On Sat, Jul 13, 2002 at 05:46:06PM +0100, Matt wrote:
> On Sat, Jul 13, 2002 at 05:16:36PM +0100, Matt wrote:
> > 
> > Anyway, I can start playing now, and the machine certainly feels quite
> > nippy for its age. I guess because the machine autoboots fine, I don't
> > have the infamous busted firmware?
> 
> Aw smeg, I mistook rebooting from the NetBSD installer for behaving
> properly. A cold boot yields the "SYS 0512" boot error, which from a
> quick google tells me this is the b0rked firmware symptom.
> 
Hmm, I'm not sure if 1.5.2 had the extra sanity check agaist misprobing
by the dz driver, although I have not found a machine to submit to a test,
so I'm not sure if that additional check was sufficient anyway...

> Is there any way to fix this?
> 
Could you make a dump of your firmware for me?

Expect to take a couple hours to do the transfer, set your comm
program to log everything, and at the >>> prompt enter:

EX/P/L/U/N:20000 20040000

This will do an ascii dump of the machine's 512KB firmware.

If your comm program were "tip", then "tip myvax |tee console.dump"
is the sort of command I have in mind.

Once you've got your "console.dump", run:

awk '/^  *P 200/ { print $NF }' < console.dump | \
perl -ne 'print pack("V", unpack("N", pack("H8", $_)));' > 4k90firm.bin

And send me the "4k90firm.bin" file.

People have sent me pristine versions from a couple machines now, and
if yours is otherwise identical, perhaps we can establish once and for
all if this corruption is in the firmware. Maybe even correct it.

Anyone else that wants to make a firmware dump available along with
the status of their machine's power on tests will be helping as well.
When I get a firmware of the same version from both a pristine and a
corrupted machine, I'll do a binary compare and report the results.

Regards,

/Hugh

> Matt
> -- 
> "Phased plasma rifle in a forty-watt range?"
> "Hey, just what you see, pal"