Adapatec - ANA6911A/TX Combo - 10Base2 problem

[Julian Highfield]

I keep getting the feeling I'm only seeing half the conversation - anything
except Donald Becker's messages in fact!

This bit is Donald, I think:
>  >I don't have a 21143 with a 10base2 port: what is the exact message?
> 
>  >I hope to release 89L tomorrow after some more testing of the
>  >10-to-100-to-10 code.  I would do it tonight, but my private version has a
>  >bunch of extra code that I need to strip out and I make errors when I've had
>  >too little sleep.
 
Bruce Robertson:
>  One thing to note.  If in /etc/conf/modules I config options to force
>  10base2, whether it is the patched driver or not, in reality the UTP
>  fails to actually work.. even though the link light is one and traffic
>  is around.
> 
>  If I connect to a hib or switch via UTP and force the option, the driver
>  starts working after about a 15 second delay..  

To quote Olaf Schnapauff (we've been trying to debug this problem too):
> I am using 89H with the timer function replaced and that works even with
> 10b2 when given the option 12, it doesnt switch away.

So at least the card can be coaxed into life on 10base2.

For 10baseT and 100baseTX I am happily using 0.89K with the following diff:

31c31
< static int options[MAX_UNITS] = {11, };
---
> static int options[MAX_UNITS] = {0, };
732,733d731
<               /*** JCH. Ensure unused bits of CSR15 are turned off. (Mailing 
list.) ***/
<               outl(inl(ioaddr+CSR15) & ~(0x800037c0),ioaddr+CSR15);


Plugging the card into 10base2 and using the modified 0.89K gets me:

Aug 12 16:04:18 scully kernel: eth0:  MII transceiver found at MDIO address 1, 
 config 3100 status 7849.
Aug 12 16:04:21 scully kernel: eth0: Using user-specified media 10base2.
Aug 12 16:04:26 scully kernel: eth0: 21143 negotiation status 000000c2, 
 10base2.
Aug 12 16:04:26 scully kernel: eth0: 21143 negotiation failed, status 
000000c2.
Aug 12 16:04:26 scully kernel: eth0: Testing new 21143 media 10baseT.
Aug 12 16:05:26 scully kernel: eth0: 21143 negotiation status 000021c6, 
 10baseT.
Aug 12 16:05:26 scully kernel: eth0: 21143 negotiation failed, status 
000021c6.
Aug 12 16:05:26 scully kernel: eth0: Testing new 21143 media AUI.
Aug 12 16:06:26 scully kernel: eth0: 21143 negotiation status 000020c6, AUI.
Aug 12 16:06:26 scully kernel: eth0: 21143 negotiation failed, status 
000020c6.
Aug 12 16:06:26 scully kernel: eth0: Testing new 21143 media 100baseTx.
Aug 12 16:07:26 scully kernel: eth0: 21143 negotiation status 000000c6, 
 100baseTx.
Aug 12 16:07:26 scully kernel: eth0: The transmitter stopped!  CSR5 is 
 f0008102, CSR6 b2420200.

Using 0.89K gets very similar results, it just doesn't try so many 
alternatives after failing 10base2.

And to add a data point, my ANA6911A-TXC card has the following
identifying marking. I've included far more than I think is
necessary, just in case...

====== board ===========

Label stuck on 21143!

ANA-6911A/TX
  COMBO
1633600B
  9823


Barcode?:
BE0F82314A9

ASSY 1619606-00

PCB label:
FAB 1619607-00 REV E 

======= chips ============

digital
21143-PA
21-44085-10
(C) DEC 1996
DC1053B
A 9710 L53191-12



(NatSemi) ES9801BP
DP83840AVCE


HALO 9735
TG01-1006N
.HIPDTTED

(NatSemi) ES9803AT
DP83223V

SEEQ
NQ83C92A

Pulse
PE-68515L
U.S.PAT.5015981
9747-C CHINA

CSI
93C46P
9748G


| 269-3
| XASH

=========================

Regards,
        Julian.