19 May
2005
19 May
'05
6:04 p.m.
Jay,
this is the one I used:
Location Content Op-code comment
001000 012700 mov #1,r0 load ro with 1
001002 000001
001004 006100 rol rotate r1 left
001006 012701 mov delay r1 load register r1 with delay
001010 007777 delay
001012 005301 dec r1 decrement register 1
001014 001376 bne -2 continue to decrement r1 until
r1=0
001016 000772 br -12 back to 001002 (dec r0)
it's not elegant, but you do get to see the lamps light, you can change the
speed by altering the "delay" value. It doesn't work to well, as the data
display will show the load of r1 as well.
Yo should be able to alter it to put the lght chaser into the console
display register (you'll have to re-calculate the value of the final
branch - it will need an extra two memory loactions, one for the mov display
register instruction, and one for the location of the register). I haven't
got this to work on my 11/45, as I get a memory address error when I try to
write to the display register, though I'm told I should be able to write to
it - I think there is a processor fault lurking in there somewhere, or
possibly a fault in the memory management module, though this is less
likely, as I can read and write to the DL11W serial port.
Jim.
Please see our website the " Vintage Communication Pages" at WWW.G1JBG.CO.UK
22 May
22 May
9:49 p.m.
Jim wrote....
Location Content Op-code comment
001000 012700 mov #1,r0 load ro with 1
001002 000001
001004 006100 rol rotate r1 left
001006 012701 mov delay r1 load register r1 with delay
001010 007777 delay
001012 005301 dec r1 decrement register 1
001014 001376 bne -2 continue to decrement r1 until
r1=0
001016 000772 br -12 back to 001002 (dec r0)
This runs on my /45, but produces a steady light display, not a cylon or
"blinking" effect. It's been decades since I looked at '11 assembler.
I'm
confused about something in the listing above. If the 012700 at location
1000 is a load r0 with 1, I'm guessing the next word (1002) is the constant
to be loaded? So, why would the branch at location 1016 go back to the data
at 1002? Obviously there's something I'm missing. Guidance? If 1002 is a DEC
R0 instruction, then where's the constant "1" for the mov#1,r0 stored in the
object? Humm?
Jay
23 May
23 May
4:05 a.m.
Jay West wrote:
It's been decades since I looked at '11
assembler
(Disclaimer: I haven't dealt with '11 assembler since ~1982 either.)
So, why would the branch at location 1016 go back to
the data
at 1002? Obviously there's something I'm missing.
The comment is incorrect. The intent is to go to 1004.
Jim wrote....
> Location Content Op-code comment
> 001000 012700 mov #1,r0 load ro with 1
^ comment should be "r0" not "ro"
> 001002 000001
> 001004 006100 rol rotate r1 left
^ comment should be "rotate r0 left"
> 001006 012701 mov delay r1 load
register r1 with delay
> 001010 007777 delay
> 001012 005301 dec r1 decrement register 1
> 001014 001376 bne -2 continue to decrement r1 until r1=0
-2 = -2
> 001016 000772 br -12 back to
001002 (dec r0)
-6 <> -12 ^
comment should be "001004",
^ comment should be
"rotate r0" not "dec r0"
At 1014 the machine code and assembler instruction are consistent, and correct for word
offsets.
At 1016 the machine code and assembler instruction are NOT consistent, but machine code is
still correct for word offsets.
(And presuming branch offsets are calculated relative to instruction following the branch
(after PC increment).)
There are some inconsistencies between the machine code / mnemonics / comments, but that
doesn't explain why it doesn't work.
I'm sure the '11 experts will chime in here...
1:46 p.m.
Brent Hilpert wrote:
Jay West wrote:
Using simh you can see what the code does.
A couple possibilities as to why it isn't.
1. Make sure console is set up to monitor R0
iirc: there are knobs on the front panel to determine what
you are looking at.
2. Running too fast? increase the value at 1010 (077777?)
Here is a badly documented run of this program in simh
** Start up simh emulator
./BIN/pdp11
PDP-11 simulator V3.3-1
** Load in the program
sim> d 001000 012700
sim> d 001002 000001
sim> d 001004 006100
sim> d 001006 012701
sim> d 001010 007777
sim> d 001012 005301
sim> d 001014 001376
sim> d 001016 000772
* Verify what we entered
sim> e 001000-001016
1000: 012700
1002: 000001
1004: 006100
1006: 012701
1010: 007777
1012: 005301
1014: 001376
1016: 000772
* Look at the disassembly
sim> e -m 001000-001016
1000: MOV #1,R0
1004: ROL R0
1006: MOV #7777,R1
1012: DEC R1
1014: BNE 1012
1016: BR 1004
** Note that the BR is to 1004, not 1002
** Load start address
sim> de pc 01000
** Set up a breakpoint
sim> br 001016
** Single step through instructions
sim> s
Step expired, PC: 001004 (ROL R0)
sim> s
Step expired, PC: 001006 (MOV #7777,R1)
sim> s
Step expired, PC: 001012 (DEC R1)
sim> s
Step expired, PC: 001014 (BNE 1012)
Step expired, PC: 001012 (DEC R1)
sim> s
Step expired, PC: 001014 (BNE 1012)
** This gets boring after a while, so we execute till the breakpoint
sim> go
Breakpoint, PC: 001016 (BR 1004)
sim> e r0
R0: 000002
sim> go
Breakpoint, PC: 001016 (BR 1004)
sim> e r0
R0: 000004
** And see where it goes from there
sim> s
Step expired, PC: 001004 (ROL R0)
sim> s
Step expired, PC: 001006 (MOV #7777,R1)
It's been decades since I looked at '11
assembler
(Disclaimer: I haven't dealt with '11 assembler since ~1982 either.)
So, why would the branch at location 1016 go back
to the data
at 1002? Obviously there's something I'm missing.
The comment is incorrect. The intent is to go to 1004.
Jim wrote....
>Location Content Op-code comment
>001000 012700 mov #1,r0 load ro with 1
>
>
^ comment should be
"r0" not "ro"
>001002 000001
>001004 006100 rol rotate r1 left
>
>
^ comment should be
"rotate r0 left"
>001006 012701 mov delay r1 load
register r1 with delay
>001010 007777 delay
>001012 005301 dec r1 decrement register 1
>001014 001376 bne -2 continue to decrement r1 until r1=0
>
>
-2 = -2
>001016 000772 br -12 back to
001002 (dec r0)
>
>
-6 <> -12 ^ comment should
be "001004",
^ comment should be
"rotate r0" not "dec r0"
At 1014 the machine code and assembler instruction are consistent, and correct for word
offsets.
At 1016 the machine code and assembler instruction are NOT consistent, but machine code is
still correct for word offsets.
(And presuming branch offsets are calculated relative to instruction following the branch
(after PC increment).)
There are some inconsistencies between the machine code / mnemonics / comments, but that
doesn't explain why it doesn't work.
I'm sure the '11 experts will chime in here...
11:38 a.m.
On Sun, 2005-05-22 at 20:49 -0500, Jay West wrote:
Jim wrote....
Here's the light chaser I use:
1 .asect
2 001000 . = 1000
3 start:
4 001000 005000 clr r0
5 001002 005200 inc r0
6 loop:
7 001004 006100 rol r0
8 001006 000005 reset
9 001010 000775 br loop
10 .end start
--
TTFN - Guy
Location Content Op-code comment
001000 012700 mov #1,r0 load ro with 1
001002 000001
001004 006100 rol rotate r1 left
001006 012701 mov delay r1 load register r1 with delay
001010 007777 delay
001012 005301 dec r1 decrement register 1
001014 001376 bne -2 continue to decrement r1 until
r1=0
001016 000772 br -12 back to 001002 (dec r0)
This runs on my /45, but produces a steady light display, not a cylon or
"blinking" effect. It's been decades since I looked at '11 assembler.
I'm
confused about something in the listing above. If the 012700 at location
1000 is a load r0 with 1, I'm guessing the next word (1002) is the constant
to be loaded? So, why would the branch at location 1016 go back to the data
at 1002? Obviously there's something I'm missing. Guidance? If 1002 is a DEC
R0 instruction, then where's the constant "1" for the mov#1,r0 stored in
the
object? Humm? 
8:15 a.m.
Location Content Op-code comment
001000 012700 mov #1,r0 load ro with 1
001002 000001
001004 006100 rol rotate r1 left
001006 012701 mov delay r1 load register r1 with delay
001010 007777 delay
001012 005301 dec r1 decrement register 1
001014 001376 bne -2 continue to decrement r1 until
r1=0
001016 000772 br -12 back to 001002 (dec r0)
This runs on my /45, but produces a steady light
display, not a cylon or
"blinking" effect. It's been decades since I looked at '11 assembler.
I'm
confused about something in the listing above. If the 012700 at location
Try increasing the delay at 1010.
1000 is a load r0 with 1, I'm guessing the next
word (1002) is the
constant to be loaded? So, why would the branch at location 1016 go back
to the data at 1002?
Obviously there's something I'm missing. Guidance?
pc = pc + (2*branch_offset)
the branch branches back to the instruction at 1004. The comment is
obviously wrong. I would suspect that the comment was for code that
this code was adapted from.
btw - the macro code at 1006 should be 'mov #delay,r1'.
If 1002 is a DEC R0 instruction, then where's the
constant "1" for the
mov#1,r0 stored in the object? Humm?
1002 is the initial value to load into R0... there is no 'DEC R0' in
the code above. The BR at the end branches back to 1004
Megan Gentry
Former RT-11 Developer
+--------------------------------+-------------------------------------+
| Megan Gentry, EMT/B, PP-ASEL,ST| email: mbg at world.std.com |
| Member of Technical Staff | megan at savaje.com |
| SavaJe Technologies, Inc. | (s/ at /@/) |
| 100 Apollo Drive | URL: http://world.std.com/~mbg/ |
| Chelmsford, MA 01824 | "pdp-11 programmer - some assembler |
| (978) 256 6521 (DEC '77-'98) | required." - mbg KB1FCA |
+--------------------------------+-------------------------------------+
7001
days inactive
7005
days old
5 comments
6 participants
participants (6)
-
ggs＠shiresoft.com -
hilpert＠cs.ubc.ca -
jim＠g1jbg.co.uk -
jwest＠classiccmp.org -
kth＠srv.net -
mbg＠theworld.com