2 Feb
2001
2 Feb
'01
1 p.m.
No, it won't show up on the bus. All transactions on that board occur
over the cable interconnect.
Megan Gentry
Former RT-11 Developer
+--------------------------------+-------------------------------------+
| Megan Gentry, EMT/B, PP-ASEL | Internet (work): gentry!zk3.dec.com |
| Unix Support Engineering Group | (home): mbg!world.std.com |
| Compaq Computer Corporation | addresses need '@' in place of '!' |
| 110 Spitbrook Rd. ZK03-2/T43 | URL: http://world.std.com/~mbg/ |
| Nashua, NH 03062 | "pdp-11 programmer - some assembler |
| (603) 884 1055 | required." - mbg KB1FCA |
+--------------------------------+-------------------------------------+
3 Feb
3 Feb
6:56 a.m.
Megan wrote:
No, it won't show up on the bus. All transactions
on that board occur
over the cable interconnect.
Jerome Fine replies:
I knew that part already - must be that way since the uVAX II can use
16 MBytes of memory whereas the Qbus can handle only 4 MBytes
of memory.
I was sort of hoping (dreaming) that someone has noticed these 8 Mbyte
boards going to sleep and the PDP-11 still looking for additional
memory. I realize it is probably impossible, but how does the 11/83
manage to use both the Qbus for disk I/O and the PMI for memory?
And if the PMI memory is below the CPU rather than on top, the 11/83
reverts to an 11/73.
If I ask enough dumb questions, maybe one of them might lead somewhere?
Didn't you once make a (very small) disk memory out of some old DEC
memory board? Or was that normal Qbus memory but at a high address
with a gap from the regular memory.
What would it take to build a small interface that looks like the HD hard
disk from E11 as far as an actual hardware interface is concerned? Could
it loop right back into this uVAX II memory board via the 50 pin header
on top? Since the Qbus interface is already there in the backplane, maybe
that interface could be added to the 8 Mbyte of memory? See, I said I
would ask dumb questions!
Since I know nothing at all about hardware design, I can ask all sorts of
questions since I don't know enough to know when something is impossible.
Sincerely yours,
Jerome Fine
2:43 p.m.
I was sort of hoping (dreaming) that someone has
noticed these 8 Mbyte
boards going to sleep and the PDP-11 still looking for additional
memory. I realize it is probably impossible, but how does the 11/83
manage to use both the Qbus for disk I/O and the PMI for memory?
And if the PMI memory is below the CPU rather than on top, the 11/83
reverts to an 11/73.
As I understand it, the VAX only communicates with the memory over
the cable. On the 11/73 and family, the original memory reference
goes out on the bus, and the initial response is over the bus, but
several additional words are then supplied over the PMI bus... so
those words are cached in expectation of them being referenced due
to proximity to the actual referenced word.
Didn't you once make a (very small) disk memory out
of some old DEC
memory board? Or was that normal Qbus memory but at a high address
with a gap from the regular memory.
What I did was take one of the Bridge86 boards (an 8086 on a Qbus
board with its own memory) and wrote a handler to access its memory
as if it were a disk).
What would it take to build a small interface that
looks like the HD hard
disk from E11 as far as an actual hardware interface is concerned?
Could it loop right back into this uVAX II memory board via the 50 pin
header on top? Since the Qbus interface is already there in the
backplane, maybe that interface could be added to the 8 Mbyte of memory?
See, I said I would ask dumb questions!
Actually, that is an idea I mentioned in mail earlier tonight, prompted
by something Zane said, I think...
I was thinking that if the MS6xx memory boards only use the Qbus for
their power, and all transactions really do happen over the cable, then
it should be possible to design a Qbus option which looks like a disk, but
which communications with one or more MS6xx boards over the PMI cable.
Is there an actual signal requirement for the MS6xx boards to be in the
PMI slots of a qbus backplane, or is it simply because they are closest
to the uVax processor board...?
Megan Gentry
Former RT-11 Developer
+--------------------------------+-------------------------------------+
| Megan Gentry, EMT/B, PP-ASEL | Internet (work): gentry!zk3.dec.com |
| Unix Support Engineering Group | (home): mbg!world.std.com |
| Compaq Computer Corporation | addresses need '@' in place of '!' |
| 110 Spitbrook Rd. ZK03-2/T43 | URL: http://world.std.com/~mbg/ |
| Nashua, NH 03062 | "pdp-11 programmer - some assembler |
| (603) 884 1055 | required." - mbg KB1FCA |
+--------------------------------+-------------------------------------+
8674
days inactive
8675
days old
2 comments
2 participants
participants (2)
-
jhfine＠idirect.com -
mbg＠world.std.com