3c59x.c V 0.99G tx_status problem (fwd)

[Allen Mcintosh]

As a data point, I hit the infamous tx_status problem today with this
NIC:

Nov 10 15:11:26 dusty1 kernel: 3c59x.c:v0.99G 9/25/98 Donald Becker http://cesdis.gsfc.nasa.gov/linux/drivers/vortex.html
Nov 10 15:11:26 dusty1 kernel: eth0: 3Com 3c905B Cyclone 100baseTx at 0xec00,  00:c0:4f:79:a2:d8, IRQ 14
Nov 10 15:11:26 dusty1 kernel:   8K byte-wide RAM 5:3 Rx:Tx split, autoselect/Autonegotiate interface.
Nov 10 15:11:26 dusty1 kernel:   MII transceiver found at address 24, status 782d.
Nov 10 15:11:26 dusty1 kernel:   Enabling bus-master transmits and whole-frame receives.

...

Nov 10 15:11:36 dusty1 kernel: eth0: transmit timed out, tx_status 00 status 6000
Nov 10 15:11:41 dusty1 kernel: eth0: transmit timed out, tx_status 00 status e000

This was with a 2.0.35 kernel. The owner wanted the machine up ASAP, so
we put in a 3C900 & disabled the 905 (it's built in).  As a result I
didn't get a chance to see if the problem was the driver or the line.
However, the 905 NIC worked just fine with the version of NT installed on the
machine (a Dell) and the 900 was perfectly happy with the same network
configuration parameters, so my suspicion is that there may still be a
problem with the driver...