Summary: Sun386i ie0 no intr problem

From: Andreas Junklewitz (ajass@rama.informatik.rwth-aachen.de)
Date: Mon Nov 01 1993 - 11:00:24 CST


Hi netters,

my original posting three weeks ago was:

>I have a strange problem with my Sun386i.
>
>If I boot, I get the message: netconfig ie0 no intr, network down;
>and any further network operation fails with a "network down" error.
>
>The configuration of the Sun (and the network) were not changed for months.
>
>If I use another transceiver or leave the network stuff away, the error
>remains.
>
>If I run the diagnostic program from Sun, it shows no failures !!!
>
>If I test the transceiver and the same ethernet connection with a Sparc, it
>works fine.
>
>I'm a little confused about the type of error occured. Should I look for
>a hardware or software fault ? In the Sun Managers Archive I have not found
>any hints, because a Sun386i is really old fashioned.

There are two different errors that can cause these effects:

a) a broken fuse for the ethernet driver on the motherboard.
   The fuse is close to the AUI port. Sun's diagnostic program does not
   detect this failure!

b) wrong EEPROM data in the NVRAM due to a weak battery-on-chip.
   The old lithium cell on the TIMEKEEPER chip cannot provide power
   anymore, if the machine is switched off and the NVRAM data get screwed up.
   This was the problem with my Sun386i.

   The problem was discussed in the sun-386i-mailing-list in June 1993.
   sun-386i@ssg.com is the mail address of the list, sun-386i-request@
   ssg.com is the mail address to join.

   Mr. Malome Khomo sent me a summary of this discussion with all necessary
   details. If anyone is interested I will mail the details. I think these
   articles are to long for a posting into the net.

   I had to buy a new TIMEKEEPER (tm) RAM, made by SGS-Thomson Microelectronics,
   with the type id MK48T02B-25. Afterwards I had to enter the correct NVRAM
   settings with the boot monitor. This process and a lot of settings are
   described in the Sun386i NVRAM Layout chapter of the PROM User's Manual.

   I made the experience that only the values at the addresses 0x110, 0x111 and
   0x112 are important in respect to the ie0-problem.
   Mr. Christopher Demke wrote in the Sun386i-mailing-list about these values:

10 110 CPU rev. level (should be 10)

03 111 CPU artwork 01=1.5 (none in field)
                               02=2.0
                               03-3.0
       ************************************************************************

       correct values for location 111 are:

       0x01 -- p1.5 cpu board
       0x02 -- p2 cpu board
       0x03 -- p3 cpu board

                       501-1241 p1.5 cpu n/a (you should not
                                                             have this board)
                       501-1241 20MHz p2 cpu 150
                       501-1324 25MHz p2 cpu 250
                       501-1414 20MHz p3 cpu 150
                       501-1413 25MHz p3 cpu 250
       ************************************************************************

00 112 ECO rev. level (should be either 00 or 02)

       ************************************************************************

       correct values for location 112 are:

       0x00 -- p1.5 cpu board
       0x00 -- p2 board with babe-01
       0x02 -- p2 board with babe-02
       0x00 -- p3 cpu board (all)

       to determine whether you have a p2 with a babe-01 or a babe-02, look at
       the revision number on the cpu card part number label.

       the babe-01 to babe-02 switch was made at different revision levels for
       20MHZ and 25MHz p2 cpu cards.

       babe-02 should be found in rev 16 and later 20MHz p2 cpu cards.
       babe-02 should be found in rev 27 and later 25MHz p2 cpu cards.
       (or you can look at the BABE and see if it is a -02)
       I couldn't find the label, but I see no babe; 0300 is my guess. -jek

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

My CPU board number is 501-1244 and 01 or 03 at address 0x111 works fine,
02 or 00 at this address causes an ie0-no-intr-error.

Thank you very much for your answers:

Mr. Thomas Brupbacher (thomas@mw.lpc.ethz.ch)

Mr. Malome Khomo (mkhomo@genesis.mcs.com)

Mr. Brian Michael Patten (patten@lumchan.IFA.Hawaii.Edu)

Mr. Chuck Smoko (csmoko@relay.nswc.navy.mil)

Mr. Larry Wagner (wagner@weru.ksu.edu)

Mr. Anthony Worrall (Anthony.Worrall@reading.ac.uk)

Bye,
       Andreas

===========================================================================
email: ajass@pool.informatik.rwth-aachen.de
               junklewitz@ggind.ind.rwth-aachen.de
Andreas Junklewitz, Institute for Signal Equipment and Data Processing,
RWTH-Aachen, 52072 Aachen, Germany



This archive was generated by hypermail 2.1.2 : Fri Sep 28 2001 - 23:08:27 CDT