[tulip] AT-2800TX, I give up

[Gert Doering]

Hi,

On Wed, Feb 20, 2002 at 12:38:39PM -0500, Donald Becker wrote:
> > I have been bugging Allied Telesyn for documentation, but the customer 
> > support is stupid and unwilling - "we no linux driver, go away".  Will
> > keep trying.
> 
> They won't have real documentation.  Anything they do have will be so
> limited and obscure that they won't want to let someone from the outside
> see it.

Ummm.  The more I think about it the more I'd say you're right.  *sigh*.

> The primary information we need to know is the type of transceiver the
> card uses.  Even the transceiver chip number would allow us to find out
> if it really is a MII transceiver or a SYM transceiver.

I just checked - there doesn't seem to be any screws or anything
"obvious" to open the card case - is there an "established" trick to 
open PCMCIA cards without damaging them?  Or is this usually impossible?

> > > > eth0: Digital DS21143-xD Tulip rev 65 at 0xc80cc000, 00:A0:D2:AF:49:60, IRQ 9.
> ...
> > > > eth0: Digital DS21143 Tulip transmit timed out, status f0260000, SIA 000000c6 ffff0000 fffbff7f 8ff10008, resetting...
> 
> We might be able to tell something by seeing if any of the status bits
> change when the link cable or dongle is disconnected.  Use 'tulip-diag
> -af' to see the bit state.

OK, here we go (with the "hacked" tulip-diag)... -> no-go :-( - no
matter whether the "dongle" is connected to the card or not, or whether
a (10BaseT) cable is connected to the dongle, the output of 
"tulip-diag -af" is always:

tulip-diag.c:v2.09b 2/20/2002 Donald Becker (becker@scyld.com)
 http://www.scyld.com/diag/index.html
Index #1: Found a Digital DS21143 Tulip adapter at 0x200.
Digital DS21143 Tulip chip registers at 0x200:
 0x00: f8a08000 ffffffff ffffffff 06983800 06983a00 f0200100 320e0000 f3fe0000
 0x40: e0000000 ffffcbf8 ffffffff 00000000 000000c6 ffff0000 fff8ffbf 8ff10008
 Port selection is MII, half-duplex.
 Transmit stopped, Receive stopped, half-duplex.
  The Rx process state is 'Stopped'.
  The Tx process state is 'Waiting for Tx to finish'.
  The transmit threshold is 128.
  The NWay status register is 000000c6.
  Internal autonegotiation state is 'Autonegotiation disabled'.

> If the reset sequence is correct, bit 0x00080000 of CSR15 is a
> general purpose input.  It normally is Input Link Status for "OnNow"
> support.

The fact that the LED for link doesn't light sounds to me as if this
whole part of the card isn't active... so it might not know whether
a link is there or not.

> > > This reports a normal MII transceiver, with a reset sequence.  This table
> > > looks as if it was written for this card.  The usual way for the designer
> > 
> > So what do you suggest how to proceed?
> 
> Last night I modified the 'tulip-diag' program to reset the transceiver using
> the EEPROM table.  Run it as 'tulip-diag  --reset -amm' to send a reset
> to the control pins, and see if the transceiver is detected.

Did so.  Output is:

$ SU /tmp/tu --reset -amm
tulip-diag.c:v2.09b 2/20/2002 Donald Becker (becker@scyld.com)
 http://www.scyld.com/diag/index.html
Index #1: Found a Digital DS21143 Tulip adapter at 0x200.
Digital DS21143 Tulip chip registers at 0x200:
 0x00: f8a08000 ffffffff ffffffff 06983800 06983a00 f0200100 320e0000 f3fe0000
 0x40: e0000000 ffffcbf8 ffffffff 00000000 000000c6 ffff0000 fff8ffbf 8ff10008
 Port selection is MII, half-duplex.
 Transmit stopped, Receive stopped, half-duplex.
  The Rx process state is 'Stopped'.
  The Tx process state is 'Waiting for Tx to finish'.
  The transmit threshold is 128.
  The NWay status register is 000000c6.
   No MII transceivers found!
  Internal autonegotiation state is 'Autonegotiation disabled'.

LEDs do not change, eth0 was "down" when I tried this.  Enabling eth0
will then lead to the usual:

Feb 20 19:09:49 mobile kernel: eth0: Tx hung, 13 vs. 9.
Feb 20 19:09:49 mobile kernel: eth0: Digital DS21143 Tulip transmit timed out, status f0260000, SIA 000000c6 ffff0000 fff8ffbf 8ff10008, resetting...
Feb 20 19:09:49 mobile kernel: eth0: transmit timed out, switching to 10baseT media.
Feb 20 19:09:58 mobile kernel: eth0: Tx hung, 9 vs. 7.
Feb 20 19:09:58 mobile kernel: eth0: Digital DS21143 Tulip transmit timed out, status f0260000, SIA 000000c6 ffff0000 fffbff7f 8ff10008, resetting...
Feb 20 19:09:58 mobile kernel: eth0: transmit timed out, switching to 100baseTx media.

If I do it while the ethernet is running, the output is:

$ SU /tmp/tu --reset -amm
tulip-diag.c:v2.09b 2/20/2002 Donald Becker (becker@scyld.com)
 http://www.scyld.com/diag/index.html
Index #1: Found a Digital DS21143 Tulip adapter at 0x200.
 * A potential Tulip chip has been found, but it appears to be active.
 * Either shutdown the network, or use the '-f' flag to see all values.
Digital DS21143 Tulip chip registers at 0x200:
 0x00: f8a08000 ffffffff ffffffff 06983800 06983a00 f0260000 b3862002 fbfffbff
 Port selection is 100mbps-SYM/PCS 100baseTx scrambler, half-duplex.
 Transmit started, Receive started, half-duplex.
  The Rx process state is 'Waiting for packets'.
  The Tx process state is 'Waiting for Tx to finish'.
  The transmit threshold is 128.
  The NWay status register is 000000c6.

but it doesn't change anything either.

> The hacked version is at
>       ftp://www.scyld.com/pub/network/tulip-diag.c

I assume you meant the one in "/pub/network/test/tulip-diag.c" (as it
knows about -reset and has a time stamp of today).

gert
-- 
Gert Doering, Munich