--- Tom Leffingwell <tom(a)sba.miami.edu> wrote:
Is it possible to boot RT-11 on an 11/23 via MOP? If so, can it
be done over ethernet or does it have to be a serial line?
Tom,
I know we've been discussing your PDP-11 off-list, but I wanted to
put forth this answer in public view because I think it's on-topic...
As others more qualified than I have said, there isn't a way to boot
RT-11 over an ethernet card; I am, however, confident you can boot RT-11
over a TU-58 and a serial port. I have never done it (I've always used
floppies or a hard disk), but with the right bootstrap on the -11, it
shouldn't be a big deal. You could then bring up your machine from an
emulated TU-58 (a DOS or Linux box, probably, or even VMS, with a little
development work to port an emulator) and tape images. It's a bit slow -
38400 bps is a standard transfer speed, but not as bad as the original
because you won't have enormous seek latency.
Also, it's possible to force-feed a bootstrap or any other memory
tidbits down the console line and ODT. We used that technique to
test COMBOARDs when I was at Software Results - the rig was software
on the VAX, test images, and a switch that gated the VT100 to be
attached only to the VAX; only to the console on the PDP-11; or
to pass data transmitted to the VAX into the console of the -11
and to echo the results back to the VT100 (using a three position
switch or a pair of ganged A-B switch boxes with some internal
re-wiring) The process went something like this...
o Set the switch to use the PDP-11
o Install the Unit Under Test into the PDP-11
o Power on the PDP-11 and issue a few ODT commands to verify the
board was alive and at the right CSR address
o Set the switch to use the VAX
o Log in and SET DEF to the directory with the test images
o Run the test program and give it the name of the test file
o Wait to see the message "Throw the Switch!"
o Set the switch to the third position so the VAX can send characters
to the PDP-11 console port.
o Watch the characters echo and see that ODT is in sync and is accepting
the data, not issuing errors (possible if you were in the middle of
an ODT command when the data started flowing from the VAX or if you
waited too long to throw the switch and the test program started
emitting characters before the PDP-11 could see them).
o When you see the flow stop, set the switch back to the first position
to use the PDP-11 from the terminal.
o Issue, I think, a G 1000, or whatever else was the right thing for
the flavor of ODT in your particular test box to execute code at
001000.
o Watch the pretty lights (if you attached the programmer's console to
the COMBOARD under test).
By using this mechanism, we were able to write programs for the -11 _or_
for the COMBOARD itself (by using ODT to stuff the data bound for our
product through its CSR and window registers) - the DMA test required
two load sessions - one for the COMBOARD and one for the PDP-11. The
software was just regular object files in a directory on the VAX; the
app we ran did the ODT conversion and formatting automatically. Today,
I would probably use a Perl script instead of a FORTRAN app.
Additionally, our test boxes were just a CPU, some nominal amount
of memory (32KW, I think) and a serial port. No programmer's console
on the -11 and no I/O besides our board and a console port; that way
*when* we blew up PDP-11s, they were cheap to repair. I have a small
box of dead Unibus RAM for the 11/34 that fell out of this process.
If I ever need to scavenge enough for an RT-11 box (as opposed to
a RSTS or Unix box), I have a place to start.
I hope this enlightens rather than obscures your options for feeding a
PDP-11 from a serial port in cases of no bootable devices or media being
available...
-ethan
__________________________________________________
Do You Yahoo!?
Try FREE Yahoo! Mail - the world's greatest free email!
http://mail.yahoo.com/