> From: William Degnan
> Tried my M7838 EIS this morning. It is bad or there is a config/jumper
> issue to investigate.
When installing the KE11-E, you have to remove a jumper on the CPU's M7233
module. See pg. 2-1 on the KE11-E/KE11-F User's Manual (EK-KE11E-OP-001),
available online. Did yours come with the three flat cable jumpers?
If it's still not working, the KE11-E/KE11-F Technical Manual
(EK-KE11E-TM-002), also available online, will undoubtly prove useful.
Noel
Hi Josh,
I saw your posting on the 23rd Jan 16 regarding Bull DPS 6.
We are based in the UK and actually run 4 working DPS6 systems in the UK.
Any system (in its entirety or parts) we may be interested in if compatible with our existing system.
Do you still have it?
Many thanks,
Julian
Julian Metcalf
Finance Director
Like Technologies Ltd
0845 519 2244 Ext 111
This email and any attachments to it may be confidential and are intended solely for the use of the individual to whom it is addressed. Any views or opinions expressed are solely those of the author and do not necessarily represent those of Like Technologies Ltd.
If you are not the intended recipient of this email, you must neither take any action based upon its contents, nor copy or show it to anyone.
Please contact the sender if you believe you have received this email in error.
> From: William Degnan
> I do not have an EIS installed. I have one however, I can try it. I am
> unsure if it's good or not, but I guess I am going to find out.
Oh yeah, without that, you're totally hosed, Unix-wise. The V6 C compiler
puts out MUL etc all over the place (e.g. for structure pointer math), so
there's absolutely no way to run vanilla V6 Unix without it.
(The GCC compiler claims to be able to compile C to machines without the
EIS; it might be an interesting hack to see if that could be used to get
V6 running on a machine without the EIS. The machine language startups
would still need work, though. And the bootstrap. :-)
> Josh Dersch
> As I mentioned above
Right, but if his bootstrap isn't working, gotta debug that first.. :-)
> you'll also need to recompile the kernel
Actually, I don't think you need to re-compile anything, just link in m40.o
instead of m45.o; I think all the C code checks for 'cputyp == 40' or
whatever, as the case may be.
Noel
> William Degnan
> It "boots" to the ! prompt at least there's that.
Yeah, but not much has to be working for that to happen! :-)
> I am unsure if one can put an M7891 into a slot that has no NPG jumper
> installed
Yes, you can - but having a slot with no NPG jumper, and either i) no board
in in it at all, or ii) a non-DMA-speaking board in it, will prevent any DMA
device _further down the bus_ from working.
> I would have to check, or whether I can put this card into a DD11-B
No. It needs a MUD slot (UNIBUS in connectors A-B, essentially; more here:
http://gunkies.org/wiki/Modified_UNIBUS_Device
although other places have info about this too). (I'm not exactly sure how
the A/B connectors on DD11-B slots 2 and 3 are wired, it's something wierd,
look at the DD11-B prints for more.)
> or a slot without the NPG installed.
The NPG jumper is IRRELEVANT to everything except DMA devices (in that slot,
and downstream).
> Typing *anything* kills the system.
Well, technically speaking, that's not entirely accurate - clearly, from the
below typing "r" doesn't crash the machine. I gather you meant 'typing
"{anything}<CR>" crashes the machine'. Hmm.
> When I type say rlunix or foobar or whatever and hit enter, the prompt
> returns to the next line and the CPU stops. The system crashes,
> unresponsive, requiring restart from the front panel.
That sounds like the bootstrap isn't running properly.
Oh, I remember an issue I had with the boostrap when first trying to bring up
Unix in Ersatz-11 - does your -11/40 have the EIS board? Is the EIS working?
If not, the bootstrap won't run - it uses the MUL instruction. (MUL is not in
the base set on an -11/40, it's an option.)
If that's not it, we'll have to debug the bootstrap... Should't be too hard,
test versions can be loaded directly into memory with GUI-11, we don't have to
write them to disk. You've got a (hopefully good) disk to have the bootstrap
ponder over...
Noel
> From: William Degnan
> 64K.
That's 64KB, right? DEC used to talk of PDP-11 memory in words, back in the
day, so it's never quite clear unless the 'B' or 'W' is specified.
Anyway, 64KB out to be enough to run most things. I have't looked to see how
big a system with just RL and DL drivers, and a minimum number of disk buffers
(that's probably the easiest way to get the size down almost as far as it will
go) is, but it's probably in the region of 30-35KB (20KB text, and 10-15 KB
data). That leaves plenty for user processes (not max sized, of course).
> I'd like to get another M7891, mine is only 1/2 populated.
Oh, so that must have been 64KW - 128KB. That will give you plenty of room for
a decent-sized kernel, and user processes. Wonder why it won't boot, then?
Anyway, those are pretty rare. I have a couple of spare M7847's; they're only
32KB, but they'd help. You've got a couple of spare hex MUD slots now in the
DD11-C, right? Speaking of which, where is the M7891 plugged in? It couldn't
go in the DD11-B, or in the CPU backplane?
> typing anything kills the CPU and I have to restart the system..
Sorry, need exact details: after the '!' prompt, anything you type kills
the system on the first character? Or only "rlunix<CR>" does? How about
"foobar<CR>"? I'd have to look at the bootstrap source, but I think
typing a non-existing file name should take you back to the top-level
prompt.
Noel
> From: William Degnan
> doesn't hurt to try I suppose
Absolutely.
> My fear is that it will not have enough RAM on top of whatever other
> issues are present.
Oooh, good point - I hadn't thought of that.
I couldn't quickly find a 'minimum memory required' in the release notes for
V6; the 'Unix Summary' says it needs "48K to 124K words".
I think that might be a bit excessive; I seem to recall our -11/40 had, to
start with, only 48KB.
Anyway, try it, and let us know what happens. How much memory does the system
have on it now?
If it doesn't work, I can do some experiments and see what's the least amount
of memory one needs.
There are a whole bunch of parameters which will reduce the size of the
resident OS; if necessary, I can turn them all down to the minimum, and see
what we get - although just reducing the number of disk buffer may do it.
As for the applications, it does swap, so there's no requirement for more than
one process to be resident at a time, so whatever the largest is - probably
the C compiler - there only needs to be enough memory left over after the OS
is loaded, for that one. It looks like the shell is about 10KB, for example.
If it won't boot, don't trash the disk: if we send you another bootable disk,
we can make it tiny (only Unix, and enough files to get it running: /etc/init,
/bin/sh, etc), and if/once it books, you can mount the disk you just wrote and
move the bootable Unix system image over, and then reboot on the current disk,
to get to the rest of the stuff.
Do you have two working RL drives on the machine?
Noel
> From: William Degnan
> Can one be made using simH to dump and set up for RL02 that can then be
> ported as a RL02 disk image to actual RL02 drive?
I see someone has already provided a pointer to someone who ha done this; I'm
not sure if that system will boot on a hardware 11/40, or if it was built with
m45.s, or some simlar issue.
It would be pretty easy for anyone with a working V6 UNIX (either on hardware,
or emulated) to do this, _iff_ they have a V6 RL driver. ('Vanilla' V6 does
not include one.) My 'Bringing up V6 under Ersatz-11' page describes one:
http://www.chiappa.net/~jnc/tech/V6Unix.html#rl
You might want to get familiar with building V6 kernels, _any_ hardware
variations you wish to support (e.g. more than one DL11 seria line) will
involve re-building the kernel. Not to worry, it's a pretty simple process, it
only takes a few commands, and a few moments (more on a real machine, of
course :-).
That page include almost all the directions on how to do so, but note that one
must also edit 'rootdev' in c.c if the root file system is to be on an RL (no,
mkconf as distributed with 'vanilla' V6 won't do any of the legwork for you,
it does not know about RL drives).
Note that running 'vanilla' V6 Unix i) has some bugs/issue (e.g. you can't
set the date to this century; the user interface is _strictly_ for printing
terminals, etc, etc, etc), some of which are handled here:
http://www.chiappa.net/~jnc/tech/V6Unix.html#Issues
so I would advise checking it out. (You can skip all the Ersatz-11-specific
stuff, like the 'DOS Device' to allow Unix to read files from the host
file-system - a very useful capability in an emulated Unix!)
> From: Angelo Papenhoff
> http://www.tom-yam.or.jp/2238/rl/
Note that his page says "bootblock is rluboot from v6_rl02_unknown, which
lacks the source code"; I disassembled and commented the code, see above
links.
Noel
I thought this question was answered recently but I can't find the answer.
I have the RK disk version of UNIX6 for PDP 11/40 but I could not find an
RL02 version. does this exist? Can one be made using simH to dump and set
up for RL02 that can then be ported as a RL02 disk image to actual RL02
drive? I can get an image onto the drive, should the image exist. My PDP
11/40 is not compatible with Unix 7, but I would be happy with 5 or 7 if
it'll run on a 64K system.
Thanks
Bill
> Here:
> http://ana-3.lcs.mit.edu/~jnc/tech/pdp11/macro.tar
> is a tar file with the whole works, along with the MACRO-11 source.
BASIC is not ready yet, and won't be for a few days. The problem is that I
can't get it to assemble into a working version, which I would like to do,
before sending it out.
I _think_ that what happened is we made some change in the toolchain, to
support other things we were doing, and one of them 'broke' re-building BASIC
- except that since nobody was actually _doing_ anything with BASIC, we had no
reason to re-build it - and so nobody noticed we had 'broken' it.
If someone's desperate to get their hands on this, let me know, and I can go
ahead and send out what I have, with the proviso that you can't build a
working version from the MACRO-11 source at this point.
> If the first person to try MACRO could send me feedback, to see if the
> whole process works, I'd be grateful.
Did anyone grab the MACRO-11 TAR and try it yet?
Noel
Hi Camiel Vanderhoeven
I nearly got all about PDOS and the TM990 system......however only on 8" floopy disks.
I also have the complete documentation and a TM990/100M system with a DRAM and a Floppy-Controller board.
best regards
Walter Schubert
--
MicroSys Electronics GmbH, M?hlweg 1, D-82054 Sauerlach
Tel.: +49 8104 801-0, Fax: +49 8104 801-110
Sitz der Gesellschaft: Sauerlach, HRB M?nchen 48340, Ust.ID No: DE129296566
Gesch?ftsf?hrer: Dipl.-Ing. Richard Loeffl, Dipl.-Ing. Dieter Pfeiffer
http://www.microsys.de