10/100 Base Card
   Daughter Card
ECA265 Defective 10/100 Ethernet MCA Cards
FC 2694 vs 2994
Drivers for 9-K

Base Card

J1,2 64 pin mini-C64 female U5 Adaptec ASIC-9060R

Y1 40.0000 MHz osc Y2 32.0000 MHz osc

U5 Adaptec ASIC-9060R BLLA741PT

There is a big cut out in the PCB at the bracket end.
Chip date stamps are '97.

Daughter Card P/N 93H3236

DS1 RCV/FDX DS2 100/LNK J1 RJ-45 P1,2 64 pin mini-C64 male T1 Pulse PE-68515

U3 ICS 1890Y U4,5 Alliance AS7C256-15JC U6 AMD AM79C971KC Y1 25.000 MHz osc Y2 R200H85JT xtal

U6 AMD AM79C971KC PCnet-Fast
Interrupt levels 10, 11, 12, 15
Busmaster 80 MB/sec data streaming
Max 1, 2 or 4 depending on the model number

ECA265 DEFECTIVE 10/100 ETHERNET MCA CARDS
   The MCA socket edgecard is NOT beveled on a few 9-K and 9-Q adapters. They can collapse the spring contacts inside the MCA socket on the planar. You can (sometimes) pry the spring contacts back into alignment (I know, I had to do it on my 7012-397).

Feature      FRU           ASM            SYSTEMS
FC2994    P/N07L6601   P/N93H7888   7012-Gxx, 7013-Jxx, 7015-RXX
FC2964    P/N07L6605   P/N93H8060   7006-4xx, 7009-Cxx, 7012-39X
                                    7013-59x, 7015-99x, 7030-3XT
 The current (fixed) FRU part numbers:  93H8022(SMP) or 93H1505(Uni) 

   This adapter can cause permanent damage to the I/O planar when  it is being installed or reseated. The slot pins on the I/O planar can become bent as the adapter is plugged into the slot.
   IMPORTANT:  This adapter should not be reseated, or reinstalled in any slot once the adapter has been removed from the system for ANY reason until the adapter has been reworked. The typical failure symptom is "failure to configure".

   Also inspect the I/O Planar slot for bent pins before installing the new adapter. If pins are bent, replace the I/O planar at the same time you are replacing the Ethernet card 

   Some of the adapters have been reworked to correct the problem, others have not.  To determine if the adapter has the problem, inspect the adapter in the tab area close to the I/O connector. The defective adapters will have a bottom edge of the chamfer along the tab that is blunt instead of beveled,(the edge of the chamfer will be flat rather than "V" shaped). See FIG. 1

BAD:                          FIG. 1
   _____________________________________________________________
   |                                                           |
   |                                                           |
   |                                                           |
   |   10/100 ETHERNET MCA CARD                                |
   |                                                           |
   |                                                           |
   |_______         _____           ___________________________|
          |         |   |           |            ____
          |         |   |           |            |  |
   large->'\_______/'   '\_________/'            |  | <---adapter
   chamfer                                       |__|
                                              ___    ___
                               I/O slot---->  | |    | |
                                              | |    | |  end
                                              |  \__/  |<-view
                                              |________|
GOOD:
   _____________________________________________________________
   |                                                           |
   |                                                           |
   |                                                           |
   |   10/100 ETHERNET MCA CARD                                |
   |                                                           |
   |                                                           |
   |_______         _____           ___________________________|
          |         |   |           |            ____
          |         |   |           |            |  |
   round->-----------   -------------            |  | <---adapter
   corner                                        |  |
                                              ___ \/ ___
                               I/O slot---->  | |    | |
                                              | |    | |  end
                                              |  \__/  |<-view
                                              |________|

  PHYSICAL CHECK:
   In addition, before removing the adapter you should inspect the metal end plate for a blue dot which indicates that the card has been reworked and should not be replaced. (see FIG. 2.)

                            FIG. 2. (end view)
                             _______
                             |     |
                             |     |
                             |  O  |<------blue dot
                             |     |
                             |     |
                             |     |
                             |     |
                             | ___ |
                             | | | |<------RJ-45 connector
                             | --- |
                             | ___ |
                             |_| |_|

Actually most of the drivers I wrote were written from scratch...the only one I modified was the NE/2 to get the D-Link card going, which uses a different way to store its MAC address.

The Adaptec interface chip is the real problem of the 9-K adapter.  The AMD Ethernet controller is well documented, but the problem is getting access to it.  I have a 9-K myself and did a few investigations.  The cards maps in a small I/O window, but this window is too small to map all the AMD's host registers.  Furthermore, since the AMD chip is a PCI chip, you wouldn't see its registers after power-on anyway.  PCI chips have to be configured by the host by assigning addresses to their PCI configuration registers.  On a 'standard' PCI-based system, this is done by the PCI BIOS.  Since this is not present on MCA-based RS/6000 systems, this job has to be done by the driver, and the Adaptec bridge must have some capability to perform PCI config space cycles.  PCI chips will remain silent as a grave in the I/O or memory space before they have
been configured (just like an MCA card disabled via the POS registers) How to  use the Adaptec chip to perform this is so far unknown (due to lack of documentation), the only soutions I could think of would be to reverse-engineer the AIX driver or to find someone inside IBM who has
access to this information.

Best regards Alfred
-- 
Alfred Arnold                   E-Mail: alfred@ccac.rwth-aachen.de