Subject: Re: PROM Pass: ?
To: NetBSD/sparc Discussion List <port-sparc@netbsd.org>
From: Rick Copeland <rickgc@calweb.com>
List: port-sparc
Date: 11/05/1998 10:10:37
At 12:46 PM 11/5/98 -0500, Greg A. Woods wrote:
>[ On Thu, November 5, 1998 at 08:41:00 (-0800), Rick Copeland wrote: ]
>> Subject: Re: PROM Pass: ?
>>
>> Sorry about the loud text! I feel that a lot of people are not really
>> listening. I can't get to the monitor.  After the system checks, the 4/110
>> tries to boot from sd(0,0,0) then bounces to the PROM:pass requirement.  I
>> have tried to access via ttya in diagnostics mode but after system checks I
>> get locked out there as well (clever little PROM).
>
>Ah ha.  That's what I was wondering about....
>
>> I do not have the
>> original disk drive that was with the system and so I can't boot into it
>> and enter as root to change the eeprom that way.  I have tried to boot a
>> disk with Solaris on it I made on my Sparc 2 but it will not boot for some
>> reason.
>
>It's probably trying to boot a different filename, perhaps "/vmunix" if
>it was running SunOS-4 previously.  I'm pretty sure SunOS-5 will try to
>boot "/ufsboot" instead.  You'll have to get the right OS, or at least
>the right filename, to match what the PROM is trying to load.
>
>>  I may end up de soldering the chip extracting the data, writing it
>> to another (sans the password) and reinstalling it (of course I would put a
>> socket in there just in case).
>
>You'll have a heck of a time extracting the data unless you can mimic
>the boot environment it expects regardless of what machine it's plugged
>into.  You could build a little circuit to attach the chip to a computer
>interface of some sort.  I suppose if you're a real cowboy you might try
>to swap the chip into an already running system and then use the
>"eeprom" command to reprogram it.  Pretty risky though.
>
>However since the Sun 4/1xx machines don't keep the host-id and the MAC
>address in the PROM you may as well just toast the whole chip and put in
>a brand new one.  It'll be quite easy to re-construct the configuration
>information by hand -- far easier than trying to read it from the old
>chip.
>
The toasting idea is the way I will most likely end up going, now if I can
only identify the prom without any doubt I will pull it (shure hate to
remove the wrong chip and trash this cpu.  

Rick Copeland