[vortex] Re: [tulip] Re: [eepro100] True on TRANSMIT ERROR TIMEOUT

Homer Wilson Smith homer@lightlink.com
Tue, 13 Jun 2000 11:47:08 -0400 (EDT)


> And there are many, many times when a wedged driver can be resurrected
> by a down/up or a rmmod/insmod.  This means that the driver _could_ have
> automtically recovered in tx_timeout, but it simply did not do so.

     What is a wedged driver?

     I have 4 linux boxes running 2.0.36 and 21140 Kingstons running
tulip.c v91g.  Running fine for *YEARS*.

     *ONE* of these boxes has taken to locking up the etherport every
night during a mirror copy of /dev/hda1 to /dev/hdb1, nothing
to do with network traffic.  ifconfig down and upping the eth0
clears it out.

     While locked, the eth0 responds to pings but won't talk.  Is
there some diagnostic I can run to find out what it is complainging about?

     It used to happen rarely, then once a week, and now once a night
during the same mirror run.  

     I replaced the card, but its still doing it.

     Is this a hardware motherboard problem?  Driver?  Stupid user, me?

     Haven't swapped the motherboard yet, but going to next.

     Homer