24 Nov
2021
24 Nov
'21
1:47 a.m.
So, with the help of you here, I was able to create OS/8 LINCtapes and to
run SerialDisk. Everything runs very fine.
Now comes the next thing: I want to have K12MIT, and it is no problem to
compile or load the program.
*But*: When I start K12MIT I don't get the prompt. I see the welcome
message, and it correctly identifies the machine as a PDP-12, but that's
it. As I've found out it apparently overwrites its code, the processor is
looping around address 3600-3620. Examining the memory reveals that the
code at 3600 has been overwritten with junk. So it's clear that it won't
run anymore. What happens? Was anyone else able to run K12MIT on a PDP-12?
BTW the same binary runs fine on a PDP-8, be it a real machine or SIMH.
Christian
24 Nov
24 Nov
1:55 a.m.
Do you have a running pdp12?
C
On 11/24/2021 4:47 AM, Christian Corti via cctalk wrote:
So, with the help of you here, I was able to create
OS/8 LINCtapes and
to run SerialDisk. Everything runs very fine.
Now comes the next thing: I want to have K12MIT, and it is no problem to
compile or load the program.
*But*: When I start K12MIT I don't get the prompt. I see the welcome
message, and it correctly identifies the machine as a PDP-12, but that's
it. As I've found out it apparently overwrites its code, the processor
is looping around address 3600-3620. Examining the memory reveals that
the code at 3600 has been overwritten with junk. So it's clear that it
won't run anymore. What happens? Was anyone else able to run K12MIT on a
PDP-12?
BTW the same binary runs fine on a PDP-8, be it a real machine or SIMH.
Christian
2:38 a.m.
Interesting. I couldn't save the pdp12 that was at my college 30 years
ago (you can't fit that rack in the back of a station wagon) but I did
grab a couple of boxes of LincTapes. I'd be happy to loan you a few if
you want to see what's on them.
C
On 11/24/2021 5:17 AM, Christian Corti via cctalk wrote:
On Wed, 24 Nov 2021, Chris Zach wrote:
Do you have a running pdp12?
Yes :-) It is completely working including the display and both LINCtape
drives.
Christian 
5:22 a.m.
On 11/24/2021 1:47 AM, Christian Corti via cctalk wrote:
So, with the help of you here, I was able to create
OS/8 LINCtapes and
to run SerialDisk. Everything runs very fine.
Yay!
Now comes the next thing: I want to have K12MIT, and
it is no problem to
compile or load the program.
*But*: When I start K12MIT I don't get the prompt. I see the welcome
message, and it correctly identifies the machine as a PDP-12, but that's
it. As I've found out it apparently overwrites its code, the processor
is looping around address 3600-3620. Examining the memory reveals that
the code at 3600 has been overwritten with junk. So it's clear that it
won't run anymore. What happens? Was anyone else able to run K12MIT on a
PDP-12?
BTW the same binary runs fine on a PDP-8, be it a real machine or SIMH.
Are you running SerialDisk on the same DP12 that K12MIT is trying to
use? That might not work well, though I'm not sure why K12MIT would
commit suicide about it.
I haven't had much use for K12MIT -- the tools that come with the latest
SerialDisk can take apart and rebuild/modify bootable/server images, so
I just copy whatever to or the directory those tools are using and
rebuild the image (or re-explode the server image and copy it out, to go
the other way).
Vince
25 Nov
25 Nov
1:15 a.m.
On Wed, 24 Nov 2021, Vincent Slyngstad wrote:
Are you running SerialDisk on the same DP12 that
K12MIT is trying to use?
That might not work well, though I'm not sure why K12MIT would commit suicide
about it.
Hmm... good point, I think that I have tried both, from LINCtape and from
SerialDisk. I know, the latter won't work, but at least I should get the
prompt, no?
I haven't had much use for K12MIT -- the tools
that come with the latest
SerialDisk can take apart and rebuild/modify bootable/server images, so I
just copy whatever to or the directory those tools are using and rebuild the
image (or re-explode the server image and copy it out, to go the other way).
I haven't understood the tools to build these disk images. There is no
documentation IIRC. But I use SIMH and have a transfer disk image that I
can mount with SerialDisk. There is one nuisance, though: SerialDisk won't
see the updated image unless restarted. The advantage of SIMH is that I
can "import" paper tape images (BIN or ASCII). That is something even PUTR
can't accomplish.
Christian
9:55 a.m.
On Thu, 25 Nov 2021, Christian Corti wrote:
On Wed, 24 Nov 2021, Vincent Slyngstad wrote:
Just tried without SerialDisk... no difference. K12MIT is broken on the
PDP-12.
Christian
Are you running SerialDisk on the same DP12 that
K12MIT is trying to use?
Hmm... good point, I think that I have tried both, from
LINCtape and from
SerialDisk. I know, the latter won't work, but at least I should get the
prompt, no?
26 Nov
26 Nov
1:43 a.m.
On Thu, 25 Nov 2021, Christian Corti wrote:
Just tried without SerialDisk... no difference. K12MIT
is broken on the
PDP-12.
I've found the bug. Charles Lasner broke PDP-12 support (and probably also
older PDP-8 models like 8/I) with his additions to support the DECmate
family. The location is at 0206, labelled SETBAUD. This should be a NOP on
standard systems, but is overlayed by an IOT to set the baud rate on a
DECmate 1. This results in a 6413 instruction (because I have assembled it
for the second serial port on 40/41). This instruction always skips (as
the transmitter done flag is always set), thus jumping into a 4411 that is
the address of a CRLF to be printed but executed as a JMS into nirvana.
The very same code runs fine on a PDP-8/E with a KL8/E interface that
apparently either ignores this instruction or, by other means, has the
flag cleared.
If patching location 0206 to 7000 (NOP), K12MIT will run on a PDP-12.
Sending files works, but receiving seems to be broken. Most of the times,
the sending Kermit aborts the transfer to the PDP-12 with a
out-of-window-NACK error (confirmed by looking at the protocol debug log).
K12MIT sends a NACK to a previous packet that it already ACK'ed *after*
receiving the next data packet.
Here's the code fragment:
1827 000200 5777 CLOOP, JMP I (INITIALIZE) /**** INITIALIZED **** CLA CLL
/024 CJL
1828 000201 4577 JMS I [CRESET] /RESET CONSOLE ROUTINE STUFF
1829 000202 3002 DCA ABFLAG /CLEAR ABORT FLAG
/044 CJL
1830 000203 3105 DCA REMWAIT /CLEAR REMOTE FLOW CONTROL WAIT
/049 CJL
1831 000204 4324 CLDMZAP,JMS DMINIT /INITIALIZE DECMATE PORT
/048 CJL
1832
1833 / THE PREVIOUS INSTRUCTION IS FOR DECMATE OPERATION
ONLY. THE DEFAULT
1834 / INSTRUCTION IS FOR KL8 INTERFACES ONLY.
1835
1836 0204 *CLDMZAP /OVERLAY DECMATE CODE
/048 CJL
1837
1838 000204 7000 CLDMZAP,NOP /DON'T INITIALIZE DECMATE PORT
/048 CJL
1839 000205 1007 TAD RATE /GET BAUD RATE
1840 000206 6363 SETBAUD,MSB /SET REMOTE BAUD RATE
/050 CJL
1841
1842 / THE PREVIOUS INSTRUCTION IS FOR DECMATE II OPERATION ONLY.
/050 CJL
1843
1844 0206 *SETBAUD /OVERLAY DECMATE II CODE
/050 CJL
1845
1846 000206 7000 SETBAUD,NOP /NOT APPLICABLE ON DECMATE I
/050 CJL
1847
1848 / THE PREVIOUS INSTRUCTION IS FOR DECMATE I OPERATION
ONLY. THE DEFAULT
1849 / INSTRUCTION IS FOR KL8 INTERFACES ONLY.
1850
1851 0206 *SETBAUD /OVERLAY DECMATE I CODE
/050 CJL
1852
1853 000206 6413 SETBAUD,RSBIOT /SET REMOTE BAUD RATE
/050 CJL
1854 000207 4576 JMS I [SCRIBE] /DO A
/025 CJL
1855 000210 4411 CRLF /<CR>, <LF>
/025 CJL
1856 000211 4576 JMS I [SCRIBE] /GIVE THEM THE
/025 CJL
1857 000212 4457 PRMTMSG /PROMPT MESSAGE
/025 CJL
1858 000213 4776 JMS I (LININP) /GET INPUT LINE FROM USER
Christian
5:48 p.m.
On 11/25/2021 1:15 AM, Christian Corti via cctalk wrote:
On Wed, 24 Nov 2021, Vincent Slyngstad wrote:
It has become clear that my documentation is inadequate. (You aren't
the first to complain.)
As mentioned in the installer/README.md, the "sdsk" script is the fancy
"does it all" installer. It requires perl, a couple of other packages
as mentioned, and the tools from the "tools" directory.
As mentioned in the tools/README.md, the tools directory is a subset of
the "8tools", a suite of software forensics tools which are targeted at
the PDP-8. (Instructions in tools/README.md on where to get those.)
The 8tools do come with a "doc" sub-directory with a write-up of each tool.
The tools, with the exception of the cross-assembler (C source
provided), are perl scripts, and contain source code but not really any
documentation.
pal is a cross-assembler, which will take PAL source and generate a
listing (.lst) and a binary (.bin). It's main use is to convert
Serialdisk source in ../handler to .bin files. (Do the "make" in
../handler first!)
bin2bn converts .bin files to proper OS/8 format by padding them with
trailer bytes to a multiple of the OS/8 block size. The result can be
included with the files used to build OS/8 images.
bincmp is a comparison tool which will diff .bin (or .bn) files.
os8xplode takes a ".dsk" file and explodes it into directories
containing the individual files for each file-system found. It also
writes an ".xml" description of the OS/8 volume so that "mkdsk" can
re-construct it.
mkdsk reads the .xml mentioned above, and the various individual files,
and reconstructs the specified OS/8 volume.
os8implode is a tools that scans the (probably modified) directories
associated with a disk volume, and writes a new .xml description
suitable for mkdsk. The magic "dot" files will be used to reconstruct
the system areas, and all files that match the OS/8 "6.2" naming
convention will be included in the new image.
installer/sdsk, then, rounds up the SerialDisk drivers from ../handler
for inclusion in the new volume, then runs os8implode and mkdsk to
construct a volume that has the .bn files for the drivers. Then it runs
SIMH (see "installer/runbuild" and "installer/pdp8.ini"). SIMH will
then be fed the commands to convert your OS/8 volume into a SerialDisk
volume.
runbuild is an expect script, which is why you need to install expect :-).
For this to work correctly, you need to install the packages mentioned
in the installer/README.md and add the tools directory to your path.
Then a simple ./sdsk should work. (You can edit sdsk to select another
image from ../disks.) Your new volume will have a name like
"diagpack2.new", and can be served with the server from ../server.
Hope that helps!
Vince
I haven't had much use for K12MIT -- the
tools that come with the
latest SerialDisk can take apart and rebuild/modify bootable/server
images, so I just copy whatever to or the directory those tools are
using and rebuild the image (or re-explode the server image and copy
it out, to go the other way).
I haven't understood the tools to build these disk images. There is no
documentation IIRC. But I use SIMH and have a transfer disk image that I
can mount with SerialDisk. There is one nuisance, though: SerialDisk
won't see the updated image unless restarted. The advantage of SIMH is
that I can "import" paper tape images (BIN or ASCII). That is something
even PUTR can't accomplish.
10:51 p.m.
On 11/26/2021 1:43 AM, Christian Corti via cctalk wrote:
I've found the bug. Charles Lasner broke PDP-12
support (and probably
also older PDP-8 models like 8/I) with his additions to support the
DECmate family. The location is at 0206, labelled SETBAUD. This should
be a NOP on standard systems, but is overlayed by an IOT to set the baud
rate on a DECmate 1. This results in a 6413 instruction (because I have
assembled it for the second serial port on 40/41). This instruction
always skips (as the transmitter done flag is always set), thus jumping
into a 4411 that is the address of a CRLF to be printed but executed as
a JMS into nirvana.
It took a bit of puzzling to follow, but I think I see. There are three
possibilities; the the DMI should have a "NOP", and the DMII should have
"MSB". DM2LIS correctly has the MSB patch for DMII, and DM1LIS has a NOP
as it perhaps should.
The problem is that the default for "KL8" interface uses an "RSBIOT".
I
don't think that will work for anything but the KL8J/M8655, though it
should be harmless on the KL8E/M8650.
I'd argue that the correct thing would be to detect an 8/e/f/m/a and
patch in "RSBIOT". The default "family of eight" thing should be the
"NOP", same as the DM1.
If patching location 0206 to 7000 (NOP), K12MIT will
run on a PDP-12.
Yep.
Sending files works, but receiving seems to be broken.
Most of the
times, the sending Kermit aborts the transfer to the PDP-12 with a
out-of-window-NACK error (confirmed by looking at the protocol debug
log). K12MIT sends a NACK to a previous packet that it already ACK'ed
*after* receiving the next data packet.
Alas, I don't know enough about Kermit protocol to comment on that.
Vince
903
days inactive
906
days old
9 comments
3 participants
participants (3)
-
cc＠informatik.uni-stuttgart.de -
cz＠alembic.crystel.com -
vincent.slyngstad＠gmail.com