VAX-11/730 and Emulex UC17 woes

[Mike Ross]

On Wed, May 25, 2016 at 7:27 AM, Josh Dersch
<JoshD at livingcomputermuseum.org> wrote:
> Hi all --
>
> I'm working on restoring a VAX-11/730 at the museum and things have been going pretty well thus far.  I've been bootstrapping the console and diagnostics from simulated TU58 (images from: https://github.com/NF6X/VAX-11-730-Console-v57).  All of the TU58-based diagnostics are passing.
>
> I'm attempting to bring up an Emulex UC17 SCSI controller for mass storage and I'm having trouble with it.  I thought I'd check with you guys to see if any of you have seen this issue or have any idea where I might be obviously going wrong before I start digging deeper into this.
>
> The current issue is that I can't get the UC17's built in diagnostic/utility (referred to as the 'FRD' in the manual) to run.  I am following all of the steps to the letter (see the manual here http://www.mirrorservice.org/sites/www.bitsavers.org/pdf/emulex/UC1751001-C_UC17_Dec90.pdf, pages 71-79 (section 4.5.7)) and I'm getting the right values back when examining the SA register during the process, but executing "S 80" halts after a second or so with:
>
> ?08  PC=00000298
>
> Which is an odd way for it to halt, 08 means "No user WCS" according to the 11/730 user's guide.
>
> Here's the full conversation, just in case:
>
>>>> I
>>>> D/L/P F26800 80000000
>>>> D/L/P F26804 80000001
>>>> D/W/P FFF46A 3003
>>>> E/W/P FFF46A
> P 00FFF46A      0100
>>>> D/W/P FFF46A 4401
>>>> E/W/P FFF46A
> P 00FFF46A      0400
>>>> S 80
>
> ?08  PC=00000298
>
> I've confirmed that the issue isn't with the card, I can run the FRD without issue on it, in an 11/44 we have here.
>
> I've done my best to ensure that everything is sane on the UNIBUS; my understanding from the 11/730 manuals is that by default none of the SPC slots have the NPG wire-wrap fitted and that any empty SPC slots need to have an NPG grant card installed.  (This makes sense given how difficult the backplane is to access, it requires pulling the power supply out first.)  Just to make sure, I have double-checked that the NPG wirewrap jumper is not present on Slot 10, where the UC17 is installed.  At the moment the grant chain should be unbroken as far as I can tell, here is the current configuration:
>
> TOP
> Slot 1 -        Empty (normally RB730 option)
> Slot 2-         Empty (normally FPA option)
> Slot 3- M8390 (DAP)
> Slot 4- M8391 (MCT)
> Slot 5- M8394 (WCS)
> Slot 6- M8750 (1mb memory)
> Slot 7- M8750 (1mb memory)
> Slot 8- M8750 (1mb memory)
> Slot 9- M8750 (1mb memory)
> Slot 10- Emulex UC17
> Slot 11- DMF32-AA
> Slot 12- M9302 terminator | G7273 grant
> BOTTOM
>
> Thanks as always for the help.
> - Josh

That sounds a HELL of a lot like the issues I was having - and which I
haven't resolved yet. Someone provided me with a hex dump of memory
starting at location 80 as you say and what I found was that the code
I was seeing that the UC17 had loaded into memory had several
locations with quite different values to those that had been dumped
from a working system. I did the obvious and manually deposited at
least the first few 'wrong' locations to the 'correct' values; the
behaviour changed. Instead of halting with an error the run light
stayed on but no utility menu was displayed; the machine was hung. I
recently obtained and very briefly tried a second UC17 with similar
results - except on this card (same firmware version) the values in
memory prior to 'S 80' were even more 'wrong'.

Haven't got further yet but will have a hack in a few days. Do you
know beyond peradventure that the NPG wire-wrap IS removed? I don't
and will have to check.

My 11/730 also  passed all diags - well until it got to the RL02 bit
where it failed due to no RL02!

Mike


http://www.corestore.org
'No greater love hath a man than he lay down his life for his brother.
Not for millions, not for glory, not for fame.
For one person, in the dark, where no one will ever know or see.'