Ethan Dicks <erd_6502(a)yahoo.com> wrote:
Is there a "toggle-in" VAX bootstrap for a
TS-11 or TU-80 or for
anything for that matter?
There are 5 different toggle-in bootstraps. TS-11 is one of them. As I
understand it, TU80 looks like a TS-11 to software, so the original enquirer
who was asking about TU80 should be all set.
Between 1984 and 1994, I installed a lot of
operating systems on PDP-11s and VAXen, and my experience is that
for VAXen, the magtapes were never bootable - you booted a console
floppy/tape/disk pack/whatever that had drivers for pre-TMSCP
tape drives (MT and MS) and you told the standalone programs which
disk to copy to and which tape to copy from (our primary machine
at work had a MASSBUS-attached TU-78 and a Unibus-attached TU80 - I
kept the TU-80 ;-)
That is how it works with DEC OSes. It is also one option with BSD. BSD,
however, offers another option, using only the magtape. Of course on the
original big VAXen the magtapes were not "bootable", but Appendix B of
Installing and Operating 4.3BSD UNIX on the VAX (SMM:1) contains little
programs in hex that you can type into memory from the console (called toggle-
in programs in reference to toggle switches that predated the console). There
is a different version for each kind of tape controller. Each program is to be
entered at 200 and run with certain registers manually set to the controller
CSR address, etc. Each time you run one of these programs, it reads one 512-
byte record from the magtape at 0 and halts.
The first file on the 4.3BSD distribution tape had 512-byte records and the
first 5 of them were secondary tape bootstraps for different tape controllers.
You would run your toggle-in program with S 200 the right number of times to
load the right secondary bootstrap and then run the latter with S 0. It would
give you a = prompt at which you type the name of the standalone program you
want to run. These standalone programs are stored in the first file on the tape
after the secondary bootstraps, and the secondary bootstrap programs load them.
Between the 5 secondary bootstraps (5 512-byte records) and the tp archive with
standalone programs (in the same tape file also with 512-byte records) there
were two unused records. I took advantage of this in 4.3BSD-Quasijarus0a to add
MicroVAX booting ability to the same tape without breaking the big VAX boot
mechanism described above. The MicroVAX tape boot program is 1024 bytes long
and neatly fit in the room that was left, but because MicroVAXen boot
automatically using their ROMs unlike the manual process above, the MicroVAX
boot program had to be first. To satisfy this requirement I simply rearranged
the order: now the MicroVAX bootblock is first, followed by the 5 big VAX
secondary bootstraps, followed by the tp archive with standalone programs.
The big VAX booting mechanism still works with the 4.3-QJ0a tape just like with
the original 4.3, the only difference is that because the secondary bootstraps
are now two records further down the tape, you type S 200 two more times than
the manual says.
Things changed a bit with MicroVAXen - early OS kits
assumed RX50, later
ones assumed TK50, but never magtape.
To the software there is absolutely no difference between TK50 and magtape as
long as both speak TMSCP. TMSCP magtapes are bootable on MicroVAXen just like
TK50s. This is the reason why I do not make separate dist tapes for big VAXen
and MicroVAXen, the same tape is used for both and you write magtapes and TK50s
in exactly the same way from the same files on my FTP site.
I've added
two blocks for MicroVAX booting in front of the first file on
the dist tape, but didn't update the Installing and Operating manual...
You will probably have to cruft up something close to that, but one
aware of the 11/750 if you are going to make a bootable magtape.
I don't need to cruft anything up: for big VAXen nothing changed except the
order of boot blocks, and I simply need to update the manual to increment by 2
the number of times one types S 200. I'll do that in the 4.3-QJ0b release.
MS