[tulip-bug] troublesome linksys NIC

[Donald Becker]

On Tue, 18 Dec 2001, Keith Warno wrote:

> I have a troublesome linksys LNE100 NIC in a linux box currently running 
> kernel 2.2.16 (yes, I admit, it's old).  The NIC likes to time out as 
> seen by the following blurb from syslog:

2.2.16 isn't old.  Some machines haven't been rebooted since then.
Now, the reports based on 1.2.13 ...

> eth0: PNIC2 transmit timed out, status e4660000, CSR6/7 01002002 / 
> effffbff CSR12 41e1d0cc, resetting...

Does the interface continue to work after this?

Did you see a messge that reported
    printk(KERN_WARNING "%s: Tx hung, %d vs. %d.\n",


> The status codes are meaningless to me, but I'm sure not to Mr. Becker. 
> :-) Can you translate them please?  I'm wondering if this is a hardware 
> or a software issue.

The status indicates
    CSR12  41e1 ....	Your link partner advertised 01e1
	It's a 10/100 H/FDX switch, and it's working properly.
    CSR6   0100 2002    100baseTx scrambler, no FDX(?), TX and Rx enabled
    CSR7   efff fbff	All normal interrupt sources enabled

Now the interesting register
    CSR5   e466 0000
     Transmitter is idle, Receiver is waiting for packets, no errors reported

So, it appears that everything is working properly, a Tx-done interrupt
was missed for some reason.  As long as the interface continues to run
you can ignore this message.

I will need to know if the "Tx hung..." message appear to know the
failure path.  I'll check for this state and emit more info in the next
driver release.

Donald Becker				becker@scyld.com
Scyld Computing Corporation		http://www.scyld.com
410 Severn Ave. Suite 210		Second Generation Beowulf Clusters
Annapolis MD 21403			410-990-9993