Memory error decoding w/ 1 Meg boards

[eichelbe at nadc.arpa]

I am currently running 4.1 BSD UNIX on a VAX 11/780 with 4 Meg of
memory in the form of 16 boards (1/4 Meg boards).  We are going to
upgrade our memory to 8 Meg of EMC memory with 64K ckips on the boards
(1 Meg per board).  I assume that the memory I currently has uses 16K
chips, and any memory diagnostics I get for soft errors always point me
to the bad board (and even the chip, if I do a little more decoding of
the error message).

The error diagnostics I am talking about are of the form:
   mcr0: soft ecc addr nnnnn syn mm
where "nnnnn" is a hex address, and "mm" is the hex value of the syndrome.

My question:  With the new memory (and controller, et al), will soft error
messages
(1) come out at all
(2) be meaningful for helping me find the failing board?

Also, if the messages are not meaningful or will not come out, is there
anything I can do to gets messages/get meaningful messages?  I have
source code!

   Thanks a lot.


          Jon Eichelberger
          eichelbe at NADC.ARPA