Link is up using internal transceiver at 100Mb/s, Full Duplex.
   
--------><------------------><------------------- 
><-------------------------

The only way I've managed to get the i/f working again is by rebooting -
module unload+load doesn't help the problem. A pain in the *ss to put  
it lightly..

I believe this is a problem with the driver not being able to reset the  
card properly,
since I ran NetBSD1.6 on this machine for some weeks without ever  
encountering this..
(would still if it didn't lack apt.. ;)

The box claims to be "Sun Ultra 1 UPA/Sbus (UltraSPARC 200MHz),  
OpenBoot 3.3,
Serial #8944688".

 /pp

--
p f p @ i k i . f i,   {040, +358-40} - 532 95 80
look like this pie lake is empty