30 Oct
2002
30 Oct
'02
3:23 p.m.
Does anyone have a WCS (writeable control store) board?
If someone does, I need a dump of the contents of the
microms of the LSI chipset... I'll be happy to write
the program to do the dump...
(If you also have the EIS/FIS chip, so much the better)
Megan Gentry
Former RT-11 Developer
+--------------------------------+-------------------------------------+
| Megan Gentry, EMT/B, PP-ASEL | email: gentry at zk3.dec.com (work) |
| Unix Support Engineering Group | mbg at world.std.com (home) |
| Hewlett Packard | (s/ at /@/) |
| 110 Spitbrook Rd. ZK03-2/T43 | URL: http://world.std.com/~mbg/ |
| Nashua, NH 03062 | "pdp-11 programmer - some assembler |
| (603) 884 1055 (DEC '77-'98) | required." - mbg KB1FCA |
+--------------------------------+-------------------------------------+
30 Oct
30 Oct
4:53 p.m.
Does anyone have a WCS (writeable control store)
board?
Yes.
If someone does, I need a dump of the contents of the
microms of the LSI chipset... I'll be happy to write
the program to do the dump...
In other words, you'll write a microprogram to dump the microms?
Normal PDP-11 code can't do it, even with a WCS board installed.
The WCS board only allows read/write access to the WCS memory; it
provides no means of accessing the contents of the MICROMs. The
RAM Address Register is the low 10 bits of 177540, and only provides
addressing for the WCS RAM. The WCS memory can only be accessed over
the Qbus while the WCS is disabled (bit 10 of 177540 set to 0). And
while the WCS is disabled, the data path to the microinstruction bus
(MIB) is completely disabled.
I'm not completely convinced that it is even possible to write a
microprogram to dump the MICROMs. There's no obvious instruction
to do it, nor is there a suitable data path in the diagrams.
The only obvious way I know of to read the MICROMs is to remove
them from their socket and wire them up to a microcontroller
programmed to act like the control chip but access the locations
sequentially and log the MIB lines.
(If you also have the EIS/FIS chip, so much the
better)
Got that too. You're probably already aware that the EIS/FIS source
code is supplied with the WCS software package. Which I haven't got
a copy of :-(
Anyhow, if you do in fact have a viable means of reading the MICROM
contents, I'll be glad to help.
The *really* tricky part will be reading the translation array, which
is mask-programmed in the control chip. The translation array can cause
microbranches without an explicit branch microinstruction, so the full
control flow of the microcode will not be entirely obvious from just
a MICROM dump. I've considered trying to write a microcode-level
simulator for the LSI-11, but the translation array seems like it will
be fairly difficult to reverse-engineer, short of decapping the control
chip, taking a photomicrograph, and spending a heck of a lot of time
studying the layout.
Eric
11:53 p.m.
Megan wrote:
Does anyone have a WCS (writeable control store)
board?
If someone does, I need a dump of the contents of the
microms of the LSI chipset... I'll be happy to write
the program to do the dump...
(If you also have the EIS/FIS chip, so much the better)
Jerome Fine replies:
I probably do have both the dual and quad 11/03 CPU
boards. Do they have the WCS? I seem to remember
that the 11/44 also has the WCS? But since you did not
mention which CPU modules have the WCS, it might
help others who do have a WCS CPU.
I think one board has all 5 chips. I suspect that means it
has the EIS/FIP chip?
Sincerely yours,
Jerome Fine
--
If you attempted to send a reply and the original e-mail
address has been discontinued due a high volume of junk
e-mail, then the semi-permanent e-mail address can be
obtained by replacing the four characters preceding the
'at' with the four digits of the current year.
31 Oct
31 Oct
7:53 a.m.
Jerome Fine replies:
I probably do have both the dual and quad 11/03 CPU
boards. Do they have the WCS? I seem to remember
that the 11/44 also has the WCS? But since you did not
mention which CPU modules have the WCS, it might
help others who do have a WCS CPU.
No PDP-11 had WCS as a standard feature. WCS options were
available for the 11/03 (LSI-11) and 11/60.
The 11/03 version is the KUV11, M8018. It's a quad-height module.
The KUV11 needs to be cabled to a specific MICROM socket on the M7264
module; it will not work on the half-height LSI-11/2 module (M7270).
The KUV11 can be loaded with the EIS/FIS microcode (equivalent to the
KEV11 MICROM), with 512 words left over for user microcode. Or you can
use all 1024 words for user microcode, subject to some restrictions.
If you have the single-hybrid-chip version of the KEV11, you can use
it with the KUV11, and strap the latter to only respond to the last
512 words of the microcode address space.
The 11/60 version is the KU116, M7870. I'm looking for one of these.
There was a third-party WCS option for the PDP-11/45; there may have
been third-party WCS options for other models.
I think one board has all 5 chips. I suspect that
means it
has the EIS/FIP chip?
It has the two-chip version of the option, so there is no socket left
for the WCS. You'd have to remove the KEV11 chips to use the WCS.
Eric
8:12 a.m.
Hi Eric,
I'm not sure how badly you want a M7870 board, but this fellow
has been able to locate some fairly rare parts for me at times.
Charle Burgess
QEI Incorporated
119 Great Road
Bedford, MA 01730
--tom
At 01:45 PM 10/31/2002 -0800, you wrote:
Jerome Fine
replies:
I probably do have both the dual and quad 11/03 CPU
boards. Do they have the WCS? I seem to remember
that the 11/44 also has the WCS? But since you did not
mention which CPU modules have the WCS, it might
help others who do have a WCS CPU.
No PDP-11 had WCS as a standard feature. WCS options were
available for the 11/03 (LSI-11) and 11/60.
The 11/03 version is the KUV11, M8018. It's a quad-height module.
The KUV11 needs to be cabled to a specific MICROM socket on the M7264
module; it will not work on the half-height LSI-11/2 module (M7270).
The KUV11 can be loaded with the EIS/FIS microcode (equivalent to the
KEV11 MICROM), with 512 words left over for user microcode. Or you can
use all 1024 words for user microcode, subject to some restrictions.
If you have the single-hybrid-chip version of the KEV11, you can use
it with the KUV11, and strap the latter to only respond to the last
512 words of the microcode address space.
The 11/60 version is the KU116, M7870. I'm looking for one of these.
There was a third-party WCS option for the PDP-11/45; there may have
been third-party WCS options for other models.
I think one board has all 5 chips. I suspect
that means it
has the EIS/FIP chip?
It has the two-chip version of the option, so there is no socket left
for the WCS. You'd have to remove the KEV11 chips to use the WCS.
Eric
12:33 a.m.
Jerome, the WCS is a separate option which plugs into one of
the sockets normally occupied by a MICROM. The only other
-11 that I know of which had WCS was the 11/60.
The specific option of which I speak is the KUV11-AA, a
quad-high qbus board which attaches to an 11/03 or 11/2.
Megan Gentry
Former RT-11 Developer
+--------------------------------+-------------------------------------+
| Megan Gentry, EMT/B, PP-ASEL | email: gentry at zk3.dec.com (work) |
| Unix Support Engineering Group | mbg at world.std.com (home) |
| Hewlett Packard | (s/ at /@/) |
| 110 Spitbrook Rd. ZK03-2/T43 | URL: http://world.std.com/~mbg/ |
| Nashua, NH 03062 | "pdp-11 programmer - some assembler |
| (603) 884 1055 (DEC '77-'98) | required." - mbg KB1FCA |
+--------------------------------+-------------------------------------+
1:35 a.m.
Megan wrote:
Jerome, the WCS is a separate option which plugs into
one of
the sockets normally occupied by a MICROM. The only other
-11 that I know of which had WCS was the 11/60.
The specific option of which I speak is the KUV11-AA, a
quad-high qbus board which attaches to an 11/03 or 11/2.
Jerome Fine replies:
The two CPU boards (I have both) are the dual and the
quad 11/03 Qbus versions. I have been told they require
a BA-11 box since they are incompatible with the
22 bit backplane found in the BA23 box.
Back in the latter part of the 1970s when the 11/03
was released, I understand that the 11/03 was far
superior to anything from Intel. Even with just V3.0x
of RT-11, users had a sophisticated operating system
that was, in my opinion, better than any other single
user OS at the time - and with all the improvements
still is.
I think I used one ONLY once since it is, of course,
unable to address more than 56 KBytes (60 KBytes
with a 2 KWord I/O page) of memory. Since I
want at least 1 MByte of memory, I use at least
an 11/23 and usually an 11/83 CPU and run
RT11XM all the time.
Of course, that is still a bit slow. Most of the time
I use Ersatz-11 on a 750 MHz Pentium III, these
days with 768 MBytes of memory. The speed
is about 15 times an 11/93 - plus I can have a
RAM: disk of 512 MBytes if I wish - 16 RT-11
partitions of 32 MBytes each. It should be even
more interesting when the 3 GHz Pentium 4 is
available at a reasonable price in a year or two -
50 times the speed of an 11/93 anyone? Even
SIMH should fly at that point. It would be nice
to have an M1 or a QED 11/93, but I can't
justify that cost and those are still slower than
Ersatz-11 with a 750 MHz Pentium III.
I did not realize that the WCS (which I had heard
was a feature on these 11/03 CPUs) was a separate
option.
Sincerely yours,
Jerome Fine
--
If you attempted to send a reply and the original e-mail
address has been discontinued due a high volume of junk
e-mail, then the semi-permanent e-mail address can be
obtained by replacing the four characters preceding the
'at' with the four digits of the current year.
8093
days inactive
8094
days old
6 comments
4 participants
participants (4)
-
eric＠brouhaha.com -
jhfinepw4z＠compsys.to -
mbg＠TheWorld.com -
uban＠ubanproductions.com