Kernel 2.3.36 + eepro100 v v1.09l on IBM Netfinity 3k

[Don Badrak]

Richard and others,

I did some heavy-duty searching of the mail archives, on this very same
problem, but on a different machine and kernel (PIII 550MHz w/ 1024MB
memory, Compaq 1850R, kernel 2.2.13ac3, onboard ethernet (tlan, not used),
and dual Compaq etherent (i82557/i82258 10/100 Ethernet)).

eepro100.c:v1.09t 9/29/99 Donald Becker
http://cesdis.gsfc.nasa.gov/linux/drivers/eepro100.html
eth0: OEM i82557/i82558 10/100 Ethernet at 0xfc9bb000, 00:08:C7:B2:F5:B7,
IRQ 15.
  Board assembly 703555-003, Physical connectors present: RJ45
  Primary interface chip i82555 PHY #1.
  Forcing 100Mbs full-duplex operation.
  General self-test: passed.
  Serial sub-system self-test: passed.
  Internal registers self-test: passed.
  ROM checksum self-test: passed (0x24c9f043).
  Receiver lock-up workaround activated.
eth1: OEM i82557/i82558 10/100 Ethernet at 0xfc9bd000, 00:08:C7:B2:F5:B8,
IRQ 15.
  Board assembly 703555-003, Physical connectors present: RJ45
  Primary interface chip i82555 PHY #1.
  Forcing 100Mbs full-duplex operation.
  General self-test: passed.
  Serial sub-system self-test: passed.
  Internal registers self-test: passed.
  ROM checksum self-test: passed (0x24c9f043).
  Receiver lock-up workaround activated.

Someone suggested trying out eepro100 v1.09t.  I did that, and all my
transmitter resets went away.  You may want to give it a shot.

I was able to force the trasmitter reset just by running ttcp a couple of
times in succession (sending).  After putting this version in, I was able
to get 89Mb/s sending (and receiving, FWIW) consistently, and NO resets
since.

On Thu, 6 Jan 2000 richard@iguana.co.nz wrote:

> PIII 500 w 64mb ram
> initrd loaded 2.3.36 linux kernel
> eepro100 1.09l modified to load properly into 2.3.36 kernel
> 
> Note I suffered an even worse problem of the same kind with the j revision
> that was in the kernel by default.
> 
> The internal ethernet card on the netfinity, an i82555, is having some
> fairly serious issues. The first is a transmit error (Which seems, from
> the mail list archives, to be a remarkably common error on the eepro's,
> and also appear to come in a large number of flavours). This does not
> appear to be causing network problems at the moment, so I'm hoping its
> benign :)
> 
> The second, which was prefixed by a similar set of information under
> 2.3.35 with the j driver, was rapidly followed by a repeating packet loss,
> approximately once every 2 minutes. This is NOT happening with .36 and the
> l driver, or at least does not appear to be at this time, so I assume the
> upgrade fixed it. I note only for those experiencing similar issues. (We
> experienced on average over 9% packet loss while using the 2.3.35/j
> default combo).
> 
> 
> Richard.
> 
> Dmesg output:

[ chopped ]

> eth2: Trying to restart the transmitter...
> eth2: Transmit timed out: status 0050  0070 at 20812/20824 command 000c0000.
> eth2: Tx ring dump,  Tx queue 20824 / 20812:

Don
-- 
Don Badrak <dbadrak@census.gov>              301.457.8263 work
Telecommunications Office                    301.457.4438 fax
U.S. Bureau of the Census
Suitland MD, USA


-------------------------------------------------------------------
To unsubscribe send a message body containing "unsubscribe"
to linux-eepro100-request@beowulf.org