19 Apr
2014
19 Apr
'14
2:58 p.m.
Kyle wrote:
I tried another version from here:
http://www.rickmurphy.net/adventure.htm land it seemed to work just fine,
although for whatever reason, typing DIR results in "No room for output file."
I'm not sure exactly what it means by that; I don't get the same error from
running in simh.
In any event, I can now run Adventure on my 8/M!
I tried the version from the link above on my 8/e. I copied it using the OS8 Disk Server,
from the RK05 image to an actual RK05 and ran it from there.
My machine has 24K (8K core, 16K MOS), 2 M8650 serial interfaces, RK8E, PT8E, and RX8E.
When I run ADVENT, it loads, but then just sits there in an apparent loop. If I type
characters, nothing happens, except if I type CONTROL-C, it does exit back to OS/8.
I've let it sit a few minutes, thinking that maybe it takes a bit to initialize, but
that clearly isn't the case.
I also tried the "R FRTS ADVENT.LD"<ESC> but with this, I get a $ prompt
which appears to be coming from FRTS, and don't know what to do from there. I did try
typing "SYS:ADVENT.LD" at the $ prompt, but it just comes back with another $
prompt.
I do note that the INT BUS in is lit. This means that an interrupt is pending, but the
ION state indicator is off. I have the serial interfaces set up to use interrupts if they
are enabled. To get CHEKMO-II (which uses interrupts for I/O to the console) to run, I
had to add some code to the interrupt service routine to ignore any pending interrupt from
the second serial interface.
Could the interrupt (output ready) from the second serial interface be causing this
problem?
I'd love to get Adventure running on the system.
Thanks,
-Rick Bensene
The Old Calculator Museum
http://oldcalculatormuseum.com
19 Apr
19 Apr
6:43 p.m.
At 05:58 PM 4/19/2014, Rick Bensene wrote:
Kyle wrote:
24K isn't enough to run it. FRTS *should* tell you it's not enough,
however.
Here's what I get when I try with 24K:
.R FRTS
*ADVENT$MORE CORE REQUIRED
(the "$" is where I hit escape.)
I tried another version from here:
http://www.rickmurphy.net/adventure.htm land it seemed to work just
fine,
although for whatever reason, typing DIR results
in "No room for
output file."
I'm not sure exactly what it means by that; I
don't get the same
error from
running in simh.
In any event, I can now run Adventure on my 8/M!
I tried the version from the link above on my 8/e. I copied it using
the OS8 Disk Server, from the RK05 image to an actual RK05 and ran it
from there.
My machine has 24K (8K core, 16K MOS), 2 M8650 serial interfaces,
RK8E, PT8E, and RX8E. When I run ADVENT, it loads, but then just sits there
in an apparent
loop. If I type characters, nothing happens, except if I type
CONTROL-C, it does exit back to OS/8. I've let it sit a few minutes,
thinking that maybe it takes a bit to initialize, but that clearly
isn't the case.
I also tried the "R FRTS ADVENT.LD"<ESC> but with this, I get a $
prompt which appears to be coming from FRTS, and don't know what to do
from there. I did try typing "SYS:ADVENT.LD" at the $ prompt, but it
just comes back with another $ prompt.
Hmm. Seems like FRTS doing something bad.
I'd try with a simple FORTRAN program and see if that works. Using
Adventure to debug is a pretty heavy diagnostic. :)
I do note that the INT BUS in is lit. This means that
an interrupt is
pending, but the ION state indicator is off. I have the serial
interfaces set up to use interrupts if they are enabled. To get
CHEKMO-II (which uses interrupts for I/O to the console) to run, I had
to add some code to the interrupt service routine to ignore any
pending interrupt from the second serial interface.
There's a way to add an additional interrupt service routine to FRTS to
handle other devices, or you can run a program that just clears the
flags on that device first.
-Rick
10:03 p.m.
Rick M. wrote in response to my query about ADVENT not running on my
machine:
however.
to handle
>My machine has 24K (8K core, 16K MOS), 2 M8650
serial interfaces,
RK8E,
PT8E, and RX8E.
24K isn't enough to run it. FRTS *should* tell you it's not enough,
Here's what I get when I try with 24K:
.R FRTS
*ADVENT$MORE CORE REQUIRED
(the "$" is where I hit escape.)
I don't get error this at all if I do this. I have the OS/8 memory
config set up to include memory through Field 5.
Does ADVENT require a full 32K to run?
If ADVENT won't run in 24K (e.g., it was never meant to run in this
small of a memory configuration), then I'll have to wait until I get my
32K Battery-Backed memory board from Vincent S. and get it assembled.
Hopefully that'll be sometime in mid-to-late May.
Hmm. Seems like FRTS doing something bad.
I'd try with a simple FORTRAN program and see if that works. Using
Adventure to debug is a pretty heavy diagnostic. :)
I have written a number of small FORTRAN programs and can successfully
compile and run them just fine.
So, FRTS seems to be healthy, as does the compiler and RALF.
>I do note that the INT BUS in is lit. This means that an interrupt
is
>pending, but the ION state indicator is off. I
have the serial
>interfaces set up to use interrupts if they are enabled. To get
>CHEKMO-II (which uses interrupts for I/O to the console) to run, I
had
>to add some code to the interrupt service routine
to ignore any
pending
interrupt from
the second serial interface.
There's a way to add an additional interrupt service routine to FRTS other devices, or you can run a program that just
clears the flags on
that
device first.
I'll write a little routine to clear the interrupt enable flag on the
second serial card, and try running it first, then running ADVENT and
see if it makes a difference.
Does anyone know where the procedure for adding an additional ISR to
FRTS is documented?
Thanks,
-Rick B.
20 Apr
20 Apr
5:59 a.m.
At 01:03 AM 4/20/2014, Rick Bensene wrote:
Rick M. wrote in response to my query about ADVENT not
running on my
machine:
however.
Yup. It's only able to fit in 32K due to a lot of bit bashing to make
it fit.
The fact that FRTS doesn't tell you that is probably an indication that
it's loading into nonexistent memory fields and tripping off into
neverland while trying to run. I think the interrupt thing is a red
herring.
If I try to run with 28K of memory, FRTS hangs, executing in field 7
(which isn't there) That's not surprising as the Fortran USR support
routines assume 32K, so in that case you've moved drivers to the
nonexistent field 7.
To add interrupts to the FRTS chain isn't easy for this case. Normally
you would call ONQI to add an interrupt to the FRTS skip chain, but
that requires code changes. You can patch the existing chain to change
an existing handler (for example, the LPT one). Documentation on how
to do all this is in the Fortran IV Software Support Manual -
www.dbit.com/pub/pdp8/doc/fivssm.doc
However, since small fortran programs run OK, I don't think that this
is what's causing your problem. Once you get a full 32K you should be fine.
-Rick
>My machine has 24K (8K core, 16K MOS), 2
M8650 serial interfaces,
RK8E,
PT8E, and
RX8E.
24K isn't enough to run it. FRTS *should* tell you it's not enough,
Here's what I get when I try with 24K:
.R FRTS
*ADVENT$MORE CORE REQUIRED
(the "$" is where I hit escape.)
I don't get error this at all if I do this. I have the OS/8 memory
config set up to include memory through Field 5.
Does ADVENT require a full 32K to run? 
22 Apr
22 Apr
5:21 a.m.
On Sat, Apr 19, 2014 at 02:58:02PM -0700, Rick Bensene wrote:
Kyle wrote:
Could the interrupt (output ready) from the second serial interface be
causing this problem?
Yes. For now you need to run after boot (assuming card at 040/041)
*200
PKIE=6405
CLL CLA
PKIE /CLEAR INTERRUPT
JMP I (7605
$
I think he will address this in the next release of the driver.
I'd love to get Adventure running on the system.
This should let yo get to insufficent memory.
10:29 a.m.
On Tue, Apr 22, 2014 at 7:21 AM, David Gesswein <djg at pdp8online.com> wrote:
I think he will address this in the next release of the driver.
That's correct; I'll have an update in two weeks or so after exams are done
with. Maybe before then if I choose to not study as much. :)
I'll also make a write-up for changes to the M8650 for running it at the
standard baud rate of 230400 baud. SerialDisk works much, much better at
that rate. Even better still is if you can procure one of Philipp's awesome
OmniUSB cards!
Kyle
4060
days inactive
4063
days old
5 comments
4 participants
participants (4)
-
djg＠pdp8online.com -
kylevowen＠gmail.com -
rick＠rickmurphy.net -
rickb＠bensene.com