DEC 11 XPTSA B D PDP Paper Tape Software Handbook

DEC-11-XPTSA-B-D PDP-11 Paper Tape Software Handbook DEC-11-XPTSA-B-D PDP-11 Paper Tape Software Handbook

User Manual: DEC-11-XPTSA-B-D PDP-11 Paper Tape Software Handbook

Open the PDF directly: View PDF .
Page Count: 280

Download
Open PDF In Browser	View PDF

PDP-11
Paper Tape
Software Handbook
Order No. DEC-11 -XPTSA-8-D

PDP-11
Paper Tape
Software Handbook
Order No. DEC-11-XPTSA-B-O

digital equipment corporation · maynard, massachusetts

First Printing, April
Revised:
March
January
February
June
April

1970
1971
1972
1973
1975
1976

The information in this document is subject to change without notice
and should not be construed as a commitment by Digital Equipment
Corporation. Digital Equipment Corporation assumes no responsibility
for any errors that may appear in this document.
The software described in this document is furnished under a license
and may be used or copied only in accordance with the terms of such
license.
Digital Equipment Corporation assumes no responsibility for the use or
reliability of its software on equipment that is not supplied by
DIGITAL.

The postage prepaid READER'S COMMENTS form on the last page of this
document requests the user's critical evaluation to assist us in preparing future documentation.

The following are trademarks of Digital Equipment Corporation:

DIGITAL
DEC
PDP
DECUS
UNIBUS
COMPUTER LABS
COMTEX
DDT
DECCOMM

7/77-34

DECsystem-lO
DECtape
DIBOL
EDUSYSTEM
FLIP CHIP
FOCAL
INDAC
LAB-8
DECsystem-20

MASSBUS
OMNIBUS
OS/8
PHA
RSTS
RSX
TYPESET-8
TYPESET-IO
TYPESET..;"ll

CONTENTS

Page
CHAPTER

PAL-lIS ASSEMBLY LANGUAGE AND ASSEMBLER

1-1

1.1
1.2
1.2.1
1. 2.2
1. 2 ~ 3
1. 2.4
1.2.S
1.3
1. 3.1
1.3.2
1. 3.3
1. 3.4
1.4
1. 4.1
1. 4.2
1.4.3
1. 4.4
loS
1.6
1.7
1. 7.1
1. 7.2
1. 7.3
1. 7.4
1. 7.S
1. 7.6
1. 7.7
1. 7.8
1. 7.9

CHARACTER SET
STATEMENTS
Label
Operator
Operand
Comments
Format Control
SYMBOLS
Permanent Symbols
User-Defined Symbols
Direct Assignment
Register Symbols
EXPRESSIONS
Numbers
Arithmetic and Logical Operators
ASCII Conversion
Mode of Expressions
ASSEMBLY LOCATION COUNTER
RELOCATION AND LINKING
ADDRESSING
. Register Mode
Deferred Register Mbde
Autoincrement Mode
Deferred AutoincrementMode
Autodecrement Mode
Deferred Autodecrement Mode
Index Mode
Deferred Index Mode
Immediate Mode and Deferred Immediate
(Absolute) . Mode
Relative and Deferred Relative Mode~
Table of Mode Forms and Codes (6-bit(A)
format only ~ see Section 1.7.12)
Instruction Forms
ASSEMBLER DIRECTIVES·
.TITLE
.GLOBL
Program Section Directives (.ASECT and
.CSECT)
.EOT
• EVEN
• END

1-2
1-2
1-3
1-3
1';"4
1-4
1-4
1-S
1-S
1-S
1-6
1-6
1-7
1-8
1-8
1-8
1-9
1-10
1-11
1-12
1-12
1-13
1-13
1-13
1-14
1-14
1-14
1-14

1. 7.10
1.7.11
1. 7 .12
1.8
1. 8.1
1. 8.2
1. 8.3
1. 8.4
1. 8. S
1.8.6
1.8.7
1. 8.8
1. 8.9
1. 8 .10
1.8.11
1. 8 .12

• WORD

. BYTE
.ASCII
.RADSO
.LIMIT
Conditional Assembly Directives
iii

I-IS
1-lS
1-16
1-17
1-18
1-18
1-18
1-19
1-20
1-20\.
1-20
1-20
1-:-21'
1,;..2L.
1-22··
1-'-23 '
1';"23

CONTENTS (CONT.)

Page

CHAPTER

1.9
1. 9.1
1. 9.2
1.9.3
1. 9.4
1. 9.5
1. 9.6
1.9.6.1
1:9.6.2
,1. 9.6.3
1.10
1.11

OPERATING PROCEDURES
Introduction
Loading PAL-lIS
Initial Dialogue
Assembly Dialogue
Assembly Listing
Object Module Output
Global Symbol Directory
Text Block
Relocation Directory
ERROR CODES
SOFTWARE ERROR HALTS

1-24
1-24
1-24
1-24
1-28
1-30
1-30
1-30
1-31
1-31
1-31
1-32

WRITING PAL-11A ASSEMBLY LANGUAGE PROGRAMS

2-1

2.1
2.2
2.2.1
2.2.2
2.2.3
2.2 .. 4
2.2.5
2.3
2.3.1
2.3.2
2.3.3
2.3.4
2.4
2.4.1
2.4.2
2.4.3
2.5
2.6
2.6.1
2.6.2
2.6.3
2.6.4
2.6.5
2.6.6
2.6.7
2.6.8
2.6.9

CHARACTER SET
STATEMENTS
Label
Operator
Operarid
Comments
Format Control
SYMBOLS
Permanent Symbols
User-Defined Symbols
Direct Assignment
Register Symbols
EXPRESSIONS
Nuinbers
Arithmetic and Logical Operators
ASCII Conversion
ASSEMBLY LOCATION COUNTER
ADDRESSING
Register Mode
Deferred Register Mode
Autoincrement Mode
Deferred Autoincrement Mode
Autodecrement Mode
Deferred Autodecrement Mode
Index Mode
Deferred Index Mode
Immediate Mode arid Deferred Immediate
(Absolute) Mode
Relative and Deferred Relative Modes
Table of Mode Forms and Codes (6-bit (A)
format only - see Section 3.7)
INSTRUCTION FORMS
ASSEMBLER DIRECTIVES
.EOT
• EVEN
• END
• WORD
• BYTE
• ASCII
OPERATING PROCEDURES
Introduction
Loading PAL-11A

2-2
2-2
2-3
2-3
2-3

2.6.10
2.6.11
2.7
2.8
2.8.1
2.8.2
2.8.3
2.8.4
2.8.5
2.8.6
2.9
2.9.1
2.9.2

"2-4

2-4
2-5
2~5

2-5
2-5
2-6
2-7
2-7
2-8
2-8
2-8
2-9
2-10
2-10
2-10
2-11
2-11
2-11
2-11
2-12
2-12
2-13
2-13
2-14
2-15
2-15
2-16
2-16
2-16
2-17
2-17
2-17
2-17
2-18

CONTENTS (CONT.)

Page
2.9.3
2.9.4
2.9.5
2.10
CHAPTER

CHAPTER

2.11

Initial Dialogue
Assembly Dialogue
Assembly Listing
ERROR CODES
SOFTWARE ERROR HALTS

2-18
2-23
2-24
2-25
2-26

LINK-llS LINKER

3-1

3.1
3.1.1
3,.1. 2
3.1. 3
3.2
3.2.1
3.2.2
3.2.3
3.3
3.3.1
3.3.1.1
3.3.2
3.3.2.1
3.3.2.2
3.3.2.3
3.3.2.4

INTRODUCTION
General Description
Absolute and Relocatab1e Program Sections
Global Symbols
INPUT AND OUTPUT
Object Module
Load Modules
Load Map
OPERATING PROCEDURES
Loading and Command String
Operational Cautions
Error Procedure and Messages
Restarting
Non-Fatal Errors
Fatal Errors
Error HALTs

3-1
3-1
3-2
3-2
3-3
3-3
3-4
3-5
3-5
3-6
3-7
3-7
3-7
3-7
3-8

EDITING THE SOURCE PROGRAM

4-1

4.1
4.2
4.2.1
4.2.2
4.2.3
4.2.4
4.2.5
4.3
4.3.1
4.3.1.1
4.3.1.2
4.3.1.3
4.3.1.4
4.3.1.5
4.3.1.6
4.3.2
4.3.2.1
4.3.2.2
4.3.2.3
4.3.3
4.3.3.1
4.3.3.2
4.3.4
4.3.4.1
4.3.4.2
4.3.4.3

COMMAND MODE AND TEXT MODE
COMMAND DELIMITERS
Arguments
The Character Location Pointer (Dot)
Mark
Line-Oriented Command Properties
The Page Buffer
COMMANDS
Input and Output Commands
Open
Read
List and Punch
Next
Form Feed and Trailer
Procedure with Low-Speed Punch
Commands to Move Dot and Mark
Beginning and End
Jump and Advance
Mark
Search Commands
Get
wHole
Commands to Modify the Text
Insert
Delete and Kill
Change and Exchange
OPERATING PROCEDURES
Error Corrections
Starting
Restarting

4-1

4.4

4.4.1
4.4.2
4.4.3

3~3

4-2
4-2
4-2
4-3
4-3
4-3
4-4

4-4
4-4
4-4
4-5
4-5
4-6
4-6
4-6
4-6
4-6

4-7
4-7
4-7
4-7
4-8
4-8
4-9
4-9
4-10
4-10
4-11

4-11

CONTENTS (CONT.)

Page

CHAPTER

4.4.4
4.4.5
4.5

Creating a Paper Tape
Editing Example
SOFTWARE ERROR HALTS

4-11
4-12
4-17

DEBUGGING OBJECT PROGRAMS ON-LINE

5-1

5.1
5.1.1
5.1. 2
5.2
5 •.2.1
5.2.1.1
5.2.1.2
5.2.1.3
5.2.1.4
5.2.1.5
5.2.1.6
5.2.2
5.2.2.1
5.2.2.2
5.2.3
5.2.4
5.2.4.1
5.2.4.2
5.2.5
5.2.6
5.3
5.3.1
5.3.1.1
5.3.1.2
5.3.1.3
5.3.2
5.3.3
5.3.4
5.4
5.5
5.5.1
5.5.2
5.5.3
5.5.4
5.6
5.6.1
5.6.2

5-1
5-1
5-2
5-3
5-4
5-4
5-4
5-5
5-5
5-5
5-6
5-6
5-6
5-7
5-7
5-8
5-8
5-9
5-9
5-10
5-10
5-10
5-11
5-11
5-11
5-11
5-12
5-12
5-13
5-14
5-14
5-14
5-18
5-19
5-20
5-20

5.6.3

INTRODUCTION
ODT-ll and ODT-lIX
ODT's Command Syntax
COMMANDS AND FUNCTIONS
Opening, Changing, and Closing Locations
The Slash U)
The LINE FEED Key
The Up-Arrow (t)
The Back-Arrow (+)
Accessing General Registers 0-7
Accessing Internal Registers
Breakpoints
Setting the Breakpoimt (niB)
Locating the Breakpoint($B)
Running the Program(niG and-niP)
Searches
Word Seach(niW)
Effective Address Search(niE)
Calculating Offsets(niO)
ODT'S Priority Level($P)
ODT-IIX
Opening, changing and Closing Locations
Open the AddressedLocation(@)
Relative Branch Offset(»
Return to Previous Sequence«)
Calculating Offsets(niO)
Breakpoints
Single-Instruction Mode
ERROR DETECTION
PROGRAMMING. CONSIDERATIONS
Functional Organization
Breakpoints
Search
Teletype Interrupt
OPERATING PROCEOURES'
Linking Procedures (LSI-II Systems Only)
Loading Procedures (non-LS'I-ll Systems
Only)
Starting and Restarting

LOADING AND DUMPING MEMORY

6-1

6.1
6.1.1
6.1. 2
6.1. 3
6.1.4
6.1. 5
6.1. 6
6.1.6.1

PAPER TAPE BOOTSTRAPS
BM792-YAPaper Tape Bootstrap ROM
BM873-YA Bootstrap Loader ROM
LSI-II Firmware Paper Tape Bootstrap
M9301-YB Bootstrap Loader
M9301-YA Bootstrap Loader
Other Bootstrap Loaders
Loading the Loader into Core

6-2
6-2
6-2
6-3
6-3
6-4
6-4
6-5

5-20
5-21

CONTENTS (CONT • )

Page
6.1.6.2
6.1.6.3
6.2
6.2.1
6.2.2
6.2.3
6.3
6.3.1
6.3.1.1

6-6
6-8
6-10
6-11
6-11
6-13
6-14
6-14

6.3.2
6.3.3

Loading Bootstrap Tapes
Bootstrap Loader Operation
THE ABSOLUTE LOADER
Loading the Loader into Core
Using the Absolute Loader .
Absolute Loader Operation
CORE MEMORY DUMPS
Operating Procedures
Using DUMPAB on Systems without Switch
Registers
Using DUMPAB and DUMPTT on Systems with
Switch Registers
Output Formats .
Storage Maps

INPUT/OUTPUT PROGRAMMING

7-1

7.1
7.1.1
7.1.2
7.1. 3
7.2
7.2.1
7.2.2
7.3
7.3.1
7.3.2
7.3.3
7.3.3.1
7.3.3.2
7.3.3.3
7.3.3.4
7.3.4
7.4
7.4.1
7.4.2
7.4.3
7.4.4
7.5
7.5.1
7.5.2
7.5.3
7.5.4
7.5.5.
7.5.6
7.5.7
7.5.8
7.5.9
7.6
7.6.1
7.6.2
7.7
7.8
7.9
7.9.1
7.9.2
7.9.3

INTRODUCTION
UsingIOX with the LSI-II Processor
Using IOX with Unibus PDP-II Processors
IOX Interrupt and Trap Vectors
THE DEVICE ASSIGNMENT TABLE
Reset
Initialization
BUFFER ARRANGEMENT IN DATA TRANSFER COMMANDS
Buffer Size
Mode Byte
Status Byte
Non~Fata1 Error Codes (Octal).
Done Bit
End-Of;';'Medium Bit
End-Of-FileBit
Byte Count
MODES
Formatted ASCII
Unformatted ASCII
Formatted Binary
Unformatted Binary
DATA TRANSFERS
Read
Wr~te
.
Device Conflicts In Data Transfer Commands
Waitr (Wait, Return)
Waitr vs. Testing the Buffer Done Bit
Single Buffer Transfer6n One Device
Double Buffering
Readr (Real-time Read)
Writr (Real-time Write)
REENABLIN<;i THE READER AND RESTARTING
Seek
.
Restart
FATAL ERRORS
EXAMPLE OF PROGRAM USING IOX
IOX INTERNAL INFORMATION
Confltct Byte/Word
Device Interrupt Table (DIT)
Device Status Table (DST)

7-1
7-3
7-3
7-3
7-3
7-4
7-4
7-4
7-5
7-5
7-6
7-6
7-7
7-7
7-7
7-8
7-8
7-8
7-10
7-10
7-11
7-11
7-11
7-12
7-12
7-13
7-13
7-14
7-15
7-15
7-16
7-16
7-16
7-16
7-17
7-17
7-19
7-19
7-20
7-21

6.3.1.2

CHAPTER

vii

6-15
6-16
6-17
6-17

CONTENTS (CONT.)

Page
7.9.4
7.9.5
7.9.5.1
7.9.5.2
7.9.5.3

Teletype Hardware Tab Facility
Adding Devices To lOX
Device Codes
Table Modification
Interrupt Routines

7-21
7-21
7-21
7-22
7-23

CHAPTER

FLOATING POINT MATH PACKAGE OVERVIEW

8-1

CHAPTER

PROGRAMMING TECHNIQUES

9-1

9.1
9.1.1
9.1. 2
9.1. 3
9.1. 4
9.1.4.1
9.1.4.2
9.1.4.3
9.2
9.3
9.3.1
9.3.2

WRITING POSITION .INDEPENDENT CODE
Position Independent Modes
Absolute Modes
Writing Automatic PIC
Writing Non-Automatic PIC
Setting Up The Stack Pointer
Setting Up A Trap or Interrupt Vector
Relocating Pointers
LOADING UNUSED TRAP VECTORS
CODING TECHNIQUES
Altering Register Contents
Subroutines

9-1
9-2
9-3
9-3
9-4
9-4
9-4
9-5
9-5
9-6
9-6
9-7

APPENDIX A

ASCII CHARACTER SET

A-I

APPENDIXB

PAL-lIS ASSEMBLY LANGUAGE AND ASSEMBLER

B-1

B.l
B.2
B.3
B.3.1
B.3.2
B.3.3
B.3.4
B.3.5
B.3.6
B.3.7
B.3.8
B.4
B.5
B.6
APPENDIX C
C.l
C.2
C.3
C.3.1
C.3.2
C.3.3
C.3.4
C.3.5
C.3.6

TERMINATORS
B-1
ADDRESS MODE SYNTAX
B-2
INSTRUCTIONS
B-2
Double Operand Instructions OP A,A
B-4
Single Operand Instructions OP A
B-4
Rotate/Shift
B-5
Operation Instructions Op
B-5
Branch Instructions Op E Where
-128 10 «E-.-2)/2<127 10
B-6
Subroutine Call JSR ER,A
B-7
B-7
Subroutine Return
Extensions for the LSI-II Version Of PAL-lIS B-7
ASSEMBLER DIRECTIVES
B-8
ERROR CODES
B-IO
B-ll
INITIAL OPERATING PROCEDURES
PAL-IIA ASSEMBLY LANGUAGE AND ASSEMBLER

C-l

SPECIAL CHARACTERS
ADDRESS MODE SYNTAX
INSTRUCTIONS
Double-Operand Instructions Op A,A
Single-Operand Instructions Op A
Rotate/Shift Instructions Op A
Operate Instructions Op
Trap Instructions Op or OpE Where
Branch InstructionsOp E where
-128 10 ~ (E-.-2)/2~12710

C-l
C-2
C-3
C-4
C-4
C-5
C-5
C-6

viii

0~E~3778

C-7

CONTENTS (CONT • )

Page
C.3.7
C.3.S
C.4
C.S
C.6

Subroutine Call Op ER, A
Subroutine Return Op ER
ASSEMBLER DIRECTIVES
ERROR CODES
INITIAL OPERATING PROCEDURES

C-7
C-S
C-S
C-S
C-9

TEXT EDITOR, ED-ll

D-l

INPUT/OUTPUT COMMANDS
POINTER-POSITIONING COMMANDS
SEARCH COMMANDS
COMMANDS TO MODIFY THE TEXT
SYMBOLS
GROUPING OF COMMANDS
OPERATING PROCEDURES
Loading
Storage Requirements
Starting
Initial Dialogue
Restarting

D-l
D-2
D-2
D-2
D-3
D-3
D-4
D-4
D-4
D-4
D-4
D-4

DEBUGGING OBJECT PROGRAMS ON-LINE, ODT-ll
AND ODT-llX

E-l

SUMMARY OF CONTENTS

E-l

LOADING AND DUMPING CORE MEMORY

F-l

THE BOOTSTRAP LOADER
Loading The Bootstrap Loader
THE ABSOLUTE LOADER
CORE MEMORY DUMPS

F-l
F-l
F-3
F-4

INPUT/OUTPUT PROGRAMMING, lOX

G-l

INSTRUCTION SUMMARY
PROGRAM FLOW SUMMARY
FATAL ERRORS

G-l
G-l
G-2

SUMMARY OF FLOATING POINT MATH PACKAGE,
FPMP-ll

H-l

OTS ROUTINES
NON-OTS ROUTINES
ROUTINES ACCESSED VIA TRAP HANDLER

H-2
H-7
H-7

APPENDIX I

TAPE DUPLICATION

I-I

APPENDIX J

ASSEMBLY AND LINKING INSTRUCTIONS

J-l

SYSTEMS WITHOUT SWITCH REGISTERS
IOX/IOXLPT
Assembling lOX
Assembling IOXLPT
Linking lOX and IOXLPT
ODTllX

J-l
J-l
J-l
J-l
J-l
J-l

APPENDIX D
D.l
D.2
D.3
D.4
D.S
D.6
D.7
D.7.l
D.7.2
D. 7.3
D.7.4
D.7.S
APPENDIX E
E.l
APPENDIX F
F.l
F.l.l
F.2
F.3
APPENDIX G
G.l
G.2
G.3
APPENDIX H
H.1
H.2
H.3

J.l
J .l.1

J.l.l.l
J.l.l.2
J.l.l.3
J .l.2

CONTENTS (CONT.)

Page
J.1.2.l
J.1.2.2
J.1. 3
J.1.3.l
J.1.3.2
J.1. 4
J.1.4.l
J.1.4.2
J.1. S
J.1.S.l
J.1.S.2
J.2
J.2.l
J.2.2
J.2.3
J.2.4
J.2.S
J.2.6

Assembling "ODTllX
Linking ODTllX
ED-II
Assembling ED-II
Linking ED-II
PAL-llS
Assembling PAL-lIS
Linking PAL-lIS,
LINK-lIS
Assembling LINK-lIS
Linking LINK-lIS
SYSTEMS WITH SWITCH REGISTERS
Assembling PAL-llA
Assembling ED~ll
ODT-ll/ODT-llX
Assembling IOX/IOXLPT
Assembling and Linking PAL-lIS
Assembling and Linking LINK-lIS

J-l
J-2
J-2
J-2
J-2
J-2
J-2
J-3
J-3
J-4
J-4
J-S
J-S
J-6
J-6
J-7
J-8
J-ll

APPENDIX K

STANDARD PDP-II ABBREVIATIONS

K-l

APPENDIX L

CONVERSION TABLES

L-l

OCTAL-DECIMAL INTEGER CONVERSIONS
POWERS OF TWO
SCALES OF NOTATION
2 x In-Decimal
10±n In Octal
n Log 2 and 10 In Decimal
Addition and Multiplication, Binary and
Octal
Mathematical Constants In Octal

L-l
L-S
L-6
L-6
L-6
L-6

L.l
L.2
L.3
L.3.l
L.3.2
L.3.3
L.3.4
L.3.S
APPENDIX M
APPENDIX N
N.l
N.2
N.3
INDEX

L-6
L-7

NOTE TO USERS OF SERIAL LA30 AND 600, 1200,
AND 2400 BAUD VTOS'S

M-l

USING THE ABSOLUTE LOADER ON PDP-II'S WITHOUT
SWITCH REGISTERS

N-l

LSI-ll
M930l-YB BOOTSTRAP LOADER
M930l-YA BOOTSTRAP LOADER

N-l
N-3
N-4
Index-l

CONTENTS (CONT . )

Page
FIGURES
FIGURE

5-1
6-1
6-2
6-3
6-4
F-l
F-2
F-3
F-4

Communication and Data Flow
Bootstrap Loader Instructions
Loading and Verifying the Bootstrap Loader
Loading Bootstrap Tapes Into Core
The Bootstrap Loader Program
Loading and Verifying the Bootstrap Loader
Loading Bootstrap Tapes into Core
Loading with the Absolute Loader
Dumping Using DUMPAB or DUMPTT

5-15
6-4

6-7
6-8
6-9
F-2
F-3
F-5
F-6

TABLES
TABLE

1-1
2-1

Instruction Operand Fields
Instruction Operand Fields

1-17
2-14

CHAPTER 1
PAL-lIS ASSEMBLY LANGUAGE AND ASSEMBLER

PAL-lIS Assembly
(Program Assembly
Language
for
the
PDP-II,
Relocatable,
Stand Alone Version)
enables you to write source
(symbolic) programs using letters, numbers, and symbols which are
meaningful to you.
The source programs,generated either on-line
using the Text Editor (ED-II), or off-line, are then assembled into
object modules which are processed by the PDP-II linker, LINK-lIS.
LINK-lIS produces a load module which is loaded by the Absolute Loader
for execution. Object modules may contain absolute and/or relocatable
code and separately assembled object modules may be linked with global
symbols.
The object module is produced after two passes through the
Assembler; an optional third pass produces a complete octal/symbolic
listing of the assembled program. This listing is especially useful
for documentation and debugging purposes.
ThLs chapter not only explains how to write PAL-lIS programs but also
how to assemble the source programs into object modules. All facets
of the assembly language are explained and illustrated with many
examples, and the chapter concludes with assembling procedures.
In
explaining how to write PAL-lIS source programs, it is necessary,
especially at the outset, to make frequent forward references.
Therefore, we recommend that you first read through the entire chapter
to get a "feel" for the language, and then reread the chapter, this
time referring to appropriate sections as indicated, for a thorough
understanding of the language and assembling procedures.
Some notable features of PAL-lIS are:
1.

Selective assembly pass functions.

Device specification for pass functions.

Optional error listing on the teleprinter.

Double buffered and concurrent I/O (provided by IOXLPT).

Alphabetized, formatted symbol table listing.

Relocatable object modules.

Global symbols for linking between object modules.

Conditional assembly directives.

Program Sectioning Directives.

The PAL-lIS Assembler requires 8K of memory and provides for about 900
user-defined symbols
(see Section 1.3.2). In addition, it allows a
line printer to be used for program listing and/or symbol table
listing.
1-1

PAL-lIS ASSEMBLY LANGUAGE AND ASSEMBLER
The following discussion of the PAL-lIS Assembly Language assumes that
you have read the PDP-II Processor Handbook with emphasis on tqose
sections which deal with the PDP-II instruct jon repertoire, forrrt-el>ts,
and timings -- a thorough knowledge of these is vital to efficient
assembly language programming.

1.1

CHARACTER SET

A PAL-lIS source program is composed of symbols, numbers, expressions,
symbolic instructions, assembler directives,-argument separators, and
line terminators written using the following ASCII) characters.
1.

The letters A through Z.
(Upper and lower case letters are
acceptable, although upon input, lower case letters will be
converted to upper case letters.)

The numbers 0 through 9.

The characters . and $.
systems use.)

The separating or terminating symbols:
%

carriage return

1.2

; (These characters are

"
tab

space

+
line feed

reserved

for

form feed

STATEMENTS

A source program is composed of a sequence of statements, where each
statement is on a single line.
The statement is terminated by a
carriage return character which must be immediately followed by either
a ~ine feed or form feed character.
Should a carriage return
character be present and not be followed by a line feed or form feed,
the Assembler will generate a Q err.or (Section 1.10), and that· portion
of the line following the carriage return will be ignored. Since the
carriage return terminator is a required statement terminator, a line
feed or form feed not immediately preceded by a carriage return will
have one inserted by the Assembler.
It should be noted that, if the Editor (ED-II) is being used to create
the source program, a typed carriage return (RETURN key) automatically
generates a line feed character.
A statement may be composed of. up to four fields which are identified
by their orde~ of appearance and by specified terminating characters
as explained below and summar ized in Appendix .B. The four fields are:
Label

Operator

Operand

Comment

The label and <;::omment fields are optional. The operator and operand
fields are inter-dependent -- either may be omitted depending upon the
contents of the other.

)

ASCII stands for American Standard Code for Information Interchange.

1-2

PAL-lIS ASSEMBLY LANGUAGE AND ASSEMBLER
1.2.1

Label

A label is a user-defined symbol (see Section 1.3.2) which is assigned
the value of the current location counter. This value may be either
absolute or relocatable depending on whether the location counter
value is absolute or relocatable.
In the latter case, the final
absolute value is assigned by the Linker, i.e.,
the value + the
relocation constant.
A label is a symbolic means of referring to a
specific location within a program.
If present, a label always occurs
first in a statement and must be terminated by a colon.
For example,
if the current location is absolute 100 the statement:
ABCD:

MOV A,B

will assign the value 100
to the label ABCD so that subsequent
reference to ABCD will be to location 100
In the above case if the
location counter were relocatable the~ the final value of ABCD would
be 100 +K, where K is the location of the beginning of the relocatable
section in which the label ABCD appears.
More than one label may
appear within a single label field;
each label within the field will
have the same value.
For example, if the current location counter is
100 I multiple labels in the statement:
ABC:

$DD:

MOV A,B

A7.7:

will equate each of the three labels ABC, $DD, and A7.7 with the value
100
($ and . are reserved for system software).
The error code M (multiple definition of a symbol) will be generated
during assembly if two or more labels have the same first six
characters.

1.2.2

Operator

An operator follows the label field in a statement, and may be an
instruction mnemonic or an assembler directive (see Section 1.8 and
Appendix B). When it is an instruction mnemonic,
it specifies what
action is to be performed on any operand(s) which follows it. When it
is an assembler directive, it specifies a certain function or action
to be performed during assembly.
The operator may be preceded only by one or more labels and may be
followed by one or more operands and/or a comment. An operator is
legally terminated by a space, tab, or any of the
following
characters:

line feed

form feed

carriage return

The use of each character above will be explained in this chapter.
Consider the following examples:
MOV -...j A,B
MOV@A,B

"'I (TAB) terminates operator MOV
;@ terminates operator MOV

When the operator stands alone without an operand or comment,
it is
terminated by a carriage return followed by a line feed or form feed
character.

1-3

PAL-lIS ASSEMBLY LANGUAGE AND ASSEMBLER
1.2.3

Operand

An operand is that part of a statement which is operated on
operator -- an instruction mnemonic or assembler directive.
may be symbols, expressions, or numbers.
When multiple
appear within a statement, each is separated from the next by
An operand may be preceded by an operator and/or label, and
by a comment.

by the
Operands
operands
a comma.
followed

The operand field is terminated by a semicolon when followed by a
comment, or by a carriage return followed by a line feed or form feed
character when the operand ends the statement. For example,
LABEL:

MOV GEORGE"BOB

iTHIS IS A COMMENT

where the space between MOV and GEORGE terminated the operator field
and began the operand field;
the comma separated the operands GEORGE
and BOB;
the semicolon terminated the operand field and began the
comment.

1.2.4

Comments

The comment field is optional and may contain any ASCII character
except null, rubout, carriage return, line feed or form feed. All
other characters, even those with special significance are ignored by
the AsSembler when used in the comment field.
The comment field may be preceded by none, any, or all of the other
three fields.
It must begin with the semicolon and end with a
carriage return followed by a line feed or form feed character.
For
example,
LABEL:

iTHIS IS A $1.00 COMMENT

CLR HERE

Comments do not affect assembly processing or program execution, but
they are useful in program listings for later analysis, checkout or
documentation purposes.

1.2.5

Format Control

The format is controlled by the space and tab characters.
They have
no effect on the assembling process of the source program unless they
are embedded within a symbol, number, or ASCII text; or are used as
the operator field terminator. Thus, they can be used to provide a
neat, readable program. A statement can be written:
LABEL:MOV(SP)+,TAGiPOP VALUE OFF STACK
or, using formatting characters it can be written:
LABEL:

MOV (SP)+,TAG

jPOP

VALUE OFF STACK

which is much easier to read.
Page size is controlled by the form feed character. A page of n lines
is created by inserting a form' feed (CTRL/FORM keys on the keyboard)
after the nth line.
If no form feed is present, a page is
automatically terminated after 56 lines.

1-4

PAL-lIS ASSEMBLY LANGUAGE AND ASSEMBLER
1.3

SYMBOLS

There are two types of symbols J permanent and user-defined.
Both are
stored in the Assembler's symbol table.
Initially, the symbol table
contains the permanent symbols, but as the source program
is
assembled, user~defined symbols are added to the table.

1.3.1

Permanent Symbols

Permanent symbols consist of the instruction mnemonics
(see Appendix
B.3)
and assembler directives (see Section 1.8). These symbols are a
permanent part of the Assembler's symbol table and need not be defined
before being used in the source program.

1.3.2

User-Defined Symbols

User-defined symbols are those defined as labels (see Section 1.2.1)
or by direct assignment (see Section 1.3.3). These symbols are added
to the symbol table as they are encountered during the first pass of
the assembly. They can be composed of alphanumeric characters, dollar
signs, and periods onlyp again $'s and. 's are reserved for system
software. Any other character is illegal and, if used, will result in
the error message I or QU (see Section 1.10). I is a low priority
error which may be flagged as QU first.
The following rules also
apply to user-defined symbols:
1.

The first character must not be a number.

Each symbol must be unique within thefLrst six characters.

A symbol may be written with more than six legal characters
but the seventh and subsequent characters are only checked
for legality, and are not otherwise recognized by the
Assembler.

Spaces and tabs must not be embedded within a symbol.

A user-defined symbol may duplicate a permanent symbol.
The value
associated with a permanent symbol that is also user-defined depends
upon its use:
1.

A permanent symbol encountered in the operator
associated with its corresponding machine op-code.

field

If a permanent symbol in the operand field
is
also
user-defined,
its user-defined value is associated with the
symbol.
If the symbol is not found to be user-defined, then
t.he corresponding machine op-code value is associated with
the symbol.

User-defined symbols are either internal or global. All symbols are
internal unless they are explicitly typed as global with the .GLOBL
assembler directive (see Section 1.8.2). Global symbols are used to
provide links between object modules.
A global symbol which is
defined (as a label or by direct assignment) in a program is called an
entry symbol or entry point. Such symbols may be referred to from
other object modules or assemblies. A global symbol which is not
defined in the current assembly is called an external symbol. Some
other assembly must define the same symbol as an entry point.

1-5

PAL-lIS ASSEMBLY LANGUAGE AND ASSEMBLER
1.3.3

Direct Assignment

A direct assignment statement associates a symbol with a value.
When
a direct assignment statement defines a symbol for the first time,
that symbol is entered into the Assembler's symbol table and the
specified value is associated with it. A symbol may be redefined by
assigning a new value to a previously defined symbol.
The newly
assigned value will replace the previous value assigned to the symbol.
The symbol takes on the relocatable or absolute attribute of the
defining expression.
However, if the defining expression is global,
the defined symbol will not be global unless previously defined as
such (see Section 1.4).
The general format for a direct assignment statement is:
symbol = expression.
The following conventions apply:
1.

An equal sign
(=)
must separate
expression defining the symbol.

the

symbol

A direct assignment statement may be preceded by a label
may be followed by a comment.

Only one symbol can be defined by anyone
statement.

Only one level of forward referencing is allowed.

direct

from

the
and

assignment

Example of two levels of forward referencing (illegal):

x
Y

Y
Z

X and Yare both undefined throughout pass 1 and will be listed on the
teleprinter as such at the end of that pass.
X is undefined
throughout pass 2, and will cause a U error message.
Examples:

c:
E:

1.3.4

A=l

;THE SYMBOL A IS EQUATED WITH THE VALUE 1

B='A-l&MASKlOW

;THE SYMBOL B IS EQUATED WITH THE EXPRESSION'S
;VAlUE

D=3
MOV livABLE

;THE SYMBOL D IS EQUATED WITH 3. THE
;LABELS C AND E ARE EQUATED WITH THE
;NUMERICAl MEMORY ADDRESS OF THE MOV
;COMMAND

The eight general registers of the PDP-II are numbered 0 through 7.
These registers may be referenced by use of a register symbol;
that
is, a symbolic name for a register. A register symbol is defined by
means of a direct assignment, where the defining expression contains
at least one term preceded by a % or at least one term previously
defined as a register symbol.
In addition, the defining expression of
a register symbol must be absolute.
For example:
1-6

PAL-lIS ASSEMBLY LANGUAGE AND ASSEMBLER

;DEFINE RO AS REGISTER 0
;DEFINEDR3 AS REGISTER 3
;DEFINE R4 AS REGISTER 4
;DEFINE "THERE" AS REGISTER 2
It is important to note that all register symbols must be defined
before they are referenced. A forward reference to a register symbol
will generally cause phase errors (see Section 1.10).
The % may be used in any expression thereby indicating a reference to
a register.
Such an expression is a register expression. Thus~ the
statement:
CLR

will clear register 6 while the statement:
CLR

will clear the word at memory address 6.
In certain cases a register
can be referenced without the use of a register symbol or register
expression. These cases are recognized through the context of the
statement and are thoroughly explained in Sections 1.7.11 and 1.7.12.
Two obvious examples of this are:

;THE FIRST OPERAND FIELD MUST ALWAYS
; DE A F~EG I STEF~
CI... F~

1.4

;ANY EXPRESSION ENCLOSED IN ( ) MUST DE
;A REGISTER. IN THIS CASE? INDEX REGISTER

EXPRESSIONS

Arithmetic and logical operators (see Section 1.4.2) may be used to
form expressions.
A term of an expression may be a permanent or
user-defined symbol (which may be absolute, relocatable or global), a
number, ASCII data, or the present value of the assembly location
counter represented by the period (see Section 1.5). Expressions are
evaluated from left to right.
Parenthetical grouping is not allowed.
Expressions
directive
truncation
expression

are evaluated as word quantities. The operands of a .BYTE
(Section 1.8.8) are evaluated as word expressions before
to the low-order eight bits.
The evaluation of an
includes the evaluation of the mode of the resultant
expression~
that is, absolute, relocatable or
external.
The
definition of the modes of ~xpression are given below in Section
1.4.4.
A missing term, expression or external symbol will be interpreted as
O.
A missing operator will be interpreted as +. The error code Q
For
(Questionable syntax) will be generated for a missing operator.
example,
A + -100
will be evaluated as A + 0
TAG ! LA 177777
will be evaluated as TAG

;OPERAND MISSING

100, and
;OPERATOR MISSING
LA+177777.
1-7

PAL-llS ASSEMBLY LANGUAGE AND ASSEMBLER
The value of an external expression will be the value of the absolute
part of the expression;
e.g., EXT+A will have a value of A.
This
will be modified by the linker to become EXT+A.

1.4.1

Numbers

The Assembler accepts both octal and decimal numbers.
Octal numbers
consist of the digits 0 through 7 only.
Decimal numbers consist of
the digits 0 through 9 followed by a decimal point.
If a number
contains an 8 or 9 and is not followed by a decimal point, the N error
code (see Section 1.10f will be printed and the number will be
interpreted as decimal.
Negative numbers may be expressed as a number
preceded by a minus sign rather
than in a
two's complement form.
Positive numbers may be preceded by a plus sign although this is not
required.
If a number is too large to fit into 16 bits, the number is truncated
from the left.
In the assembly listing the statement will be flagged
with a Truncation (T) error.
Numbers are always considered to be
absolute quantities (that is, not relocatable).

1.4.2

Arithmetic And Logical Operators

The arithmetic operators are:
+

indicates addition or a positive numhe.r
indicates subtraction or a negative number

The logical operators are:
indicates the logical AND operation

indicates the logical inclusive OR operation
AND
0
0
1
1

1.4.3

& 0
& 1
& 0
& 1

=
=
=
=

0
0
0
1

0
0
1
1

!
!
!
!

0
1
0
1

=
=
=
=

0
1
1
1

ASCII Conversion

When preceded by an apostrophe,
any ASCII character
(except null,
rubout,
carriage return,
line feed,
or form feed) is assigned the
7-bit ASCII value of the character (see Appendix A).
For example,

'A
is assigned the value lOIS'
When preceded by a quotation mark, two ASCII characters (not including
null,
rubout,
carriage return, line feed, or form feed) are assigned
the 7-bit ASCII values of each of the characters to be used.
Each
7-bit value is stored in an S-bit byte and the bytes are combined to
form a word.
For example "AB will store the ASCII value of A in the
low-order (even) byte and the value of B in the high-order (odd) byte:
l-S

PAL-llS ASSEMBLY LANGUAGE AND ASSEMBLER

B's value

high-order byte
102

-- --

low-order byte
101

--------..---.-~

100

001

"AB=041101

= A's value

--.. --..

-- -000

001

ASCII text is always absolute.

1.4.4

Mode of Expressions

The mode of an expression may be absolute, relocatable or external
defined below:

A term of an expression is absolute, relocatable or external depending
on whether its definer
(i.e., number, symbol, etc.) is absolute,
relocatable or external. Numbers, permanent symbols and generated
data are always treated as absolute.
An absolute expression is defined as:
1.

Absolute term (one whose value is defined at assembly
preceded optionally by a single plus or minus sign, or

Relocatable expression minus a relocatable term, or

Absolute expression followed by an operator
absolute expression.

followed

time)

A relocatable expression is defined as:
.1.

Relocatable term (one whose value is
time), or·

not

Relocatable expression followed by an
followed by an absolute expression, or

3.-

Absolute_ express.ion.followed by
relocatable express.ion.

a~.plus

known

until

arithmetic

link

operator

oper-ator f_ollowed by a

An external expression is defined as:
1.

External term
program), or

(one

whose

value

defined

External expression followed by
followed by an absolute term, or

arithmetic

Absolute expression followed by a plus operator
an external expression.

In the following examples:
ABS is an absolute symbol,
REL is a relocatable symbol,·
EXT is an external symbol.

1-9

outside

the

operator

followed

PAL-lIS ASSEMBLY LANGUAGE AND ASSEMBLER
Examples:
The following are valid expressions:
EXT + ABS

;External expression

REL+REL-REL

;Relocatable expression

ABS+REL-REL & ABS

;Absolute expression

The following are illegal expressions:
EXT+REL
REL+REL
ABS-EXT

1.5

ASSEMB'LY LOCATION COUNTER

The period (.) is the symbol for the assembly location counter.
(Note
difference of Program Counter.
~PC.
See Section 1.7.) When used
in the operand field of an instruction, it represents the address of
the first word of the instruction. When used .inthe .operand field of
an ,assembler directive, it represents'the a'ddress of the current byte
or word. For example,
A:

MOV#: .,RO

;.refers to location A,
;i.e., the address of the
;MOV instruction

(#: is explained in Section 1.7.9.)
At the beginning of each assembly pass, the Assembler clears the
location counter. Normally, consecutive memory locations are assigned
to each byte of object data generated. However, the location where
the object data is stored may be changed by a direct assignment
altering the location counter:
.=expression
Similar to other symbols, the location counter symbol "." has a mode
associated with it.
However, the mode cannot be external. Neither
can one change the existing mode of the location counter by using a
defining expression of a different mode.
The mode of the location counter symbol can be changed by the
the .ASECT or .CSECT directive as explained in Section 1.8.3.

use

The expression defining the location counter"must not contain
references or symbols that vary from one pass to another.

forward

Examples:
.ASECT
.=500
;SET LOCATION COUNTER TO ABSOLUTE 500
MOV .+10,COUNT ;THE LABEL FIRST HAS THE VALUE 500
;(OCTAL) .+10 EQUALS 510 (OCTAL), THE
;CONTENTS OF LOCATION 510 (OCTAL) WILL
_BE DEPOSITED IN LOCATION COUNT.
+ ::::~520

;THE ASSEMBLY LOCATION COUNTER NOW
;HAS A VALUE OF ABSOLUTE 520 (OCTAL),

1-10

PAL-lIS ASSEMBLY LANGUAGE AND ASSEMBLER

SECOND:

MOV

.,INDEX

;THE LABEL SECOND HAS THE
;(OCTAL), THE CONTENTS OF
;THAT IS, THE BINARY CODE
;INSTRUCTION ITSELF, WILL
tIN LOCATION INDEX.

VALUE 520
LOCATION 520
FOR
BE DEPOSITED

<.cnECT
;nET LOCATION COUNTER TO RELOCATABLE

.::::.+20

;20.
THIF~D

;THE LABEL THIRD HAS THE VALUE OF
; F~ELOCf~TABLE 20.

• WDFm ()

Storage area may be reserved by advancing the location counter.
For
example,
if the current value of the location counter is 1000, the
direct assignment statement
.=.+100
will reserve 100 bytes of storage space in
instruction will be stored at 1100.

1.6

the

program.

The

RELOCATION AND LINKING

The output of the relocatable assembler is an object module which must
be processed by the PDP-II Linker, LINK-lIS, before loading and
execution.
The Linker essentially fixes (i.e., makes absolute)
the
values of external or relocatable symbols and creates another module
(load module)
which contains the binary data to be loaded anq
executed.
To enable the Linker to fix the value of an expression the assembler
issues certain directives to the Linker together with the required
parameters.
In the case of relocatable expressions the Linker adds
the base of the relocatable section
(the location in memory of
relocatable 0) to the value of the relocatable expression provided by
the Assembler.
In the case of an external expression the value of the
external term in the expression is determined by the Linker (since the
external symbol must be defined in one of the other object modules
being linked and adds it to the value of the external expression
provided by the Assembler.
All instructions that are to be modified as described above will be
marked by a single apostrophe in the assembly listing. Thus the
binary text output will look as follows for the given examples:
0()~:506~3.'

CLF~

EXTEF~NAL (~:;)

;VALUE OF EXTERNAL nYMBOL
;AnSUMED ZERO~ WILL BE
;MODIFIED BY THE LINKER.

000000
()O~506~:).'

CI..J~

EXTEF~NAI...Ni (~3)

eLI:;:

I:;:EI... DCATf~BLE (5)

OO()OOt.l
O()~::j06~3'

OO()04()

1-11

;ASSUMING WE ARE IN THE
;ABSOLUTE SECTION AND
ITHE VALUE OF RELOCATABLE
tIS RELOCATABLE 40

PAL-lIS ASSEMBLY LANGUAGE AND ASSEMBLER
1.7

ADDRESSING

The Program Counter (register 7 of the eight general registers) always
contains the address of the next word to be fetched; i.e., the
address of the next instruction to be executed, or the second or third
word of the current instruction.
In order to understand how the address modes operate and how they
assemble, the action of the Program Counter must be understood. The
key rule is:
Whenever the processor implicitly uses the Program
Counter to fetch a word from memory, the Program
Counter is automatically incremented by two after
the fetch.
That is, when an instruc~ion is fetched, the PC is incremented by two,
so that it is pointing to the next word in memory; and, if an
instruction uses indexing (see sections 1.7.7, 1.7.8 and 1.7.10), the
processor uses the Program Counter to fetch the base from memory.
Hence, using the rule above, the PC increments by two, and now points
to the next word.
The following conventions are used in this section:
1.

Let E be any expression as defined in Section 1.4.

Let R be a register expression.
This is any expression
containing a term preceded by a % character of a symbol
previously equated to such a term.

Examples:
I~O

._. %0

F~ :L

.... Rt
(ROt3>t

CLR
CLR
CLR

;CLEAR WORDS AT ADDRESSES
;CONTAINED IN REGISTERS 0,3, AND 2
'AND INCREMENT REGISTER CONTENTS
;BY TWO.

(2)+

NOTE
Both JMP and JSR instructions using mode
2
(non-deferted autoincrement mode),
execute differently on different PDP-II
processors.
Avoid
use
of
these
instructions with mode 2 addressing.
Double operand instructions
of
the
addressing f.o.rm %R,
(R) + or %R, -(R)
wbere'
the·
source
and
destination
registers are the same, give different
results on different PDP-II processors,
and should be avoided.

1.7.4

Deferred .Autoincrement Mode,

The register 'containsthe' pointer to the address of .the operand.
contents of the register are incremented after being used.
Format:

@(ER)+

Example:

CLR

@(3)t ;CONTENTS OF REGISTER 3 POINT
;TO ADDRESS OF WORD TO BE CLEARED
;BEFORE BEING INCREMENTED BY TWO

1-13

" The

PAL-lIS ASSEMBLY LANGUAGE AND ASSEMBLER
1.7.5

Autodecrement Mode

The contents of the register are :decremented before being used as
address of the operand (see note in Section 1.7.3).
Format:

the

-(ER)

Examples:

CLR -(RO) 9DECREMENT CONTENTS OF REGISTERS
CLR -(RO+3) ;0, 3 AND 2 BEFORE USING
CLR -(2)
9AS ADDRESSES OF WORDS TO BE CLEARED

1.7.6

Deferred Autodecrement Mode

The contents of the register are decremented before being used as
pointer to the address of the operand.
Format:

the

@-(ER)

Example:

;DECREMENT CONTENTS OF REG. 2
9BEFORE USING AS POINTER TO ADDRESS
;OF WORD TO BE CLEARED.

1.7.7

Index Mode
Format:

.E(ER)

The value of an expression Eis stored as the second or third word of
the instruction. The effective address is calculated as the value of
E plus the contents of register ER. The value E is called the base.
Examples:

1.7.8

CLRX+2(Rl)

9EFFEGTIVE ADDRESS IS X+2 PLUS
;THE CONTENTS OF REGISTER 1

CLR -2(3)

9EFFECTIVE ADDRESS IS -2 PLUS
9THE CONTENTS OF REGISTER 3

Deferred Index Mode

An expression plus the contents of a register gives the pointer to the
address of the operand.
F6rmat:

@E(ER)

Example:

CLR @14(4) ;IF REGISTER 4 HOLDS 100, AND LOCATION
;114 HOLDS 2000, LOC.2000 IS CLEARED.

1-14

PAL-lIS ASSEMBLY LANGUAGE AND ASSEMBLER
1.7.9. Immediate Mode .and Deferred Immediate (Absolute) Mode
The immediate mode allows the oper~nd itself to be stored as the
It is assembled as an
second or third word of the instruct.ion.
autoincrement of register 7, the pc.
Format:·

,'4/:E

Examples: .

MOV :1J:100,RO
MOV :ft:X,RO

;MOV~

AN OCTAL 100 TO REGISTER 0
THE VALUE OF SYMBOL X TO
;REGISTER O.
.
~MOVE

The operation of this mode is explained as follows:
The statement MOV 4/:100,R3 assembles as two words.

These are:

012 703
00 0 1 0 0
Just b~fore this instruction is fetched and executed, the pc points to
the fir~t word of the instruction. The processor fetches ~he first
word and increments the pc by two. The source operand mode is 27
(autoincrement the PC). Thus the PC is used as a pointer to fetch the
operand (the second word of the instruction) before being incremented
by two, to point to the next instruction.
If the #E is preceded by @, E specifies an absolute address.

1.7.10

Relative and Deferred Relative Modes

Relative mode is the normal mode for memory references.
Format:

Examples:

CU';: 100
MOV X,y

~CLEAR LOCATION 100
;MOVE CONTENTS hF LOCATION
HJJCATION Y.

X TG

This mode is assembled as Index mode, using 7, the PC, as
register. The base of the address calculation, which is stored in
second or third word of the instruction, is not the address of
operand.
Rather, it is the number which, when added to the
becomes the address of tbe operand. Thus, the· base is X-PC.
operation is explained as follows:
If the statement MOV lOO,R3 is assembled at absolute location 20
the assembled code is:
Location 20:
Location 22

the
the
the
PC,
The
then

01670 3
o 0 0 054

The processor fetches the MOV instruction and adds two to the PC so
that it points to location 22. The source operand mode is 67~ that
is; indexed by the PC . . To.pick up the base, the processor fetches the
word. pointed to by the PC~nd adds two to the PC. The PC now points
to location 24. To calculate the address of the source operand, the
base is added to the designated register. That is, BASE+PC=54+24=100,
the operand address.
1-15

PAL~llS

ASSEMBLY LANGUAGE AND ASSEMBLER

Since the Assembler conside~s "." as the address of the first word
the instruction, an equivalent statement would be

MOV 100 -.- 4(PC),R3
This mode is called relative because the operand address is calculated
relative to the current pc.
The base is the distance (in bytes)
between the operand and the current pc.
If the operator and its
operand are moved in memory so that the distance b~tween the operator
and data remains constant, the instruction will operate correctly.
If E is preceded by @ the expression's value is
address of the operand.

1.7.11

Table of Mode
Forms
Section 1.7.12)

and

Codes

the

(6-bi~(A)

pointer

format only

the

see

Each instruction takes at least one word. Operands of the first six
forms listed below, do not increase the length of an instruction.
Each operand in one of the other modes, however, increases the
instruction length by one word.

None
of
these
Iorms
increases
the
instruction
leng'th.

Form

Mode

Meaning

@-(ER)

On
In
2n
3n
4n
5n

Form

Mode

Meaning

E(ER)
@E(ER)
:JI:E
@#E

Index
Index deferred
Immediate
Absolute memory reference
Relative
Relative deferred reference

@R or· (ER)
(ER) +
@(ER)+
';"fER)

Any of these
forms adds a
word to the
instruction
length.

~AutoincIement

. Autoi-ncrement deferred
Autodecrement
Autodecrement deferred

7n
27
37
67
77

Notes:
1.

An alternate form for @R is (ER).
is equivalent to @O(ER).

However, the

form

@O(ER)

The form @:JI:E differs from the form E in that the second or
third word of the instruction contains the absolute address
of the operand rather than the relative distance between the
operand and the PC. Thus, the statement CLR @:JI:100 will clear
location 100 even if the instruction is moved from the point
at which it was assembled.

The Assembler is not particular about left and right and dangling +
and
signs in address fields. The following are some examples of
incorrect syntax that assemble as indicated, without an
error
indication.

1-16

PAL-lIS ASSEMBLY LANGUAGE AND ASSEMBLER
Form

Assembles As:

Form

Assembles As:

(R2)A
A- (R2)
A(Rw)+
+(R2)

A.(R2)
A(R2) or A-O(R2)
A(R2)
(R2)+

(R2)@ (R2)A
A(R2)+B

- (R2)
@ A(R2)
A+B (R2)

1.7.12

Instruction Forms

The instruction mnemonics are given in Appendix B.
This section
defines the number and nature of the operand fields for these
instructions.
In the table that follows, let R, E, and ER represent
defined in Sections 1.4 and 1.7 and let A be
specification of the forms:
E

@R or

(ER) +

@(ER)+

- (ER)
E (ER)
#E

(R)

expressions as
a 6-bit address

@ -(ER)
@ E(ER)
@ :fI:E

Table 1-1
Instruction Operand Fields
Instr.uction

Example

Form

MOV (R6)+, ;@Y
CLR-(R2)
HALT
BR X+2
BLO .-4

Double. :6,p.arand
Singl'eOp-erand
OPERATE
Branch

Qp ·AiA
Qp .A
OP
Op E

Subroutine Call
Subroutine Return
EMT/TRAP

where -128«E-.-2)/2<127
JSR ER,A
RTS ER
Op or Op'E
where O377 S ) it is truncated to eight bits ~nd a
Truncation (T) error occurs.
Do not try to micro-progr~m the condition cod~ operators (see Appendix
B, B.4).
This makes sense in the PDP-II hardwarep however, the
current PAL-lIS Assembler does not support this capability. Thus:
CLC!CLV
results in a Q error (see
assembled as CLC.

Appendix

B.5)

and

the

statement

Expressions in the Assembler 'do, however, allow logical operators and
the use of instruction mnemonics. Thus, the proper ways to write the
above statement:
; 0 F' f.H' af"Ji.:I (:If • warm
; Ol"'erand of d(f!fal..llt
; DF·e r'i~nd fJf defi31..11t

.WORD CLC!
+CLC!CLV
!CLC!CLV

1.S

.WORD
.WORD

ASSEMBLER DIRECTIVES

Assembler directives (sometimes called pseudo-ops) direct the assembly
process and may generate data.
Assembler directives may be preceded bya label and followed by a
comment.
The assembler directive occupies the operator field. Only
one directive may be placed in anyone statement. . One or more
operands may occupy the operand field or it may be void -- allowable
~perands vary from directive to directive.
.

1.S.1

.TITLE

The .TITLE directive is used to name the object module. The name is
assigned by the first symbol following the directive • If there is no
. TITLEstatement the default name 'assigned is ~MAIN. ".
II

1. S • 2

. GLOBL

The .GLOBL directive is used to declare a symbol as bei~g global.
It
may be an entry symbol, in which case i t i s defined in the program, or
it may be a external symbol, in which case it should be defined in
another program which will be linked with this program by the linker.
The form of the .GLOBL directive is
.GLOBL

NAMA, NAMB, .•. ,NAMN

I-IS

PAL-llS ASSEMBLY LANGUAGE AND ASSEMBLER
NOTE
A symbol cannot be declared global by
defining' it as a global,expression in a
direct assignment statement.
If an illegal character is detected in the operand field of a .GLOBL
statement, an error message is not generated; and the Assembler may
ignore the remainder of the statement. Thus:.
GLOBL A,B,@C,D
as~embles

without error as:
.GLOBL A,B

1.8.3

Program Section Directives (.ASECT

and~CSECT)

The relocatable assembler provides for two program sections, an
absolute section declared by an .ASECT directive and a relocatable
section declared by a .CSECT directive.
These directives therefore
enable the programmer to specify that parts of his program be
assembled in the absolute section and others in a relocatable section.
The scope of each directive extends until a directive to the contrary
is given. The Assembler initially starts in the relocatable section:
Thus, if the first statement of a program were
A:

.ASECT

the label "A" would be a relocatable symbol which is assigned the
value of relocatable zero. The absolute value of A will be calculated
by the Linker by adding the value of the base of the relocatable
section.
Example:

'. ASECT
.::::1000
CL.R X
.CBECT
JMP A
.END

;ASSEMBLER IN ABSOLUTE SECTION
9PC = 1000 ABSOLUTE
;A = 1000 ABSOLUTE
9ASSEMBLE IN RELOCATABLE SECTIO~
;x=o RELOCATABLE

The absolute and/or relocatable section may be discontinued
(by
~witching to the alternate section) and,then continued where they left
off by using another .ASECT or .CSECT statement.
Example:

.CBECT

.wmm 0,;J,,2

.ASECT
.WORD 0,l.,2
.CSECT
.WORD 0
.END

;ASSEMBLE[I AT RELOCATABLE 0, 2
.; ASSEMBLED AT ABSOLUTE 0, 2

and 4

9ASSEMBLED AT RELOCATABLE 6.

If a label is defined twice, first in an absolute section and then in
a relocatable section, the symbol will be relocatable but its value
will be as defined in the absolute section.

},-19

PAL-lIS ASSEMBLY LANGUAGE AND ASSEMBLER
1. 8.4

.EOT

The .EOT directive indicates the physical End Of Tape though not. the
logical end of the program. ,If the .EOT is followed by a single line
feed or form feed, the Assembler will still read to the end of the
tape, but will not process anything past the .EOT directive.
If .EOT
is followed by at least two line feeds or form feeds,
the Assembler
will stop before the end of the tape.
Either case is proper, but it
should be understood that even though it appears as if the Assembler
has read too far, it actually hasn't.
I f a .EOT is embedded in a tape, and more information to be assembled
follows it,
.EOT must be immediately followed by at least two line
feeds or form feeds.
Otherwise, the first line following the
.EOT
will be lost.

Any operands following a .EOT directive will be ignored.
The
.EOT
directive allows several physically separate tapes to be assembled as
one program. The last tape should be terminated by a
.END directive
(see Section 1.8.6) but may be terminated with
• EDT
(spe .END
emulation in Section 1.9.4).

1. 8. 5

. EVEN

The .EVEN directive ensures that the assembly location counter is even
by adding one if it is odd. Any operands following a .EVEN directive
will be ignored.

1. 8.6 '. END

The .END directive indicates the logical and physical end of the
source program.
The
.END directive may be followed by only one
operand, an expression indicating the prog,r,aIIl's tr,ansfer addresa.
At loadct±me, the load module wEI be loaded and program execution
will begin at the transfer address indicated by the .END directive.
If the address is not specified, the loader will halt after reading in
the load moduler

1. 8.7

• WORD

- The .WORD assembler directive may have one or more operands', separated
by commas.
Each operand is stored ina word of the object program.
If there is more than one operand, they are stored in successive
words.
The operands may be any legally formed expression. For
example,

.:::::1.420
SAL~-::O

.WORD 177535,.+4,SAL ;STORED IN WORDS 1420, 1422 AND
;1424 WILL BE 177535, 1426, AND 0
Values exceeding 16 bits
length.

wil~

be truncated

1-20

from

the

left,

word

PAL-lIS ASSEMBLY LANGUAGE AND ASSEMBLER
A .WORD directive followed by one or more void operands separated
commas will store zeros for the void operands. For example,

.=1430
.WORD ,5,

;ZERO, FIVE, AND ZERO ARE STORED
;IN WORDS 1430, 143~, AND 1434

An. operator field left blank will be interpreted as the
. WORD
directive if the operand field contains one or more expressions. The
first term of the first expression in the operand field must not be an
instruction mnemonic or assembler directive unless preceded by a +, or one of the logical operators, ! or &. For example,

.::::440
LABEL: +MOV,LABEL

;THE OP-CODE FOR MOV, WHICH IS 010000,
;IS STORED IN LOCATION 440. 440 IS
;STORED IN LOCATION 442.

Note that the default .WORD will occur whenever there is a leading
arithmetic or logical operator, or whenever a leading symbol is
encountered which is not recognized as an instruction mnemonic or
assembler
directive.
Therefore, if an instruction mnemonic or
assembler directive is misspelled, the .WORD directive is assumed and
errors will result. Assume that MOV is spelled incorrectly as MaR:
MaR A,B
Two error codes can result: A Q will occur because an expression
operator is missing between MaR and A, and a U will occur if MaR is
undefined. Two words will be generated; one for MaR A and one for B.

1. 8.8

.BYTE

The .BYTE assembler directive may have one or more operands separated
by commas.
Each operand is stored in a byte of the object program.
If multiple operands are specified, they are stored in successive
bytes.
The operands may be any legally formed expression with a
result of 8 bits or less.
For example,

SAM:=:5
• :::410
• BYTE 48.,SAM

;STORED IN LOCATION 410 WILL BE
;060 (THE OCTAL EQUIVALENT OF 48) •
;IN 411 WILL BE'005 •

If the expression has a result of more than 8 bits, it will be
truncated to its low-order 8 bits and will be flagged as a Terror.
If an operand after the .BYTE directive is left void, it will be
interpreted as zero. For example,

• """420
• BYTE , ,

;ZERO WILL BE STORED IN
;BYTES 420, 421 AND 422 •

If the expression is relocatable, a warning flag, A, will be given.

1.8.9

.ASCII

The .ASCII directive translates strings of ASCII characters into their
7-bit ASCII codes with the exception of null, rubout, carriage return,
line feed and form feed.
The text to be translated is delimited by a
character at the beginning and the end of the text. The delimiting
character may be any printing ASCII character except colon and equal
sign and those used in the text string.
The 7-bit ASCII code
generated for each character will be stored in successive bytes of the
object program. For example,
1-21

PAL-lIS ASSEMBLY LANGUAGE AND ASSEMBLER

.=500

.ASCII IYESI
.ASCII 15+3/21

;THE ASCII CODE FOR Y WILL BE
;STORED IN 500~ THE CODE FOR E
;IN 501, THE CODE FOR S IN 502.
9THE DELIMITING CHARACTER OCCURS
9AMONG THE OPERANDS. THE ASCII
;CODES FOR 5 , + , AND 3 ARE
;STORED IN BYTES 503, 504, AND
;505. 21 IS NOT ASSEMBLED.

The .ASCII directive may be terminated by any legal terminator except
for
and
.• Note that if the text delimiter is also a terminator,
the leading text delimiter can also serve as the
.ASCII directive
terminator.
For example,

.ASCII IABCDI
• ASCII+ABCD+

1.8.10

9THE SPACE IS REQUIRED
;BECAUSE I IS NOT A TERMINATOR.
9NO SPACE IS REQUIRED •

.RAD50

PDP-II system programs often handle symbols in a specially coded form
caled "RADIX 50"
(this form is sometimes referred to as "MOD40").
This form allows 3 characters to be packed into 16 bits;
therefore,
any symbol can be held in two words, the form of the directive is:
.RAD50 /CCC/
The single operand is of the form /CCC/ where the slash
(the
delimiter)
can be any printable character except for = and:. The
delimiters enclose the characters to be converted which may be A
through Z,
0 through 9, dollar ($), dot (.) and space ( ) .
If there
are fewer than 3 characters they are considered to be left-justified
and trailing spaces are assumed.
Any characters following the
t~ailing delimiter are ignored and no.error results.
Examples:

.RAD50 IABCI
.RAD50 IABI
.1:;:AD~50

I I

;PACK ABC INTO ONE WORD
;PACK AB (SPACE) INTO ONE WORD;
;PACK 3 SPACES INTO ONE WORD

The packing algorithm is as follows:
A.
Each character is translated into
indicated in the following table:
Character
(SPACE)
A-Z
$
0-9

its

RADIX

equivalent

RADIX 50 Equivalent (octal)

o
1-32
33
34
36-47

Note that another character can be defined for code 35.

B. The RADIX 50 equivalents for characters 1 through 3 (Cl,C2,C3) are
combined as follows:
RESULT=((Cl*50)+C2)*50+C3

1-22

PAL-lIS ASSEMBLY LANGUAGE AND ASSEMBLER
1.8.11

.LIMIT

A program often wishes to know the boundaries of the relocatable code.
The .LIMIT directive generates two words into which the linker puts
the low and high addresses of the relocated code.
The low address
(inserted into the first word) is the address of the first byte of
code. The high address is the address of the first free byte
following the relocated code.
These addresses will always be even
since all relocatable sections are loaded at even addresses and if a
relocatable section consists of an odd number of bytes the linker adds
one to the size to make it even.

1.8.12

Conditional Assembly Directives

Conditional assembly directives provide the programmer with the
capability to conditionally include or not include portions of his
source code in the assembly process. In what follows, E denotes an
expression and S(i) denotes a symbol. The conditional directives are:
.IFZ
.IFNZ
.IFL
.IFLE
.IFG
.IFGE
.IFDF
.IFNDF

E
E
E
E
E
E
S (1 )
S (1)

iIF
iIF
iIF
iIF
i IF
;IF
[ I , &] S
[ I , &] S

E=O
E*O
EO
E~O

(2)
(2)

[1,&] ... [1,&] S(N)
[1,&] ..• [1,&] S(N)

(I =OR, &=AND)

If the condition is met, all statements up to the matching .ENDC are
assembled.
Otherwise, the statements are ignored until the matching
.ENDC is detected.
In the above,.IFDF and .IFNDF mean "if defined" and "if undefined"
respectively. The scan is left to right, no parentheses permitted.
E"xample:

.IFDF S!T&U

Means assemble if either S or T
defined and U is defined

.IFNDF T&LJ!S

Means assemble if both T and U
undefined or if S is undefined

is
are

General Remarks:
An errored or null expression takes the default value 0 for purposes
of the conditional test. An error in syntax, e.g., a terminator other
than ;, I, &, or CR results in the undefined situation for
.IFDF and
.IFNDF, as does an errored or null symbol.
All conditionals must end with the .ENDC diIective. Anything in the
operand field of .ENDC is ignored. Nesting is permitted up to a depth
of 127
Labels are permitted on conditional directives, but the
scan is purely l"eft to right. For example:
A:

.IFZ 1
. ENDC

A is ignored.
A:

.IFZ 1
.ENDC

A is entered in the symbol table.

1-23

PAL-lIS ASSEMBLY LANGUAGE AND ASSEMBLER
If a .END is encountered while inside a satisfied conditional, a Q
flag will appear, but the .END directive will still be processed
normally.
If more .ENDC's appear than are required, Q flags appear on
the extras.

1.9

1.9.1

OPERATING PROCEDURES

Introduction

The Assembler enables you to assemble an ASCII tape containing PAL-II
statements into a relocatable binary tape (object module). To do
this, two or three passes are necessary.
On the first pass, the
Assembler creates a table of user-defined symbols and their associated
values, and a list of' undefined symbols is printed on the teleprinter.
On the second pass the Assembler assembles the program and punches out
an absolute binary tape and/or outputs an assembly listing.
During
the third pass
(this pass is optional), the Assembler punches an
absolute binary tape or outputs an assembly listing. The symbol table
(and/or a list of errors) may be output on any of these passes. The
input and output devices as well as various options are specified
during the initial dialogue
(see Section 1.9.3).
The Assembler
initiates the dialogue immediately after being loaded and after the
last pass of an assembly.

1.9.2

Loading PAL-lIS

PAL-lIS is loaded by the Paper Tape Software Absolute Loader.
Note
that on systems with hardware switch registers, the start address of
the Absolute Loader must be in the Switch Register when loading the
Assembler.
This is because the Assembler tape has an initial program
which clears all of core up to the address specified in the Switch
Register, and jumps to that address to start loading the Assembler.

1.9.3

Initial Dialogue

After being loaded, the Assembler prints its name and version and then
initiates dialogue by printing on the teleprinter

*S
meaning "What is the Source symbolic input device?" The
be

.J
H
L
T

response

may

use Low-speed reader (.Jdenotes typing the RETURN key)
meaning High-speed reader
meaning Low-speed reader
meaning Teleprinter keyboard

The device specification is terminated, as is all
typing the RETURN key.

user

response,

1-24

PAL-lIS ASSEMBLY LANGUAGE AND ASSEMBLER
After the *S question and response, the Assembler prints:
*B
meaning "What is the Binary output device?" The responses
similar to those for *S:
H
L
./

are

meaning High-speed punch
meaning Low-speed punch
meaning do not output binary tape (./denotes typing
the RETURN key)

In addition to I/O device specification, various options may be
chosen.
The binary output will occur on the second pasS unless /3
(indicating the third pass) is typed following the H or L.
Errors
will be listed on the same pass if /E is typed.
If /E is typed in
response to more than one inquiry, only the last occurrence will be
honored.
It is strongly suggested that the errors be listed on the
same pass as the binary output, since errors may vary from pass to
pass.
If both /3 and /E are typed, /3 must
terminated by typing the RETURN key.

precede /E.
Examples:

The

response

L/E

Binary output on the low-speed punch and
the errors on the teleprinter, both
during the second pass.

H/3/E

Binary output on the high-speed punch
and the errors on the teleprinter during
the third pass.

The RETURN key alone will cause
Assembler to omit binary output

the

After the *B question and response, the Assembler prints:
*L
meaning "What is the assembly Listing output device?" The response
*L may be:
L

H
T
P

meaning
meaning
meaning
meaning
meaning

Low-speed punch
High-speed punch
Teleprinter
Line Printer
do not output listing (.Jdenotes typing RETURN)

After the I/O device £pecification, pass and error list options
similar to those for *B may be chosen. The assembly listing will be
output on the third pass unless /2 (indicating the second pass)
is
typed following H, L, T, or P.
Errors will be listed on the
teleprinter during the same pass if /E is typed.
If both /2 and /E
are typed, /2 must precede /E. The response is terminated by typing
the RETURN key. Examples:
*L

L/2/E

Listing on low-speed punch and errors on
teleprinter during second pass.

*.L

Listing on high-speed punch
pass

~uring

The RETURN key alone will cause
Assembler to omit listing output.
1-25

third
the

PAL-lIS ASSEMBLY LANGUAGE AND ASSEMBLER
After the *L ,question and response, the final question is
the teleprinter:

printed

*T
meaning "What is the symbol Table output device?" The
device
specification is the same as for *L question. The symbol table will
be output at the end of the first pass unless /2 or /3 is typed in
response to *T.
The first tape to be assembled should be placed in
the reader before typing the RETURN key because assembly will begin
upon typing RETURN to the *T question.
The /E option is not a
meaningful response to *T. Example
*T

Symbol table output
end of third pass.

T/3

teleprinter

Typing the RETURN key alone will
the
Assembler to omit symbol
output.

cause
table

The symbol table is printed alphabetically, three symbols per line.
Each symbol printed is followed by its identifying characters and by
its value.
If the symbol is undefined, six asterisks replace its
value.
The identifying characters indicate the class of the symbol;
that is, whether it is a label, direct assignment,
register symbol,
etc. The following examples show the various forms.
ABCDEF
R3
DIRASM

XYZ
R6
LABEL

001244
%000003
177777
******
%******
******

(Defined Label)
(Register Symbol)
(Direct Assignment)
(Undefined direct assignment)
(Undefined register symbol)
(Undefined label)

Generally, undefined symbols and external symbols will be listed as
undefined
direct assignments.
Multiply-defined symbols are not
flagged in the symbol table printout but are flagged wherever they are
used in the program.
If the symbol is relocatable or global or
will be followed by an R, a G or both.

both,

the

symbol's

value

It is possible to output both the binary tape and the assembly listing
on the same pass, thereby reducing the assembly process to two passes
(see Example 1 below). This will happen automatically unless the
binary device and the listing device are conflicting devices or the
same device (see Example 2 below). The only conflicting devices are
the teleprinter and the low-speed punch. Even though the Assembler
deduces that three passes are necessary, the binary and listing can be
forced on pass 2 by including /2 in the responses to *B and *L (see
Example 3 below).
Example 1.
*S
*B
*L
*T

Runs 2 passes:
H
H
P
T

High-speed reader
High-speed punch
Line Printet
Teleprinter

Example 2. Runs 3 passes:
H
H
H

High-speed reader
High-speed punch
High-speed punch
Teleprinter
1-26

PAL-llS ASSEMBLY LANGUAGE AND ASSEMBLER
Example 3. Runs 2 passes:
H

H/2
H/2
T

High-speed reader
High-speed punch on pass 2
High-speed punch on pass 2
Teleprinter

Note that there are several cases
intermixed with ASCII output:

where

the

binary

output

can

listing

*B
*L

H/2
H/2

Binary and listing to punch on pass 2.

L/E

Binary to low-speed punch and error
teleprinter (and low-speed punch).

*B
*L

L/2/E
T/2

Binary, error listing, and
listing to low speed punch.

The object module so generated is acceptable to the Linker as long as
there are no CTRL/A characters in the source program. The start of
~very block on the binary tape is indicated by a 001
and the Linker
19nores all information until a 001 is detected. Thus, all source
and/or error messages will be ignored if they do not contain any
CTRL/A characters (octal 001).
If a character other than those mentioned is typed in reponse to a
question,
the Assembler will ignore it and print the question again.
Example:
*S
*B
*B

High-speed reader
Q is not a valid response
The question is repeated

If at any time you wish to restart the Assembler, type CTRL/P.
If the
low-speed reader is the source input device, turn it off before typing
CTRL/P.
When no passes are omitted or error options specified,
performs as follows:

the

Assembler

PASS 1:
Assembler creates a table of user-defined symbols and their associated
values to be used in assembling the source to object program.
Undefined symbols (not including external globals) are listed on the
teleprinter at the end of the pass. The symbol table is also listed
at this time.
If an illegal location statement of the
form
.=expression is encountered, the line and error code will be printed
out on the teleprinter before the assembly proceeds. An error in a
location statement is usually a fatal error in the program and should
be corrected.
PASS 2:
Assembler punches the object module, and prints the pass
and undefined location statements on the teleprinter.

error

count

PASS 3:
Assembler prints or punches the assembly program listing,
undefined
location statements, and the pass error count on the teleprinter.

1-27

PAL-lIS ASSEMBLY LANGUAGE AND ASSEMBLER
The functions of passes 2. and 3 will occur simultaneously on pass 2 if
the binary and listing devices are different, and do not conflict with
each other
(the
low-speed
punch
and
teleprinter
conflict).
Furthermore, if the binary object module is not requested, the listing
will be produced on pass 2.
.
The following table summarizes the initial dialogue questions:
PRINTOUT

INQUIRY

*S
*B
*L
*T

What
What
What
What

is
is
is
is

the
the
the
the

input device of the Source symbolic tape?
output device of the Binary object tape?
output device of th~ assembly Listing?
·output deviqe of the symbol Table?

The following table summarizes the legal responses:
CHARACTER

RESPONSE INDICATED
Teleprinter keyboard
Low-speed reader or punch
High-speed reader or punch
Line Printer
Pass 1
Pass 2
Pass 3
Errors listed on same pass
(not meaningful
response to *S or *T)
Omit function (except in response .to *S).

T
L

H
P

II
12
13
IE

Typical examples of complete initial dialogues:
For minimal PDP-II configuration:
*S
*B

L
LIE

*L
*T

T
T

Source input on low-speed reader
Binary output on low-speed punch
errors during same (second) pass
Listing on teleprinter during pass 3
Symbol table on teleprinter at end of pass 1

For a PDP-II with high-speed IIO devices:

1.9.4

*S
*B

*L
*T

HIE
T/2

Source input on high-speed reader
Binary output on high-speed punch
errors during same (second) pass
No listing
Symbol table on teleprinter at end of pass 2.

Assembly Dialogue

During assembly, the Assembler wiil pause to print on the teleprinter
various messages to indicate that you must respond in some way before
the assembly process can continue. You may also type CTRLIP, at any
time, if you wish to stop the assembly process and restart the initial
dialogue, as mentioned in the previous section.
When a .EOT assembler directive is read on
prints

the

tape,

the

Assembler

and pauses. Our ing this pa"use, the next tape is placed in the reader,
and RETURN is typed to continue the assembly.
1-28

PAL-lIS ASSEMBLY LANGUAGE AND ASSEMBLER
If the specified assembly listing output device is the high-speed
punch and if it is out of tape, or if the device is the Line Printer
and is out of paper, the Assembler prints on the teleprinter

and waits for tape or paper to be placed in the device.
Type the
RETURN key when the tape or paper has been replenished; assembly will
continue.
Conditions causing the EOM?
are:
HSP

messages for an assembly listing

device

LPT

No power
No tape

No power
Printer drum gate open
Too hot
No paper

There is no EOM if the line printer is switched off-line, although
characters may be lost for this condition as well as for an EOM.
If the binary output device is the high-speed punch and if it
of tape, the Assembler prints:

The assembly process is aborted and
again.

the

When a .END assembler directive is read on
prints:

initial
the

dialogue
tape,

the

out

begun

Assembler

END ?
and
and
the
for

pauses.
During the pause the first tape is placed in the reader,
the RETURN key is typed to begin the next pass. On the last pass,
.END directive causes the Assembler to begin the initial dialogue
the next assembly.

If you are starting the binary pass and the binary is to be punched on
the low-speed punch, turn the punch on before typ.ing the RETURN key
for starting the pass. The carriage return and line feed characters
will be punched onto the binary tape, but the Linker will ignore them.
If the last tape ends with a
.EOT, the Assembler may be told to
emulate a
.END assembler directive by responding with E followed by
the RETURN key.
The Assembler will then print
END ?
and wait for another RETURN before starting the next pass.
EOF?
END ?

Example:

E...J

Note that forcing a .END in this manner causes the error counter to be
incremented by one.

1-29

PAL-lIS ASSEMBLY LANGUAGE AND ASSEMBLER
1.9.5

Assembly Listing

PAL-lIS produces a side-by-side assembly listing of symbolic source
statements, their octal equivalents, assigned addresses, and error
codes, as follows:
EELLLLLL OOOOOOASSS ...•... S
000000
000000
The E's represent the error field.
The L's represent the address.
The a's represent the object data in octal. The SiS represent the
source statement.
"A" represents a single apostrophe which indicates
that either the second, third or both words of the instruction will be
modified by the Linker. While the Assembler accepts 72
characters
per line on input, the risting is reduced by the 16 characters to the
left of the source statement.
The above represents a three-word statement.
The second and third
words of the statement are listed under the command word. No
addresses precede the second and third words since· the address order
is sequential.
The third line is o~itted for a two-word statement~
third lines are omitted for a one-word statement.

both· second

and

For a .BYTE directive, the object data field is three octal digits.
For a direct assignment statement, the value of the
defining
expression is given in the object code field although it is not
actually part of the code of the object program.
The .ASECT and .CSECT directives cause the current value of the
appropriate location counter (absolute or relocatable) to be printed.
Each page of the listing is headed by a page number (octal).

1.9.6

Object Module Output

The output of the assembler during the binary object pass is an object
module which is meaningful only to the linker. What follows gives an
overview of what the object module contains and at what stage each
part of it is produced.
The binary object module consists of three main types of data block:
a) Global symbol directory
b) Text blocks
c) Relocation Directory

(GSD)
(TXT)
(RLD)

1.9.6.L Global Symbol Directory - As the name suggests, the
GSD
contains a list of all the global symbols together with the name of
the object module. Each symbol is in Radix-50 form and contains
information regarding its mode and value whenever known.
The GSD is created at the start of the binary object pass.

1-3Q

,PAL-lIS ASSEMBLY LANGUAGE AND ASSEMBLER
1.9.6.2 Text Block - The text blocks consist entirely of the binary
object data as shown in the listing.
The operands are in the
unmodified form.

1.9.6.3 Relocation Directory - The RLD blocks consist of directives
to the Linker which may reference the text block preceding the RLD.
These directives control the relocation and linking process.
Text and RLD blocks are constructed during the binary object pass.
Outputting of each block is done whenever either the TXT or RLD buffer
is full and whenever the location counter needs to be modified.

1.10

ERROR CODES

The error codes printed beside the octal and
assembly listing have the following meanings:

symbolic

code

the

Error Code

Meaning

Addressing error. An address within the instruction
is incorrect. Also may indicate a relocation error.

Bounding error.
Instructions or word data are being
assembled at an odd address in memory.
The location
counter is updated by +1.

Doubly-defined symbol referenced.
Reference was
made to a symbol which is defined more than once.

Illegal character detected.
Illegal
characters
which are also non-printing are replaced by a? on
the listing.

Line buffer overflow. Extra characters
(more than 72 ) are ignored.

Multiple definition of a label.
A label
was
encountered which was equivalent (in the first six
characters) to a previously encountered label.

Number containing 8 or 9 has decimal point missing.

Phase error. A label's definition or
from one pass to another.

Questionable syntax. There are missing arguments or
the instruction scan was not completed or a carriage
return was not immediately followed by a line feed
or form feed.

Symbol table overflow.
When the
quantity
of
user-defined symbols exceeds the allocated space
available in the user's symbol table, the assembler
outputs the current source line with the S error
code, then returns to the initial dialogue.

1-31

value

line

varies

PAL-lIS ASSEMBLY LANGUAGE AND ASSEMBLER

Truncation error. A number generated more than 16
bits of significance or an expression gener?ted more
than 8 bits of significance during the use of the
.BYTE directive.

Undefined
symbol.
An
undefined
symbol
was
encountered during the evaluation of an expression.
Relative to the expression, the undefined symbol is
assigned a value of zero.

1.11

SOFTWARE ERROR HALTS

PAL-lIS loads all of its unused trap vectors with the code
.WORD .+2,HALT
so that if the trap does occur, the processor will halt in the second
word of the vector. The address of the halt, displayed in the console
address register, therefore indicates the cause of the halt.
Address of Halt (octal)

Meaning

12
16

Reserved instruction executed
Trace trap occurred
Power fail trap
EMT executed

A halt at address 40 indicates an IOXLPT detected error.
(displayed in the console lights) contains an identifying code:
Code in RO

0
1

2
3
4
5

Meaning
Illegal memory reference, SP overflow or
illegal instruction.
Illegal lOX command.
Slot number out of range.
Device number illegal
Referenced slot not INITed.
Illegal Data Mode •

. IOXLPT also sets Rl as follows:
If the error code is 0, Rl contains the PC at the time of the error.
If the error code is 1-5, Rl points to some element in the lOT
argument list or to the instruction following the argument list,
depending on whether IOXLPT has finished decoding all the arguments
when it detects the error.

1-32

CHAPTER 2
WRITING PAL-IIA ASSEMBLY LANGUAGE PROGRAMS'

PAL-IIA (Program Assembly Language for
the
PDP-II's
Absolute
Assembler)
enables you to write source
(symbolic) programs using
letters, numbers, and symbols which are meaningful to you. The source
programs, generated either on-line using the Text Editor (ED-II), or
off-line, are then assembled into object programs (in absolute binary)
which are executable by the computer. The object program is produced
after two passes through the Assembler;
an optional third pass
produces a complete octal/symbolic listing of the assembled program.
This listing is especially useful for documentation and debugging
purposes.
This chapter explains not only how to write PAL-IIA programs but also
how·· to assemble the source programs into computer-acceptable object
programs. All facets of the assembly language are explained and
illustratea with many examples, and the chapter concludes with
assembling procedures. In explaining how to write PAL-IIA source
programs it is necessary, especially at the outset, to make frequent
forward references. Therefore, we recommend that you first read
through the entire chapter to get a "feel" for the language, and then
reread the chapter, this time referring to appropriate sections as
indicated, for a thorough understanding of the language and assembling
procedures.
Some notable features of PAL-IIA are:
1.

Selective assembly pass functions

Device specification for pass functions

Optional error listing on Teletype

Double buffered and concurrent I/O (provided by IOX)

Alphabetized, formatted symbol table listing

The PAL-IIA Assembler is available in two versions:
an 8K version.

a 4K version

and

The assembly language applies equally to both versions.
The 4K
version provides symbol storage for about 176 user-defined symbols,
and the 8K version provides for about 1256 user-defined symbols
(see
Section 2.3).
In addition, the 8K version allows a line printer to be used
program listing and/or symbol table listing.
lpAL-llA is not currently available for PDP-II systems without
registers.
2-1

for

the

switch

WRITING PAL-IIA ASSEMBLY LANGUAGE PROGRAMS
The following discussion of the PAL-IIA Assembly Language assumes that
you have read the PDP-II Processor Handbook, with emphasis on those
sections which deal with the PDP-II instruction set, formats, and
timings
a thorough knowledge of these
is vital to efficient
assembly language programming.

2.1

CHARACTER SET

A PAL-IIA source program is composed of symbols, numbers, expressions,
symbolic instructions, assembler directives, arguments separators, and
line terminators written using the following ASCII l characters.
1.

The letters A through Z.
(Upper and lower case letters are
acceptable, although upon input, lower case letters will be
converted to upper case letters.)

The numbers 0 through 9.

The characters.

The separating or terminating symbols:
: = % # @ ( ) ,

carriage return

2.2

and $ (reserved for system software).

tab

+ _ '" !
space

line feed

form feed

STATEMENTS

A source program is composed of a sequence of statements, where each
statement is on a single line.
The statement is terminated by a
carriage return character and must be immediately followed by either a
line feed or form feed character. Should a carriage return character
be present and not be followed by a line feed or form feed, the
Assembler will generate a Q error (Section 2.10) and that portion of
the line following the carriage return will be ignored.
Since the
carriage return is a required statement terminator, a line feed or
form feed not immediately preceded by a carriage return will have one
inserted by the Assembler.
It should be noted that, if the Editor (ED-II) is being used to create
the source program (see Section 4.4.4), a typed carriage return
(RETURN key) automatically generates a line feed character.
A statement may be composed of up to four fields which are identified
by their order of appearance and by specified terminating characters
as explained below and summarized in Appendix B. The four fields are:
Label

Operator

Operand

Comment

The label and comment fields are optional. The operator and operand
fields are interdependent -- either may be omitted depending upon the
contents of the other.

ASCII stands for American Standard Code for Information Interchange.
2-2

WRITING PAL-llA ASSEMBLY LANGUAGE PROGRAMS
2.2.1

Label

A label is a user-defined symbol (see Section 3.3.2) which is assigned
the value of the current location counter. It is a symbolic means of
referring to a specific location within a program.
If present, a
label always occurs first in a statement and must be terminated by a
colon. For example, if the current location is 100(octal), the
statement
ABCD:

MOV A,B

will assign the value 100(octal) to the label ABCD so that subsequent
reference to ABCD will be to location 100(octal). More than one label
may appear within a single label field;
each label within the field
will have the same value.
For example, if the current location is
100, multiple labels in the statement
ABC:

$DD:

A7.7:

MOV A,B

will equate each of the three labels ABC, $DD, and A7.7 with the value
100(octal).
($ and
are reserved for system software.)
The error code M (multiple definition of a symbol) will be generated
during assembly if two or more labels have the same first six
characters.

2.2.2

Operator

An operator follows the label field in a statement, and may be an
instruction mnemonic or an assembler directive (see Appendix B). When
it is an instruction mnemonic, it specifies what action is to be
performed on any operand(s) which follows it. When it is an assembler
directive, it specifies a certain function or action to be performed
during assembly.
The operator may be preceded only by one or more labels and may be
followed by one or more operands and/or a comment. An operator is
legally terminated by a space, tab, or any of the
following
characters.
# +
@
line feed

form feed

carriage return

The use of each character above will be explained in this chapter.
Consider the following examples:
MClV
AvB
MmJ@A,B

i~(TAB) terminates operator MOV
i@ terminates operator MOV

When the operator stands alone without an operand or comment, it is
terminated by a carriage return followed by a line feed or form feed
character.

2.2.3

Operand

An operand is that part of a statement which is operated on by the
operator -- an instruction mnemonic or assembler directive. Operands
may be symbols, expressions, or numbers.
When multiple operands
appear within a statement, each is separated from the next by a comma.
2-3

WRITING PAL-IIA ASSEMBLY LANGUAGE PROGRAMS
An operand may be preceded by an operator and/or label,
by a comment.

and

followed

The operand field is terminated by a semicolon when followed hy a
comment, or by a carriage return followed by a line feed or form feed
character when the operand ends the statement. For example,
LABEL:

MOV GEORGE,BOB

;THIS IS A COMMENT

where the space between MOV and GEORGE terminated the operator field
and began the operand field;
the comma separated the operands GEORGE
and BOB; the semicolon terminated the operand field and began the
comment.

2.2.4

Comments

The comment field is optional and may contain any ASCII character
except null, rubout, carriage return, line feed or form feed.
All
other characters, even those with special significance are ignored by
Assembler when used in the comment field.
The comment field may be preceded by none, any, or all of the other
three fields.
It must begin with the semicolon and end with a
carraige return followed by a line feed or form feed character.
For
example,
LABEL:

CLR HERE

iTHIS IS A $1.00 COMMENT

Comments do not affect assembly processing or program execution, but
they are useful in program listings for later analysis, checkout or
documentation purposes.

2.2.5

Format Control

The format is controlled by the space and tab characters.
They have
no effect on the assembling process of the source program unless they
are embedded within a symbol, number, or ASCII text; or are used as
the operator field terminator. Thus, they can be used to provide a
neat, readable program. A statement can be written
LABEL:MOV(SP)+,TAGjPOP VALUE OFF STACK
or, using formatting characters it can be written
LABEL:

MOV (SP)+,TAG

;POP VALUE OFF STACK

which is much easier to read.
Page size is controlled by the form feed character. A page of n lines
is created by inserting a form feed (CTRL/FORM keys on the keyboard)
after the nth line.
If no form feed is present, a page is terminated
after 56 lines.

2-4

WRITING PAL-IIA ASSEMBLY LANGUAGE PROGRAMS
2.3

SYMBOLS

There are two types of symbols, permanent and user-defined. Both are
stored in the Assembler's symbol table.
Initially, the symbol table
contains the permanent symbols, but as the source program
is
assembled, user-defined symbols are added to the table.

2.3.1

Permanent Symbols

Pe'rmanent symbols consist of the instruction mnemonics
(see Appendix
B.3)
and assembler directives (see Section 2.8). These symbols are a
permanent part of the Assembler's symbol table and need not be defined
before being used in the source program.

2.3.2

User-Defined Symbols

User-defined symbols are those defined as labels (see Section 2.2.1)
or by direct assignment (see Section 2.3.3). These symbols are added
to the symbol table as they are encountered during the first pass of
the assembly.
They can be composed of alphanumeric characters, dollar
signs, and periods only; again, dollar signs and periods are reserved
for use by the system software. Any other character is illegal and,
if used, will result in the error message I (see Section 2.11).
The
following rules also apply to user-defined symbols:
1.

The first character must not be a number .

.2.

Each .symbol must be unique within the first s.ixcharacters.

A symbol may be written with more than six legal characters
but the seventh and subsequent characters are only checked
for legality, and are not otherwise recognized by the
Assembler.

Spaces and tabs must not be embedded within a symbol.

A user-defined symbol may duplicate a permanent symbol.
The value
associated with a permanent symbol that is also user-defined depends
upon its use:

2.3.3

A permanent symbol encountered in the operator
associated with its corresponding machine op-code.

field

If a permanent symbol in the operand field
is
also
user-defined,
its user-defined value is associated with the
symbol.
If the symbol is not found to be user-defined,
then
the corresponding machine op-code value is associated with
the symbol.

Direct Assignment

2-5

WRITING PAL-IIA ASSEMBLY LANGUAGE PROGRAMS
The general format for a direct assignment statement is
symbol

expression

The following conventions apply:
I.

An equal sign
(=)
must separate
expression defining the symbol.

the

symbol

A direct assignment statement may be preceded by a label
may be followed by a comment.

Only one symbol can be defined by anyone
statement.

Only one level of forward referencing, is allowed.

direct

from

the
and

assignment

Example of the two levels of forward referencing (illegal):

x
Y

Y
Z

X and Yare both undefined throughout pass 1 and will be listed on the
printer as such at the end of that pass. X is undefined throughout
pass 2, and will cause a U error message.
Examples:
A ..• :I.

;THE SYMBOL A IS EQUATED WITH THE VALUE :I.

;THE SYMBOL B IS EQUATEP WITH THE EXPRES;GION'B VAL.UE.

c~
E:

2.3.4

'A,,· :1. &MASI\LOW
D :::: :3

11DV :H::I.? ABLE

;THE GYMBOL D IB EQUATED WITH 3. THE
;LABELG C AND E ARE EQUATED WITH THE
;NUMERICAL MEMORY ADDRESS DF THE MOV
; COM1'iAND •

The eight general registers of the PDP-II are numbered 0 through 7.
These registers may be referenced by use of a register symbol, that
is, a symbolic name for a register. A register symbol is defined by
means of a direct assignment, where the defining expression contains
at least one term preceded by a % or at least one term previously
defined as a register symbol.
F~O::::%O

R3:::: I:;: 0 +3
f~4:::::I.+%:3
THEF~E::::%2

;DEFINE
;DEFINE
;DEFINE
;DEFINE

RO AS REGIGTER 0
R3 AS REGIGTER 3
R4 AS REGISTER 4
"THERE" AS REGIBTER 2

It is important to note that all register symbols must be defined
before they are referenced. A forward reference to a register symbol
will generally cause phase errors (see Section 2.10).

2-6

WRITING PAL-IIA ASSEMBLY LANGUAGE PROGRAMS
The % may be used in any expression thereby indicating a reference to
a register.
Such an expression is a register expression. Thus, the
statement
CLR %6
will clear register 6 while the statement
CLR 6
will clear the word at memory address 6.
In certain cases a register
can be referenced without the use of a register symbol or register
expression. These cases are recognized through the context of the
statement and are thoroughly explained in Sections 2.6 and 2.7.
Two
obvious examples of this are:

JSR
CLR

2.4

;THE FIRST OPERAND FIELD MUST
;ALWAYS BE A REGISTER.
X(2)

;ANY EXPRESSION ENCLOSED IN
;() MUST BE A REGISTER. IN
;THIS CASEp INDEX REGISTER 2.

EXPRESSIONS

Arithmetic and logical operators (see Section 2.4.2) may be used to
form expressions.
A term of an expression may be a permanent or
user-defined symbol, a number, ASCII data, or the present value of the
assembly location counter represented by the period. Expressions are
evaluated from left to right.
Parenthetical grouping is not allowed.
Expressions are evaluated as word quantities. The operands of a .BYTE
directive
(Section 2.8.5)
are evaluated as word expressions before
truncation to the low-order eight bits.
A missing term or expression will be interpreted as O.
A missing
operator will be interpreted as +. The error code Q (Questionable
syntax) will be generated for a missing operator. For example,
A +

-100

will be evaluated as A +

;OPERAND MISSING

a -

100, and

TAG! LA 177777 ;OPERATOR MISSING
will be evaluated as TAG! LA+177777.

2.4.1

Numbers

The Assembler accepts both octal and decimal numbers.
Octal numbers
consist of the digits a through 7 only. Decimal numbers consist of
the digits a through 9 followed by a decimal point.
If a number
contains an 8 or 9 and is not followed by a decimal point, the N error
code (see Section 2.10) will be printed and the number interpreted as
decimal.
Negative numbers may be expressed as a number preceded by a
minus sign rather than in a two·s complement form.
Positive numbers
may be preceded by a plus sign although this is not required.
If a number is too large to fit into 16 bits, the number is truncated
from the left.
In the assembly listing the statement will be flagged
with a Truncation (T) error.
2-7

WRITING PAL-IIA ASSEMBLY LANGUAGE PROGRAMS
2.4.2

Arithmetic and Logical Operators

The arithmetic operators are:

indicates addition or a positive number
indicates subtraction or a negative number

The logical operators are defined and illustrated below.
indicates the logical AND operation

indicates the logical inclusive OR operation
AND

& 0 = 0

0&1

1 = 1

1 & 0

1 & 1 = 1

2.4.3

ASCII Conversion

When preceded by an apostrophe, any ASCII character
(except null,
rUbout, carriage return, line feed, or form feed) is assigned the
7-bit ASCII value of the character (see Appendix A). For example,

'A
is assigned the value 101(octal).
When preceded by a quotation mark, two ASCII characters (not including
null, rubout, carriage return, line feed, or form feed) are assigned
the 7-bit ASCII values of each of the characters to be used.
Each
7-bit value is stored in an 8-bit byte and the bytes are combined to
form a word. For example, "AB will store the ASCII value of A in the
low-order (even) byte and the value of B in the high-order (odd) byte:
high-order byte
BI

value

low-order byte
1

o--------~-100
001
001
000
o

"AB

2.5

-1

= A's

value

001
1

041101

ASSEMBLY LOCATION COUNTER

The period (.) is the symbol for the assembly location counter.
(Note
difference of Program Counter.
. = PC. See Section 2.6.) When used
in the operand field of an instruction, it represents the address of
the first word of the instruction. When used in the operand field of
an assembler directive, it represents the address of the current byte
or word. For example,
2-8

WRITING PAL-IIA ASSEMBLY LANGUAGE PROGRAMS

MOV •• ,RO

I. REFERS TO LOCATION A, I.E.,
ITHE ADDRESS OF THE MOV INSTRUCTION

(# is explained in Section 2.6.9).
At the beginning of each assembly pass, the Assembler clears the
location counter. Normally, consecutive ~emory locations are assigned
to each byte of object data generated. However, the location where
the object data is stored may be changed by a direct assignment
altering the location counter .
. =expression
The expression defining the period must not contain forward references
or symbols that vary from one pass to another. Examples:

.=500
FIRST:

MOV

.+10,COUNT ITHE LABEL FIRST HAS THE VALUECOCTAL)
l.tl0 EQUALS 510COCTAL). THE CONTENTS
10F THE LOCATION 510COCTAL) WILL BE DE;POSITED IN LOCATION COUNT.
ITHE ASSEMBLY LOCATION COUNTER NOW
IHAS A VALUE OF 520(OCTAL).

.~520

SECOND: MOV

.,INDEX

ITHE LABEL SECOND HAS THE
ITHE CONTENTS OF LOCATION
ITHAT IS, THE BINARY CODE
IINSTRUCTION ITSELF, WILL
lIN LOCATION INDEX.

VALUE 520(OCTAL).
520(OCTAL),
FOR THE
BE DEPOSITED

Storage area may be reserved by advancing the location counter.
For
example, if the current value of the location counter is 1000, the
direct assignment statement
.=.+100
will reserve 100(octal) bytes of storage space in
next instruction will be stored at 1100.

2.6

the

program.

The

ADDRESSING

The Program Counter (register 7 of the eight general registers) always
contains the address of the next word to be fetched;
i.e., the
address of the next instruction to be executed, or the second or third
word of the current instruction.
In order to understand how the address modes operate and how they
assemble
(see Section 2.6.11), the action of the Program Counter must
be understood. The key rule is:
(PC)
to
Whenever the processor implicitly uses the Program Counter
fetch a word from memory, the Program Counter is automatically
incremented by two after the fetch.
That is, when an instruction is fetched, the PC is incremented by two,
so that it is pointing to the next word in memory; and, if an
instruction uses indexing (see Sections 2.6.7, 2.6.8, and 2.6.10), the
processor uses the Program Counter to fetch the base from memory.
Hence, using the rule above, the PC increments by two, and now points
to the next word.
2-9

WRITING PAL-llA ASSEMBLY LANGUAGE PROGRAMS
The following conventions are used in this section:
a.

Let E be any expression as defined in Section 3.4.

This is any expression
Let R be a register expression.
containing a term preceded by a
% character or a symbol
previously equated to such a term.
Examples:
RO
Rl
R2

%0
RO + 1
1 + %1

iGENERAL REGISTER 0
; GENERAL REGISTER 1
iGENERAL REGISTER 2

Let ER be a register expression or an expression in the range
0 to 7 inclusive.

Let A be a general address specification which produces
6-bit address field as described in the PDP-II Handbook.

The addressing specification, A, may now be explained in terms of E,
R, and ER as defined above.
Each will be illustrated with the single
operand instruction CLR or double operand instruction MOV.

2.6.1

The register contains the operand.
Format:

Example:
RO

CLR

2.6.2

iDEFINE RO AS REGISTER 0

iCLEAR REGISTER 0

Deferred Register Mode

The register contains the address of the operand.
Format:

@R or

(ER)

Example:

Cl Y'
CI...F~

2.6.3

(:J.)

;CLEAR THE WORD AT THE
;ADDRESS CONTAINED IN
; F~EG I

STEF~

:J..

Autoincrement Mode

The contents of the register are incremented immediately
used as the address of the operand. l

after

being

~a.
Both JMP and JSR instructions using mode 2 may increment the
register before or after its use, depending on what PDP-II processor
is being used.
This mode should be avoided.
b.
In double operand instructions of the addressing form %R, (R)+ or
%R,-(R)
where the source and destination registers are the same, the
results may be different when executed on different PDP-II processors.
The use of these forms should be avoided!

2-10

WRITING PAL-IIA ASSEMBLY LANGUAGE PROGRAMS
(ER) +

Format:
Examples:

CLR
CLR
CLR

2.6.4

(ROH
(ROt3)t
(2)+

;CLEAR WORDS AT ADDRESSES
;CONTAINED IN REGISTERS 0, 3, AND 2 AND
;INCREMENT REGISTER CONTENTS
ilBY TWO.

Deferred Autoincrement Mode

The register contains the pointer to the address of the operand.
contents of the register are incremented after being used.
Format:

The

@(ER)+

Example

CLR

2.6.5

@(3)t

;CONTENTS OF REGISTER 3 POINT
ilTO ADDRESS OF WORD TO BE CLEARED
;BEFORE BEING INCREMENTED BY TWO

Autodecrement Mode

The contents of the register are decremented before being used as
address of the operand. 1
Format:

the

- (ER)

Examples:

--(RO)
- ( ROt3 )
CLR -(2)

CLR
CLF~

2.6.6

ilDECREMENT CONTENTS OF REGilISTERS 0, 3, AND 2 BEFORE USING
ilAS ADDRESSES OF WORDS TO BE CLEARED

Deferred Autodecrement Mode

The contents of the register are decremented before being used as
pointer to the address of the operand.
Format:

the

@-(ER)

Example:
. CLI:;;

2.6.7

@·-(2)

;DECREMENT CONTENTS OF REG. 2
ilBEFORE USING AS POINTER TO ADDRESS
;OF WORD TO BE CLEARED

Index Mode

Format:

E(ER)

The value of an expression E is stored as the second or third word of
the instruction. The effective address is calculated as the value of
E plus the contents of register ER. The value E is called the base.

'See previous footnote.

2-11

WRITING. PAL-llA ASSEMBLY LANGUAGE PROGRAMS

Examples:

2.6.8

CLR

Xt2(R1)

;EFFECTIVE ADDRESS IS Xt2 PLUS
;THE CONTENTS OF REGISTER 1

CLR

-2(3)

;EFFECTIVE ADDRESS IS -2 PLUS
;THE CONTENTS OF REGISTER 3

Deferred Index Mode

An expression plus the contents of a register gives the pointer to the
address of the operand.
Format:

@E(ER)

Example:
CLR @14(4)

2.6.9

;IF REGISTER 4 HOLDS 100, AND LOCA;TION 114 HOLDS 2000, LOC. 2000 IS
;CLEARED

Immediate Mode and Deferred Immediate (Absolute) Mode

The immediate mode allows the operand itself to be stored as the
It is assembled as an
second or third word of the instruction.
autoincrement of register 7, the PC.
Format:

Examples:
MOV

$100, RO

;MOVE AN OCTAL 100 TO REGISTER 0

MOV

IX, RO

;MOVE THE VALUE OF SYMBOL X TO
;REGISTER 0

The operation of this mode is explained as follows:
The statement MOV #IOO,R3 assembles as two words.

These are:

Just before this instruction is fetched and executed, the PC points to
the first word of the instruction. The processor fetches the first
word and increments the PC by two. The source operand mode is 27
(autoincrement the PC).
Thus, the PC is used as a pointer to fetch
the operand (the second word of the instruction) before being
incremented by two, to point to the next instruction.
If the #E is preceded by @, E specifies an absolute address.

2-12

WRITING PAL-llA ASSEMBLY LANGUAGE PROGRAMS
2.6.10

Relative and Deferred Relative Modes

Relative Mode is the normal mode for memory references.
Format

Examples:

CLI:;:

:LOO

~CLEAR

MOV

X,Y

;MOVE CONTENTS OF LOCATION X TO
PLOCATION Y

LOCATION :LOO

This mode is assembled as Index Mode, using 7, the PC, as
register. The base of the address calculation, which is stored in
second or third word of the instruction, is not the address of
operand.
Rather, it is the number which, when added to the
PC.
becomes the address of the operand. Thus, the base is X
operation is explained as follows.

the
the
the
PC,
The

If the statement MOV 100,R3 is assembled
assembled code is:

the

location

20,

then

Location 20:

o 1

703

Location 22:

o 0

054

The processor fetches the MOV instruction and adds two to the PC so
that it points to location 22. The source operand mode is 67; that
is, indexed by the PC. To pick up the base, the processor fetches the
word pointed to by the PC and adds two to the PC. The PC now points
to location 24. To calculate the address of the source operand, the
base is added to the designated register. That is, Base + PC = 54 +
24 = 100, the operand address.
Since the Assembler considers . as the address of the
the instruction, an equivalent statement would be
MOV

first

word

100-.-4(PC) ,R3

This mode is called relative because the operand address is calculated
relative to the current PC.
The base is the distance (in bytes)
between the operand and the current PC.
If the operator and its
operand are moved in memory so that the distance between the operator
and data remains constant, the instruction will operate correctly.
If E is preceded by @, the expression's value is the
address of the operand.

2.6.11

Table of Mode
Section 3.7)

Forms

and

pointer

the

Codes (6-bit (A) format only - see

Each instruction takes at least one word. Operands of the first six
forms listed below do not increase the length of an instruction. Each
operand in one of the other forms however, increases the instruction
length by one word.
Form
R
None of these
@R or (ER)
( (ER) +
forms increase
the instruction
@(ER)+
length.
- (ER)
@-(ER)

Mode

Meaning

In
2n
3n
4N
5N
2-13

WRITING PAL-IIA ASSEMBLY LANGUAGE PROGRAMS

Any of these
forms adds a
word to the
instruction
length

E (ER)
@E(ER)
#E
@#E

6n
7n
27
37

E
@E

67
77

Index
Index deferred
Immediate
Absolute memory
reference
Relative
Relative deferred
reference

Notes:

2.7

An alternate form for @R is (ER).
equivalent to @O(ER).

However, the form @(ER) is

The form @#E differs from the form E in that the second or
third word of the instruction contains the absolute address
of the operand rather than the relative distance between the
operand and the pc. Thus, the statement CLR @#100 will clear
location 100 even if the instruction is moved from the point
at which it was assembled.

INSTRUCTION FORMS

The instruction mnemonics are given in Appendix B.
This section
defines the number and nature of the operand fields for these
ihstructions.
In the table that follows, let R, E, and ER represent expressions as
defined in Section 3.4, and let A be a 6-bit address specification of
the forms:
E
@E
R
@R or
(R)
(ER)+ @(ER)+
-(ER) @-(ER)
E(ER) @E(ER)
#E
@#E
Table 2-1
Instruction Operand Fields
Instruction

Form

Example

Double Operand

Op A,A

MOV

(R6)+,@Y

Single Operand

Op A

CLR

- (R2)

Operate

HALT

Branch

Op E

BR
BLO

X+2
.-4

where -128

~(E-.-2)/2~127

Subroutine Call

JSR ER,A

JSR

PC,SUBR

Subroutine Return

RTS ER

RTS

EMT/TRAP

EMT
or

Op E

EMT 31

where O!>E!>.377 (octal)

2-14

WRITING
PAL-llA ASSEMBLY LANGUAGE PROGRAMS
,
The branch instructions are one word instructions.
The high byte
contains the op code and the low byte contains an 8-bit signed offset
(7 bits plus sign) which specifies the branch address relative to the
PC. The hardware calculates the branch address as follows:
a)

Extend the sign of the offset through bits 8-15.

Multiply the result by 2.
than a byte offset.

Add the result to the PC to form the final branch address.

This creates a word offset

rather

The Assembler performs the reverse operation to form the byte offset
from the specified address. Remember that when the offset is added to
the PC, the PC is pointing to the word following the branch
instruction; hence the factor -2 in the calculation.
Byte offset
Since PC

(E-PC)/2 truncated to eight bits.

.+2, we have

Byte offset

(E-.-2)/2 truncated to eight bits.

The EMT and TRAP instructions do not use the low-order byte of the
word.
This allows information to be transferred to the trap handlers
in the low-order byte.
If EMT or TRAP is followed by an expression,
the value is put into the.low-order byte of the word. However, if the
expression is too big (>377(octal)) it is truncated to eight bits and
a Truncation (T) error occurs.

2.8

ASSEMBLER DIRECTIVES

Assembler directives (sometimes called pseudo-ops direct the assembly
process and may generate data. They may be preceded by a label and
followed by a comment. The assembler directive occupies the operator
field. Only one directive may be placed in anyone statement. One or
more operands may occupy the operand field or it may be void
allowable operands vary from directive to directive.

2.8.1

.EOT

The .EOT directive indicates the physical End-of-Tape though not the
logical end of the program. If the .EOT is followed by a single line
feed or form feed, the Assembler will still read to the end of the
tape, but will not process anything past. the .EOT directive. If .EOT
is followed by at least two line feeds or form feeds, the Assembler
will stop before the end of the tape. Either case is proper, but it
should be understood that even though it appears as if the Assembler
has read too far, it actually hasn't.
If a

.EOT is embedded in a tape; and more information to be assembled
follows it,
.EOT must be immediately followed by at least two line
feeds or form feeds. Otherwise, the first line following the
.EOT
will be lost.
Any operands following a .EOT directive will be ignored.
The .EOT
directive allows several physically separate tapes to be assembled as
one program. The last tape is normally terminated by a .END directive
(see Section 3.8.3)
but may be terminated with .EDT
(see .END
emulation in Section 3.9.4).
2-15

WRITING PAL-IIA ASSEMBLY LANGUAGE PROGRAMS
2.8.2

. EVEN

The .EVEN directive ensures that the assembly location counter is even
by adding one if it is odd. Any operands following a .EVEN directive
will be ignored.

2.8.3

.END

The .END directive indicates the logical and physical end of the
source program.
The
.END directive may be followed by only one
operand, an expression indicating the program's entry point.
At load time, the object tape will be loaded and program execution
will begin at the entry point indicated by the .END directive.
If the
entry point is not specified, the Loader will halt after
reading in
the object tape.

2.8.4

. WORD

The .WORD assembler directive may have one or more operands, separated
by commas.
Each operand is stored in a word of the object program.
If there is more than one operand,
they are stored in successive
words.
The operands may be any legally formed expressions.
For
example,

, :::::1.420
SAL::::O
,WORD :l.77535 •• t4,SAL ;STORED IN WORDS :1.420, :1.422, AND
;:1.424 WILL BE :1.77535, :1.426, AND O.
Values exceeding 16 bits will be truncated
length.

from

the

left,

A .WORD directive followed by one or more void operands separated
commas will store zeros for the void operands.
For example,
.
, :::::1.4JO
,WOFW .~5,

word
by

PZERO, FIVE, AND ZERO ARE STORED
;IN WORDS :1.430. :1.432, AND :1.434.

An operator field left blank will be interpreted as the
.WORD
directive if the operand field contains one or more expressions.
The
first term of the first expression in the operand field must not be an
instruction or assembler directive unless preceded by a +, -, or one
of the logical operators! or &.
For example,

• ::::440
LABEL:

tHOV,LABEL

;THE OP-CODE FOR HOV, WHICH IS 010000,
;IS STORED IN LOCATION 440. 440 IS
;STORED IN LOCATION 442.

Note that the default .WORD will occur whenever there is a leading
arithmetic or logical operator, or whenever a leading symbol is
encountered which is not recognized as an instruction mnemonic or
assembler
directive.
Therefore,
if an instruction mnemonic or
assembler directive is misspelled, the .WORD directive is assumed and
errors will result. Assume that MOV is spelled incorrectly as MaR:
MaR A,B
Two error codes can result:
a Q will occur because an expression
operator is missing between MOR and A, and a U will occur if MOR is
undefined.
Two words will be generated; one for MaR A and one for B.
2-16

WRITING PAL-IIA ASSEMBLY LANGUAGE PROGRAMS
2.8.5

. BYTE

SAM=5
.=410
• BYTE 48.,SAM

;STORED IN LOCATION 410 WILL BE
;060 (THE OCTAL EQUIVALENT OF 48) •
;IN 411 WILL BE 005.

.=420
.BYTE , ,

2.8.6

;ZERO WILL BE STORED IN
;BYTES 420, 421 AND 422.

.ASCII

The .ASCII directive translates strings of ASCII characters into their
7-bit ASCII codes with the exception of null, rubout, carriage return,
line feed, and form feed. The text to be translated is delimited by a
character at the beginning and the end of the text. The delimiting
character may be any printing ASCII character except colon and equal
sign and those used in the text string.
The 7-bit ASCII code
generated for each character will be stored in successive bytes of the
object program. For example,

.=500
.ASCII IYESI

;THE ASCII CODE FOR "Y" WILL BE
;STORED IN 500, THE CODE FOR "E"
;IN 501, THE CODE FOR as" IN S02 •

• ASCII 15t3/21

;THE DELIMITING CHARACTER OCCURS
;AMONG THE OPERANDS. THE ASCII
;CODES FOR as", "to, AND "3" ARE
;STORED IN BYTES 503, 504, AND
;505. 21 IS NOT ASSEMBLED.

The ASCII directive must be terminated by a space or a tab.

2.9

2.9.1

OPERATING PROCEDURES

Introduction

The Assembler enables you to assemble an ASCII tape containing PAL-IIA
statements into an absolute binary tape. To do this, two or three
passes are necessary. On the first pass the Assembler creates a table
of user-defined symbols and their associated values, and a list of
undefined symbols is printed on the teleprinter. On the second pass
the Assembler assembles the program and punches out an absolute binary
tape and/or outputs an assembly listing. During the third pass
(this
pass is optional)
the Assembler punches an absolute binary tape or
outputs an assembly listing. The symbol table
(and/or a list of
errors) may be output on any of these passes. The input and output
devices as well as various options are specified during the initial
dialogue
(see Section 3.3.9). The Assembler initiates the dialogue
immediately after being loaded and after the last pass of an assembly.
2-17

WRITING PAL-IIA ASSEMBLY LANGUAGE PROGRAMS
2.9.2

Loading PAL-IIA

PAL-IIA is loaded by the Absolute Loader (see Chapter 6 for operating
procedures) .
Note that the start address of the Absolute Loader must
be in the Switch Register when loading the Assembler.
This is because
the Assembler tape has an initial portion which clears all of core up
to the address specified in the Switch Register, and
jumps to that
address to start loading the Assembler.

2.9.3

Initial Dialogue

After being loaded, the
the teleprinter:

Assemb~er

initiates dialogue

printing

*S
meaning "Whpt is the Source symbolic input device?" The
be:
H

meaning High-speed reader

meaning Low-speed reader

meaing Teletype keyboard

response

may

If the response is T, the source program must be typed at the terminal
once for each pass of the assembly and it must be identical each time
it is typed.
The device specification is terminated, as is all
typing the RETURN key.

user

response,

If an error is made in typing at any time, typing the RUBOUT key will
erase the immediately preceding character if it is on the current
line.
Typing CTRL/U will erase the whole line on which it occurs.
After the *S question and response, the Assembler prints:
*B
meaning "What is the Binary output device?" The responses
similar to those for *S:
H

meaning High-speed punch

meaning Low-speed punch

meaning do not output binary tape
( --" denotes typing the RETURN key)

are

In addition to I/O device specification, various options may be
chosen.
The binary output will occur on the second pass unless /3
(indicating the third pass) is typed following the H or L.
Errors
will be listed on the same pass if /E is typed.
If /E is typed in
response to more than one inquiry, only the last occurrence will be
honored.
It is strongly suggested that the errors be listed on the
same pass as the binary output, since errors may vary from pass to
pass.
If both /3 and /E are ,typed, /3 must precede /E.
The response
is terminated by typing the RETURN key.
Examples:

2-18

WRITING PAL-IIA ASSEMBLY LANGUAGE PROGRAMS
*BL/E

Binary output on the low-speed punch and the
errors on the the teleprinter, both during
the second pass.

Binary output on the high-speed punch and the
errors on the teleprinter; both during the
third pass.

H/3/E

Typing just the RETURN key will
Assembler to omit binary output.

cause

the

After the *B question and response, the Assembler prints:
*L
meaning "What is the assembly Listing output device?" The response
*L may be:

meaning Low-speed punch (outputs a tab as a tab-rubout)

meaning High-speed punch

meaning Teleprinter (outputs a tab as multiple spaces)

meaning line Printer

meaning do not output listing
(-/ denotes typing the RETURN key)

(8K version only)

After the I/O device specification, pass and error list options
similar to those for *B may be chosen. The assembly listing will be
output to the third pass unless /2 (indicating the second pass)
is
typed following H, L, T, or P.
Errors will be listed on the
teleprinter during the same pass if /E is typed.
If both /2 and /E
are typed, /2 must precede /E. The response is terminated by typing
the RETURN key.
Examples:
*L

L/2/E

Listing on low-speed punch and errors
on teleprinter during second pass.

Listing on high-speed punch during
third pass.

..J

The RETURN key alone will cause the
Assembler to omit listing output.

After the *L question and response, the final question is
the teleprinter:

printed

*T
meaning "What is the symbol Table output device?" The
device
specification is the same as for the *L question. The symbol table
will be output at the end of the first pass unless /2 or /3 is typed
in response to *T. The first tape to be assembled should be placed in
the reader before typing the RETURN key because assembly will begin
upon typing the RETURN key in response to the *T question. The /E
option is not a meaningful response to *T. Example:
*T

T/3

Symbol table output on teleprinter at
end of third pass.

..J

Typing just the RETURN key will cause the
Assembler to omit the symbol table output.
2-19

WRITING PAL-llA ASSEMBLY LANGUAGE PROGRAMS
The symbol table is printed alphabetically, four symbols per line.
Each symbol printed is followed by its identifying characters and by
its value. If the symbol is undefined, six asterisks replace its
value.
The identifying characters indicate the class of the symbol;
that is, whether it is a label, direct-assignment, register symbol,
etc. The following examples show the various forms:
ABCDEF

=
=

R3
DIRASM

XYZ

001244

(Defined label)

%000003

(Register symbol)

177777

(Direct assignment)

******

(Undefined direct assignment)

%******

(Undefined register symbol)

LABEL

******

(Undefined label)

Generally, undefined symbols (including
undefined direct assignments.

labels)

will

Multiply-defined symbols are not flagged in the symbol table
but they are flagged wherever they are used in the program.

listed

printout

It is possible to output both the binary tape and the assembly listing
on the same pass, thereby reducing the assembly process to two passes
(see Example 1 below). This will happen automatically unless the
binary device and the listing device are conflicting devices or the
same device (see Example 2 below). The only conflicting devices are
the teleprinter and the low-speed punch. Even though the Assembler
deduces that three passes are neceSsary, the binary and listing can be
forced on pass 2 by including /2 in the responses to *B and *L (see
Example 3 below).
Example l .

Runs 2 passes:

High-speed reader

High-speed punch

Line Printer

Teleprinter

Example 2.

Runs 3 passes:

High-speed reader

High-speed punch

Teleprinter

Example 3.

Runs 2 passes:

High-speed reader

H/2

High-speed punch on pass 2

H/2

High-speed punch on pass 2

Teleprinter
2-20

WRITING PAL-llA ASSEMBLY LANGUAGE PROGRAMS
Note that there are several cases
intermixed with ASCII output:

where

the

binary

*B
*L

H/2
H/2

Binary and
listing to punch on pass 2

L/E

Binary to low-speed punch and
error listing to teleprinter
(and low-speed punch)

*B
*L

L/2/E
T/2

Binary, error listing, and
listing to low-speed punch.

output

can

The binary so ge~erated isloadable by the Absolute Loader as long as
there are no CTRL/A characters in the source program. The start of
every block on the binary tape is indicated by a 001 and the Absolute
Loader ignores all information until a 001 is detected. Thus, all
source and/or error messages will be ignored if they do not contain
any CTRL/A characters (octal 001).
If a character other than those mentioned is typed in response to a
question, the Assembler will ignore it and print the question again.
Example:

High-speed reader

Q is not a valid response

The question is repeated

If at any time you wish to restart the Assembler, type CTRL/P.
When no passes are omitted or error options specified,
performs as follows:

the

Assembler

PASS 1:

Assembler creates a table of user-defined symbols and their
associated values to be used in assembling the source to
object program.
Undefined symbols are listed
on
the
teleprinter at the end of the pass. The symbol table is also
listed at this time.
If an illegal location statement of the
form .=expression is encountered, the line and error code
will be printed out on the teleprinter before the assembly
proceeds.
An error in a location statement is usually a
fatal error in the program and should be corrected.

PASS 2:

Assembler punches 'the object tape, and prints the pass error
count and undefined location statements on the teleprinter.

PASS 3:

Assembler prints or punches the
undefined location statements,
the teleprinter.

assembly program listing,
and the pass error count on

The functions of passes 2 and 3 will occur simultaneously on pass 2 if
the binary and listing devices are different, and do not conflict with
each other (low-speed punch and Teleprinter conflict).

2-21

WRITING PAL-IIA ASSEMBLY LANGUAGE PROGRAMS
The following table summarizes the ini Hal dialogue questions:
Printout

Inquiry

What is the input device of the Source symbolic tape?

What is the output device of the Binary object tape?

What is the output device of the assembly Listing?

What is the output device of the symbol Table?

The following table summarizes the legal responses:
Character

Response Indicated

Teletype keyboard or printer

Low-speed reader or punch

High-speed reader or punch

Line Printer

Pass 1

Pass 2

Pass 3

Errors listed on same pass (not meaningful in response to *S
or *T)

(8K version only)

Omit function

Typical examples of complete initial dialogues:
For minimal PDP-II configuration:
*S

Source input on low-speed reader

L/E

Binary output on low-speed punch
Errors during same (second) pass

Listing on teleprinter during pass 3

Symbol table on teleprinter at end of pass 1

For a PDP-ll with high-speed I/O devices:
*S

Source input on high-speed reader

H/E

Binary output on high-speed punch,
Errors during same (second) pass.
No listing

*L
*T

T/2

Symbol table on teleprinter at end of pass 2

2-22

WRITING PAL-IIA ASSEMBLY LANGUAGE PROGRAMS
2.9.4

Assembly Dialogue

During assembly, the Assembler will pause to print on the teleprinter
various messages to indicate that you must respond in some way before
the assembly process can continue. You may also type CTLR/P, at any
time, if you wish to stop the assembly process and restart the initial
dialogue, as mentioned in the previous section.
When a .EOT assembler directive is read on
prints:

the

tape,

the

assembler

EOF ?
and pauses. During this pause, the next tape is placed in the reader,
and RETURN is typed to continue the assembly.
If the specified assembly listing output device is the high-speed
punch and if it is out of tape, or if the device is the Line Printer
and is out of paper, the Assembler prints on the teleprinter:
EOM ?
and waits for tape or paper to be placed in the device.
Type the
RETURN key when the tape or paper has been replenished; assembly wil
continue.
Conditions causing the EOM?
are:

message for an

assembly

listing

device

HSP

LPT

No power

No tape

Printer drum gate open
Too hot
No paper

There is no EOM if the line printer is switched off-line, although
If
characters may be lost for this condition as well as for an EOM.
the binary output device is the high-speed punch and if it is out of
tape, the Assembler prints:
EOM ?

The assembly process is aborted and
again.

the

When a .END assembler directive is read on
prints:

initial
the

dialogue
tape,

the

begun

Assembler

END ?
and
and
the
for

pauses. During the pause the first tape is placed in the reader,
the RETURN key is typed to begin the next pass. On the last pass,
.END directive causes the Assembler to begin the initial dialogue
the next assembly.

If you are starting the binary pass and the binary is to be punched on
the low-speed punch, turn the punch on before typing the RETURN key
for starting the pass. The carriage return and line feed characters
will be punched onto the binary tape, but the Absolute Loader will
ignore them.
2-23

WRITING PAL-llA ASSEMBLY LANGUAGE PROGRAMS
If the last tape ends with a .EOT, the Assembler may be told to
emulate a
.END assembler directive by responding with E followed by
the RETURN key. The Assembler will then pr-int:
END ?
and wait for another RETURN before starting the next pass.
EOF?
END ?

Example:

NOTE
When a .END directive is emulated with
an E response to the EOF? message, the
error counter is incremented.
To avoid incrementing the error counter,
place a paper tape containing only the
line .END in the reader and press the
RETURN
key instead of using the E
response.

2.9.5

Assembly Listing

PAL-llA produces a $ide-by-side assembly listing of symbolic source
statements, their octal equivalents, assigned absolute addresses, and
error codes as follows:
EELLLLLL 000000 888 •••••• 8
000000
000000

The E's represent the error field.
The L's represent the absolute
address.
The O's represent the object data in octal. The S's
represent the source statement.
While
the
Assembler
accepts
72(decimal) characters per line on input, the listing is reduced by
the 16 characters to the left of the source statement.
The above represents a three-word statement.
The second and third
words of the statement are listed under the command word. No
addresses precede the second and third word since the address order is
sequential.
The third line is omitted for a two-word statement;
third lines are omitted for a one-word statement.

both

second

and

2-24

WRITING PAL-IIA ASSEMBLY LANGUAGE PROGRAMS
2.10

ERROR CODES

The error codes printed beside the octal and
assembly listing have the following meanings:
Error Code

symbolic

code

the

Meaning

Addressing error.
incorrect.

Bounding error.
Instructions or word data
are
being
assembled at an odd address in memory. The location counter
is updated by +1.

Doubly-defined symbol referenced. Reference was
,symbol which is defined more then once.

address

within

the

instruciton

made

Illegal character detected.
Illegal
also non-printing are replaced by a ?

Line buffer overflow. Extra characters on a line (more
72(decimal)) are ignored.

Multiple definition of a label.
which was equivalent
(in the
previously encountered label.

Number containing 8 or 9 has no decimal point.

Phase error. A label's definition or value varies
pass to another.

Questionable syntax. There are missing arguments or the
instruction scan was not completed or a carriage return was
not immediately followed by a line feed or form feed.

Symbol table overflow. When the quantity of user-defined
symbols exceeds the allocated space available in the user's
symbol table, the assembler outputs the current source line
with the S error code, then returns to the initial dialogue.

Truncation error. A number generated more than 16 bits of
significance or an expression generated more than .8 bits of
significance during the use of the .BYTE directive.

Undefined symbol. An undefined symbol was encountered during
the evaluation of an expression. Relative to the expression,
the undefined symbol is assigned a value of zero.

2-25

characters which
on the listing.

are
than

A label was encountered
first sit char~cters) to a

from

reference

one

WRITING PAL-IIA ASSEMBLY LANGUAGE PROGRAMS
2.11

SOFTWARE ERROR HALTS

PAL-IIA loads all unused trap vectors with the code
. WORD

.+2,HALT

so that if the trap does occur, the processor will halt in the second
word of the vector.
The address of the halt, displayed in the console
address register, therefore indicates the cause of the halt.
In
addition to the halts which may occur in the vectors, the standard lOX
error halt at location 40 may occur (see Chapter 7).
Address of Halt

Meaning

Reserved instruction executed

Trace trap occurred

Power fail trap

EMT executed

lOX detected error

See Appendix B for summaries of PAL-IIA features.

2-26

CHAPTER 3
LINK-llS LINKER

3.1

INTRODUCTION

3.1.1

General Description

LINK-lIS (stand alone) is a PDP-II system program designed to link and
relocate programs previously assembled by PAL-lIS.
The user can
separately assemble the main program and each of its
various
subroutines without assigning an absolute load address at assembly
time. The binary output of assembly
(called an object module)
is
processed by LINK-lIS to:
1.

Relocate each object mo'dule and assign absolute addresses.

Link the modules by correlating global symbols defined in one
module and referenced in other modules.

Print a load
addresses.

Punch a load module which can subsequently be loaded (by
Absolute Loader) and executed.

map

which

displays

the

assigned

absolute
the

Seme of the advantages of using PAL-lIS and LINK-lIS are:
1.

The program is divided into segments
(usually subroutines)
which are assembled separately.
If an error is discovered in
one segment, only that segment needs to be reassembled.
The
new object module is then linked with the other object
modules.

Absolute addresses need not be assigned at assembly time.
The Linker automatically assigns absolute addresses. This
keeps programs from overlaying each other. This also allows
subroutines to change size without influencing the placement
of other routines.

Separate assemblies allow the total number of
exceed the number allowed in a single assembly.

Internal symbols (symbols which are not global) need not be
unique among object modules. Thus, naming rules are required
only for global symbols when separate programmers prepare
separate subroutines of a single program.

Subroutines may be provided for general use in object
form to be linked into the user's program.

3-1

symbols

module

LINK-llS LINKER
LINK-lIS is designed to run on an 8K PDP-II with an ASR-33.
A PCll
(high. speed paper tape reader and punch) and an LPll (line printer)
may be used if available.
The PCll significantly speeds up the
linking process.
An LPll provides a fast device for the loa'tl map
listing.

3.1.2

Absolute and Relocatable Program Sections

A program assembled by PAL-lIS may consist of an absolute program
section, declared by the .ASECT assembler directive, and a relocatable
program section, declared by the .CSECT assembler directive.
(If a
program has neither an .ASECT or
.CSECT directive, the assembler
implicitly assumes a .CSECT directive.) The program and data in the
absolute section are assigned absolute addresses as specified by the
location counter setting statements
(.=x).
The Linker
assigns
absolute addresses to the program and data in the relocatable section.
Addressses are normally assigned such that the relocatable section is
at the high end of memory.
The assignment of addresses may be
influenced by command string options (see Section 3.3.2).
The Linker appropriately modifies all instructions and/or data
necessary to account for the relocation of the control section.

LINK-lIS can handle object
modules
containing
named
control
(relocatable) sections as generated by PAL-llR. However, PAL-lIS can
create only the unnamed control section (which has the special default
name of 6 blanks) and the absolute section (with the special name
. ABS.). The unnamed control section is internal to each object
module.
That is, every object module may have an unnamed control
section (each with the name 6 blanks) but the Linker treats them
independently.
Each is assigned an absolute address such that they
occupy mutually exclusive areas of memory. Named control sections, on
the other hand, are treated globally. That is, if different object
modules each have control sections with the same name, they are all
assigned the same absolute load address and the size of the area
reserved or loading of the section is the maximum of the sizes of each
section.
Thus, named control sections allow the sharing of data
and/or programs among object modules. This is very similar to the
handling and function of labelled COMMON in FORTRAN IV. A restriction
of LINK-lIS is that the name of a control section must not be the same
as the name of a global entry symbol, as this results in multiple
definition errors.

3.1.3

Global Symbols

Global symbols provide the links for communication between object
modules
(or assemblies). Global symbols are created with the .GLOBL
assembler directive~ Symbols which are not global are called internal,
symbols.
If the global symbol is defined
(as a label or direct
assignment) in an object module it is called an entry symbol, and
other object modules may reference it. If the global symbol is not
defined in the object module it is an external symbol. It is assumed
to be defined (as an entry symbol) in some other object module.
As the Linker reads the object modules it records all the global
symbol definitions and references.
It then modifies the instructions
and/or data that reference the global symbols.

3-2

LINK-lIS LINKER
3.2

3.2.1

INPUT AND OUTPUT

Object Module

Input to LINK-lIS is the object module. This is the output of PAL-lIS
(or any other program which can create an object module). The Linker
reads each object module twice; that is, it is a two-pass processor.
On pass 1, the Linker reads each object module to gather enough
information to assign absolute addresses to all reloc~table sections
and absolute values to all globals. This information appears in the
global symbol directory (GSD) of the object module.
On pass 2, the Linker reads all of each object module and produces the
load module
(see Section 3.2.2). The data gathered on pass 1 guides
the relocation and linking process on pass 2.

3.2.2

Load Modules

The normal output of the Linker is a load module which may
and run.

loaded

A load module consists of formatted binary blocks holding absolute
load addresses and object data as specified for the Paper Tape System
Absolute Loader and the PDP-II Disk Monitor. The first few words of
data are the communications directory (COMD) and have an absolute load
address equal to the lowest relocated address of the program.
The
absolute loader loads the COMD at the specified address but the
program subsequently overlays it. l The disk monitor loader expects the
COMD and loads it where the monitor wants it. The end of the load
module is indicated by a TRA block; that is, a block containing only
a load address. The byte count in the formatted binary block is 6 on
this block; on all other blocks the byte count is larger than 6. The
TRA (transfer address) is selected by the Linker to be the first even
transfer address seen.
Thus, if four object modules are linked
together and if the first and second had a .END statement, the third
had a .END A and the fourth had a .END B, the transfer address would
be A of module three.
lThe overlaying of the COMD by the relocated program is a trick to
allow the Absolute Loader to handle load modules with a COMD.
However, a problem arises if a load module is to be loaded by the
absolute loader and either of the following conditions exists:
a.

The object modules used to construct
contained no relocatable code; or

the

The total size of the relocatable code is
(decimal) bytes (the size of the COMD).

load
less

module
than

In either case, there is not enough relocatable code to overlay the
COMD which means the COMD will load into parts of memory not intended
to be altered by the user. The COMD's load address, selected by the
Linker in the above cases, is such that it will be up against the
current top of memory (see *T .option in section 3.3.1).
If the top
happens to be very low, the Linker does not allow the COMD to be
loaded below address 0; it loads it at O.

3-3

L~NK-llS

3.2.3

LINKER

Load Map

The load map provides several types of information concerning the load
module's make-up.
The map begins with an indication of the low and
high limits of the relocatable code and the transfer address.
Then
there is a section of the map for each object module included in the
linking process. Each of these sections begins.with the module name
followed by a list of the control sections and the entry points for
each control section. For each control section, the base of the
section
(its low address) and its size (in bytes) is printed to the
right of the section name (enclosed in angle brackets).
Following
each section name is a list of entry points and their addresses.
After all information has been printed for each object module, any
undefined symbols are listed. Note that modules are loaded such that
if modules A, Band C are linked together, A is lowest and C is
highest in memory.
The format is quite self-explanatory as can be seen from the following
example:
LOAD MAP
TRANSFER ADDRESS,: 037434
LOW LIMIT: 037406
HIGH LIMIT: 037460

**********

MODULE MODt
SECT I ON ENTRY
<,0 ABS.>
>
<
X3
X4
X5
X7

ADDRESS SIZE
000000 .000000
037406 000044
037452
037440
037450
037430

**********

MODULE MOD2
SECTION ,ENTRY
<

Xl
X2

ADDRESS SIZE
037452 000006
037452
037452

**********
**********

UNDEFINED REFERENCES
X6
PASS 2

3-4

LINK-llS LINKER
3.3

3.3.1

OPERATING PROCEDURES

Loading and Command String

The Linker is loaded by the Absolute Loader and is self-starting.
It
uses a simple command dialogue which allows the object module, load
module and load map devices to be specified. During pass 1 and pass
2, the Linker asks for each object module individually.
Operation begins by the linker typing its name and version.
This
followed by the input option printed as *I~.
The responses are:

...J

Read object module from HSR.
Read object module from HSR.
Read object module from LSR.

H./
L...J

The input option is followed by the output option *O/:'.
are:

../

The

responses

Punch load module on HSP.
Punch load module on HSP.
Punch load module on LSP.

H./
L...J

LINK-llS asks i f a load map is desired by typing *M/:'.
The legal
responses are ..J for no map,
T../ or H...,.) or P../for a map on the
teleprinter, high-speed punch, or line printer, respectively.
The next two options concern the placement of the relocated object
program in memory.
The standard version of the Linker assumes it is
linking for an 8K machine.
It relocates the program such that it is
as high as possible in 8K but leaves room for the Absolute and Boot
Loaders.
These assumed values may be changed by altering parameters
HGHMEM
(highest legal memory address +1) and ALODSZ (number of bytes
allocated for Absolute Loader and Boot Loader)
and reassembling the
Linker.
The user may control where a program is relocated to with the
*T and *B options. After the option *T~ has been typed, the user may
respond as follows:
Relocate so that program is up against the
current top of memory.
If the top has not
been changed, then the top is the assembled-in
top
(HGHMEM-ALODSZ).
The standard assumption
is 16272 decimal (16384-112) or 37460 octal.
N is an octal number
a new top address.

(unsigned) which

defines

If a new top is specified, the *B option is suppressed.
After the option

*B~

has been printed the user may respond as follows:
Use current top of memory.

N../

N is an unsigned octal number which defines
the bottom address of the program.
That is, a
new top of memory is calculated so that the
bottom of the program corresponds with N.

Once a top of memory has been calculated (by *T or *B), that value
used until it is changed.

3-5

LINK-llS LINKER
LINK-lIS indicates the start of pass one by typing PASS 1. The
input
is requested by the Linker, one tape at a time, by typing *~.
The
legal responses are:
Read a tape and request more input.
List all undefined globals on the
and request more input.

teleprinter

End of input.
If there are undefined globals,
list them on the teleprinter and request more
input.
Otherwise print the load map,
if
requested, and enter pass-2.

C...J

End of input.
Assign 0 to any undefined
globals, print the load map (if requested),
and enter pass 2.

The Linker indicates the start of pass 2 by typing PASS
requests each input tape as in pass 1.

then

A carriage return is the only useful response to * on pass 2.
The
modules must be read on pass 2 in the same order as pass 1. When the
the load
last module has been read the Linker automatically finishes
module and restarts itself.
Leader and trailer are punched on the load module.
If the LSP is being used for the load module output,
it should be
turned on before pass 2 begins.
Thus, turn it on before typing E...Jor
c-l The echo of these characters (and the load map, if printed on the
TTY)
is punched on the load module but may be easily removed since
leader is punched on the load module.
In any case, ASCII information
in a load module is ignored by the Absolute and Disk Monitor loaders.
However, the LSP can be turned on while leader is being punched (after
the linker has typed PASS 2) to keep the load map, etc., from being
punched onto the tape.
Note:
On all command string options, except for
*T and *B,
the linker
examines only the last character typed preceding the carriage return.
Thus,
ABCDEFGH...J
is equivalent to H-l

3.3.1.1 Operational Cautions ~ The Linker does not give a warning if
a program is linked so low in memory that it goes below address O.
However, this case is easily seen by examining the low and high limits
which are always printed (on the load map or on the teleprinter).
The Linker reads object modules until an end of medium is detected.
Object modules from the DEC Program Library contain a special checksum
at the end of the tape which must be removed before they are linked.
Failure to remove this checksum can result in fatal Linker errors.

3-6

LINK-lIS LINKER
3.3.2

Error Procedure and Messages

p ) is used for two
3.3.2.1 Restarting - CTRL/P (symbolized as
purposes by LINK-II.
If a Ap is typed while a load map is being
printed, the load map is aborted and the Linker continues. A Ap typed
at any other time causes the Linker to restart itself.
A

3.3.2.2
1.

Non-Fatal Errors Non-unique object module name - this error is detected during
pass 1;
an error message is issued and the module is
rejected. The message is:

?MODULE NAME xxxxxx NOT UNIQUE
The Linker then asks for more input.
2.

Load map device EOM - this error allows the user to fix the
device and continue or abort the map listing. The Linker
prints:

?MAP DEVICE EOM.
TYPE (CR) TO CONTINUE
Any response, terminated by .J or
~
causes the
continue. A t p causes the map to be aborted.
3.

Linker

A byte relocation error - the Linker tries to relocate
link byte quantities. However, relocation usually·fails
linking may fail.
Failure is defined as the high byte of
relocated value (or the linked value) not being all zero.
such a case, the value is truncated to 8 bits and
following message is printed:

and
and
the
In
the

?BYTE RELOC ERROR AT ABS ADDRESS xxxxxx.
The Linker automatically continues.
4.

If the object modules are not read in the same order on pass
2 as pass 1, the Linker indicates which module should be
loaded next by typing:
?LOAD xxx xxx NEXT!
The linker then asks for more input.

Multiply-Defined Globals - this
error message during pass 1:

results

the

following

?xxxxxx MULTIPLY DEFINED BY MODULE xxxxxx.
The second definition is ignored and the Linker continues.

3.3.2.3 Fatal Errors - Each of the following errors causes
the
indicated error message to be printed and the Linker to be restarted.
1.

Symbol Table overflow - the message is:

?SYMBOL TABLE OVERFLOW - MODULE xxxxxx, SYMBOL xxxxxx
3-7

LINK-llS LINKER
2.

System Errors - this class of errors prints:

1SYSTEM ERROR xx
where xx is an identifying number as follows:
Number

Meaning

Unrecognized symbol table entry was found.

A relocation directory references a global
name which cannot be found in the symbol
table.

A relocation directory contains a location
counter modification command which is not
last.

Object module does not start with a GSD.

The first entry in the GSD is not the
name.

An RLD references a section name which cannot
be found.

The
TRA
specification
non-existent module name.

references

The
TRA
specification
non-existent section name.

references

An internal jump table index is out of range.

A checksum
module.

An object module binary. block is too
(more than 64 decimal words of data).

A device error occurred on
output device.

error

occurred

the

load

module

object
big

module

All system errors except for numbers 10 and 12 indicate a program
failure either in the Linker or the program which generated the object
module. Error 05 can occur if a tape is read which is not an object
module.

3.3.2.4 Error HALTs - LINK-lIS loads all of its unused
with the code:
. WORD

trap

vectors

.+2, HALT

so that if the trap occurs, the processor halts in the second word of
the vector. The address of the halt, displayed in the console lights,
therefore indicates the cause of the halt.

3-8

LINK-llS LINKER
3.3.2

Error Procedure and Messages

p)
3.3.2.1 Restarting - CTRL/P
(symbolized as
is used for
two
purposes by LINK-II.
If a
Ap is typed while a load map is being
printed, the load map is aborted and the Linker continues.
A Ap typed
at any other time causes the Linker to restart itself.
A

3.3.2.2

Non-Fatal Errors Non-unique object module name - this error is detected during
pass 1;
an error message is issued and the module is
rejected.
The message is:

1MODULE NAME xxxxxx NOT UNIQUE
The Linker then asks for more input.
2.

Load map device EOM - this error allows the user to fix
the
device and continue or abort the map listing.
The Linker
prints:

'i'MAP DE\) I CE EOM.
TYPE (CR) TO CONTINUE
Any response, terminated by ~ or
+ causes the
continue. A t p causes the map to be aborted.
3.

Linker

A byte relocation error - the Linker tries to relocate
link byte quantities.
However, relocation usually fails
linking may fail.
Failure is defined as the high byte of
relocated value (or the linked value) not being all zero.
such a case, the value is truncated to 8 bits and
following message is printed:

and
and
the
In
the

'i'BYTE RELOC ERROR AT ABS ADDRESS xxxxxx.
The Linker automatically continues.
4.

If the object modules are not read in the same order on pass
2 as pass 1,
the Linker indicates which module should be
loaded next by typing:
?LOAD xxx xxx NEXT!
The linker then asks for more input.

Multiply-Defined Globals - this
error message during pass 1:

results

the

following

1xxxxxx MULTIPLY DEFINED BY MODULE xxxxxx.
The second definition is ignored and the Linker continues.

3.3.2.3 Fatal Errors - Each of the following errors causes
the
indicated error message to be printed and the Linker to be restarted.
1.

Symbol Table overflow - the message is:

1SYMBOL TABLE OVERFLOW - MODULE
3-7

XXXXXX9

SYMBOL xxxxxx

LINK-lIS LINKER
2.

System Errors - this class of errors prints:

?SYSTEM ERROR xx
where xx is an identifying number as follows:
Number

Meaning

Unrecognized symbol table entry was found.

A relocation directory references a global
name which cannot be found in the symbol
table.

A relocation directory contains a location
counter modification command which is not
last.

Object module does not start with a GSD.

The first entry in the GSD is not the
name.

An RLD references a section name which cannot
be found.

The
TRA
specification
non-existent module name.

references

The
TRA
specification
non-existent section name.

references

An internal jump table index is out of range.

A checksum
module.

An object module binary block is too
(more than 64 decimal words of data).

A device error occurred on
output device.

error

occurred

the

load

module

object
big

module

All system errors except for numbers 10 and 12 indicate a program
failure either in the Linker or the program which generated the object
module .. Error 05 can occur if a tape is read which is not an object
module.

3.3.2.4 Error HALTs - LINK-lIS loads all of its unused
with the code:
. WORD

trap

vectors

.+2, HALT

so that if the trap occurs, the processor halts in the second word of
the vector. The address of the halt, displayed in the console lights,
therefore indicates the cause of the halt.

3-8

LINK-lIS LINKER
Meaning

Address of HALT (octal)

Reserved instruction executed.
Trace trap occurred.
Power fail trap.
~MT executed.

12
16
26

A halt at address 40 indicates an IOXLPT detected error.
(displayed in the console lights) contains an identifying code:
Meaning

Code in RO
0
1
2
3
4
5

Illegal memory reference, SP overflow or
illegal instruction.
Illegal lOX command.
Slot number out of range.
Device number illegal.
Referenced slot not INIT ed.
Illegal data mode.

IOXLPT also sets Rl as follows:
If the error code is 0, Rl contains the PC at the time of the error.
If the error code is 1-5, Rl points to some element in the lOT
argument list or to the instruction following the argument list,
depending on whether IOXLPT has finished decoding all the arguments
when it detects the error.

3-9

CHAPTER 4
EDITING THE SOURCE PROGRAM

The PDP-II Text Editor program
source program (or any text)
additions to it, and punch all
tape. This is accomplished by
the keyboard.

(ED-II) enables you to display your
on the teleprinter, make corrections or
or any portion of the program on paper
typing simple one-character commands on

The Editor commands can be grouped according to function:
1.

input/output;

searching for strings of characters;

positioning the current character location pointer;

inserting, deleting, and exchanging text portions.

All input/output functions are handled by lOX, the PDP-II Input/Output
Executive (see Chapter 7).

4.1

COMMAND MODE AND TEXT MODE

Whenever ED-II prints an * on the teleprinter, you may type a command
to it.
(Only one command per line is acceptable.) The Editor is then
said to be in Command Mode. While most commands operate exclusively
in this mode, there are five ED-II commands t.hat require additional
information in order for the commands to be carried out.
The Editor
goes into Text Mode to receive this text.
Should a nonexistent command be typed or a command appear in incorrect
format, ED-II prints a?
This is followed by an * at the beginning
of a new line indicating that the Editor is in Command Mode.
Editor processing begins in Command Mode. When you type a command, no
action occurs until you follow it by typing the RETURN key (symbolized
as..) ).
If the command is not a text-type command, typing the RETURN
key initiates the execution of the command and ED-II remains in
Command Mode. However, if the command is a text-type command (Insert,
eXchange, Change, Get, or wHole), typing the RETURN key causes the
Editor to to into Text Mode. At this time you should type the text to
be operated on by the command. This can include the non-printing
characters discussed below, as well as spaces and tabs
(up to eight
spaces generated by the CTRL/TAB keys).
Note that typing the RETURN key always causes the physical return of
the Teletype print element to the beginning of the line, and
automatically generates a line feed, thereby advancing the carriage to
a new line.
In Text Mode, the RETURN key not only serves these
4-1

EDITING THE SOURCE PROGRAM
mechanical functions, allowing you to continue typing at the beginning
of a new line, but at the same time it enters a carriage return and
line feed character into the text.
(A carriage return not followed by
a line feed cannot, therefore, be entered from the keyboard.)
RETURN and LINE FEED are both counted as characters and can be edited
along with the printing characters (as can the form feed, discussed in
Section 4.2.5). When you wish to terminate Text Mode and reenter
Command Mode, you must type the LINE FEED key symbolized as ~
). A
typed LINE FEED is not considered to be part of the text unless it is
the first character entered in Text Mode.

4.2

COMMAND DELIMITERS

4.2.1

Arguments

Some ED-II commands require an argument to specify the particular
portion of text to be affected by the command or how many times to
perform the command.
In other commands this specification is implicit
and arguments are not allowed.
The ED-II command arguments are described as follows:
1.

n stands for any number from 1 through 32767
(decimal)
may, except where noted, be preceded by a + or If no sign precedes n, n is assumed to be a positive
Where an argument is acceptable, its absence
argument of 1 (or -1 if a - is present).
The role of n varies according to the command with
is associated.

0 refers to the beginning of the current line.

@ refers to a marked
Section 4.2.3).

(designated)

character

number.

implies

which

location

(see

refers to the end of text in the Page Buffer.

The roles of all arguments are
explained
corresponding commands which qualify them.

4.2.2

and

further

with

the

The Character Location Pointer (Dot)

Almost all ED-II commands function with respect to a movable reference
point, Dot.
This character pointer is normally located between the
most recent character operated upon and the next character and, at any
given time, can be thought of as "where the Editor is" in your text.
There are commands which move Dot anywhere in the text, thereby
redefining the "current location" and allowing greater facility in the
use of the other commands.

4-2

EDITING THE SOURCE PROGRAM
4.2.3

Mark

In addition to Dot, a secondary character pointer known as Mark also
exists in ED-II. This less agile pointer is used with great effect to
mark or "remember" a location by moving to Dot and conditionally
remaining there while Dot moves on to some other place in the text.
Thus, it is possible to think of Dot as "here" and Mark as "there".
Positioning of Mark, which is referenced by means of the argument @,
is discussed below in several commands.

4.2.4

Line-Oriented Command Properties

ED-II recognizes a line as a unit by detecting a line terminator in
the text.
This means that ends of lines (line feed or form feed
characters) are counted in line-oriented commands. This is important
to know, particularly if Dot, which is a character location pointer,
is not pointing at the first character of a line.
In such a case, an argument n does not affect the same number of lines
(forward)
as its negative (backward). For example, the argument -1
applies to the character string beginning with the first character
following the second previous end-of-line character and ending at Doti
argument +1 applies to the character string beginning at Dot and
ending at the first end-of-line character. If Dot is located, say, in
the center of a line, notice that this affects 1-1/2 lines back or 1/2
line forward, respectively:
Example of List Commands -lL and +lL:
Text
CMPB ICHAR,#033
BEQ
$ALT
CMPB ~CHAR' #175
BNE
PLACE
Dot is here

4.2.5

Command

Printout
BEQ $ALT
CMPB I

*-lL

CHAR'#~Dot

*+lL

remains

~here

The Page Buffer

The Page Buffer holds the text being edited. The unit of source data
that is read into the Page Buffer from a paper tape, is the page.
Normally a page is terminated, and therefore defined, by a form feed
(CTRL/FORM)
in the source text wherever a page is desired.
(A form
feed is an acceptable Text Mode character.) Overflow, no-tape, or
reader-off conditions can also end a page of input (as described in
Section 4.3.1.2). Since more than one page of text can be in the
buffer at the same time, it should be noted that the entire contents
of the Page Buffer are available for editing.

4-3

EDITING THE SOURCE PROGRAM
4.3

COMMANDS

4.3.1

Input and Output Commands

Three commands are available for reading in a page of text.
The Read
command
(Section 4.3.1.2)
is a specialized input command;
the Next
command (Section 4.3.1.4) reads in a page after punching out the
previous page;
and the wHole command (Section 4.3.3.2) reads in and
punches out pages of text as part of a search for
a specified
character string.
Output commands either list text or punch it on paper tape.
The List
command causes specified lines of text to be printed at the terminal
so that they may be examined.
Paper tape commands
(Next and wHole
also perform input) provide for the output of specified pages, lines,
form feeds (for changing the amount of data that constitutes a given
page), and blank tape.
Note that the process of outputting text does
not cause Dot to move.

4.3.1.1 Open - The Open command (0) should be typed whenever a new
tape is put in the reader.
This is used when the text file being
edited is on more than one paper tape.
Note also that if the reader is off at the time an input command
given, turning the reader on must be followed by the Open command.

4.3.1.2 Read - One way of getting a page of text into the Page Buffer
so that ~ can be edited is the Read (R) command.
The R command
causes a page of text to be read from either the low-speed reader or
high-speed reader
(as specified in the starting dialogue, Section
4.4.2), and appended to the contents (if any) of the Page Buffer.
Text is read in until either:
1.

A form feed character is encountered;

The page buffer is 128 characters from being filled,
or a
line feed
is encountered after the buffer has become 500
characters from being filled;

The reader is turned off, or runs out of paper tape
command, Section 4.3.1.1).

Following execution of an R command, Dot and Mark are located
beginning of the Page Buffer.

(see Open
at

the

A 4K system can accommodate about 4000 characters of text.
Each
additional 4K of memory provides space for about 8000 characters.
NOTE
An attempt to overflow the storage area
causes the command (in this case, R) to
stop executing. A? is then printed,
followed by an * on the next line
indicating that a command may be typed.
No data is lost.

4-4

EDITING THE SOURCE PROGRAM
4.3.1.3 List and Punch - Output commands List (L) and Punch
(P) can
be described together, as they differ only in that the device
addressed by the former is the terminal, and the device addressed by
the latter is the paper tape punch.
Dot is not moved by these
commands.
the character string beginning at Dot and
ending with the nth end-of-line

nL
nP

Lists
}
Punches

-nL
-nP

Lists
Punches

}

the character string beginning with the
first character following the (n+l)th previous end-of-line and terminating at Dot

OL
OP

Lists
Punches

}

the character string beginning with the
first character of the current line and
ending at Dot

@L
@P

Lists
Punches

}

the character string between Dot and the
Marked location

Lists
Punches

}

the character string beginning at Dot and
ending with the last character in the Page
Buffer

In addition to the above List commands, there are three special List
commands that accept no arguments. The current line is defined as the
line containing Dot, i.e., from the line feed (or form feed) preceding
Dot to the line feed (or form feed) following Dot.

Lists the entire line containing Dot

Same as -lL. If Dot is located at the
beginning of a line, this simply lists
the line preceding the current line

Lists the line following the current line

Examples:
TEXT
CMPB ICHAR,#033
BEQ
$ALT
CMPB/CHAR,#175
BNE
LACE

PRINTOUT

COMMANDS

CMPB
BEQ
CMPB
BNE
Dot remains here.

Dot is here.

4.3.1.4 Next - Typing nN punches out the entire contents of the Page
Buffer (followed by a trailer of blank tape if a form feed is the last
character in the buffer), deletes the contents of the buffer, and
reads the Next page into the buffer.
It performs this sequence n
times.
If there are fewer than the n pages specified, the command is
executed for the number of pages actually available, and a? is
printed out. Following execution of a Next, Dot and Mark are located
at the beginning of the Page Buffer.

4-5

EDITING THE SOURCE PROGRAM
4.3.1.5

Form Feed and Trailer -

Punches out a Form feed character and four inches
tape

blank

Punches out four inches of Trailer (blank) tape n times

4.3.1.6 Procedure with Low-Speed Punch - If the low speed punch is
the specified output device
(see Section 4.4.2), the Editor pauses
before executing any tape command just typed
(Punch, Form feed,
Trailer, Next, wHole).
The punch must be turned on at this time,
after which typing the SPACE bar initiates the execution of the
command.
Following completion of the operation, the Editor pauses
again to let you turn the punch off. When the punch has been turned
off, typing the SPACE bar returns ED-II to Command Mode.

4.3.2

Commands to Move Dot and Mark

4.3.2.1

Beginning and End -

Moves Dot to the Beginning of the Page Buffer

Moves Dot to the End of the Page Buffer
and /A below)

4.3.2.2

(see

also

Jump and Advance-

Jumps Dot forward past n characters

-nJ

Moves Dot backward past n characters

Advances Dot forward past n ends-of-lines
beginning of the succeeding line

-nA

Moves Dot backwards across
n
ends-of-lines
and
positions Dot immediately after n+l ends-of-lines,
i.e., at the beginning of the -n line.

OJ or OA

Moves Dot to the beginning of the current line

@J or @A

Moves Dot to the Marked location

/J or /A

Moves Dot to the end of the Page
above)

Buffer

(see

also

the

Notice that while n moves Dot n characters in the Jump command, its
role becomes that of a line counter in the Advance command. However,
because 0, @, and /
are absolute, their use with these commands
overrides line/character distinctions.
That is, Jump and Advance
perform identical functions if both have either 0, @ or /
for an
argument.

4-6

EDITING THE SOURCE PROGRAM
4.3.2.3 Mark - The M command marks ("remembers") the current position
of Dot for late.r reference in a command using the argument @. Note
that only one position at a time can be in a marked state.
Mark is
also affected by the execution of those commands which alter the
contents of the Page Buffer:
C

4.3.3

Search Commands

4.3.3.1 Get - The basic search command nG starts at Dot and Gets the
nth occurrence of the specified text in the Page Buffer.
If no
argument is present, it is assumed to be 1.
When you type the
command, followed by the RETURN key, ED-II goes into Text Mode. The
character string to be searched for must now be typed.
(ED-II will
accept ~ search object of up to 42 characters.) Typing the LINE FEED
key terminates Text Mode and initiates the search.
This command sets Dot to the position immediately following the found
character string, and a OL listing is performed by ED-II.
If a
carriage return, line feed, or form feed is specified as part of the
search object, the automatic OL displays only a portion of text -- the
part defined as the last line. Where any of these characters is the
last character of the search object, the OL of course yields no
printout at all.
If the search is unsuccessful, Dot is at the end of the Page Buffer
and a? is printed out. The Editor then returns to Command Mode.
Examples:
Text

MOV
ADD
( CLR
TST
BEQ

Command

@RMAX,@RS
#6,(RS)+
$CK3
R2
CKCR

2G./
CK

CMPB
BEQ
BR

BEQ CK

Dot is no\

Dot was here.
2.

Printout

G..}

ICHAR,#RUBOUT
SITE
PUT

TE./
BR-I-

4.3.3.2 wHole - A second search command, H, starts at Dot and looks
through the wHole text file for the next occurrence of the character
string you have specified in Text Mode. It combines a Get and a Next
such that if the search is not successful in the Page Buffer, the
contents of the buffer are punched on tape, the buffer contents are
deleted, and a new page is read in, where the .search is continued.
This continues until the search object is found or until the complete
source text has been searched.
In either case, Mark is at the
beginning of the Page Buffer.

4-7

EDITING THE SOURCE PROGRAM
If the search object is found, Dot is located immediately following
it, and a OL is performed by ED-II. As in the Get command, if the
search is not successful Dot is at the end of the buffer and a ?
appears on the teleprinter.
Upon completion of the command, the
Editor will be in Command Mode. No argument is allowed. Note that an
H command specifying a nonexistent search object can be used to close
out an edit, i.e., copy all remaining text from the input tape to the
output tape.

4.3.4

Commands to Modify the Text

4.3.4.1 Insert - The Insert command (I) allows text to be inserted at
Dot. After I is typed (followed by the typing of the RETURN key), the
Editor goes into Text Mode to receive text to be inserted. Up to 80
characters per line are acceptable. Execution of the command occurs
when the LINE FEED key (which does not Insert a line feed character
unless it is the first key typed in Text Mode) is typed terminating
Text Mode. At this point, Dot is located in the position immediately
following the last inserted text character. If the Marked location
was anywhere after the text to be Inserted, Dot becomes the new Marked
location.
.
During an insert, it sometimes happens that the user accidentally
types' CTRL/P rather than SHIFT/P (for @), thus deleting the entire
insert (see Section (.4.1). To minimize the effect of such a mistake,
the insert may be terminated every few lines and then continued with a
new Insert command.
As with the Read command, an attempt to overflow the Page Buf£er
causes a ?
to be printed out followed by an * on the next line
indicating that a command may be typed. Allor part of the last line
typed
may be lost.
All previously typed lines are inserted.
Examples:
Text
l.

MOV #8. ,EKO,

Dot
I -

MOV~~::~;F
Dot

Inserting a carriage return (and automatic line feed) :
CLR Rl,LR R2

I..-J

..-J

Dot
3.

Effect

Command

CLR Rl
CLR R2

.j.

Inserting a single line feed:
I~

LOOK

Dot/"

HAPPENS HERE

+
+

LOOK WHAT
,APPENS HERE
Dot

4-8

EDITING THE SOURCE PROGRAM
4.3.4.2 Delete and Kill - These commands are closely related to each
other;
they both erase specified text from the Page Buffer. The
Delete command (D) differs from the Kill command (K) only in that the
former accepts an argument, n, that counts characters to be removed,
while the latter accepts an argument, n, that counts lines to be
removed.
0, @, and / are also allowed as arguments. After execution
of these commands, Dot becomes the Marked location.
nO

Deletes the following n characters

-nO

Deletes the previous n characters

Kills the character string beginning at Dot and
at the nth end-of-line

-nK

Kills the character string beginning with the first
(n+l)th previous end-of-line
character following the
and ending at Dot

00 or OK

Removes the current line up to Dot

@D or @K

Removes the character string bounded by Dot and Mark

/0 or /K

Removes the character string beginning at Dot
ending with the last character in the Page Buffer
Text

Command

Effect

-20

iCHECK THE MODE

jCHECK THE MOZXDE

00/

ending

and

Dot

; IS IT A TAB OR
i IS IT A CR \

iIS IT

Dot/

Dot

4.3.4.3 Change and exchange - The Change
(C)
and
eXchange
(X)
commands
can
be
thought of as two-phase commands combining,
respectively, an Insert followed by a Delete, and an Insert followed
by a Kill. After the Change or eXchange command is typed, ED-II goes
into Text Mode to receive the text to be inserted. If n is used as
the
argument,
it
is
then
interpreted
as
in
the Delete
(character-oriented) or Kill (line-oriented), and accordingly removes
the indicated text. 0, @, and / are also allowed as arguments.
nC
xxxx

Changes the following
n characters

xxxx
-nC
xxx
nX
xxxx
xxxx

Changes the previous
n characters
eXchanges the character
string beginning at Dot and
ending at the nth end-of-line

-nX
xxx

eXchanges the character
the
first
character
string beginning
with
following the
(n+l)th previous end-of-line and
ending at Dot

4-9

EDITING THE SOURCE PROGRAM

OC or OX
xxxx xxxx
xxxx xxx x

Replaces the current line up to Dot

@C
xxx
xxx

@X
xxx
xxx

Replaces the character string bounded by Dot
and the Marked location

Replaces the character string beginning at Dot
and ending with the last character in the Page
Buffer.

xxx

Again, the use of absolute arguments 0, and @, and /
overrides the
line/character distinctions that nand -n produce in these commands.
If the Insert portion of a Change or eXchange is terminated because of
attempting to overflow the Page Buffer, data from the latest line may
have been lost, and text removal does not occur. Such buffer overflow
might be avoided by separately executing a Delete or Kill followed by
an Insert, rather than a Change or eXchange, which does an Insert
followed ·by a Delete or Kill. Examples:
Text
;A LINE FEED \IS HERE
;THIS
;IS ON
·FOUR
CLINES

Dot

Command
-9C~

Effect
IS HERE

TAB 42X..j
PAPER

Dot

4.4

OPERATING PROCEDURES

4.4.1

Error Corrections

During the course of editing a page of the program, it may become
necessary to correct mistakes in the commands themselves. There are
four special commands which do this:
1.

Typing the RUBOUT key removes the preceding typed character,
if it is on the current line. Successive RUBOUTs remove
preceding characters on the line (including the SPACE), one
character for each RUBOUT typed.

The CTRL/U combination (holding down the CTRL key and
U) removes all the characters in the current line.

CTRL/P cancels the current command in its entirety.
This
includes all the current command text just typed, if ED-II
was in Text Mode. Do not use another CTRL/P before typing a
line terminator as this will cause an ED-II restart (see 4.
below).
If CTRL/P is typed while,a found search object of a
Get or wHole is being printed out, the normal position of Dot
(just after the specified search object) is not affected.

typing

CTRL/P should not be used while a punch operation is in
progress as it is not possible to know exactly how much data
will be output.
4-10

EDITING THE SOURCE PROGRAM
4.

Two CTRL/P's not interrupted by a typed line terminator
restart ED-II, initiating the dialogue described in Section
4.4.2.

After removing the incorrect command data, the user can directly
in the desired input.

4.4.2

type

Starting

The Editor is loaded by the Absolute Loader (see Chapter 6, Section
6.2.2) and starts automatically. Once the Editor has been loaded, the
following sequence occurs:
ED-II Prints

User Types

(if the low-speed Reader is to
for source input)

used

(if the high-speed Reader is to be
for source input)

used

L...J

(if the low-sp,eed Punch is
for edited output)

used

H.,.}

(if the high-speed Punch is to
for edited output)

used

If all text is to be entered from the keyboard (i.e., via
command) i either L or H may be specified for !nput.

the

Insert

If the output device is the high-speed punch (HSP), the Editor enters
Command Mode to accept input. Otherwise, the sequence continues with:
LSP OFF

(when low-speed Punch (LSP) is off)

Upon input of ~ from the keyboard, the Editor enters Command Mode and
is ready to accept input.

4.4.3

Restarting

To restart ED-II, type CTRL/P twice.
This initiates the normal
starting dialogue described in Section 4.4.2.
If the Low-speed Reader
The text to be
(LSR) is in operation it must first be turned off.
edited should be loaded (or reloaded) at this time.

4.4.4

Creating a Paper Tape

Input commands assume that text is to be read from a paper tape by the
low-speed reader or high-speed reader. However, the five commands
that go into Text Mode enable the user to input from the keyboard.
The Insert command, in particular (Section 4.3.4.1) can be useful for
entering large quantities of text not on paper tape.
The Page Buffer
can thus be filled from the keyboard, and a paper tape actually
created by using a command to punch out the buffer contents.

4-11

EDITING THE SOURCE PROGRAM
4.4.5

Editing Example

The following example consists of three parts:
1.

The marked up source program listing indicating
changes.

the

The ED-II commands to implement those changes (with
on the editing procedure).

desired
comments

NOTE
Typing the RETURN key terminates Command Mode in all
cases.
In commands which then go into Text Mode,
typing the LINE FEED key (symbolized as +) produces
the terminator.
3.

Part I

Th~

edited text.

Original Source for Edit

;COMMON INPUT ROUTINE FOR USE BY NON FILE DEVICES
UNPI.J'T": ADC
CLR
MOV
MOV
~;CI\MODE!

BITB
BNE

!I;CKNUL: TSTB
BEt~

$CI

take contents of opened location as relative branch
instruction and open referenced location (ODT-llX only)

return to sequence prior to last @, >, or
open succeeding location (ODT-IlX only)

$n/

open general register n (0-7)
separates commands from command
alphabetic commands below)

niB

remove Breakpoint(s)
(see
version for particulars)

command and

arguments

description

address

(used
of

each

with
ODT

set Breakpoint at location n

lThe circumflex appears on some keyboards and printers in place of the
up-arrow.
2The underline appears on some keyboards and printers in place of
back-arrow.

5-2

the

DEBUGGING OBJECT PROGRAMS ON-LINE
n;rB

set Breakpoint r at location n (ODT-IIX only)

;rB

remove r(th) Breakpoint

n;E

search for
address n

n;W

search for Words with bit patterns which match n

inS

enable Single-instruction mode (n can have any
and is not significant); disable breakpoints

disable Single-instruction mode

n;G

Go to location n and start program run

Proceed with program execution from breakpoint;
stop
when next breakpoint is encountered or at end of
program

(ODT~llX

instructions

that

only)
reference

In Single-instruction mode only (ODT-IIX),
execute next instruction only
niP

value

Proceed

to
stop

Proceed with program execution from breakpoint;
after encountering the breakpoint n times.
In single-instruction mode only (ODT-IIX),
execute next n instructions.

n/(word)n;O

Effective

Proceed

calculate Offset from location n to location m

$B/

open Breakpoint status word (ODT-ll)
open BREAKPOINT D STATUS WORD (ODT-IIX)

$M/

open search Mask

$S/

open location containing user program's Status register

$P/

open location containing ODT's Priority level

With ODT-ll, location references must be to even numbered 16-bit
words.
With ODT-IIX, location references may be to 16-bit words or
8-bit bytes.
The semicolon in the above commands is ignored by ODT-ll, but is used
for the sake of consistency, since similar commands to ODT-IIX require
it.

5.2

COMMANDS AND FUNCTIONS

When ODT is started as explained in section 5.6, it indicates its
readiness to accept commands by printing an asterisk (*) on the left
margin of the terminal paper. In response to the asterisk, you can
issue most commands;
for example, you can examine and, if desired,
change a word, run the object program in its entirety or in segments,
or even search core for certain words or references to certain words.
The discussion below first explains some elementary features, and then
covers the more sophisticated features.
All commands to ODT are issued using the characters and symbols
above in Section 5.1.2.

5-3

shown

DEBUGGING OBJECT PROGRAMS ON-LINE
5.2.1

Opening, Changing, and Closing Locations

An open location is one whose contents ODT has
printed
for
examination, and whose contents are available for change. A closed
location is one whose contents are no longer available for ch~nge.
Any even-numbered location may be opened using ODT-ll.
The contents of an open location can be changed by typing the new
contents followed by a single character command, which requires no
argument (i.e. -} t RETURN + @ > <). Any command typed to open a
location when another location is already open causes the currently
open location to be closed.

5.2.1.1 The Slash (/) - One way to open a location
address followed by a slash:

type

its

:'1000/012746
Location 1000 is open for examination and is available for change.
Note that in all examples ODT's printout is underlined~ your typed
input is not.
Should you not wish to change the contents of an open location, merely
type the RETURN key and the location will be closed; ODT prints
another asterisk and waits for another command. However, should you
wish to change the word, simply type the new contents before giving a
command to close the location.
*1000/012746

012345

In the example above, location 1000 now contains 012345 and is closed
since the RETURN key was typed after entering the new contents, as
indicated by ODT's second asterisk.
Used alone, the slash reopens the last location opened:
*1000/012345

~/002340

2340

As shown in the example above, an open location can be closed by
typing the RETURN key.
In this case, ODT changed the contents of
location 1000 to 002340 and then closed the location before printing
the *
A single slash then directed ODT to reopen the last location
opened. This allowed us to verify that the word 002340 was correctly
stored in location 1000.
(ODT supplies the leading zeroes if not
given.)
Note again that opening a location while another is currently open
automatically closes the currently open location before opening the
new location.

5.2.1.2 The LINE FEED Key - If the LINE FEED key is typed when a
location is open, ODT closes the open location and opens the next
sequential location:
*1000/002340 -}
001002/012740

( + denotes typing the LINE FEED key)

5-4

DEBUGGING OBJECT PROGRAMS ON:....LINE·
In this example, the LINE FEED key instructed DDT to print the" addres~
of the next location along with its contents and to wait for furth~r
instructions. After the above operation, location 1000 is closed and
1002 is open.
The open location may be modified by typin~ the new
contents.

5.2.1.3 The Up-Arrow(t) - The up-arrow (or curcumflex)
symbol is
effected by typ1ng the SHOFT and N key combination. If the up-arrow
is typed when a location is open, ODT closes the open location and
opens the previous location (as shown by continuing" from the example
above) :
001002/012740
001000/002340

t (t is printed by typing SHOFT and N)

Now location 1002 is closed and 1000 is open.
be modified by typing-the new contents.

The

o~en

location

may

5.2.1.4 The Back-Arrow(+) - The back-arrow (or underline) symbol is
effected by typing the SHIFT and 0 key combination. If the back-arrow
is typed to an open location, ODT interprets the contents of the
currently open location as an address indexed by the Program Counter
(PC) and opens the location so addressed~
*1006/000006
001016/100405

( + is printed by typing SHIFT and 0)

Notice in this example that the open location(1006) was indexed by the
PC as if it were the operand of an instruction with address mode 67 as
explained in Chapter 3.
A modification to the opened location can be made before a + , t , or +
is typed.
Also, the new contents of the location will be used for
address calculations using the
command. Example:
*100/000222
4+
000102/000111 6t
000100/000004
100+
000202/(contents)

(modify to 4 and open next location)
(modify to 6 and open previous location)
(change to 100 and open location indexed
by PC)

5.2.1.5 Accessing General Registers 0-7 - The
program's
general
registers 0-7 can be opened using the following command format:
~$n/

where n is the integer representing the desired register (in the range
o through 7). When opened, these registers can be examined or changed
by typing in new data as with any addressable location. For example:
~$0/000033

(RO was examined and closed)

*
and
*$4/000474 464
1<

(R4 was opened, changed, and closed}

DEBUGGING. OBJECT PROGRAMS- ON-LINE
The example above can be verified by typing a
ODT's ast.er isk:

slash

response

~/000464

The .j. , t , + , or @ commands may be used when a register is open (the
is an ODT-IIX command).

5.2.1.6 Accessing Internal Regist~rs. - The program's Statu& Register
contains the condition codes of the most recent operational results
and the interrupt priority level of the object program. It is opened
using the following command:
~$S/000311

where $S represents the address of the Status Register~
In response
to $S/ in the example above, ODT printed the 16~bit (of which only the
low-order 8 bits are meaningful): Bits 0-3 indicate whether a carry,
overflow, z~ro, or negative (in that order) has resulted, and bits 5-7
indicate the interrupt priority level (in the range 0-7) of the object
program.
The $ is used to open certain other internal locations:
$B

internal breakpoinb status word (see section 5.2.2.2)

mask location for specifying which bits are to be
examined during a bit pattern search (see section
5.2.4)

.location defining the operating priority
section 5.2.6)

containing the condition codes (bits
interrupt priority level (bits 5-7)

locatio~

ODT
0~3)

(see
and

Breakpoints

5.Z~2

The breakpoint feature facilitates monitoring the progress of program
execution.
A breakpoint may be set at any instruction which is not
referenced by the program for data. When a breakpoint is set, ODT
replaces
the contents of the breakpoint location with a trap
instruction. Thus, when the program is executed and the breakpoint is
encountered, program exeeution is suspended, the original contents of
the breakpoint location are restored, and ODT regains control.

only
one
5.2.2.1 .Setting the Breakpoint(n:B) - ODT-11
provides
breakpoint:
ODT-IIX provides eight. Breakpoint(s) may be changed at
any time. A breakpoint is set by typing the address of the desired
location of the breakpoint followed by :B. For example::
~1020:B

*
sets a breakpoint at location 1020.
to location 1120 as shown below.
*1020·B
,

-*1120 ,·B
*

5-6

The breakpoint above. is

changed

DEBUGGING OBJECT PROGRAMS ON-LINE
Breakpoints should not be set at locations referenced by the program
for data, nor at an lOT, EMT, or TRAP instruction. This restriction
is explained in section 5.5.2.
The breakpoint is removed by typing ;B without an argument,
below.
~1l20

*;B
"*

shown

(sets breakpoint at location 1120)
(removes breakpoint)

5.2.2.2 Locating the Breakpoint($B) - The command $B/ causes ODT-ll
to print the address of the breakpoint (see also section 5.3.3 on $B
in ODT-IIX):
~$B/001120

The breakpoint was set at location 1120. $B represents the address
containing ODT-ll's breakpoint location. Typing the RETURN 'key in the
example above leaves the breakpoint at locatidn 1120 and returns
control to ODT-ll.
The breakpoint could be changed to a different
location:
*$B/001120
"*$B/0011l4

1114

*
The breakpoint was found in location 1120, changed to
and the change was verified.

location

1114,

If no breakpoint is set, $B contains an address internal to ODT-ll.

5.2.3

Running the Ptogram(n;G and niP)

Program execution is under control of ODT. There are two commands for
running the program: n;G and niP. The n;G command is used to start
execution (GO) and niP to .continue (Proceed) execution after halting
at a breakpoint. For example:
*IOOO;G
starts execution at location 1000.
The program runs until it
encounters a breakpoint or until program completion.
If the program
enters an infinite loop,
it must be restarted or reentered as
explained in section 5.6.2.
When a breakpoint is encountered, execution stops and ODT-ll prints B;
followed by the address of the breakpoint. Desired locations can then
be examined for expected data. For example:
*IOIO;B
*IOOO;G
B;OOIOIO

(breakpoint is set at location 1010)
(execution started at location 1000)
(execution stopped at location 1010)

*
To continue program execution from the breakpoint, type ;P in response
to ODT-ll's last *.
When a breakpoint is set in a loop, it may be desirable to allow the
program to execute a certain number of times through the loop before
5-7

DEBUGGING OBJECT PROGRAMS

ON~LINE

recognizing the breakpoint. This may be, done by typing the n;P
commaI;ldand specifying the number of times the breakpoint is to be
encountered before program execution is suspended
(on then(th)
encounter).
(See section 5.3.3 for ODT-IIX interpretation of this
command When more than one breakpoint is set in a loop.)
Example:
B;OOlOlO

(execution halted at breakpoint)
(set breakpoint at location 1250)
(continue execution. loop through
breakpoint 3 times and halt on the
4(th) occurrence of the breakpoint)

~1250;B

-*4'P
,

B;001250

The breakppint repeat count caI;l be inspected by typing $B/ followed by
LINE FEED.
The repeat count is then printed. This also provides an
alternative way of specifying the count. Since the location is open,
its contents can be 'modified in the usual manner by typing new
contents followed by the RETURN key.
*$B/001114 +
nnnnnn/000003

(address of breakpoint .is 1114)
.(repeat count wa$ 3, changed to 6)

Breakpoints are inserted when performing an n;G or n;P command.
Upon
execution of the n;G or n;P command, the general registers 0-6 are set
to the values in the locations specified as $0-$6 and the processor
status register is set to the value in the location specified as $S.

5.2.4

Searches

wi th ODT yOI,l can ,search all or any specyfied portion of core ,memory
for any specific bit pattern or for references to a specific location.
The location represented by $M is used to specify the mask of the
search. The next two sequential locations contain the lower and upper
limits of the search. Bits set to 1 in the mask are examined during
the search; other bits are ignored. For example,
*$M/OOOOOO

177400
1000
nnnnnn/OOOOOO 1040
~nnnnn/OOOOOO

(+ denotes typing LINE FEED)
(starting address of search)
(last address in search)

*
where nnnnnn represents some location in ODT.
This location varies
and,is meaningful only for reference purposes. Note that in the first
line above, the slash was used to open $M which now contains 177400,
and that the LINE FEEDs opened the next two sequential locations which
now contain the lower and upper limits of the search.

5.2.4.1 Word Search(n;W) - Before initiating a word search, the mask
and search, limits must be specified as explained above. Then the
search object and the initiating comm~nd are given using the n;W
command where n is the search object. When a match is found, the
address of the unmasked matching word is printed. For example:
*$M/000000177400 +
1000 +
nnnnnn/OOOOOO 1040
*400;W
001010/000770
001034/000404
~nnnnn/OOOOOO

(test high order eight bits)
(initiating word search)

5-8

DEBUGGING OBJECT PROGRAMS ON-LINE
the
In the search process, the word currently being examined and
search object are exclusive ORed (XORed), and the result is ANDed to
the mask.
If this result is zero, a match has been found,
and
is
reported at the terminal.
Note that if the mask is zero, all
locations within the limits are printed.

5.2.4.2 Effective Address Search(njE) - ODT enables you to search for
words which address a specified location.
After specifying the search
limits (section 5.2.4), type niE (where n is the effective address) to
initiate the search.
Words which are either an absolute address
(argument n
itself) I
a
relative address offset, or a relative branch to the effective address
are printed after their addresses.
For example:
'::$M/177400
1nnnnnn/OOlOOO
nnnnnn/00l040
*1034,E
001016/001006
001054/002767
*1020'E
001022/177774
001030/001020
!

1010 11060
(initiating search)
(relative branch)
(relative branch)
(initiating a new searhc)
(relative address offset)
(absolute address)

Particular attention should be given to the reported references to the
effective address because a word may have the specified bit pattern of
an effective address without actually being so used.
ODT will
report
these as well.

5.2.5

Calculating Offsets(n,O)

Relative addressing and branching use an offset - the number of wrods
or bytes forward or backward from the current location of the
effective address.
During the debugging session it may be necessary
to change a
relative address or branch reference by replacing one
instruction offset with another.
ODT calculates the offsets
in
response to the n;O command.
The command niO causes ODT to print the l6-bit and 8-bit offsets from
the currently open location to address n.
In ODT-ll, the 8-bit offset
is printed as a 16-bit word.
For example:
*346/000034
*/000022
*20/000046
~20/000067

414,0

000044

000022

200,0

000156

000067

In the first example, location 346 is opened and the offsets from that
location to location 414 are calculated and printed.
The contents of
location 346 are then changed to 22 and verified on the next line.
The l6-bit offset is printed followed by the 8-bit offset.
In the
example above, 000156 is the 16-bit offset and 000067 is the 8-bit
offset.
The 8-bit offset is printed only if the 16-bit offset is even,
as
in
the case above. with ODT-ll only, the user must determine whether the
8-bit offset is out of the range 177600 to 000177 (-128 decimal to 127
decimal) .
The offset of a relative branch is calculated and modified
as follows:
5-9

DEBUGGING OBJECT PROGRAMS ON-LINE
*1034/103421

1034;0

177776

177777

103777

Note that the modified low-order byte 377 must be combined with the
unmodified high-order byte.
Location 1034 was still open after the
calculation, thus typing 103777 changed its contents;
the location
was then closed.

5.2.6

ODT'S Priority Level ($P)

$P represents a location in ODT that contains the priority level at
which ODT operates.
If $P contains the value 377, ODT operates at the
priority level of the processor at the time ODT is entered. Otherwise
$P may contain a value between 0 and 7 corresponding to the fixed
priority at which ODT operates.
To set ODT to the desired priority level, open
present contents, which may then be changed:
*$P/000006

$P.

ODT

prints

the

377

*
If $P is not specified, its value is seven.
Breakpoints may be set in routines at different priority levels.
For
example, a program running at a low priority level may use a device
service routine operating at a higher priority level.
If a breakpoint
occurs from a low priority routine, if ODT operates at a low priority,
and if an interrupt does occur from a high priority routine, then the
breakpoints in the high priority routine will not be executed since
they have been removed.

'5;.3

ODT-llX

ODT-IIX. has all the commands and features of ODT-ll
section 5.2, p~us the following.

5.3.1

explained

Opening, Changing and Closing Locations

In addition to operating on words, ODT-IIX operates on bytes.
One way to open a byte is to type the address of the byte followed
a backslash:
~1001/025

(\ is printed by typing SHIFT and L)

A backslash typed alone reopens the last open byte.
If
previously open, the backslash reopens its even byte.

word

was

~1002/000004\004

The LINE FEED and up-arrow (or circumflex) keys operate on bytes if
byte is open when the command is given. For example:
*1001\025
001002\004
001001\025

+
t

:!:,

5-10

DEBUGGING OBJECT PROGRAMS ON-LINE
5.3.1.1 Open the Addressed Location(@) - The symbol
@ optionally
modifies, closes an open word, and uses its contents as the address of
the location to open next.
*1006/001024
001024/000500
*1006/001024
002100/177774

(open location 1024 next)

2100 @

(modify to 2100 and open
location 2100)

5.3.1.2 Relative Branch Offset(» - The right angle
bracket,
>,
optionally modifies, closes an open word, and uses its even byte as a
relative branch offset to the next word opened.
*1032/000407
000636/000010

(modify to 301 and interpret
as a relative branch)

301

doubled

Note that 301 is a negative offset
(-77).
The offset is
before it is added to the PC; therefore, 1034 + -176 = 636.

5.3.1.3 Return to Previous Sequence«) - The left angle bracket, <,
optionally modifies, closes an open location, and opens the next
location of the previous sequence interrupted by a + , @, or > command.
Note that + , @, > cause a sequence change to the word opened.
If a
sequence change has not occurred, < simply opens the next location as
a LINE FEED does. The command operates on both words and bytes.
*1032/000407 301
000636/000010 <

(> causes a sequence change)
«causes a return to original
sequence)
(@ causes a sequence change)
« now operates on byte)
« acts like -I- )

001034/001040 @
001040/000405\005 <
001035\ 002 <
001036\ 004

5.3.2

Calculating Offsets (n;O)

The command n;O causes ODT to print the 16-bit and 8-bit offsets from
the currently open location to address n. The following examples,
repeated from the ODT-ll section describing this command (see section
5.2.5), show a difference only in printout format:
*346/000034

~/000022

*1034/103421
YI03777

414;0
1034;0

~0~00~0~4~4~~0~2~2

177776

377\021

377

Note that the modified low-order byte 377 must be
unmodified high-order byte.

5-11

combined

with

the

DEBUGGING OBJECT PROGRAMS ON-LINE
5.3.3

Breakpoints

with ODT-IIX you can set up to eight breakpoints concurrently,
numbered 0 through 7.
The niB command used in ODT-ll to set the
breakpoint at address n sets the next available breakpoint in ODT-IIX.
Specific breakpoints may be set or changed by the n:mB command where m
is the number of the breakpoint.
For example:
~1020;B

(sets breakpoint 0)
(sets breakpoint 1)
(sets breakpoint 2)
(resets breakpoint 1)

~1030iB

*1040;B
*1032;lB

The ;B command used in ODT-ll to remove the only breakpoint removes
all breakpoints in ODT-IIX.
To remove only one of the breakpoints,
use the inB command, where n is ~he number of t~e break~oint.
For
example:
(removes the second breakpoint)

*'2B
-'

The $B/ command opens the location containing the address
of
breakpoint O.
The next seven locations contain the addresses of the
other breakpoints in order, and thus can be opened using the LINE FEED
key.
(The next location is for single-instruction mode, explained in
the next section.) Example:
*$B/00I020 +
nnnnnn/00l032 +
nnnnnn/(address internal to ODT)
In this example, breakpoint 2 is not set. The contents are an address
internal to ODT.
After the table of breakpoints is the table ot
Proceed command repeat counts for
each breakpoint and for
t~e
single-instruction mode (see Section 5.3.4).

nnnnnn/00l036
nnnnnn/nnnnnn
nnnnnn/OOOOOO
nnnnnn/OOOOOO

+
+
+
15

(address of breakpoint 7)
(single-instruction address)
(count for breakpoint 0)
(count for breakpoint 1)

It should be noted that a repeat count in a Proceed command refers
only to the most recent breakpoint.
Execution of other breakpoints
encountered is determined by their own repeat counts.

5.3.4

Single-Instruction Mode

With this mode you can specify the number of instructions you wish
executed before suspension of the program run.
The Proceed command,
instead of specifying a repeat count for
a breakpoint encounter,
specifies the number of succeeding instructions to be executed. Note
that breakpoints are disabled when single-instruction
mode
is
operative. Commands for single-instruction mode follow:
inS

Enables single-instruction mode (n can have any value
and serves only to distinguish this form from the form
is) i
breakpoints are disabled.

5-12

DEBU~GING

OBJECT PROGRAMS

ON-L~NE

niP

Proceeds with program run for next.n instructions
before reeritering OPT (if n is missing. it is assumed
to be 1). . (Trap instructionsrand' associated' handlers
can affect the Proceed repeat count.
See section
5.5.2.)

Disables single-instruction mode

When the repeat count for single-instruction mode is exhausted and the
program suspends execution, ODT prints:

where n is the address of the next instruction to be executed. The $B
breakpoint table contains this address following. that of, breakpoint 7.
However, unlike the table entries for breakpoints 0-7, the B8 entry is
not affected by direct modification.
Similarly, the repeat count for single-instruction mode follows the
repeat count for breakpoint 7.
This table entry, however, may be
directly modified, and thus is an alternative way of setting the
single-instruction mode repeat. count. In such a case, iP implies the
argument set in the $B repeat count table rather than the argument 1.

5.4

ERROR DETECTION

ODT-ll and ODT-IIX inform you of two types of errors:
unrecognizable command and bad breakpoint entry.

illegal

Neither ODT-ll nor ODT-IIX checks for the legality of an address
commanded to open a location for examination or modification.

or
when

Thus, the command
177774/
references nonexistent memory, and causes a trap through the vector at
location 4.
If this vector has not been properly initialized (by IOX~
or the user program if IOX is not used), unpredictable results occur.
Similarly, a command such as
$20/
which references an address eight times the value represented
may cause an illegal (nonexistent) memory reference.
Typing other than a legal command causes ODT to
print

ignore

the

$2,

command,

*
and wait for another command. Therefore, to cause ODT to ignore a
command just typed, type an illegal character (such as 9 or RUBOUT)
and the command will be treated as an error, i.e., ignored.
ODT suspends program execution whenever it encounters a breakpoint,
i.e., a trap to its breakpoint routine. If the breakpoint routine is
entered and no known breakpoint caused the entry, ODT prints:
BE001542

5-13

DEBUGGING OBJECT PROGRAMS ON-LINE
and waits for another command.
In the example above, BE001542 denotes
Bad Entry from location 001542.
A bad entry may be caused by an
illegal .trace trap instruction, setting the T-bit in the status
register, or by a jump to the middle of ODT.

5.5

PROGRAMMING CONSIDERATIONS

Information in this section is not necessary for the efficient use of
ODT.
However, its content does provide a better understanding of how
ODT performs some of its functions.

5.5.1

Functional Organization

The internal organization of ODT is almost totally modularized into
independent subroutines.
The internal structure consists of three
major functions: command decoding, command execution, and various
utility routines.
The command decoder interprets the individual commands, checks for
command errors, saves input parameters for use in command execution,
and send control to the appropriate command execution routine.
The command execution routines take parameters saved by the command
decoder and use the utility routines to execute the specified command.
Command execution routines exit either to the object program or back
to the. ".command decoder.
The utility routines are common routines such as SAVE-RESTORE and I/O.
They are used by both the command decoder and the command executers.
Communication and data flow are illustrated in Figure 5-1.

5.5.Z

Breakpoints

The function of a breakpoint is to pass control to ODT whenever the
user program tries to execute the instruction at the selected address.
Upon encountering a breakpoint, the user can utilize all of the ODT
commands to examine and modify his program.
When a breakpoint is
executed,
ODT
removes
the
breakpoint
instruction(s)
from the user's code so that the locations may be
examined and/or altered. ODT then types a message to the user, in the
form Bn(Bm;n for ODT-IIX), where n is the breakpoint address (and m is
the breakpoint number). The breakpoints are automatically restored
when execution is resumed.
A major restriction in the use of breakpoints is that the word

5-14

DEBUGGING OBJECT PROGRAMS ON-LINE

MANUAL
ENTRY

BREAKPOINT
HANDLER

COMMAND

L--rD_E_C.,.O_DE_R..,.---,~

~--------~I~
I ________~

PROGRAM
EXAMINATION
MODIFICATION
COMMANDS

PROGRAM
ACTION
COMMANDS

USER

INTERNAL
TABLE MAINPULATION
COMMANDS

--~--------~---~

r-_=t_--

ODT
INTERNAL
TABLES

PROGRAM

-------~--------~-

UTILITY
_ _ (I/O,
_
RO_U_T_I_N_E_S
ETC.) __
L..

~ O~i~UT

ODr

USER ENVIRONMENT
LEGEND

Flow of control - - Flow of dolo
11-0065

Figure 5'-1

Communication and Data Flow

5-15

DEBUGGING OBJECT PROGRAMS ON-LINE
where a breakpoint has been set must not be referenced by the program
in any way since ODT has altered the word.
Also, no breakpoint should
be set at the location of any instruction that clears the T-bit.
For
example:
MOV :It 2 4 a , 1 77 77 6

;SET PRIORITY TO LEVEL 5.

A breakpoint occurs when a trace trap instruction (placed in the user
program by ODT) is executed. When a breakpoint occurs, ODT takes the
following steps:

Set processor priority to seven
instruction) .

Save

If internal T-bit trap flag is set, go to step 13.

Remove breakpoint(s).

Reset processor
priori ty.

Make sure a breakpoint or Single-instruction mode caused
interrupt.

If the breakpoint did not cause the interrupt, go to step 15.

Decrement repeat count.

Go to setp 18 if non-zero, otherwise reset count to one.

registe~s

(automatically

set

and set up stack.

priority

ODT's

priority

10.

Save Teletype status.

11.

Type
message
to
user
about
Single-instruction mode interrupt.

12.

Go to command decoder.

13.

Clear T-bit in stack and internal T-bit flag.

14.

Jump to the "GO" processor.

15.

Save Teletype status.

16.

Type "BE"

17.

Clear the T-bit, if set, in the user status
the command decoder.

18.

Go to the "Proceed", bypassing the TTY restore routine.

the

user's

breakpoint

the

(Bad Entry) followed by the address.

Note that steps 1-5 inclusive take
during which time interrupts are
running at level 7).
When a proceed

trap

and

proceed

approximately 100 microseconds
not permitted to occur (ODT is

(;P) command is given, the following occurs:

The proceed is checked for legality.

The processor priority is set to seven.

The T-bit flags

(internal and user status) are set.

DEBUGGING OBJECT PROGRAMS ON-LINE
4.

The user registers, status, and Program Counter are restored.

Control is returned to the user.

When the T-bit trap occurs, steps 1, 2, 3, 13, and 14 of the
breakpoint sequence are executed, breakpoints are restored,
and program execution resumes normally.

When'a breakpoint is placed on an lOT, EMT, TRAP, or
causing a trap, the following occurs:

any

instruction

When the
entered.

When iP is typed, the T-bit is set and the lOT, EMT, TRAP, or
other trapping instruction is executed.

The current PC and status
pushed on the stack.

The new PC and status (no T-bit set) are
respective trap vector.

The whole trap
breakpoints.

When an RTI is executed, the saved PC and PS
(including the
T-bit)
are
restored.
The
instruction following the
trap-causing instruction is executed.
If this instruction is
not another trap-causing instruction, the T-bit trap occurs,
causing the breakpoints to be reinserted in the user program,
or
the
Single-instruction
mode
repeat
count to be
decremented.
If the following instruction is a trap-causing
instruction, this sequence is repeated, starting at step 3.

breakpoint

occurs

service

(with

routine

described

the

above,

T-bit

included)

obtained

executed

ODT

are

from

the

without

any

NOTE
Exit from the trap handler must be via the RTI
instruction.
Otherwise, the T-bit will be lost. ODT
will not gain control again since the breakpoints
have not been reinserted yet.
In ODT-ll, the iP command is illegal if a breakpoint has not occurred
(ODT responds with?).
In ODT-IIX, iP is legal after any trace trap
entry.
WARNING
Since ODT-ll ignores all semicolons,
typing the ODT-IIX form of breakpoint
command number to ODT-ll, specifying a
breakpoint
number
n,
causes
the
following error:

100iB

(sets the breakpoint at location
100)
breakpoint
at
100iOB (sets
the
location
1000)
100i4B (sets the breakpoint at location
1004)

5-17

DEBUGGING OBJECT PROGRAMS ON-LINE
The internal breakpoint status words for
format:

ODT-ll

have

the

following

The first word contains the breakpoint address.
If this
location points to a location within ODT, it is assumed no
breakpoint is set for the cell (specifically, ODT has set a
dummy breakpoint within itself).

The next word contains the breakpoint repeat count.

For ODT-IIX (with eight breakpoints) the formats are:
1.

The first eight words contain the breakpoint addresses for
breakpoints 0-7.
(The ninth word contains the address of the
next instruction to be executed in Single-instruction mode.)

The next eight words contain the respective repeat counts.
(The
following
word
contains
the
repeat count for
Single-instruction mode.)

The user may change these words at will,
either
breakpoint commands or by direct manipulation with $8.

using

the

When program runaway occurs (that is, when the program is no longer
under ODT control, perhaps executing an unexpected part of the program
where a breakpoint has not been placed) ODT may be given control by
pressing the HALT key to stop the machine, and restarting ODT (see
Section 5.6.2). ODT prints *, indicating that it is ready to accept a
command.
If the program being debugged uses the terminal for input or output,
the program may interact with ODT to causes an error since ODT also
uses the terminal.
This interactive error does not occur when the
program being debugged is run without ODT.

5.5.3

If the terminal output interrupt is
ODT break routine,
and no output
ODT is entered, ODT is entered, ODT
interrupt when returning control to

enabled upon entry to the
interrupt is pending when
generates an unexpected
the program.

If the interrupt of the terminal input
(the keyboard)
is
enabled upon entry to the ODT break routine, and the program
is expecting to receive an interrupt to input a character,
both the expected interrupt and the character will be lost.

If the terminal input (keyboard) has just read a character
into the reader data buffer when the ODT break routine is
entered, the expected character in the input data buffer will
be lost.

The word search allows the user
to search for bit patterns in
specified sections of memory.
Using the
$M/ command,
the user
specifies a mask, a lower search limit ($M+2),
and an upper search
limit
($M+4).
The search object is specified in the search command
itself.
The word search compares selected bits (where ones appear in the mask)
in the word and search object.
If all selected bits are equal, ODT
prints the unmasked word.

5-18

DEBUGGING OBJECT PROGRAMS ON-LINE
The search algorithm is:
1.

Fetch a word at the current address.

XOR (exclusive OR)

AND the result of step 2 with the mask.

If the result of step 3 is zero,
type the address of the
unmasked word and its contents.
Otherwise, proceed to step
5.

Add two to the current address.
If the current address is
greater than the
upper limit,
type
* and return to the
command decoder, otherwise go to step 1.

the word and search object.

Note that if the mask is zero,
ODT prints every word between the
limits, since a match occurs every time (i.e., the result of step 3 is
always zero).
In the effective address search, ODT interprets every word
in the
search range as an instruction which is interrogated for a possible
direct relationship to the search object.
The algorithm for the effective address search is
contents of x, and k denotes the search object):

(where

(x)

denotes

Fetch a word at the current address X.

If (x)+x+2=k
5.

If (x) is a relative branch to k, print contents.

Add two to the current address.
If the current address
is
greater than the upper limit, perform a carriage return/line
feed and return to the command decoder;
otherwise, go to
step 1.

5.5.4

(x)=k

[direct reference], print contents and go to step 5.
[indexed by PC]. print contents and go

Teletype Interrupt

Upon entering the TTY SAVE routine, the following occurs:
1.

Save the LSR status register

(TKS).

Clear interrupt enable and maintenance bits in the TKS.

Clear interrupt enable and maintenance bits in the TPS.

To restore the TTY:
1.

wait for completion of any I/O from ODT.

Restore the TKS.

Restore the TPS.

5-19

step

DEBUGGING OBJECT PROGRAMS ON-LINE
NOTES
If the TTY printer interrupt is enabled
upon entry to the ODT break routine, the
following may occur:
1.

If no output interrupt is pending
when ODT is entered, an additional
interrupt always occurs when ODT
returns control to the user.

If an output interrupt is pending
upon entry,
the expected interrupt
occurs
when
the
user
regains
control.

If the TTY reader (keyboard) is busy or
done,
the expected character in the
reader data buffer will be lost.
If the TTY reader
(keyboard)
interrupt
is enabled upon entry to the ODT break
routine, and a character is pending, the
interrupt
(as well as the character)
will be lost.

5.6

OPERATING PROCEDURES

This section describes procedures for linking ODT on LSI-II machines,
and for loading ODT on other PDP-II machines.
It describes starting,
restarting, error recovery, and setting the priority level of ODT.

5.6.1

Linking Procedures (LSI-II Systems Only)

For LSI-II systems, ODT-IIX is supplied on relocatable object tapes.
Binary tapes are produced by linking the ODT-IIX object tape with the
object tapes of the program to be debugged
(using LINK-lIS).
The
ODT-IIX tape should be the first tape processed by LINK-lIS;
in this
manner, ODT-IIX is started first when the binary tape is loaded.

5.6.2

Loading Procedures (non-LSI-ll Systems Only)

For all systems other than LSI-II, ODT is supplied on source and
binary tapes.
Appendix N explains assembly instructions for source
tapes. Binary tapes are loaded with the Absolute Loader.
Since ODT
is started as soon as it is loaded, the program to be debugged should
be loaded prior to ODT.
When supplied on binary tape, ODT-ll loads beginning at location
13026, and occupies about 533
(decimal) words of memory.
ODT-IIX
loads beginning at location 12054, and requires about 800
(decimal)
words of memory.

5-20

DEBUGGING OBJECT PROGRAMS ON-LINE
5.6.3

Starting and Restarting

The Absolute Loader starts ODT automatically after loading it into
core. ODT indicates its readiness to accept input by printing an *
The starting address for ODT-ll on binary tape is 13026; the starting
address for ODT-llX on binary tape is 12054. If ODT is reassembled
using PAL-IIA, the starting address in indicated in the symbol table
as the value of the symbol O.ODT.
If ODT is linked using LINK-lIS,
the starting address is indicated in the link map as the value of the
global symbol O.ODT.
When ODT is started at its start address, the SP register is set to an
ODT internal stack, registers RO-R5 are left untouched, and the trace
trap vector is initialized.
If ODT is started after breakpoints have
been set in a program, ODT ignores the breakpoints and leaves the
program modified, i.e., the breakpoint instructions are left in the
program.
There are two ways to restart ODT:
1.

Restart at start address+2

Reenter at start address+4

To restart, key in the start address+2, press LOAD ADDRess and then
START.
A restart saves the general registers, removes all the
breakpoint instructions from the user program and then ignores all
breakpoints, i.e., simulates the ;B command.
To reenter, key in the load address+4, press LOAD ADDRess and then
START.
A reenter saves the general registers, removes the breakpoint
instructions from the user program, and types the BE (Bad Entry) error
message.
ODT remembers which breakpoints were set and resets them on
the next ;G command (iP is illegal after a Bad Entry).

5-21

CHAPTER 6
LOADING AND DUMPING MEMORY

This chapter describes procedures for loading programs into memory
(using the Bootstrap Loader and Absolute Loader) and for dumping the
contents of memory (using the DUMPAB and/or DUMPTT programs).
The Bootstrap Loader, which loads short paper tape programs
(162 or
fewer octal words), appears on one of three forms, depending upon the
system configuration:
1.

Hardware - on some CPUs, the Bootstrap Loader is present as a
ROM chip.

Software - on some CPUs, the Bootstrap Loader must be toggled
in via console switches.

Firmware - on LSI-lIs, the Bootstrap Loader
loader, present as a programmable ROM chip.

firmware

Once familiar with the operation of the Bootstrap Loader, the user can
load other programs (such as the Absolute Loader, DUMPAB, and DUMPTT).
The Absolute Loader (see section 6.2) is a system program that enables
the user to load data punched on paper tape in absolute binary format
into any available memory bank.
It is used primarily to load the
paper tape system software, binary programs assembled with PAL-llA,
and binary tapes produced by LINK-lIS from object tapes produced by
PAL-llS.
The loader programs are loaded into the upper-most area of available
core and are available for use with system and user programs.
Programs should not use the locations used by the loaders without
restoring their contents;
otherwise,
the loaders must be reloaded
since they will have been altered by the object program.
Core memory dump programs
(see section 6.3)
print or punch the
contents of specified areas of core.
For example, when developing or
debugging user programs it is often necessary to get a copy of the
program or portions of core. There are two dump ptograms supplied in
the paper tape software system: DUMPTT, which prints or punches the
octal representation of specified portions of core, and DUMPAB, which
punches specified portions of core in absolute binary format suitable
for loading with the Absolute Loader.

6-1

LOADING AND DUMPING MEMORY
6.1

PAPER TAPE BOOTSTRAPS

Procedures for operating the various PDP-II paper tape bootstraps
described below:

6.1.1

are

BM792-YA Paper Tape Bootstrap ROM
1.

Set the console ENABLE/HALT switch to HALT.

Place the bootstrap tape in the desired paper tape reader
with the special bootstrap leader code over the reader
sensors (under the reader station).

If the low-speed reader
(ASR-33)
is to be used,
and a
high-speed reader
is, present on the system, turn the high
speed reader OFF.
If the high-speed reader is to be used,
turn it ON.

Set the console ENABLE/HALT switch to ENABLE.

Set the console switch register to 773000.

Press the console START switch.
The contents of
the
bootstrap tape will be loaded into the highest locations of
memory.

The bootstrap transfers control to the program
Typically, this program halts.

6.1.2

just

loaded.

BM873-YA Bootstrap Loader ROM
1.

Set the console ENABLE/HALT switch to HALT.

Place the bootstrap tape in the desired paper tape reader
with the special bootstrap leader code over the reader
sensors (under the reader station).

and a
If the low-speed reader
(ASR-33)
is to be used,
turn the
high-speed
reader
is present on the system,
high-speed reader OFF.
If the high-speed reader
is to be
used, turn it ON.

Set the console ENABLE/HALT switch to ENABLE.

Sa. If the low-speed reader is to be used, set the console switch
register to 773210.
Sb. If the high-speed reader is
switch register to 773312.

used,

set

the

console

Press the console START switch.
The contents of
the
bootstrap tape will be loaded into the highest locations of
memory.

The bootstrap transfers control to the program
Typically, this program halts.

6-2

just

loaded.

LOADING AND DUMPING MEMORY
6.1.3

LSI-II Firmware Paper Tape Bootstrap
1.

Press the front panel BOOT/INIT switch.
micro-ODT;
an @ prints at the terminal.

Place the bootstrap tape in the desired paper tape reader
with the special bootstrap leader code over the reader
sensors (under the reader station).

If the low-speed reader
(ASR-33)
is to be used,
and a
high-speed
reader
is present on the system,
turn the
high-speed reader OFF.
If the high-speed reader is to be
used, turn it ON.

Type the command/status register address of the input
followed by L to load the tape.
For example, when
reader, type:

from

This

the

enables

console

the

device
terminal

@ 177560L
After reading the contents of the
tape,
the
LSI-II
microprocessor starts the program, which typically halts.
In
this case, the micro-ODT automatically restarts and prints @
followed by the address of the instruction after the HALT
instruction. For example, after loading the Absolute Loader
on an 8K system, the micro-ODT prints:
@375000
~

The starting address of the Absolute Loader in this
375000.

6.1.4

case

M9301-YB Bootstrap Loader
lao If the system does not have a
front panel BOOT/INIT switch.

switch

press

the

lb. If the system does not have a BOOT/INIT switch, set the
console switch register to 773000; press LOAD/ADDR;
then
press START.

Four numbers are printed at the terminal, followed by a
$.
These numbers are the contents of the general registers RO,
R4, R6, and R5,
respectively.
For CPUs without switch
registers
(such as the 11/04), R5 contains the contents of
the program counter (PC) at the time BOOT/INIT was pressed.
For example:

007740

012450

00546

004054

Place the bootstrap tape in the desired paper tape reader
with the special bootstrap leader code over the reader
sensors (under the reader station).

6-3

LOADING AND DUMPING MEMORY
4.

Type the device code
(PR for high-speed reader,
terminal reader), and type RETURN, as follows:
or

for

~TT..J

After reading the contents of the tape, the Bootstrap Loader
transfers control to the program just loaded. Typically,
this program halts.

6.1.5

M9301-YA Bootstrap Loader

If a console terminal is available, boot instructions for the M9301-YA
Bootstrap Loader are the same as for the M9301-YB Bootstrap Loader
(Section 6.1.4).
If no console terminal is available, the auto-boot feature of
M9301-YA
must be used.
See the M9301 Maintenance Manual
instructions on placing'the appropriate paper tape bootstrap in
M9301 module micro-switch. Then follow the procedure below:

6.1.6

the
for
the

Place the bootstrap tape in the desired paper tape reader
with the special bootstrap leader code over the reader
sensors (under the reader station).

Set the console HALT/CaNT switch to CaNT.

Press the console BOOT/INIT switch.
After reading the
contents of the tape, the Bootstrap Loader transfers control
to the program just loaded. Typically, this program halts.

Other Bootstrap Loaders

This section is for users without any of
above.

the

bootstrap

aids

listed

The Bootstrap Loader should be loaded (toggled) into the highest core
memory bank.
The locations and corresponding instructions of the
Bootstrap Loader are listed and explained below.
Instruction

Location
xx7744
xx7746
xx7750
xx7752
xx7754
xx7756
xx7760
xx7762
xx7764
xx7766
xx7770
xx7772
xx7774
xx7776
Figure 6-1

016701
000026
012702
000352
005211
105711
100376
116162
000002
xx7400
005267
177756
000765
yyyyyy

Bootstrap Loader Instructions

6-4

LOADING AND DUMPING MEMORY
In Figure 6-1, xx represents the highest available memory bank.
For
example, the first location of the Loader would be one of the
following, depending on memory size, and xx in all subsequent
locations would be the same as the first.
Location

Memory Bank

Memory Size

017744
037744
057744
077744
117744
137744
157744

0
1
2
3
4
5
6

4K
8K
12K
16K
20K
24K
28K

Note also in Figure 6-1 that the contents of location
reflect the appropriate memory bank in the same
location.

xx7766
manner

should
as the

The contents of location xx7776 (yyyyyy in the Instruction column of
Figure 6-1)
should contain the device status register address of the
paper tape reader to be used when loading the bootstrap formatted
tapes. Either paper tape reader may be used, specified as follows:
Teletype Paper Tape Reader
High-Speed Paper Tape Reader

177560
177550

6.1.6.1 Loading the Loader Into Core - Toggle in the Bootstrap Loader
as explained below.
1.

Set xx7744 in the Switch Register (SR) and press LOAD ADDRess
(xx7744 is displayed in the ADDRESS REGISTER).

Set the first instruction, 016701, in the SR and lift DEPosit
(016701 is displayed in the DATA register).
NOTE
When DEPositing data into consecutive words, the
DEPosit automatically increments the ADDRESS REGISTER
to the next word.

Set the next instruction, 000026, in the SR and lift
(000026 is displayed in the DATA register).

DEPosit

Set the next instruction in the SR, press DEPosit, and
continue depositing subsequent instructions
(ensure that
location xx7766 reflects the proper memory bank) until after
000765 has been deposited in location xx7774.

Deposit the desired device status register address
in
location xx7776, the last location of the Bootstrap Loader.

It is good programming practice to verify that all instructions
stored correctly. This is done by proceeding at step 6 below.
6.

Set xx7744 in the SR and press LOAD ADDRess.

6-5

are

LOADING AND DUMPING MEMORY
7.

Press EXAMine (the octal instruction in location xx7744 is
displayed in the DATA register so that it can be compa~,ed to
the correct instruction, 016701.
If the instruction is
correct, proceed to step 8; otherwise go to step 10.

Press EXAMine (the instruction of the location displayed in
the ADDRESS REGISTER is displayed in the DATA register;
compare the DATA register contents to the instruction for the
displayed location.

Repeat step 8 until all instructions have been verified or go
to step 10, whenever the correct instruction is not displayed.

When an incorrect instruction is displayed, it
performing steps 10 and 11.

can

corrected

10.

With the desired location displayed in the ADDRESS REGISTER,
set the correct instruction in the SR and lift DEPosit (the
contents of the SR are deposited in the displayed location).

11.

Press EXAMine to ensure that the instruction was
stored (it is displayed in the DATA register).

12.

Proceed at step 9 until all instructions have been verified.

The Bootstrap Loader is now loaded into core.
The
are illustrated in the flowchart of Figure 6-2.

correctly

procedures

above

6.1.6.2 Loading Bootstrap Tapes - Any paper tape punched in bootstrap
format is referred to as a bootstrap tape (see Section 6.1.3) and is
loaded into core using the Bootstrap Loader.
Bootstrap tapes begin
with about two feet of special bootstrap leader code (ASCII code 351,
not blank leader tape as required by the Absolute Loader).
With the Bootstrap Loader in core, the bootstrap tape is loaded into
core starting anywhere between location xx7400 and location xx7743,
i.e., 162 (octal) words. The paper tape input device used is that
which is specified in location xx7776 (see section 6.1.6.1).
Bootstrap tapes are loaded into core as explained below.
1.

Set the ENABLE/HALT switch to HALT.

Place the bootstrap tape in the specified reader with the
special bootstrap leader code over the reader sensors (under
the reader station).

Set the console switch register to xx7744
(the starting
address of the Bootstrap Loader) and press LOAD ADDRess.

Set the ENABLE/HALT switch to ENABLE.

Press START. The bootstrap tape passes through the reader as
data is being loaded into core.

The bootstrap tape stops after the last frame of data
(see
Figure 6-5)
has been read into core. The program on the
bootstrap is now in core.

The procedures above are illustrated in the flowchart of Figure 6-3.

6-6

LOADING AND DUMPING MEMORY

INITIALIZE

SETSR TO
xx7744

PRESS
LOADADDR

LOAD

VERIFY

SETSRTO

PRESS EXAM

016701

LIFT DEP

SETSRTO
NEXT
INSTRUCTION

SETSR TO
CORRECT
INSTRUCTION
NO

LIFT DEP
LIFT DEP
FINISHED

Figure 6-2

Loading and Verifying the Bootstrap Loader

6-7

LOADING AND DUMPING MEMORY

WITH BOOTSTRAP
LOADER IN CORE
(SEE FIGURE 6-2)

SET ENABLE/HALT
TO ENABLE

SET ENABLE/HALT
TO HALT

PRESS START

PLACE BOOTSTRAP
TAPE IN SPECIFIED
READER (CODE 351
MUST BE OVER
READER SENSORS)

i
TAPE READS IN
AND STOPS AT
ENDOF DATA

SETSR TO
xx7744

DATA IS
IN CORE

~
PRESS
LOADADDR

I
Figure 6-3

Loading Bootstrap Tapes Into Core

Should the bootstrap tape not read in immediately after depressing the
START switch, one of the following conditions may exist:
1.

Bootstrap Loader not correctly loaded.

Wrong input device used.

Code 351 not directly over the reader sensors.

Bootstrap tape not properly positioned in reader.

6.1.6.3 Bootstrap Loader Operation - The Bootstrap
Loader
source
program 1S shown below. The starting address in the example denotes
that the Loader is to be loaded into memory bank zero (a 4K system).

6-8

LOADING AND DUMPING MEMORY
1
2

3
4
5
6
7
8

00 (lJ(lJ0(lJ
000001 Rl
000"''''2 R2
0(lJ0(lJ07 PC
01.7400 LCAD

=
=
..

017744

LCAD+344 ; STARTING ADDRESS

10 17744 13167131 START:
01H:'I(lJ26
11 17750 0127"'2 LCOP:
12 17752 0013352 DSPMNT:
13
14
15 17754 005211
16 17756 105711 WAIT:
17 17760 1013~76
18 17762 116162
·130"'002
0174"''''
19 17770 "'05267
177756
20
21 17774 000765 BRNCH:
22 17776 177560 DEVICE:
23
24
25
"'00"'01'

.ASECT

%7
17400

Mo.V DE"I CE.. Rl

; PC I NTER TO. DE'TI CE ADDRESS
;LCAD ADDRESS DISPLACEMENT
; PROGRAM COUNTER
;DATA CANNOT BE LOADED BELo.W
;THIS ADDRESS.

; Co.PY DEVI CE ADDRESS

Mo.V CPC)+ .. R2
+.-LCAD

;Co.PY ADDRESS DISPLACEMENT
; INITIALLY o.FFSET TO. THIS Lo.C
;NCTE THAT THIS LOC IS PART o.F
;PREVIo.US INSTRUCTION
INC @Rl
; START THE PAPER TAPE READER
TSTB @Rl
;FRA~E READY?
BPL\OlAIT
;BRIFNOT
MOVB 2(Rl) .. LOADCR2) ; STORE FRAME READ IN MEMORY
INC DSPMNT

; INCREMENT DISPLACEMENT TO. NEXT

BR LCCP
177560

; LO CATION
; READ NEXT BYTE
;ADDRESS OF INPUT DEVICE .. MAY BE
;17755'" IF HIGH SPEED READER

• END

Figure 6-4

The Bootstrap Loader Program

The program above is a brief example of the PAL-IIA Assembly
which is explained in Chapter 2.

Language

Bootstrap tapes are coded in the following format.
351
Special bootstrap leader code (at least two feet
in length)
351
xxx
AAA
BBB
CCC

Load .offset (see text below)
Program to be loaded (up to 162
frames)

words or 344

ZZZ
301
035
026.
000
302
025
373
yyy

Boot overlay code, as shown.
Jump offset (see text below)
Figure 6-5

Bootstrap Tape Format

6-9

LOADING AND DUMPING MEMORY
The Bootstrap Loader starts by loading the device status register
address into RI and 3528 into R2. The next instruction indicates a
read operation in the device and the next two instructions form a loop
to wait for
the read operation to be completed.
When data is
encountered it is transferred to a location determined by the sum of
the index word (xx7400) and the contents of R2.
Because R2 is initially 3528' the first word is moved to location
xx7752,
and it becomes the immediate data to set R2 in the next
execution of the loop. This immediate data is then incremented by one
and the program branches to the beginning of the loop.
The leader code, plus the increment, is equal in value to the data
placed in R2 during the initialization;
therefore, leader code has no
effect on the loader program.
Each time leader code is read the
processor executes the same loop and the program remains unmodified.
The first code other than leader code, however, replaces the data to
be loaded into R2 with some other value which acts as a pointer to the
program starting location (loading address).
Subsequent bytes are
read not into the location of the immediate data but into consecutive
core locations. The program will thus be read in byte by byte.
The
INC instruction which operates on the data for R2 puts data bytes in
sequential locations, and requires that the value of the leader code
and the offset be one less than the value desired in R2.
The boot overlay code overlays the first two instructions of the
Loader, because the last data byte is placed in the core location
immediately preceding the Loader. The first instruction is unchanged
by the overlay, but the second instruction is changed to place the
next byte read, jump offset,
into the lower byte of the branch
instruction.
By changing the offset in this branch instruction, the
Loader can branch to the start of the loaded program or to any point
within the program.
The Bootstrap Loader is self-modifying, and the program loaded by the
Loader restores the Loader to its original condition by restoring the
contents of locations xx7752 and xx7774 to 000352 and
000765
respectively.

6.2

THE ABSOLUTE LOADER

The Absolute Loader is a system program that enables the user to load
data punched on paper tape in absolute binary format into any
available memory bank.
It is used primarily to load the paper tape
system software, binary programs assembled with PAL-IIA, and binary
tapes produced by LINK-lIS from object tapes produced by PAL-lIS. The
major features of the Absolute Loader include:
1.

Testing of the checksum on the input tape to assure complete,
accurate loads.

Starting the loaded program upon completion of loading
without additional user action, as specified by the .END in
the program just loaded.

Specifying the load bias of position independent programs at
load-time rather than at assembly time, by using the desired
Loader switch register option.

LOADING AND DUMPING MEMORY
6.2.1

Loading the Loader Into Core

The Absolute Loader is supplied on punched paper tape in bootstrap
format.
Therefore, a Bootstrap Loader is used to load the Absolute
Loader into core. It occupies locations xx7474 through xx7743, and
its starting address is xx7500.
The Absolute Loader program is
72
words long, and is loaded adjacent to the Bootstrap Loader as
explained in section 6.1.6.2.

6.2.2

Using the Absolute Loader

Paper tapes punched in absolute binary format are also called absolute
tapes, binary tapes, or .LDA tapes. These are the tapes loaded by the
Absolute Loader.
In the following discussion, reference is made to a "switch register."
For
systems without switch registers
(such as the LSI-II and
PDP-llj04) , this term refers to a software switch register, which is a
memory location internal to the Absolute Loader for systems without
hardware switch registers. The location within the Absolute Loader is
xxx516, where xxx reflects memory size as follows:
Memory

xxx

4K
8K
12K
16K
20K
24K
28K

017
037
057
077
117
137
157

When text indicates that a value be placed in a switch register, users
without hardware switch registers must use either the M9301 console
emulator or the LSI-II micro-ODT, as appropriate, to store the switch
register value in location xxx516. Once this value has been stored,
the user starts the Absolute Loader at location xxx500.
Once the
Absolute Loader is loaded, it initializes the value of location xxx516
to O. This value changes only when modified by the user.
A normal load occurs when data is loaded into memory according to the
load addresses on the binary tape. The user must set bit 0 of the
switch register to 0 immediately before starting the load.
There are two types of relocated loads:
1.

Loading to continue from where the loader left off after
previous load -

the

This is used, for example, when the object program being
loaded is contained on more than one tape. It is specified
by setting the switch register to 000001 immediately before
starting the load.
2.

Loading into a specific .area of core This is normally used when loading position independent
programs. A position independent program is one which may be
loaded and run anywhere in available core.
The program is
written using the position independent instruction format
(see Chapter 9). This type of load is specified by setting
the switch register to the load bias and adding 1 to it
6-11

LOADING AND DUMPING MEMORY
(i.e., setting bit 0 to 1). The effect of this is to adq the
value in the switch register to the start address c>n the
tape.
Optional switch
listed below.

~egister

settings for the three

types

loads

are

Switch Register
Bits 1-14
Bit 0

Type of Load
Normal

(ignored)

Relocated - cont-inue
loading where left off

Relocated - load in
specified area of core

nnnnn
(specified
address)

The absolute tape may be loaded using either of the paper tape
readers.
The desired reader is specified in the last word of
available core memory (xx7776), the input device status word, as
explained in section 6.1.6.
The input device status word may be
changed at any time prior to loading the absolute tape.
With the Absolute Loader in core as explained
absolute tapes are loaded as explained below.
1.

section

6.1.6.2,

Set the ENABLE/HALT switch to HALT.
To use an input device different. from that used when loading
the Absolute Loader, change the address of the device status
word (in location xx7776)
to reflect the desired device,
i.e., 177560 for the Teletype reader or 177550 for the
high~speed reader.

Set the switch register to xx7500 and press LOAD ADDR.

Set the switch register to reflect the desired type
(Figure E-3 in Appendix E).

Place the absolute tape in the proper reader
leader tape directly over the reader sensors.

Set ENABLE/HALT to ENABLE.

Press START. The absolute tape begins passing through
reader station as data is being loaded into core.

with

load
blank

the

If the absolute tape does not begin passing through the reader
station, the Absolute Loader is not in core correctly. Reload the
Loader and start over at step 1 above.
If it halts in the middle of
the tape, a checksum error occurred in the last block of data read in.
Normally, the absolute tape stops passing through the reader station
when it encounters the transfer address as generated by tne statement,
.END, denoting the end of a program.
If the system halts after
loading, check that the low byte of the DATA register is zero.
If so,
the tape is correctly loaded.
If not zero, a checksum error
(explained later) has occurred in the block of data just loaded,
indicating that some data was not correctly loaded. Reload the tape
starting at step 1 above.

6-12

LOADING AND DUMPING MEMORY
When loading a continuous relocated load, subsequent blocks of data
are loaded by placing the next tape in the appropriate read~r and
pressing the CONTinue switch.
The Absolute Loader may be restarted at any time by starting at step 1
above.•

6.2~3

Absolute Loader Operation

The Loader uses the eight general registers
(RO-R7)
and does not
preserve or restore their previous contents.
Therefore, caution
should be taken to restore or load these registers when necessary
after using the Loader.
A block of data punched on paper tape in absolute
the following format.
FRAME

1
2
3
4
5
6

001
000
xxx
xxx
yyy
yyy

zzz

binary

format

has

start frame
null frame
byte count (low 8 bits)
byte count (high 8 bits)
load address (low 8 bits)
load address (high 8 bits)
d.ata is
placed
here
last frame contains a block checksum

A program on paper tape may consist of one or more blocks of data.
Each block with a byte count (frames 3 and 4) greater than six causes
subsequent data to be loaded into core
(starting at the address
specified in frames 5 and 6 for a normal load). The byte count is a
positive integer denoting the total number of bytes in the block,
excluding the checksum.
When the byte count of a block is six, the
specified load address is checked to see whether the address is to an
even or to an odd location. If even, the Loaded transfers control to
the address specified. Thus the loaded program runs upon completion
of loading. If odd, the loader halts.
The transfer address (TRA) may be explicitly specified in the source
program by placing the desired address in the operand field following
the .END statement. For example,
.END ALPHA
specifies the symbolic location ALPHA as the TRA, and
.END
causes the Loader to halt.

with

.END nnnnnn
the Loader also halts if the address (nnnnnn) is odd.
The checksum is displayed in the low byte of the DATA register of the
computer console. Upon completion of a load, the low byte of the DATA
register should be all zeros (unlit). Otherwise, a checksum error has
occurred, indicating that the load was not correct. The checksum is
the low-order byte of the negation of the sum of all the previous
bytes in the block. When all bytes of a block including the checksum
are added together, the low-order byte of the result should be zero.
6-13

LOADING AND DUMPING MEMORY
If not, some data was lost during the load or erroneous data was
picked up;
the load was incorrect.
When a checksum error is
displayed, the entire program should be reloaded, as explained in the
previous section. The loaders occupy core memory as illustrated
below.

xx7776

I/O DEVICE WORD

xx7744

BOOTSTRAP LOADER

xx7500

ABSOLUTE LOADER

xx7474

LOADER STACK

USER AND
SYSTEM
PROGRAMS

6.3

CORE MEMORY DUMPS

A core memory dump program is a system program which enables the user
to dump (print or punch) the contents of any specified portion of core
memory onto the Teletype printer anp/or punch, line printer or
high-speed punch. There are two dump programs available in the Paper
Tape Software System:
1.

DUMPTT 1 , which dumps the octal representation of the contents
of specified portions of core onto the teleprinter, low-speed
punch, high-speed punch, or line printer.

DUMPAB, which dumps the absolute binary code of the contents
of specified portions of core onto the low-speed punch or
high-speed punch.

Both dump programs are supplied on punched paper tape in bootstrap and
absolute binary formats.
The bootstrap tapes are loaded over the
Absolute Loader as explained in section 6.1.6.3, and are used when it
would be undesirable to alter the contents of user storage (below the
Absolute Loader). The absolute binary tapes are position independent
and may be loaded and run anywhere in core as explained in section
6.2.2.
DUMPTT and DUMPAB are similar in function, and differ primarily in the
type of output they produce.

6.3.1

Operating Procedures

Neither dump program punches leader or trailer tape, but DUMPAB always
punches ten blank frames of tape at the start of each block of data
dumped.

DUMPTT is not available for systems without switch registers.
6-14

LOADING AND DUMPING MEMORY
6.3.1.1 Using DUMPAB on Systems without Switch Registers - Operating
procedures for DUMPAB on systems without switch registers are as
follows:
1.

Select either the absolute binary or the bootstrap version of
DUMPAB and place it in the reader specified by location
xx7776 (see section 6.1).

2a. If using a bootstrap tape, load the tape using the procedure
outlined in section 6.1. When the computer halts, go to step
3.
2b. If using an absolute binary tape, load the tape using the
procedure outlined in section 6.2.2, relocating as follows:
a.

Select the address to which the program is to be
relocated.
The relocation offset is then equal to the
loading address.
For example, if the desired relocation
address is 000400, the relocation offset is 000401.

Deposit the relocation offset with bit 0 set in the
Absolute Loader's software switch register. Using the
example from the previous step, the user would deposit
000401 into location xxx516.

Start the Absolute Loader.
3.

When the program halts,
find the address in the program
counter.
For LSI-II machines, the value is printed at the
console terminal by the micro-ODT.
For UNIBUS
PDP-II
machines,
the user must press the BOOT/INIT switch to obtain
register values at the console terminal (see section 6.1.4).
The last of the four values displayed is the PC contents.
Add 2 to the value of the PC.
(For example, the PC contents
for
the bootstrap version of DUMPAB are xxx516; adding 2 to
this value gives xxx520.) This new value is the address of
the first of these succeeding parameters, described in
subsequent steps.

Deposit the address of the first byte to be dumped into the
first parameter (whose address was determined in the previous
step) .

Deposit the address of the last byte to be dumped
second parameter (next sequential location).

The third parameter contains the value 177564
(a default
specifying the ASR-33 punch).
If this is the first time this
step is executed and the high-speed reader
is the desired
output device,
change the value of the third parameter to
177554.

into

the

7a. If using the LSI-II, type P to proceed.
7b. If using a UNIBUS PDP-II, restart the program (at xxx510
bootstrap tape); press CONT when the program halts.
8.

DUMPAB dumps the specified segment of memory and halts.

Repeat steps 4 through 8 until all
have been dumped.

6-15

desired

memory

segments

LOADING AND DUMPING MEMORY
10.

A transfer block for DUMPAB must be generated to terminate
the dump.
This value mUst be deposited in the first
parameter (step 4) to terminate DUMPAB. If the tape is not
to be self-starting, use 000001 as the transfer address.
Under no conditions can 000000 be used as the transfer
address.

11.

Deposit 000000 in the second parameter (as in step 5).

12.

Repeat step 7a or 7b, as appropriate, to punch
block.

6.3.1.2

the

transfer

Using DUMPAB and DUMPTT on Systems with Switch Registers -

Select the dump program desired and place it in
specified by location xx7776 (see Section 6.1).

the

reader

If a bootstrap tape is selected, load it using the Bootstrap
Loader, section 6.1.6.2. When the computer halts go to step
4.
.

If an absolute binary tape is selected, load it using
Absolute Loader (section 6.2.2), relocating as desired.

the

Place the proper start address in the switch register, press
LOAD ADDRess and START.
(The start addresses are shown in
section 6.3.3).
4.

When the computer halts, enter the address of the desired
output device status register in the switch register and
press CONTinue
(low-speed punch and teleprinter = 177564;
high-speed punch = 177554; line printer = 177514).

When the computer halts, enter in the switch register the
address of the first byte to be dumped and press CONTinue.
This address must be even when using DUMPTT.

When the computer halts again enter in the switch register
the address of the last byte to be dumped and press CONTinue.
When using the low-speed punch, set the punch to ON before
pressing CONTinue.

Dumping proceeds on the selected output device.

When dumping is complete, the computer halts.

If further dumping is desired, proceed to step 5. It is not necessary
to respecify the output device address except when changing to another
output device. In such a case, proceed to the second paragraph of
step 3 to restart.
If DUMPAB is being used, a transfer block must be generated as
described below. If a tape read by' the Absolute Loader does not have
a transfer block, the loader will wait in an input loop.
In such a
case, the program may be manually initiated. However, this practice
is not recommended, as there is no ,guarantee that load errors will not
occur when the end of the tape is read.
The transfer block is generated by performing step 5 with the transfer
address in the Switch Register, and step 6 with the transfer address
minus 1 in the Switch Register.
If the tape is not to
be
self-starting, an odd-numbered addr~ss must be specified in step 5
(000001, for example).
6-16

LOADING AND DUMPING MEMORY
The dump programs use all eight general registers and do not
their
original contents.
Therefore, after a dump the
registers should be loaded as necessary prior to their
subsequent programs.

6.3.2

restore
general
use by

Output Formats

The output from DUMPTT is in the following format:
xxxxxx>yyyyyy yyyyyy yyyyyy yyyyyy yyyyyy yyyyyy yyyyyy yyyyyy
where xxxxxx is the octal address of the first location printed or
punched, and yyyyyy are words of data, the first of which starts at
location xxxxxx. This is the format for every line of output.
There
will be no more than eight words of data per line, but there will be
as many lines as are needed to complete the dump.
The output from DUMPAB is in absolute binary, as explained in
6.2.3.

6.3.3

section

Storage Maps

The DUMPTT program is 87 words long.
storage map is:

When used in absolute format the

xx7776
BOOTSTRAP LOADER

xx7744
ABSOLUTE LOADER

xx7500
xx7474

LOADER STACK SPACE

xxxxxx+256

DUMPTT

xxxxxx

lWO-WORD STACK SPACE

xxxxxx = desired load address = start address

When used in bootstrap format the storage map is:

6-17

LOADING AND DUMPING MEMORY
xx7776
BOOTSTRAP LOADER

xx7744
DUMPTT

start address

xx7440
xx7434

TWO-WORD STACK SPACE

The DUMPAB program (for systems with a switch register)
is
words long. When used in absolute format the storage map is:
xx7776
BOOTSTRAP LOADER

xx7744

ABSOLUTE LOADER

xx7500
xx7474

LOADER STACK SPACE

xxxxxx+244

DUMPAB

xxxxxx
THREE-WORD STACK SPACE

xxxxxx = desired load address = start address

When used in bootstrap format the storage map is:

6-18

65{lO)

LOADING AND DUMPING MEMORY
xx7776
BOOTSTRAP LOADER

xx7744

DUMPAB

start address = xx7510
xx7500

THREE-WORD STACK SPACE

The DUMPAB program (for systems without a switch register)
is
words long. When used in absolute format the storage map is:

xx7776
BOOTSTRAP LOADER

xx7744

ABSOLUTE LOADER

xx7500
xx7474

LOADER STACK SPACE

xxxxxx+202

DUMPAB

xxxxxx

TWO-WORD STACK SPACE

xxxxxx = desired load address = start address

6-19

82(10)

LOADING AND DUMPING MEMORY
When used in bootstrap format the storage map is:

xx7776
BOOTSTRAP LOADER

xx7744

DUMPAB

start address = xx7500
xx7474

TWO-WORD STACK SPACE

6-20

CHAPTER 7
INPUT/OUTPUT PROGRAMMING

7.1

INTRODUCTION

The PDP-II Input/Output executive. (lOX), frees the user from dealing
directly with I/O devices.
It provides programming formats that allow
programs written for the paper tape software system to be used later
in a monitor environment with only minor coding changes.
lOX
provides
asynchronous
non-file-oriented devices:

I/O

service

for

the

following

Teletype keyboard, printer, and tape reader and punch

High-speed paper tape reader and punch

For line printer handling, in addition to all lOX
is available.

facilities,

IOXLPT

Simple I/O requests can be made, specifying devices and data forms for
interrupt-controlled data transfers, that can occur concurrently with
the execution of a user program. Multiple I/O devices can run singleor double-buffered I/O processing simultaneously.
Real-time capability is provided by allowing user programs to be
executed at device priority levels upon completion of a device action
or data transfer.
Communication with lOX is accomplished by lOT
(Input/Output Trap)
instructions in the user's program. Each lOT is followed by two or
three words consisti~g of one of the lOX commands and its operands.
The lOX commands can be divided into two categories:
1.

those concerned with establishing necessary conditions for
performing input and output (mainly initializations), and

those concerned directly with the transfer of data.

When transfer of data is occurring, lOX is operating at the priority
level of the device. The calling program runs at its priority level,
either concurrent with the data transfer, or sequentially.
Programming format for commands is:
lOT
.WORD .(an address)
.BYTE (a command code),

(a slot number)

Before using the data transfer commands, two preparatory tasks must be
performed:

7-1

INPUT/OUTPUT PROGRAMMING
1.

Since device specifications are made by referring to "slots"
in lOX's Device Assignment Table (DAT) rather than devices
themselves, the slots specified in the code must have devices
assigned to them.

The buffer, whose address is specified in the code,
set up with information about the data.

must

In those non-data-transfer commands where an address or slot number
does not apply, a 0 must be used. Addresses or cod~s indicated can,
of course, be specified symbolically.
The
following
program
segment
illustrates
input-process-output sequence.
It includes:
1.

The setting up of a single buffer

All necessary initializations

A formatted ASCII read into the buffer

A wait for completion of the read

.5.

Processing of data just read

A write command from the buffer.

simple

;ASSIGN lOX COMMAND CODES

F"~ESET::::2
F"~EAD::::ll

WAITR::::4
WRITE::::12

WAIT:

lOT
.WORD 0
.BYTE RESET,O

9IOX RESET TO DO NECESSARY
9INITIALIZATIONS INCLUDING
9INITING SLOT 0 FOR KBD, AND 1 FOR TTY

lOT
• WOF"W BUFFER
.BYTE REA[I,()

;TRAP TO lOX
;SPECIFY BUFFER
9READ FROM KBD (SLOT 0) TILL
;LINE FEED OR FORM FEED

roT

;TRAP TO lOX
;BUSY RETURN ADDRESS WHILE WAITING
9FOR KBD TO FINISH
;WAIT FOR KBD (SLOT 0) TO FINISH

.WOFW WAIT

• BYTE

0
BUFFEFO

WAITI~"

(p rocess

roT

;TRAP TO lOX
;SPECIFY BUFFER
9WRITE TO TELEPRINTER (SLOT 1)

100

;BUFFER SIZE IN BYTES
;CODE FOR FORMATTED ASCII MODE
9IOXWILL SET HERE THE NUMBER OF BYTES READ
9STORAGE RESERVED FOR 100 BYTES

.WORD BUFFER
• BYTE WRITE, 1

BUFFER:

o
o

.::::.+100

In more complex programming it is likely that more than one buffer
will be set up for the transfer of data, so that data processing can
occur concurrently rather than sequentially, as here. Note too, that
there are five lOX commands not used in this example that will help
meet the requirements of I/O problems not as straightforward as this.

7-2

INPUT/OUTPUT PROGRAMMING
7.1.1

Using lOX With The LSI-II Processor

lOX (IOXLPT) is supplied on source and relocatable object tapes.
It
is thus unnecessary to assemble lOX unless the program is to be
modified. User object tapes can be linked with the lOX object tape
(using LINK-lIS)
to produce an absolute binary tape. Appendix J
describes assembly procedures for source tapes.
lOX requires approximately 633
(decimal)
requires approximately 724 (decimal) words.

7.1.2

words

core;

IOXLPT

Using lOX with Unibus PDP-II Processors

Appendix J
lOX (IOXLPT) is supplied on source and binary tapes.
describes assembly procedures for source tapes. Binary tapes are
loaded prior to user programs by the Absolute Loader.
After lOX is
loaded, the Absolute Loader halts.
IOXLPT is used instead of lOX if the program uses a line printer.
lOX is supplied on an absolute binary tape with a loading address of
15100;
the load address for IOXLPT is 34600. If the user desires
different load addresses, the programs must be reassembled
as
described in Appendix J.
lOX requires approximately 634
(decimal)
requires approximately 725 (decimal) words.

7.1.3

words

core;

IOXLPT

lOX Interrupt and Trap Vectors

lOX (IOXLPT) loads the following interrupt and trap vectors:
Console terminal
high speed reader and punch
timeout and other errors
lOT
line printer (IOXLPT only)

7.2

THE DEVICE ASSIGNMENT TABLE

The Device Assignment Table (OAT) makes programs device-independent by
allowing the user to .refer to a slot to which a device has been
assigned, rather than a specific device itself.
Thus, changing the
input or output device becomes a simple matter of reassigning a
different device to the slot indicated in the program.
The OAT is created by means of the Reset and/or Init commands.
The
lOX codes for devices (listed in the description of the Init command
below) are assigned to the slots.

7-3

INPUT/OUTPUT PROGRAMMING
7.2.1

Reset
IOT
.WORD 0
.BYTE 2,0

This command must be the first IOX command issued by a user program.
It clears the DAT, initializes IOX, resets all devices to their state
at power-up, enables keyboard interrupts, and initializes DAT slots 0
and 1 for the keyboard and teleprinter, respectively.

7.2.2

Initialization
IOT
.WORD (address of device code)
.BYTE 1, (slot number)

The device whose code (stored as a byte) is found at the specified
address is associated with the specified slot (numbered in the range
0-7). The device interrupt is turned off when necessary.
(The
keyboard interrupt always remains enabled.) There is no restriction on
the number of slots that can be initialized to the same device.
DEVICE
CODE

DEVICE
Teletype Keyboard

(KBD)

Teletype printer

(TTY)

Low-Speed Reader

(LSR)

Low-Speed Punch

(LSP)

High-Speed Reader

(HSR)

High-Speed Punch

(HSP)

Line Printer
(IOXLPT only)

(LPT)

Note that a device code is used only in the Initialization
(INIT)
command.
All other commands that refer to a device do so by means of
a slot. Example:
INIT:=::L
10T
.WOI:;:D Hm~C(JD

.BYTE 1N:l:T,3

HSF~CCJD

7.3

BYTE

vTRAF' TO 10X

v1NITSLOT :3
jl

FIJI:;: H81:;:

jl HSF~

CODE

BUFFER ARRANGEMENT IN DATA TRANSFER COMMANDS

Use of data-transfer commands (Read, Write, Real-time Read, Real-time
Write)
requires the creation of at least one buffer. This buffer is
used not only to store data for processing, but to hold information
regarding the quantity,
form, and status of the data. The non-data
7-4

INPUT/OUTPUT PROGRAMMING
portion of the buffer is called the buffer header,
and precedes the
data portion.
In data transfer commands, it is the address of the
first word of the buffer header that is specified in the word
following the lOT of the command.
NOTE
lOX uses the
buffer
header
while
transferring data.
The user's program
must not chaQge or reference it.
The buffer format is:
Location

BUFFER
HEADER

Contents

Buffer

Maximum number
integer)

data

Buffer+2

Mode of data (byte)

Buffer+3

Status of data (byte)

Buffer+4

Number of data bytes
(unsigned integer)

Buffer+6

Actual data begins here

{unsigned

bytes

involved

transfer

BUFFER SIZE (IN BYTES)
STATUS

MODE

BYTE COUNT
DATA

7.3.1

Buffer Size

The first word of the buffer contains the size (in bytes) of the data
portion of the buffer as specified by the user.
lOX will not store
more than this many data bytes on input. Buffer size has no meaning
on output.
7.3.2

Mode Byte

The low-order byte of the second word holds information concerning the
mode or transfer. A choice of four modes exists:
Coded as
1.

Formatted ASCII

Formatted Binary

Unformatted ASCII

Unformatted Binary

3
7-5

(or 200 to suppress echo)

(or 202 to suppress echo)

INPUT/OUTPUT PROGRAMMING
The term echo applies only to the
devices never involve an echo.

KBD.

Data

transfers

from

other

MODE BYTE

NO ECHO

ECHO

7.3.3

Bits

UN FOR- '
BINARY
MATTED

FORMATTED

ASCii

Status'Byte

The high-order byte of the second word of the buffer header
information set by IOX on the status of the data transfer:

contains

Bits 0-4

contain the non-fatal error codes (coded octally)

Bit

End-of-Medium has occurred (see Section 7.3.3.3)

Bit

Done (Data Transfer complete)

End-Of-File has occurred (attempt at
after an End-Of-Medium)

reading

data

STATUS BYTE

7
1

DONE

EOM

1=
EOF

2
I

SEE CODES

NON-FATAL ERRORS

7.3.3.1

Non-Fatal Error Codes (Octal) 2

checksum error

truncation of a long line

an improper mode

A checksum error can occur only on a
(see Section 7.4.3).

7-6

Formatted

Binary

read

INPUT/OUTPUT PROGRAMMING
2.

Truncation of a long line can occur on either a Formatted
Binary or Formatted ASCII read (Section 7.4.1). This error
occurs when the binary block or ASCII line is bigger than the
buffer size specified in the buffer header.
In both cases,
lOX continues reading characters into the last byte in the
buffer until the end of the binary block or ASCII line is
encountered.

An improper mode can occur only on a Formatted Binary read.
Such occurrence means that the first non-null character
encountered was not the proper starting character for a
Formatted Binary block (see Section 7.4.3)

7.3.3.2 Done Bit - When the data transfer to or from
complete, the Done Bit is set by lOX.

the

buffer

7.3.3.3 End-Of-Medium Bit - The following conditions cause the EOM
bit to be set in the buffer Status byte associated with a data
transfer command. An EOM occurrence also sets the Done Bit.
HSR

HSP

LSR

LPT

No tape

Timeout
detected

No paper

Off line
No power

No power

No power
Printer drum gate open
Overtemperature condition

An End-Of-Medium condition on an output device is cleared by a manual
operation such as putting a tape in the high-speed punch.
lOX does
not retain any record of an EOM on an output device.
However, an EOM
on an input device is recorded by lOX so that succeeding attempts to
read from that device will cause an End-Of-File (see Section 7.3.3.4).
To reenable input the device must be manually readied and a Seek
command (Section 7.6) executed on the proper slot. The INIT and RESET
commands will also clear the EOM condition for the device.
See Section 7.5.3 for information on detection of
LSR timeouts.

conditions

causing

When an End-Of-Medium has occurred on a Read, there may be data in the
buffer.
If an EOM has occurred on a Write, there is no way of knowing
how much of the buffer was written.

7.3.3.4 End-Of-File Bit - An EOF condition appears in the Status byte
if an attempt to read is made after an EOM has occurred. EOF cannot
occur on output. When an EOF has occurred, no data is available in
the buffer.

7-7

INPUT/OUTPUT PROGRAMMING
7.3.4

Byte Count

The third wdrd contains the Byte Count:

7.4

Input:

In unformatted data modes, lOX reads as many data bytes
as the user has specifie~.
In formatted modes, lOX
inserts here the number of data bytes available in the
buffer.
In all modes, if an EOM occurs, lOX will set
the Byte Count equal to the number of bytes actually
read.
If an EOF occurs, Byte Count will be set to O.

Output:

Byte Count determines the number of bytes output, for
all modes.
An HSP end-of-tape or LPT out-of-paper
condition will also terminate output, and EOM will be
set in the Status byte.
lOX does not modify the Byte
Count on output.

MODES

7.4.1

Formatted ASCII

A Formatted ASCII read transfers 7-bit characters (bit 8 will be zero)
until a line feed or form feed is read.
lOX sets the Byte Count word
in the buffer header to indicate the number of characters in the
buffer.
If the line is too long, characters are read and overlaid
into the last byte of the buffer until an end-of-line (a line feed or
form feed) or EOM is detected. Thus, if there is no error, the buffer
will always contain a line feed or form feed.
A Formatted ASCII write transfers the number of 7-bit characters
specified by the buffer Byte Count. Bit 8 will always be output as
zero.
Device-Dependent Functions
Keyboard
Seven-bit characters read from the keyboard are entered in the
and are echoed on the teleprinter except as follows:

buffer

Null

- Ignored.
This
character
transferred to the buffer.

Tab
(CTRL/TAB
keys)

- Echoes as spaces up to the next tab stop.
are located at every 8th carriage position.

RUBOUT

- Deletes the previous character on the current line
and echoes as a backslash
(\). If there are no
characters to delete, RUBOUT is ignored.

CTRL/U

- Deletes the current line and echoes as tU.

not

echoed

"Stops"

Carriage
- Echoes as a carriage return followed by a line feed.
Return
Both characters enter the buffer.
(RETURN key)
CTRL/P

Echoes as +P and causes a jump
address, if non-zero (see 7.6.2).

the

restart

The echo may be suppressed by setting bit 7 of the buffer header
byte.
7-8

Mode

INPOTjOUTPUT PROGRAMMING
If the buffer overf19ws, only the characters which fit into the buffer
are echoed.
Of course, characters which are deleted by RUBOUT or
CTRLjU do not read into the buffer even though they are echoed. If a
carriage return causes an overflow, or is typed after an overflow has
occurred, a carriage return and line feed will be echoed but only the
line feed will enter the buffer.
In the following For.matted ASCII exa.mples:
1.

assume there is room for five characters

.J indicates:
in left column, the RETURN key
in center column, the execution of a carriage return
in right column, the ASCII code for carriage return

RUB
OUT

indicates the RUBOUT key

CTRL

indicates the CTRL and U keys.

indicates:
in center column, the execution of a line feed
in right column, the ASCII code for line feed

Typed
ABC..)
ABCD..)

ABC..) +

ABCDEF~

ABCD..J4o
ABCD\..)4o

ABC.J+
ABCD+
ABCD+
ABC..)+

ABCD~4o

ABCDEF RUB..)
OUT
CTRL
U

Entered Buffer

Echoed

RUB..)
OUT

+U..)+

ABCDEF RUB RUB..)
OUT OUT

ABCD\ \..)+

ABCDEF RUB RUB RUB x..)
OUT OUT OUT

ABCD \ \ \X.J4o

AX..) +

Low-Speed Reader and High-Speed Reader
All characters are transferred to the buffer
ruboutsare ignored.

except

that

nulls

and

Teleprinter
Characters are printed from the buffer as they appear except that
nulls are ignored and tabs are output as spaces up to the next tab
stop.
Low-Speed Punch and High-Speed Punch
Characters are punched from the buffer as they appear
nulls are ignored and tabs are followed by a rubout.

7-9

except

that

INPUT/OUTPUT PROGRAMMING
Line Printer (IOXLPT only)
Characters are printed from
follows:

the

buffer

they

appear

except

Nulls

Ignored

Tab

Output as spaces up to the next tab stop.

Carriage
Return

Ignored.
It is assumed that a line feed or form
feed follows.
These characters cause the line
printer "carriage" to advance.

All characters beyond the 80th are ignored except a line feed or
feed.

7.4.2

form

Unformatted ASCII

Unformatted ASCII transfers the number of 7-bit
by the header Byte Count.

characters

specified

Device-Dependent Functions
Keyboard
Characters are read and echoed except as follows;

7.4.3

Tab

Echoes as spaces up to the next tab stop.

CTRL/P

Echoes as Ap and causes a jump to
address, if non-zero (see 7.6.2).

the

restart

Formatted Binary

Formatted Binary is used to transfer checksummed binary data
(a-bit
characters) in blocks. A Formatted Binary block appears as follows:
Byte (Octal)

Meaning

001
000

Start of block
Always null

~}
xxx

Block Byte Count
(low-order followed by
high-order). Count includes data and
preceding four bytes.

DDD
DDD
Data bytes
DDD
DDD
CCC

Checksum. Negation of the sum of all
preceding bytes in the block.
7-10

INPUT/OUTPUT PROGRAMMING
IOX creates the block on output, from the
The Byte Count word in the buffer header
bytes following, which are to be output.
output is four larger than the header
output, IOX calculates the checksum which
data byte.

buffer and buffer header.
specifies the number of data
Note that the Byte Count
Byte Count. As the block is
is output following the last

On Formatted Binary reads, IOX ignores null characters until the first
non-null character is read.
If this character is a 001, a Formatted
Binary block is assumed to follow and is read from the device under
control of the. Byte Count value. If the first non-null character is
not 001, the read is immediately terminated and error code 4 is set in
the Status byte.
As the block is read a checksum is calculated and
compared to the checksum following the block.
If the checksum is
incorrect, error code 2 is set in the Status byte of the buffer
header.
If the binary block is too large (Byte Count less 4, larger
than the Buffer Size specified in the header), the last byte of the
buffer is overlaid until the last data byte has been read; error code
3 is set in the Status byte.
Device-Dependent Functions
None. Eight-bit data characters are transferred
device and buffer exactly as they appear.

7.4.4

and

from

the

Unformatted Binary

This mode transfers 8-bit characters with no formatting or character
conversions of any kind. For both input and output, the buffer header
Byte Count determines the number of characters transferred.
Device-Dependent Functions
None

7.5

7.5.1

DATA TRANSFERS

Read
IOT
.WORD (address of first word of the buffer header)
.BYTE 11, (slot number)

This command causes IOX to read from the device associated with the
specified slot according to the information found in the buffer
header. IOX initiates the transfer of data, clears the Status byte,
and returns control to the calling program. If the device on the
selected slot is busy, or a conflicting device (see Section 7.5.3)
is
busy, IOX retains control until the data transfer can be initiated.
Upon completion of the Read, the appropriate bits in the Status byte
are set by IOX and the Byte Count word indicates the number of bytes
in the data buffer. Note that use of the KBD while an LSR Read is in
progress
will
intersperse
KBD
characters
into
the
buffer
unpredictably.

7-11

INPUT/OUTPUT PROGRAMMING
7.5.2

Write
lOT
.WORD (address of first word of the buffer header)
.BYTE 12, (slot number)

lOX writes on the device associated with the specified slot according
to the information found in the buffer header. Transfer of data
occurs in the amount specified by Byte Count (Buffer+4).
lOX returns
control to the calling program as soon as the transfer has been
initiated.
If the device on the selected slot is busy, or a
conflicting device is busy, lOX retains control until the transfer can
be initiated. Upon completion of the Write, lOX will set the Status
byte to the latest conditions.
If a write causes an EOM condition,
the user has no way of determining how much of his buffer has been
written (the Byte Count remains the same.)

7.5.3

Device Conflicts In Data Transfer Commands

Because there is a physical association between the devices on the ASR
Teletype, certain devices cannot be in use at the same time.· When a
data transfer command is given,
lOX simultaneously checks for two
conditions before executing the command:
1.

Is the device requested already in use?

Is there some other device in use that
operational conflict?

and,
would

result

lOX resolves both conflict situations by waiting until the first
device is no longer busy, before allowing the requested device to
start functioning.
(This is an automatic Waitr command.
See next
section.)
For example, if the LSR is in use, and either a KBD request
or a second request for the LSR itself is made, lOX will wait until
the current LSR read has been completed before returning control to
the calling program.
In the particular case of the LSR,
lOX also
performs a timeout check while waiting for it to become available.
When a Read command has been issued for the LSR, lOX waits about 100
milliseconds for each character to be read.
If no character is
detected by this time (presumably because the LSR is turned off, or
out of tape), a timeout is declared and lOX sets EOM in the
appropriate buffer Status byte.
The following is a table listing the devices. Corresponding to each
device on the left is a list of devices (or the echo operation) which
would conflict with it in operation.

Device

All Possible Conflicting
Devices or Operations

KBD
CHO, KBD, TTY, LSR, LSP
TTY
Echo, KBO, TTY, LSP
LSR
KBD, LSR
LSP
Echo, KBD, TTY, LSP
HSR
HSR
HSP
HSP
LPT (IOXLPT only)LPT

7-12

INPUTjOUTPUTPROGRAMMING
7.5.4

Waitr

(Wait, Return)

lOT
.WORD (busy return address)
.BYTE 4, (slot number)
Waitr, like device conflict resolution, causes lOX to test the status
of the device associated with the specified slot.
If the device (or
any possible conflicting device) is not transferring data, control is
passed to the instruction following the Waitr.
Otherwise,
lOX
transfers program control to the busy return address.
If it is
desired to continuously test for completion of data transfer on the
device, the busy return address of the immediately preceding lOT
instruction can be specified, effecting a Wait loop.
If a slot is inited to any device other than the LSR, control is
returned to the calling program about 150 microseconds after execution
of a Waitr.
For the LSR, however, the time is about 100 milliseconds.
Note that a not-busy return from Waitr normally means the device is
available.
However, in the case of a Write, this only means that the
last character has been output to the device. The device is still in
the process of printing or punching the character.
Thus, care must be
exercised when performing an lOX Reset, hardware RESET, or HALT after
a Write-Waitr sequence, since these may prevent the last character
from being physically output.

7.5.5

waitr vs. Testing the Buffer Done Bit

Since lOX permits you to have device-independent code, it may not be
known,
from run to run, what devices will be assigned to the slots in
your program. Waitr tests the status, not only of the device it
specifies, but also of all possible conflicting devices.
This means that when Waitr indicates that the device is not busy,
the
data transfer on the device of interest may have been done for some
time.
Depending on the program and what devices are assigned to the
slots for a given run, the Waitr could have been waiting an additional
amount of time for a conflicting device to become free.
Where this possibility exists and buffer availabili,ty is what is of
interest, testing the Done bit of the Status byte (set when buffer
whereas waitr
transfer is complete) would be preferable to Waitr;
would be preferable if device availability is what is of interest.
This distinction is made in order to
In the example below:

write

device-independent

code.

If the devices at slots 2 and 3 could be guaranteed always to
be conflicting, neither Waitr nor testing the Done bit would
the
be necessary, because lOX would automatically wait for
busy device to finish before allowing the other device to
begin.

If these devices could be guaranteed never to be conflicting,
it wouldn't matter which of these methods was used, because
Waitr couldn't be waiting extra time for a conflicting device
(of no interest) to become free.

7-13

INPUT/OUTPUT PROGRAMMING
Example:

PROGRAM A

PROGRAM B

JOT
.WORD BUF2
.BYTE F'~EAD , SUJT2

lOT
.WORD BUF2
.BYTE READ, SUJT2

lOT
• WOF'W BUF:L
.BYTE READ,

JOT
.WORD BUn
.BYTE READ, SLOT2

SLOT~.~

JOT
.WORD BUF2
.BYTE WRITE, SLOT3

101'

.WORD BUF2
.BYTE WRITE, SI.. OT:3
DUNTST:

TSTB BUF1+~~
BPL DUNTST

(IE~VTST

lOT
.WORD DEVTST
.BYTE WA ITR, SL.OT~;~
JOT
.WORD SLOT2DEV
.BYTE INIT, SLOT4

Programs A and B do two successive reads from the same device into two
different buffers. Since the devices are the same, lOX waits for the
first read to finish before allowing the second to begin.
In Program A, we wish to process buffer 1. To have issued a Waitr for
the device associated with slot 2 could have meant waiting also for
the device at slot 3 if that device were in conflict. Hence, testing
the Done bit in the buffer header is the proper choice.
In Program B, we wish control of the device at slot 2, so that it can
be assigned to another slot and so we must know its availability.
Therefore, Waitr is appropriate.

7.5.6

Single Buffer Transfer on One Device

BUSY:

lOT
.WOF'W BUF:L
.BYTE READ,SLOT3

; TF~AP TO lOX
; SF'ECIFY BUFFER
9READ FROM DEVICE AT
;SLOT 3 INTO BUFFER

IOT
.WORD BUSY
.BYTE WAITR,SLOT3

;TRAP TO lOX
;SPECIFY BUSY RETURN ADDRESS
;WAIT FOR DEVICE AT SLOT
;3 TO FINISH READING

(process buffer :L)

JMP A
The program segment above includes a waitr which goes to a Busy Return
address that is its own lOT -- continuously testing the device at slot
3 for availability. In this instance, involving only a single device
and a single buffer, a Done condition in the Buffer 1 Status byte can
be inferred from the availability of the device at slot 3.
This
knowledge assures us that all data requested for Buffer 1 is available
for processing.
Testing the Done Bit of Buffer 1 might have been used instead, but was
not necessary with only one device operating. Moreover, a Waitr,
unlike a Done Bit test, would detect a timeout on the LSR if that
device happened to be associated with slot 3.
7-14

INPUT/OUTPUT PROGRAMMING
7.5.7

Double Buffering

lOT
.WORD BUFI
.BYTE READ,SLOT3

;TRAP TO lOX
;SPECIFY BUFFER 1
;READ FROM DEVICE AT
;SLOT 3 INTO BUFFER 1

lOT
.WORD BUF2
.BYTE READ,SLOT3

;TRAP TO lOX
;SPECIFY BUFFER 2
READ FROM DEVICE AT SLOT
;3 INTO BUFFER 2

(process BUFI concurrent with Read into BUF2)
lOT
.WORD BUFI
.BYTE READ,SLOT3

;TRAP TO lOX
;SPECrFY BUFFER 1
;READ FROM DEVICE AT
;SLOT 3 INTO BUFFER 1

(process BUF2 concurrent with Read into BUF1)
JMP A
The example above illustrates a time-saving double-buffer scheme
whereby data is processed in Buffer 1 at the same time as new data is
being read into Buffer 2; and, sequentially, data is processed in
Buffer 2 at the same time as new data is being read into Buffer 1.
Because lOX ensures that the requested device is free
before
initiating the command, the subsequent return of control from the rOT
at A implies that the read prior to A is complete;
that is, that
buffer 1 is available for processing.
Similarly, the return of
control from the rOT at B implies that buffer 2 is available. Wai~r's
are not required because lOX has automatically ensured the device's
availability before initiating each Read.

7.5.8

Readr (Real-time Read)
lOT
.WORD (address of first word of the buffer header)
.BYTE 13, (slot number)
.WORD (done-address)

The Readr command functions as the Read except that upon completion of
the data transfer, program control goes to the specified Done-address
at the priority level of the device. Readr is used when you wish to
execute a segment of your program immediately upon completing the data
transfer.
lOX goes to the Done address by executing a JSR R7,
Done-address.
The general registers, which were saved when the last character
interrupt occurred, are on the SP stack in the order indicated below:
(SP)~

Return address to lOX
R5
R4
R3
R2
Rl

7-15

INPUT/OUTPUT PROGRAMMING
Return to lOX is accomplished by an RTS R7 instruction.
lOX will then
restore all registers and return to the interrupted program
Care
should be taken in initiating another data transfer if the specified
device can conflict with device requests at other priority levels.
Waitr cannot be used to resolve conflict situations between priority
levels.

7.5.9

writr

(Real-time Write)

lOT
.WORD (address of first word of the buffer header)
.BYTE 14, (slot number of device)
.WORD (done address)
The Writr command functions as the Write except that, upon completion
of
the
data transfer, program control goes to the specified
Done-address at the priority level of the device.
lOX goes to the
Done-address by executing a JSR R7, Done-address. The condition of
the general registers and the return to lOX are the same as for Readr.
Writr is used when you wish to execute a segment of your program
immediately upon completing the data transfer.
As in the Readr, care should be taken in initiating another data
transfer if the specified device can conflict with device requests at
the priority level of the calling program.

7.6

7.6.1

REENABLING THE READER AND RESTARTING

Seek
lOT
.WORD
.BYTE 5,

(slot number of LSR or HSR)

The Seek command clears lOX's internal End-Of-Medium
(EOM)
indicator
on the LSR or HSR, making possible a subsequent read on those devices.
With no EOM, an EOF cannot occur.
The device associated with the
specified slot remains Inited.

7.6.2

Restart
lOT
.WORD (address to restart)
.BYTE 3,0

This command designates an address at which to restart your program.
After this command has been issued, typing CTRL/P on the KBD will
transfer program control to the restart address, providing there is no
LSR read in progress.
In such a case, the LSR must be turned off
(causing a timeout) before typing a CTRL/P.
If the Restart address is
designated as 0, the CTRL/P Restart capability is disabled.
It is the
The Restart command does not cancel any I/O in progress.
program's responsibility in its restart routine to clean up any I/O by
executing a RESET command and ensuring that the stack pointer is
reset.
7-16

INPUT/OUTPUT PROGRAMMING
7.7

FATAL ERRORS

Fatal errors result in program termination and a jump to location
408
(loaded with a HALT by lOX), with RO set to the error code and Rl set
as follows:
If the fatal error was due to an illegal memory reference
0), Rl will contain the PC at the time of the error.

(code

If the fatal error was due to an error coded in the range 1-5, Rl
will point to some element in the lOT argument list or to the
instruction following the argument list, depending on whether lOX
has finished decoding the arguments when it detects the error.
Fatal Error Code

Reason

Illegal Memory Reference,
instruction

Illegal lOX command

Slot out of range

Device out of range

Slot not inited

Illegal data mode

overflow,

illegal

Note that the SP stack contains the value of the registers at the time
of the error, namely
(SP)-+ R5
R4
R3
R2
Rl
RO
PC
Processor Status

(PS)

(See Section 7.3.3.1 for a discussion of non-fatal errors.)

7.8

EXAMPLE OF PROGRAM USING lOX

This program is used to duplicate paper tape. Note that it could be
altered by changing the device code at RDEV or PDEV.
For instance,
the program could easily be made to list a tape.
RO::::%O
Rl::::%l
R2::::%2
r~3:::::;('3

1:;:4::::%4
1:~6::::%6

KBI...OT::::0
TSl..ClT:::: :1.
f~SI...OT::::3

PBI...DT::::4
f~E8ET::::2

RESTRT:::::3

I NIT:::::/.

7-17

INPUT/OUTPUT PROGRAMMING

WAITR=4
READ=ll
WRITE=12
EOF=20000
CR=15
LF=12
MSG1:

9CR ASSIGNED ASCII CODE FOR CARRIAGE RETURN
9LF ASSIGNED ASCII CODE FOR LINE FEED

.=1000

o
o

;CANNED MESSAGE
;FORMATTED ASCII
;BYTE COUNT

9DEVICE CODE FOR HSR
;DEVICE CODE FOR HSP

MSG1BC: END1-MSG1BC-2
.BYTE
CR,LF
.ASCII / PLACE TAPE IN READER/
.BYTE
CR,LF
.ASCII / STRIKE CR WHEN READY/
END1:
.EVEN
BUF3:
2
9BUFFER SIZE
o
;FORMATTED ASCII MODE
o
9BC
;CR LF
o
RDEV:
PDEV:
BUF1:

100
3

BUF2:

100
.=.+100
100
3

BEGIN:

100
.=.+100
MOV
$500,R6

9BUFFER SIZE
9CODE FOR UNFORMATTED BINARY
;SPECIFIES NUMBER OF BYTES FOR TRANSFER
;RESERVES STORAGE FOR DATA
9BUFFER SIZE
;CODE FOR UNFORMATTED BINARY
;SPECIFIES NUMBER OF BYTES FOR TRANSFER
9RESERVES STORAGE FOR DATA
9SPECIFY ADDRESS FOR BOTTOM OF STACK

lOT

.BYTE

RESET,O

9INITIALIZATION

lOT
BEGIN
.BYTE
MOV
MOV

RESTRT,O
$100,BUF1+4
*100,BUF2+4

9"BEGIN" SPECIFIED AS RESTART
9ADDRESS FOR CTRL P
98ET UP INITIAL BC ON BUFI
;SET UP INITIAL BC ON BUF2

lOT
MSGI
.BYTE WRITE,TSLOT

9TYPE OUT DIRECTIONS

lOT
BUF3
.BYTE READ,K8LOT
At

lOT

;WAIT FOR HIM TO TYPE A CARRIAGE RETURN,
;LINE FEED

A
.BYTE WAITR,KSLOT
lOT
RDEV
.BYTE INIT,RSLOT

jINIT READER

lOT
PDEV
.BYTE INIT,PSLOT

9INIT PUNCH

7-18

INPUT/OUTPUT PROGRAMMING

IOT
BUn
• BYTE
LOOP:

;START FIRST READ
F~EAD, F~SLOT

;READ INTO 2ND BUFFER

IOT
BUF2
• BYTE READ, F~SI... OT
BIT
BNE

~~EOF

;END OF FILE'?
nES
;NO

BUF 1 +2

BEGIN

;WRITE OUT THIS BUFFER

IOT
BUF1
~BYTE

WRITE,PSLOT
;WAIT TILL DEVICE HAS FINISHED

IOT
C

.BYTE WAITR,PSLOT
;READ INTO 1ST BUFFER

roT
BUF1
.BYTE READ,RSLOT
BIT
BNE

;END OF FILE?

:II:EOF, BUF2+2
BEGIN

IOT
BUF2
.BYTE WRITE,PSLOT

;WRITE OUT BUFFER 2

HlT

;WAIT TILL DEVICE HAS FINISHED

.BYTE WAITR,PSLOT
BR
LOOP
.END
BEGIN

7.9

7.9.1

lOX INTERNAL INFORMATION

Conflict Byte/Word

The lOX Conflict byte (in IOXLPT, Conflict Word) contains the status
(busy or free) of all devices as well as whether or not an echo is in
progress. Bit 0 is the echo bit, bits 1-6 (and 8 in IOXLPT) refer to
the corresponding codes for devices:
I f Bit is Set

Bit
Bit

}

Echo in progress

KBD busy

TTY busy

LSR busy

Device
Bit
Device

}

Bit

}

Device.

7-19

INPUT/OUTPUT PROGRAMMING
Bit
Device
Bit
Device
Bit
Device
Bit
Device

}
}
}
}

LSP busy

HSR busy

HSP busy

LPT busy

lOS

In IOXLPT, the Conflict Byte is expanded to a word in order to
accommodate the line printer, there being no bit 8 to correspond with
that device's code of 10 (octal) (the lowest available code for an
output device - see Section 7.9.5.1).
All possible
Conflicting Devices

Device

Conflict
Number

KBD

Echo, KBD, TTY, LSR, LSP

TTY

Echo, KBD, TTY, LSP

LSR

KBD, LSR

LSP

Echo, KBD, TTY, LSP

HSR

HSP

100

LPT

400

For each of the devices in the left hand column, all the possible
conflicts are listed along with their respective conflict numbers.
These numbers, representing bit patterns of the devices listed in
column two above, are used to resolve any conflicting requests for
devices. The appropriate number is masked with the conflict byte.
If
the result is zero, there are no conflicts and the device being tested
has its bit set allowing data transfer to begin.

7.9.2

Device Interrupt Table (DIT)

Each device interrupt handler has associated with it a Device
Interrupt Table (DIT) containing information that the handler needs:
DIT

Checksum

DIT+2

Byte size from buffer header

DIT+4

Address of Mode byte in buffer header

DIT+6

Byte Location Pointer

DIT+lO

Byte Count

DIT+l2

Device code
7-20

INPUT/OUTPUT PROGRAMMING
DIT+14

Real time done-address

DIT+16

Address of device's data buffer register

The device interrupt routines gain access to the proper data by means
of the DIT entry.
When a transfer is complete, they set the
appropriate bits in the buffer header pointed to by the DIT contents.

7.9.3

Device Status Table (DST)

The Device Status Table
(DST)
is used by lOX to check for EOF
conditions.
This table contains a word for each device indicating an
EOM condition with a 1. When an EOM condition is recognized on input,
lOX not only sets the appropriate bit in the buffer status byte
associated with the data transfer, it also records this occurrence in
the DST.
When a data transfer command is given, lOX checks the DST
for the EOM condition. If the appropriate word has a value of 1, lOX
sets EOF in the Status byte of the current-command buffer. Since EOF
is only possible for the LSR (code 3), and HSR (code 5), the words
corresponding to those devices are the only ones that can ever be set
to 1.

7.9.4

Teletype Hardware Tab Facility

If the Teletype model has a hardware tab facility, teleprinter
can be speeded up by:

output

For lOX, deleting the code from I.TTYCK+6 through I.TAB3+3.

For IOXLPT, skipping the code from I.IOLF through I.TAB3+3
(for the teleprinter only - not the line printer).

7.9.5

Adding Devices To lOX

In order to add a device to lOX the following tasks must be done:
1.

Assign a legal code- to the device

Modify the lOX tabies

Provide an interrupt routine to handle data for the device.

The line printer (in IOXLPT) will be used
this discussion.

example

throughout

7.9.5.1 Device Codes - The numbers from 7 to 17 (octal) are available
for new-device codes, with the exception of 10 (octal) in the IOXLPT
version. This code has been assigned to the line printer. The device
code must be odd for an input device and even for an output device.
This is so a check can be made for command/device correspondence;
i.e., for a Read from an input device or a write to an output device.
If the newest device was assigned a number that is higher than the
codes of all the other devices, I.MAXDEV must be redefined to that
value. This is so an out-of-range device specification in an Init
command can be detected. In IOXLPT, I.MAXDEV=lO.
7-21

INPUT/OUTPUT PROGRAMMING
Since each device code functions as an index in several word tables,
the entries relating to a given device must be placed at the same
relative position in each appropriate table. That is, the code~~mber
must indicate how many words into the table the entry for that device
will be found.
This, of course, means accounting for any unused space
preceding the entry, if the codes are not assigned in strict sequence.
Table entries for the line printer are found at the 10th (octal)
word
past the table tag, i.e., at Table+20.

7.9.5.2

Table Modification -

I.FUNC - Each entry is the octal value of the bit pattern in
the
device
Control/Status
Register
that enables the
corresponding device and/or any interrupt facility it has.
Bit setting this number
into the device's Control/Status
register turns the device on; bit clearing turns it off.
Determine this value for the device to be added, and place
the entry in the appropriate device position in the table.
For example, the line printer Control/Status Register has an
Interrupt Enable facility in bit 6. This pattern of 100 is
the LPT entry, and is located at I.FUNC+20.

I.SCRTAB - This table contains the addresses of the device
Control/Status registers.
The line printer entry I.LPTSCR
has the value 177514, and is located at I.SCRTAB+20.

I.DST - (Refer to Section 7.9.3.) Create an entry of 0 for
the device in the proper table location.
Inserting a word of
o at I.DST+20 created a device status entry for the line
printer.

I.CONSIT - An entry in this table is used to set or clear a
device's busy/free bit in the Conflict Byte (Conflict Word in
IOXLPT).
(See Section 7.9.1, and 5. below.) Each value is
obtained
by
setting
one bit only
the bit number
corresponding to the device number. The line printer, being
device 10(octal), has a value of 400(octal) (bit 10 set) and
is located at I.CONSIT+20.
In the lOX version without the line printer, entries to this
table are found
in the high-order bytes of Table I.CONFLC.
One more input device entry can be added to it.
In IOXLPT,
however,
I.CONSIT is a separate word table, allowing eight
more devices (four input and four output) to be added.
Byte
operations in the lOX I.CONSIT became word operations in
IOXLPT to adapt to this expansion.

I.CONFLC - (Refer to Section 7.9.1 on Conflict Byte/Word.)
Entries are bit patterns of conflicting devices.
Since the
line printer can only conflict with itself,
the I.CONFLC
entry is equal to the I.CONSIT entry. As in the I.CONSIT
table, byte operations were changed to word operations for
I.CONFLC in IOXLPT.

Create a DIT for the device
(refer to Section 7.9.2)
by
assigning a DIT label and seven words of O.
If it is an
output device, the address of the Device Buffer Register must
be added as an eighth word.

7-22

INPUT/OUTPUT PROGRAMMING
7.

I.INTAB - This is a table of DIT addresses. Place the label
of the DIT (mentioned in 6. above) in the correct position
in the table.
I.INTAB+20 contains the line printer entry
I.LPTDIT.

7.9.5.3 Interrupt Routines - write
interrupt routine for the device to:

(and

assign

label

to)

Get a character

Check for
register

Do character interpretation according to the device and mode

Get a character in or out of the buffer

Update lOX's Byte Count

Compare lOX's Byte Count to User's Byte Count and Buffer size
specification

Return for next character

errors

means

the

device

Control/Status

Place the label of the interrupt routine at the address of the device
vector, and follow it with the value of the interrupt priority in bits
7, 6, and 5.
I.LPTIR, the address of the line printer interrupt
routine,
is at location 200.
Location 202 contains the value 200
(indicating priority level 4).
If the device to be added is similar to the other single-character
devices, steps 3-7 above can be performed by lOX as indicated below:
There are two routines, I.INPUT and I.OUTPUT, that are called from the
interrupt routines.
These routines mainly perform common functions
for input and output devices. They are called as follows:
JSR

R5,I.INPUT

JSR

and

R5,I.OUTPUT

At the location following one of these calls is the DIT for the proper
device.
The routine is thus able to use R5 to reference the DIT
entries.
I.INPUT and I.OUTPUT also contain device-dependent code to perform
functions such as tab counters for the teleprinter and line printer,
and deletion of carriage returns in Formatted ASCII mode for the line
printer.
The device index value is used to identify the device.
For
the line printer, a symbol I.LPT, has been assigned the value 20 for
convenient reference to the device index.

7-23

CHAPTER 8
FLOATING POINT MATH PACKAGE OVERVIEW]

The new Floating-Point Math Package, FPMP-ll, is designed to bring the
2/4 word floating point format' of the FORTRAN environment to the paper
tape software system of the PDP-II. The numerical r6utines in FPMP-ll
are the same as those of the DOS/BATCH FORTRAN Operating Time System
(OTS) .
TRAP and error handlers have been included to aid in
interfacing with the FORTRAN routines.
FPMP-ll provides an easy means of performing basic
arithmetic
operations such as add, subtract, multiply, divide, and compare.
It
also provides transcendental functions
(SIN, COS,
etc.),
type
conversions
(integer to floating-point, 2-word to 4-word, etc.), and
ASCII conversions (ASCII to 2-word floating-point, etc.).
Floating-point notation is particularly useful for
computations
involving numerous multiply and divide operations where operand
magnitudes may vary widely. FPMP-ll stores very large and very small
numbers by saving only the significant digits and computing an
exponent to account for leading and trailing zeros.
To conserve core space in a small system, FPMP-ll can be tailored to
include only those routines needed to run a particular user program.
For more information on FPMP-ll, refer to the FPMP-ll
(DEC-II-NFPMA-A-D) and to Appendix H of this manual.

User's

FPMP is not currently available for the LSI-II (PDP-ll/03).
8-1

Manual

CHAPTER 9
PROGRAMMING TECHNIQUES

This chapter presents various programming techniques.
They can be
used to enhance your programming and to make optimum use of the PDP-II
processor. The reader is expected to be familiar with the PAL-II
assembly language (Chapters 1 & 2).

9.1

WRITING POSITION INDEPENDENT CODE

When a standard program is available for different users, it often
becomes useful to be able to load the program into different areas of
core and to run it there. There are several ways to do this:
1.

Reassemble the program at the desired location.

Use a relocating loader which accepts specially coded
from the assembler.

Have the program relocate itself after it is loaded.

Write code which is position independent.

binary

On small machines, reassembly is often performed. When the required
core is available, a relocating loader
(usually called a linking
loader) is preferable. It generally is not economical to have a
program relocate itself since hundreds or thousands of addresses may
need adjustment. Writing position independent code is usually not
possible because of the structure of the addressing of the object
machine. However, on the PDP-II, position independent code
(PIC)
is
possible.
PIC is achieved on the PDP-II by using addressing modes which form an
effective memory address relative to the Program Counter (PC). Thus,
if an instruction and its object(s) are moved in such a way that the
relative distance between them is not altered, the same offset
relative to the PC can be used in all positions in memory. Thus, PIC
usually references locations relative to the current location. PIC
may make absolute references as long as the locations referenced stay
in the same place while the PIC is relocated. For example, references
to interrupt and trap vectors are absolute, as are references to
device registers in the external page and direct references to the
general registers.

9-1

PROGRAMMING TECHNIQUES
9.1.1

position Independent Modes

There are three position independent modes or forms
They are:

instructions.

Branches
the conditional branches,
as well as
the
unconditional branch, BR, are position independent since the
branch address is computed as an offset to the PC.

Relative Memory References
of the form

any relative

memory

reference

CLR X
MOV X,Y
JMP X
is position independent because the assembler assembles it as
an offset indexed by the PC.
The offset is the difference
between the referenced location and the PC.
For example,
assume the instruction CLR 200 is at address 100:

:1.001
:1.021

;FIRST WORD OF CLR 200
;OFFSET = 200-:J.04

()O~5067

000074

The offset is added to the PC.
The PC contains
the address of the word following the offset.

104,

i.e.,

Although the form CLR X is position independent, the form CLR
@X is not.
Consider the following:

;CLEAR LOCATION A

CU~

(l1X

.wmw

• wmm 0

; PO I NTEF, TO A

The contents of location X are used as the address of the
operand in the location labeled A.
Thus, if all of the code
is relocated, the contents of location X must be altered to
reflect the new address of A.
If A, however, was the name
associated with some fixed location
(e.g.,
trap vector,
device register), then statements S and X would be relocated
and A would remain fixed.
Thus,
the following code is
position independent.

;ADDRESS OF SECOND WORD OF
; n~AP VECTOI:::
;CLEAR LOCATION A

A'"" 36
CL.F~

x:
3.

(i?X

; POINTEF, TO A

• WClFm A

Immediate Operands
The assembler addressing form #X
specifies immediate data,
that is,
the operand is in the
instruction.
Immediate data is position independent since it
is
a part of the instruction and
is moved with the

9-2

PROGRAMMING TECHNIQUES
instruction.
Immediate data is fetched using the PC
autoincrement mode.

the

As with direct memory references, the addressing form @#X is
not position independent.
As before, the final effective
address is absolute and points to a fixed location not
relative to the PC.

9.1.2

Absolute Modes

Any time a memory location or register is used as a pointer
to data,
the reference is absolute.
If the referenced data is fixed in memory,
independent of the position of the PIC (e.g., trap-interrupt vectors,
device registers),
the absolute modes must be used. l If the data is
relative to the PIC, the absolute modes must not be used unless the
pointers involved are modified.
The absolute modes are:
@X

Location X is a pointer

@#X

The immediate word is a pointer

( R)

The register is a pointer

(R)+

and - (R)

The register is a pointer

@(R)+

and @- (R)

The register points to a pointer

X (R)

R;i6 or 7

The base, X, modified by (R)
the address of the operand

@X(R)

The base, modified by (R), is a
pointer

The non-deferred index modes and stack operations require a
little
clarification.
As described in Sections 3.6.10 and 9.1.1, the form
X(7) is the normal mode to reference memory and is a relative mode.
Index mode, using a stack pointer (SF or other register) is also a
relative mode and may be used conveniently in PIC.
Basically
the
stack pointer points to a dynamic storage area and index mode is used
to access data relative to the pointer.
The stack pointer may be
initially set up by a position independent program as shown in Section
9.1.4.1.
In any case, once the pointer is set up,
all data on the
stack is referenced relative to the pointer.
It should also be noted
that since the form 0 (SP) is considered a relative mode so is its
equivalent @SP.
In addition, the forms (SP)+ and -(SP) are required
for stack pops and pushes.

9.1.3

Writing Automatic PIC

Automatic PIC is code which requires no alteration of addresses or
pointers.
Thus, memory references are limited to relative modes
unless the location referenced is fixed
(trap-interrupt vectors,
etc.).
In addition to the above rules,
the following must be
observed:
1.

Start the program with .=0 to allow easy relocation using the
Absolute Loader (see Chapter 6).

lWhen PIC is not being written, references to fixed locations
performed with either the absolute or relative forms.
9-3

may

PROGRAMMING TECHNIQUES
2.

All location setting statements must be of the form .=.±X or
.- function of tags within the PIC. For example, .=A+IO
where A is a local label.

There must not be any absolute location setting statements.
This means that a block of PIC cannot set up trap and/or
interrupt vectors at load time with statements such as:

.=34
.WORD

TRAPH,340

;TRAP VECTOR

The Absolute Loader, when it is relocating PIC, relocates all
data by the load bias (see Chapter 6). Thus, the data for
the vector would be relocated to some other place.
vectors
may be set at execution time (see Section 9.1.4).

9.1.4

Writing Non-Automatic PIC

Often it is not possible or economical to write totally automated PIC.
In these -cases, some relocation may be easily peformed at execution
time'. Some of the required methods of solution are presented below.
Basically, the methods operate by examining the PC to determine where
the PIC is actually located. Then a relocation factor can be easily
computed.
In all examples, it is assumed that the code is assembled
at zero and has been relocated somewhere else by the Absolute Loader.

9.1.4.1 Setting Up The Stack Pointer - Often the first task of a
program is to set the stack pointer
(SP).
This may be done as
follows:

BEG:

MOV
TST

PC,SP
-CSP)

;BEG IS THE FIRST INSTRUCTION OF
;THE PROGRAM
;SP=ADR BEG+2
;DECREMENT SP BY 2.
;A PUSH ONTO THE STACK WILL STORE
;THE DATA AT BEG-2.

9.1.4.2 Setting Up A Trap or Interrupt Vector - Assume the first word
of the vector is to point to location INT which is in PIC.

Mov PC,RO
ADD tINT-X-2,RO
MOV RO,@tVECT

;RO = ADR X+2
;ADD OFFSET
;MOVE POINTER TO VECTOR

The offset INT-X-2 is equivalent to INT-(X+2) and X+2 is the value of
the PC moved by statement X.
If PC is the PC that was assumed for
th~ program when loaded at 0, and if PC
is the current real PC, then
the calculation is:

Thus, the relocation factor, PC~-PCo' is added to the assembled
of INT to produce the relocated value of INT.

9-4

value

PROGRAMMING TECHNIQUES
9.1.4.3 Relocating Pointers - If pointers must be used, they may be
relocated as shown above.
For example, assume a list of data is to be
accessed with the instruction
ADD (RO)+,Rl
The pointer to the list, list L, may be calculated at
as follows:

MOV
ADD

PC,RO
tL-M-2,RO

execution

time

;GET CURRENT PC
;ADD OFFSET

Another variation is to gather all pointers into a table.
The
relocation factor may be calculated once and then applied to all
pointers in the table in a loop.

MOV
SUB
MOV
ADD
MOV
ADD
DEC
BGE

LOOP!

PC,RO

;RELOCATE ALL ENTRIES IN PTRTBL
;CALCULATE RELOCATION FACTOR
:fI:PTRTBL,Rl ;GET AND RELOCATE A POINTER
I:;:O,IU
TO PTRTBL
tTBLLEN,R2 ;OET LENGTH OF TABLE
RO,(IUH
;RELOCATE AN ENTRY
;COUNT
R2
;BRANCH IF NOT DONE
LOOP
:II:X+:;!,RO

Care must be exercised when restarting a program which relocates a
table of pointers.
The restart procedure must not include the
relocating again, i.e., the table must be relocated exactly once after
each load.

9.2

LOADING UNUSED TRAP VECTORS

One of the features of the PDP-II is the ability to trap on various
conditions such as illegal instructions, reserved instructions, power
failure, etc. However, if the trap vectors are not loaded with
meaningful information, the occurrence of any of these traps will
cause unpredictable results. By loading the vectors as indicated
below, it is possible to avoid these problems as well as gain
meaningful information about any unexpected traps that occur.
This
technique, which makes it easy to identify the source of a trap, is to
load each unu"sed trap vector with:
.=trap address
.WORD .+2,HALT
This will load the first word of the vector with the address of the
second word of the vector (which contains a HALT). Thus, for example,
a halt at location 6 means that a trap through the vector at location
4 has occurred. The old PC and status m"ay be examined by looking at
the stack pointed to by register 6.
The trap vectors of interest are:
Vector
Location

Halt At
Location

Meaning
Bus Error; Illegal Instruction;
Stack Overflow; Nonexistent
Memory; Nonexistent Device; Word
Referenced at Odd Address
Reserved Instruction
9-5

PROGRAMMiNG TECHNIQUES

9.3

Trace Trap Instruction {000003) or
T-bH. Set in Status Word (used by
ODT)

IOT Executed (used by IOX)

Power Failure or Restoration

EMT Executed (used by FPP-ll)

TRAP Executed

CODING'TECHNIQUE~

Because of the great flexibility in' PDP-II coding,
timeand
space-saving ways.' of performing operations may not be immediately
apparent.
Some comparisons follow.

9.3.1

Altering Register Contents

The techniques described in this section take advantage of the
automatic stepping feature of auto increment and autodecrement modes
when used especially in TST andCMP instructions. These instructions
do not alter operands.
However, it is important to make note of the
following:
•

These alternative ways of altering register
affect the condition codes differently.

contents

•

These techniques work properly only if the registers are
pointing to an. existing memory location; otherwise, a
trap is generated.

1 . . Adding 2 to a register might be accomplished

by ADD #2.,RO.
this takes two words, whereas TST (RO)+ which also
adds 2 to a register, takes only one word.

How~ver,

Subtracting 2 from a register can
complementary instructions SUB #2~~0
same conditions as in adding)2.

This can be extended to adding or
different registers, or 4 from
single-word instruction:

CMF-'
CMF-'
CMF-'
CMF-'
CMF-'

subtracting 2 from two
the same register, in one

;A[I[I 4 TO
;SUBTRACT
;A[lD 2 TO
;SUBTRACT
;AD[I 2 TO

(FW)·h (ROH
-·(Rl) ,-(R:L)
(ROH"-(F~:L>

-(R3)'-(R:L>
(R3H, (ROH

9-6

be
done
by
the
or TST -(RO) with the

RO
4 FROM Rl
RO, SUBTi=i:ACT 2 FROM Rl
2 FROM BOTH R3 AND Rl
BOTH F~3 AND RO

PROGRAMMING TECHNIQUES
4.

Variations of the examples above can be employed if the
instructions operate on bytes and one of the registers is the
Stack Pointer. These examples depend on the fact that the
Stack Pointer
(ai well as the PC) is always auto incremented
or autodecremented by 2, whereas registers RO-R5 step by 1 in
byte instructions.

CMPB (SP)+,(R3)t
CMPB -(R3),-(SP)
CMPB (R3)t,~(SP)
5.

9ADD 2 TO SP ~ND 1 TO R3
9SUBTRACT 1 FROM R3 AND 2 FROM SP
9ADD 1 TO R3, SUBTRACT 2 FROM SP

Popping an unwanted word off the processor stack (adding' 2 to
register 6)
and testing another value can be two separate
instructions or one combined instruction:

TST (SP)t
TST COUNT

;POP WORD
;SET CONDITION CODES FOR COUNT

MOV

;POP WORD & SET CODES FOR COUNT

or
COUNT~(SP)t

The differences are that the TST instructions take three
words and clear the Carry bit, and the MOV instruction takes
two words and doesn't affect the Carry bit.

9.3.2

Subroutines
1.

Condition codes set within a subreutine can be used to
conditionally branch upon return to the calling program,
since the RTS instruction does not affect condition codes.

JSR PC,X
BNE ABC

;CALL SUBROUTINE X
;BRANCH ON CONDITION SET
;IN SUBROUTINE X

·;SUBROUTINE ENTRY

CMP R2,DEF
RTS PC
2.

;TEST CONDITION
;RETURN TO CALLING PROGRAM

When a JSR first operand register is not the PC, data stored
following a subroutine call can be accessed within the
subroutine by referencing the register.
(The
register
contains the return address.)

JSR R5,Y
.WORD HIGH
.WORD LOW
;LATEST R5 VALUE WILL POINT HERE
Y:

MOV (RS)t,R2
MOV (RS)+,R4

;VALUE OF HIGH ACCESSED
;VALUE OF LOW ACCESSED

RTS RS

;RETURN TO LOCATION
;CONTAINED IN RS

9-7

PROGRAMMING TECHNIQUES
Another possibility is:

JSR R5,SUB
BR PSTARG

;LOW-ORDER BYTE
;ADDRESS, WHICH
;ADDRESS OF ARG
9ADDRESS OF ARG
9ADDRESS OF ARG

.WORD A
.WORD B
.WORD C

PSTARG:

SUB:

IS OFFSET TO RETURN
EQUALS NO. Qf ARGS.
A
B
C

;RETURN ADDRESS

MOVD@R5,COUNT

;GET NO. OF ARGS FROM LOW BYTE
90F BR (IF DESIRED).
;E.G., GET 6TH ARGUMENT
;GET 3RD ARGUMENT

MOV @14(R5),R2
NOV @6(R5),Rl

RTS R5
In the example above, the
advantages:

;RETURNS TO BRANCH WHICH JUMPS PAST
;ARG LIST TO REAL RETURN ADDRESS
branch

instruction

contributes

two

main

If R5 is unaltered when the RTS is executed, return will
always be to the branch instruction. This ensures a return
to the proper location even if the length of the argument
list is shorter or longer than expected.

The operand of the branch, being an offset past the
list, provides the number of arguments in the list.

argument

Arguments can be made sharable by separating the data from the main
code.
This is easily accomplished by treating the JSR and its return
as a subroutine itself:

CALL:

ARGLST:
JSR PC,ARGLST

JSR R5,SUB
DR PSTARG
.WORD A

•
3.

The examples above all demonstrate the calling of subroutines
from a non-reentrant program. The called subroutine can be
either reentrant or non-reentrant in each
case.
The
following example illustrates a method of also allowing
calling programs to be reentrant. The arguments and linkage
are first placed on the stack, simulating a JSR R5,SUB, so
that arguments are accessed from the subroutine via X(R5).
Return to the calling program is executed from the stack.

9-8

PROGRAMMING TECHNIQUES

CALL:

MOV R5,-(SP)
MOV JSBR,-(SP)

x:
RET:

MOV
MOV
JSR
MOV

BRN,-(SP)
SP,R5
PC, SUB
(SP >+, R5

;SAVE R5 ON STACK.
;PUSH INSTRUCTION JSR R6,@R5 ON
;STACK. PUSH ADDRESSES OF ARGU;MENTS ON STACK IN REVERSE ORDER
; (SEE BELOW).
;PUSH BRANCH INSTRUCTION ON STACK
;MOVE ADDRESS OF BRANCH TO R5.
9CALL SUB AND SAVE RETURN ON STACK.
9RESTORE OLD R5 UPON RETURN.
;DATA AREA OF PROGRAM.

JSBR:
BRN:

~JSR R6,@R5
BR .+N+N+2

;BRANCH PAST N WORD ARGUMENTS

The address of an argument can be pushed
several ways. Three are shown below.
a.

the

The arguments A, B, and C are read-only constants
in memory (not on the stack):

MOV :IJ:C,·· (SP )
MOV :g:B, - (SP )
MOV :fI:A,-(SP)
b.

stack
which

in
are

;PUSH ADDRESS OF C
;PUSH ADDRESS OF B
9PUSH ADDRESS OF A

Arguments A, B, and C have their addresses on the stack
the Lth, Mth, and Nth bytes from the top of the stack.

MOV N(SP),-(SP)
MOV M+2(SP),-(SP)
MOV L+4(SP),-(SP)

;PUSH ADDRESS OF C
;PUSH ADDRESS OF B
;PUSH ADDRESS OF A

Note that the displacements from the top of the stack are
adjusted by two for each previous push because the top of the
stack is being moved on each push.
c.

Arguments A, B, and C are on the stack at the Lth, Mth,
Nth bytes from the top but their addresses are not.

MOV
ADD
MOV
ADD
MOV
ADD

:fI:N+~~, _. (SP)
SP,@SP
:Jf:M+4 , .. (SF')
SP,@SP
:IJ:L. +6, ." ( SP )
SP,@SP

9-9

;PUSH DISPLACEMENT TO ARGUMENT
;CAL.CULATE ACTUAL ADm~ESS OF (".,
9ADDRESS OF B
;ADDREBS OF A

and

PROGRAMMING TECHNIQUES
When subroutine SUB is entered, the stack appears as follows:

RET
BR .+N+N+2
A
B

JSR R6.@R5

;BRANCH IS TO HERE

OLD R5

Subroutine SUB returns by means of an RTS R5, which places R5
into the PC and pops the return address from the stack into
R5. This causes the execution of the branch because R5 has
be"en loaded
(at location X) with the address of the branch ..
The JSR branched to then returns control to the calling
program, and in so doing, moves the current PC value into the
SP, thereby removing everything above the old R5 from the
stack. Upon return at RET, this too is popped, restoring the
original R5 and SP values.
4.

The next example is a recursive subroutine
(one that calls
itself).
Its function is to look for a matching right
parenthesis for every left parenthesis encountered.
.The
subroutine is called by JSR PC,A whenever a left parenthesis
is enco~ntered (R2 points to the character following it).
When a right parenthesis is found, an RTS PC is executed, and
if the right parenthesis is not the la&t legal one, another
is searched for.
When the final matching parenthesis is
found, the RTS returns control to the main program.
Al

MOVB (R2)f,RO
CMF'B :If: , ( , RO
BNE B
JSR F'C,A

;GET SUCCESSIVE CHARACTERS.
;LOOK FOR LEFT PARENTHESIS.
; FOUND'i'
;LEFT F'AREN FOUND, CALL SEL.
;GO LOOK AT NEXT CHARACTER

BF: A

CMF'B :J:'),RO

;LEFT F'AREN NOT FOUND, LOOK FOR
HUGHT F'AREN.
;FOUND'i' IF NOT, GO TO A.
;RETURN F'AREN FOUND. IF NOT LAST,
;GO TO B. IF LAST, GO TO MAIN PROGRAM.

BNE A

Ins F'C
5.

The example below illustrates the use of co-routines, called
by JSR PC,@(SP)+. The program uses double buffering on both
input and output, performing as follows:
Write 01
}
Read II
Process 12

concurrently

Write 02
}
Read 12
Process II

concurrently

JSR PC,@(SP)+ always performs a jump to the address specified
on top of the stack and replaces that address with the new
return address.
Each time the JSR at B is executed, it jumps
to a different location;
initially to A and thereafter to
the location following the JSR executed prior to the one at
B. All other JSR's jump to B+2.

9-10

PROGRAMMING TECHNIQUES

PC==%7
BEGIN:

(do 1/0 resets, inits, etc.)

lOT
;READ INTO 11 TO START PROCESS
.WORD 11
.BYTE READ,INSLOT
MOV *A,-(6)
;INITIALIZE STACK FOR FIRST JSR
iiDO 1/0 FOR 01 AND 11 OR 02 AND 12
JSR PC,@(6)+
perform

BR B

processin~

iiMORE I/O
fEND OF MAIN lOOP
; 1/0 CO··-ROUTINES
iiREAD INTO 12

lOT
.wmm 12
.BYTE READ,INSLOT

set parameters to process 11, 01

JSR PC,@(6H
iiRETURN TO PROCESS AT B+2
lOT
ii WRITE FROM 01
.WORD 01
.BYTE WRIT~,OUTSLOT
lOT
iiREAD INTO 11
.wmm 11
.BYTE READ,INSLOT
set parameters to process 12, 02
lOT

iiRETURN TO PROCESS AT B+2
WR ITE FROM O~~

.WORD 02
.BYTE WRITE, OUTSlOT
BR A
.READ INTO 12
6.

The trap handler, below, simulates a two-word JSR instruction
with a one-word TRAP instruction.
In this example, all TRAP
instructions in the program take an operand, and trap to the
handler address at location 34.
The table of subroutine
addresses (e.g., A, B, ... ) can be constructed as follows;

TABLE:

CALA::::. --TABLE
.WORD A

fCALLED BY:

TRAP CALA

CALB::: • _. TABLE
.wmm B

iiCALLED BY:

TRAP CALB

Another way to construct the table:

TABLE:
CALA~.-TABLE+TRAP

.WORD A

leALLED BY:

9-11

CAL A

PROGRAMMING TECHNIQUES
The TRAP handler for either of the above methods follows:

TRAP34:

MOV @SP,2(SP)
;REPLACE STACKED PS WITH Per.
SUB t2,@SP
;GET POINTER TO TRAP INSTRUCTION.
MOV @(SP)+,-(~P);REPLACE ADDRESS OF TRAP WITH
9 TRAP INSTRUCTION ITSELF.
ADD tTABLE-TRAP,@SP ;CALCULATE SUBROUTINE ADDR.
MOV @(SP)+,PC
9JUMP TO SUBROUTINE.

In the example above, if the third instruction had been written
MOV @(SP), (SP)
it would have taken an extra word since @(SP) is in
Index Mode and assembles as @O(SP].
In the final instruction, a jump
was executed by a MOV @(SP)+,PC because no equivalent JMP instruction
exists.
Following are some JMP and MOV equivalences (note that
affect condition codes).
JMP (R4)

MOV R4,PC

JMP @(R4)
(2 words)

MOV (R4) ,PC
(1 word)

none
JMP -(R4)

=
=

JMP @(R4)+

not

none
MOV (R4) ,PC

MOV -(R4) ,PC

none

MOV @(R4)+,PC

none

MOV @-(R4) ,PC

JMP X

MOV #X,PC

none

does

MOV @(R4) ,PC

JMP @-(R4)

JMP @X

JMP

MOV X,PC
MOV @X,PC

The TRAP handler can be useful, also, as a patching technique.
Jumping out to a patch area is often difficult because a two-word jump
must be performed. However, the one-word TRAP instruction may be used
to dispatch to patch areas. A sufficient number of slots for patching
should first be reserved _in the dispatch table of the TRAP handler.
The jump can then be accomplished by placing the address of the patch
area into the table and inserting the proper TRAP instruction where
the patch is to be made.

,-.iReplacing the saved PS loses the T-bit status. If a breakpoint has
been set on the TRAP instruction, ODT will not gain control again to
reinsert the breakpoints because the T-bit trap will not occur.

9-12

APPENDIX A
ASCII CHARACTER SET

EVEN
PARITY
BIT

7-BIT
OCTAL
CODE

o
1

000
001

NUL
SOH

002

STX

003

ETX

004

EOT

o
1
1

005
006
007
010

ENQ
ACK
BEL
BS

o
o

011
012

HT
LF

013
014

VT
FF

015

016

017

CHARACTER

o
o

020
021

DLE
DCl

022

DC2

023

DC3

024

DC4

025

NAK

026
027

SYN
ETB

030
031
032
033

CAN
EM
SUB
ESC

o
1
1

REMARKS
NULL, TAPE FEED, CONTROL SHIFT P.
START OF HEADING; ALSO SOM, START OF
MESSAGE, CONTROL A,
START OF TEXT; ALSO EOA, END OF ADDRESS,
CONTROL B,
END OF TEXT: ALSO EOM, END OF MESSAGE
CONTROL C,
END OF TRANSMISSION (END): SHUTS OFF TWX
MACHINES, CONTROL D,
ENQUIRY (ENQRY); ALSO WRU, CONTROL E,
ACKNOWLEDGE. ALSO RU, CONTROL F.
RINGS THE BELL. CONTROL G.
BACKSPACE: ALSO FEO, FORMAT EFFECTOR.
BACKSPACE SOME MACHINES, CONTROL H.
HORIZONTAL TAB. CONTROL I.
LINE FEED OR LINE SPACE (NEW LINE) :
ADVANCES PAPER TO NEXT LINE, DUPLICATED
BY CONTROL J.
VERTICAL TAB (VTAB). CONTROL K.
FORM FEED TO TOP OF NEXT PAGE (PAGE).
CONTROL L.
CARRIAGE RETURN TO BEGINNING OF LINE.
DUPLICATED BY CONTROL M.
SHIFT OUT: CHANGES RIBBON COLOR TO RED.
CONTROL N.
SHIFT IN: CHANGES RIBBON COLOR
TO BLACK. CONTROL 0
DATA LINK ESCAPE. CONTROL P (DCO).
DEVICE CONTROL 1, TURNS TRANSMITTER
(READER) ON, CONTROL Q (XON).
DEVICE CONTROL 2, TURNS PUNCH OR AUXILIARY ON. CONTROL R (TAPE, AUX ON).
DEVICE CONTROL e, TURNS TRANSMITTER
(READER) OFF, CONTROL S (XOFF).
DEVICE CONTROL 4. TURNS PUNCH OR AUXILIARY OFF. CONTROL T (TAPE, AUX OFF)
NEGATIVE ACKNOWLEDGE: ALSO ERR. ERROR.
CONTROL U.
SYNCHRONOUS IDLE (SYNC). CONTROL V.
END OF TRANSMISSION BLOCK: ALSO LEM.
LOGICAL END OF MEDIUM. CONTROL W.
CANCEL (CANCL). CONTROL X.
END OF MEDIUM. CONTROL Y.
SUBSTITUTE. CONTROL Z.
ESCAPE. PREFIX.

A-I

ASCII CHARACTER SET
1
0
0
1
1
0
0
1
0
1
1
0
0
1
1
0
1
0
0
1
0

1
1
0
1

b
0
1
1
O.

0
1
0
1
1
0
1
0
0
1
0
1
1
0
0
1
1
0
1
0
0
1
0
1
1
0
1
0
0
1
1
0
0
1

034
035
036
037
040
041
042
043
044
045
046
047
050
051
052
053
054
055
056
057
060
061
062
063
064
065
066
067
070
071
072
073
074
075
076
077
100
101
102
103
104
105
106
107
110
111
112
113
114
115
116
117
120
121
122
123
124
125
126
127
130
131
132
133

FS
GS
RS
US
SP

FILE SEPARATOR. CONTROL SHIFT L.
GROUP SEPARATOR. CONTROL SHIFT M.
RECORD SEPARATOR. CONTROL SHIFT N.
UNIT SEPARATOR. CONTROL SHIFT O.
SPACE.

#
$
%
&

ACUTE ACCENT OR APOSTROPHE.
(
)

1
2
3
4
5
6
7
8
9
i

<
>
?
@

A
B

C
D
E
F
G
H
I
J

K
L
M

N
0

P
Q

R
S
T
U
V
W
X
y
Z

[

SHIFT K
A-2

ASCII CHARACTER SET
0
1
1
0
0
0
0
1

134
135
136
137
140
175
176
177

]
t

SHIFT L
SHIFT M
SHIFT N

-+-

DEL

ACCENT GRAVE.
THIS CODE GENERATED BY ALT MODE.
THIS CODE GENERATED BY ESC KEY (IF PRESENT)
DELETE, RUB OUT.
LOWER CASE ALPHABET FOLLOWS (TELETYPE
MODEL 37 ONLY).

1
1
0
1
0
0
1
1
0
0
1
0
1
1
0
1
0
0
1
0
1
1
0
0
1
1
0
1

141
142
143
144
145
146
147
150
151
152
153
154
155
156
157
160
161
162
163
164
165
166
167
170
171
172
173
174

a
b

c
d

e
f
9
h

i
j
k
1
m
n
0

P
q
r

s
t
u

v
w

x
Y

A-3

APPENDIX B
PAL-lIS ASSEMBLY LANGUAGE AND ASSEMBLER

B.l

TERMINATORS

The list below defines all characters which are considered to be
terminators.
The order of the list implies the descending hierarchy
of significance.
Character

Function

CTRL/FORM

Source line terminator.

LINE FEED

Source line terminator.

RETURN

Source line terminator
Label terminator
Direct assignment delineator
Register term delineator

TAB

--?\

Item terminator
Field terminator

BLANK or
SPACE

Item terminator
Field terminator

Immediate expression field indicator

Deferred addressing indicator
Initial register field indicator
Terminal register field indicator
Operand field separator
Comments field delimiter

Arithmetic addition operator
Arithmetic subtraction operator

Logical AND operator
Logical OR operator

Double ASCII text indicator
Single ASCII text indicator.

B-1

PAL-lIS ASSEMBLy LANGUAGE AND ASSEMBLER
B.2

ADDRESS MODE SYNTAX

r is an integer between 0 and 7.
R is a register expression, E is an expression, ER is either a
register expression or an absolute expression in the range of 0 to 7.
Address Address Mode
Mode
Name
Number

Symbol in
Operand
Field

Meaning

Deferred Register @R or (R)Register
address.

Autoincrement

(ER) +

The contents
of
the
register
specified
by ER is incremented
after being used as the address of
the operand.

Deferred
Autoincrement

@(ER)+

ER contains the pointer
address of the operand.
incremented after use.

Autodecrement

-(ER)

The contents of register ER is decremented before it is used as the
address of the operand.

Deferred
Autodecrement

@-(ER)

Index by the
register
Specified

The contents of register ER is decremented before it is used as the
pointer to the address of the operand.

E(ER)

Deferred index
by the register
specified

E plus the contents of the register
specified, ER, is the address of
the operand-.-

@E(ER)

E. added to ER gives the pointer
the address of the operand.

Immediate Operand #E

E is the operand.

Absolute address @#E

E is the operand address.

Relative address

E is the address of the operand.

Deferred relative address.

B.3

E
@E

contains

the. operand

to the
ER is

~ is the pointer to the address
the operand.

INSTRUCTIONS

The tables of instructions which follow are grouped according to the
operands they take and according to the bit patterns of their
op-codes.

B-2

PAL-lIS ASSEMBLY LANGUAGE AND ASSEMBLER
In the representation of op-codes, the following symbols are used:
SS

Source operand

specified by a 6-bit
address mode

Destination operand

specified by a 6-bit
address mode

8-bit offset to a
location

(branch instructions)

Integer between 0 and 7

representing a general
register

Symbols used in the description of instruction operations are:
SE

Source effective address

Destination effective address

()

contents of

becomes

Processor Status word

The condition codes in the processor status word (PS) are affected by
the instructions;
these condition codes are represented as follows:
N

Begative bit:

set i f the result is negative

Zero bit:

set i f the result is zero

OVerflow bit:

set i f the result had an overflow

~arry

bit:

set i f the result had a carry

In the representation of the instruction's
codes, the following symbols are used:

effect

the

condition

Conditionally set
Not affected

Cleared

Set

To set conditionally means to
determine the state of the code.

use

the

instruction's

result

Logical operators are represented by the following symbols:
Inclusive OR
Exclusive OR
&

AND
(used over a symbol) NOT (i.e., l's complement)

B-3

PAL-lIS ASSEMBLY LANGUAGE AND ASSEMBLER
B.3.1

OP A,A

Double Operand Instructions

Op-code

MNEMONIC

Stands for

01ssdd
llssdd

mov
movb

move
move Byte

02SSDD
12SSDD

CMP
CMPB

03SSDD
13SSDD

Operation
(SE)+DE

Condition Codes
N Z V C

CoMPare
(SE)-(DE)
CoMPare Byte

BIT
BITB

BIt Test
(SE)&(DE)
BIt Test Byte

04SSDD
14SSDD

BIC
BICB

BIt Clear
(SE) & (DE)-+DE
BIt Clear Byte

05SSDD
15SSDD

BIS
BISB

BIt Set
(SE)
BIt Set Byte

(DE)-+DE

06SSDD
16SSDD

ADD
SUB

ADD
SUBtract

(SE) + (DE)-+DE
(DE) - (SE)-+DE

*
*

B.3.2

Single Operand Instructions

Op-code

MNEMONIC

Stands for

0050DD
1050DD

CLR
CLRB

CLeaR
CLeaR Byte

0051DD
1051DD

COM
COMB

0052DD
1052DD

*
*

OP A
Operation:

O-+-DE

Condition Codes
N Z V C
0

COMplement (DE)-+-DE
COMplement Byte

INC
INCB

(DE) + l-+-DE
INCrement
INCrement Byte

0053DD
1063DD

DEC
DECB

DECrement
(DE) -l-+-DE
DECrement Byte

0054DD
1054DD

NEG
NEGB

NEGate
NEGate Byte

0055DD
1055DD

ADC
ADCB

ADd Carry
(DE) + (C)-+DE
ADd Carry Byte

0056DD
1056DD

SBC
SBCB

SuBtract Carry (DE) - (C)-+DE
SuBtract Carry Byte

0057DD
1057DD

TST
TSTB

TeST
TeST Byte

(DE) + l-+-DE

(DE) - O-+DE

B-4

PAL-llS ASSEMBLY LANGUAGE ANO ASSEMBLER
B.3.3

Rotate/Shift
C

006000

ROR

ROtate Right

106000

RORB

ROtate Right
Byte

006100

ROL

ROtate Left

106100

ROLB

ROtate Left
Byte

006200

ASR

Ari thmetic
Shift Right

1 '"

'" '" '"
'" '" '" '"

~ '" '" '" '"
'"
'"

r£H

'" '" '" '"

106200

ASRB

Ari thmetic
Shift Right
Byte

'" '" '" '"

006300

ASL

Arithmetic
Shift Left

'" '" '" '"

106300

ASLB

Arithmetic
Shift Left
Byte

000100

JMP

JuMP

0003DO

SWAB

SWAp Bytes

B.3.4

Operation Instructions

ict

,ll~o

'" '" '"

OE-PC

C{5J

'" '"

Op
Condi tion Codes
N Z V C

Op-Code

MNEMONIC

Stands for

Operation

000000

HALT

The computer stops
all functions.

000001

WAIT

The computer stops and waits for an
interrupt.

000002

RTI

ReTurn from
Interrupt

The PC and PS are
popped off the SP
stack:
( (SP) ) .... PC
(SP)+2 .... SP

( (SP) ) .... PS

'" '" * '"

000003

breakpoint
trap

Trap to location 14 •
This is used to
call OOT-ll.

000004

IOT

Input/Output
Trap

Trap to location
20. This is used
to call IOX.

000005

RESET

Returns all I/O
device handlers to
power-on state.
B-5

'" *

PAL.,...llS ASSEMBLY LANGUAGE"AND ASSEMBLER
Trapping

Op or Op

E where OO

Assemble what follows up to the
terminating
.ENDC,
if
the
expression E is greater than O.

.IFGE

E~O

Assemble what follows up to the
terminating
.ENDC,
if
the
expression E is greater than or
equal to o.

.IFDF

NAME

IF NAME
defined

Assemble what follows up to the
terminating .ENDC if the symbol
NAME is defined.

.IFNDF

NAME

IF NAME
undefined

Assemble what follows up to the
terminating .ENDC if the symbol
NAME is undefined.

. EN DC

none

End of
Conditional

Terminates the range of a conditional directive.

B-9

50
ASCII

the
the

PAL-llS ASSEMBLY LANGUAGE AND ASSEMBLER
B.S

ERROR CODES

Error Code

Meaning

Addressing
error.
Tnstruction
is
relocation errors.

Bounding error. Instructions or word data
being assembled at an odd address in memory.

Doubly-defined symbol referenced.
Reference was
made to a symbol which is defined more than once.

Illegal character detected.
Ilegal characters
which are also non-printing are replaced by a ?
on the listing.

Line buffer overflow.
are ignored.

An
address
within
the
incorrect.
Also
includes
are

All extra characters beyond

Multiple definition of a label.
A label was
encountered which was equivalent (in the first six
characters) to a previously encountered label.
Number containing an 8 or 9 was not terminated
point.

a decimal

Phase error. A label's definition or value varies
from one pass to another.

Questionable syntax. There are missing arguments
or the instruction scan was not completed, or a
carriage return was not followed by a linefeed or
form feed.

~egister-type

Symbol table overflow.
When the quantity of
user-defined symbols exceeds the allocated space
available in the user's
symbol
table,
the
assembler outputs the current source line with the
S error code, then returns to the command string
interpreter to await the next command string to be
typed.

Truncation err.or. More than the allotted number
of bits were input so the leftmost bits were
truncated. T error does not occur for the result
of an expression.

Undefined symbol.
An
undefined
symbol
encountered
during
the
evaluation
of
expression.
Relative to the expression,
undefined symbol is assigned a value of zero.

error.
An invalid use
reference to a register has been made.

B-IO

was
an
the

PAL-lIS ASSEMBLY LANGUAGE AND ASSEMBLER
B.6

INITIAL OPERATING PROCEDURES

Loading:

Use Absolute Loader. The start address
Loader must be in the console switches.

the

Storage Requirements:PAL-llS uses 8K of memory.
Starting:

Immediately upon loading, PAL-lIS
control and initiate dialogue.

will

Ini tial Dialogue:
Printout

Inquiry

What is the input device of the

What is the output device of the .!!inary object tape?

What is the output device of the assembly

What is the output device of the symbol Table?

~ource

Each of these questions may be answered by anyone
characters:
Character

the

~isting?

following

Answer Indicated

!eleprinter keyboard

~ow-speed

gigh-speed reader or punch

Line Printer

Each of these answers
indicating options:
Option Typed

symbolic tape?

reader or punch

may

followed

the

other

characters

Function to be performed

on pass 1

on pass 2

on pass 3

errors to be listed on the Teletype on the same
(meaningful only for *B or *L) .

Each answer is terminated by typing the RETURN key.
answer will delete the function.

pass

A RETURN alone as

Dialogue During Assembly:
Printout

Response

EOF ?

A .END
Place next tape in reader and type RETURN.
statement may be forced by typing E followed by
RETURN.

END ?

Start next pass by placing first tape
typing RETURN.

B-11

reader

and

PAL-lIS ASSEMBLY LANGUAGE AND ASSEMBLER
EOM ?

If the end-of-medium is on the listing device, the
device
may be readied and the assembly may be
continued by typing RETURN.
If the end-of-medium is on the binary device, the
assembler will discontinue the assembly and restart
itself.

Restarting:

Type CTRL/P.
again.

The initial

B-12

dialogue

will

started

APPENDIX C
PAL-IIA ASSEMBLY LANGUAGE AND ASSEMBLER

C.l

SPECIAL CHARACTERS

Character

Function

form feed

Source line terminator

line feed

Source line terminator

carriage return

Source statement terminator
Label terminator
Direct assignment indicator

tab

Item terminator
Field terminator

space

Ite"m terminator
Field terminator

Immediate expression indicator

Deferred addressing indicator
Initial register indicator
Terminal register indicator
Operand field separator
Comment field indicator

Arithmetic addition operator
Arithmetic subtraction operator

Logical AND operator
Logical OR operator

Double ASCII character indicator
Single ASCII character indicator
Assembly location counter

C-1

PAL-IIA ASSEMBLY LANGUAGE AND ASSEMBLER
C.2

ADDRESS MODE SYNTAX

n is an integer between 0 and 7 representing a register.
R is a
register expression, E is an expression, ER is either a register
expression or an expression in the range 0 to 7.

Format

Address
Mode
Name

Address
Mode
Number

Meaning

@R or (ER)

Deferred Register

(ER)+

Autoincrement

The contents
of
the
register specified by ER
are incremented
after
being
used
as
the
address of the operand.

@(ER)+

Deferred Autoincrement

ER contains the pointer
to the address of the
operand.
ER
is
incremented after use.

-(ER)

Autodecrement

The contents of register
ER
are
decremented
before being used as the
address of the operand.

Deferred 'Autodecrement

The contents of register
ER
are
decremented
before being used as the
pointer to the address
of the operand.

E (ER)

Index

E plus the contents of
the register specified,
ER, is the address of
the operand.

@E (ER)

Deferred Ind,ex

E added to ER gives the
pointer to the address
of the operand.

Immediate

E is the operand.

@#E

Absolute

E is the address of
operand.

the

Relative

E is the address of
operand.

the

Deferred Relative

E is the pointer to the
address of the operand.

(ER)

C-2

the

PAL-IIA ASSEMBLY LANGUAGE AND ASSEMBLER
C.3

INSTRUCTIONS

The instructions which follow are grouped according
they take and the bit patterns of their op-codes.

the

operands

In the representation of op-codes, the following symbols are used:
SS

Source operand specified by a 6-bit address mode.

Destination operand specified by a
mode.

8-bit offset to a location (branch instructions)

Integer between 0 and
register.

6-bit

representing

address

general

Symbols used in the description of instruction operations are:
SE

Source Effective address

Destination Effective address

()

Contents of
Is transferred to

Processor Status word

The condition codes in the processor status word (PS) are affected by
the instructions. These condition codes are represented as follows:
set i f the result is negative

~egative

Zero bit:

set i f the result is zero

oVerflow bit:

set i f the operation caused an
overflow

.£arry bit:

set i f the operation caused a
carry

bit:

In the representation of the instruction's
codes, the following symbols are used:

effect'

the

condition

Conditionally set
Not affected

Cleared

Set

To set conditionally means to use the instruction's result to
determine the state of the code (see the PDP-II Processor Handbook).
Logical operations are represented by the following symbols:
Inclusive OR
Exclusive OR
&

AND
(used over a symbol) NOT (i.e., l's complement)
C-3

PAL-llA ASSEMBLY LANGUAGE AND ASSEMBLER
C.3.l

Op A,A

Double-Operand Instructions

Op-Code

MNEMONIC

Stands for

OlSSDD
llSSDD

MOV
MOVB

MOVe
MOVe Byte

(SE) -.. DE

02SSDD
l2SSDD

CMP
CMPB

CoMPare
CoMPare Byte

_(SE) - (DE)

03SSDD
13SSDD

BIT
BITB

BIt Test
BIt Test Byte

(SE)

(DE)

04SSDD
14SSDD

BIC
BICB

BIt Clear
BIt Clear Byte

(SE)

(DE) -.. DE

05SSDD
l5SSDD

BIS
BISB

BIt Set
BIt Set Byte

(SE) ! (DE) -.. DE

06SSDD
l6SSDD

ADD
SUB

ADD
SUBtract

(SE) + (DE) -.. DE
(DE) - (SE) -.. DE

*
*

C.3.2

Operation

Status Word
Condition Codes
N
Z
V
C

*
*

Op A

Single-Operand Instructions

Op-Codes MNEMONIC

Stands for

Operation

0050DD
lO50DD

CLR
CLRB

CLeaR
CLeaR Byte

005lDD
lO5lDD

COM
COMB

COMplement
(DE)
COMplement Byte

0052DD
lO52DD

INC
INCB

INCrement
INCrement Byte

(DE)+l

0053DD
lO53DD

DEC
DECB

DECrement
DECr.ement Byte

0054DD
lO54DD

NEG
NEGB

0055DD
lO55DD

Status Word
Condition Codes
C
N
Z
V
0

(DE)-l

NEGate
NEGate Byte

(DE)+l

ADC
ADCB

ADd Carry
ADd Carry Byte

(DE) + (C) -.. DE

0056DD
lO56DD

SBC
SBCB

SuBtract Carry
(DE)-(C) -.. DE
SuBtr.act Carry Byte

0057DD
lO57DD

TST
TSTB

TeST
TeST Byte

DE
DE

(DE)-O -"DE

C-4

PAL-11A ASSEMBLY LANGUAGE AND ASSEMBLER
C.3.3

Op-Code

Rotate/Shift Instructions

MNEMONIC

Op A

Stands for

Operation
15

0060DD

1060DD

0061DD

1061DD

ROR

RORB

ROL

ROLB

ROtate Right

1062DD

0063DD

1063DD

ASR

ASRB

ASL

ASLB

~2,o *

cD='

ROtate Right
even or odd byte
Byte
~,

cD~

ROtate Left

ROtate Left
Byte

rD.J
C

0062DD

Status Word
Condition Codes
Z
N
V
C

~l~
c

even or odd byte

Arithmetic
Shift Left
Byte

0003DD

SWAB

SWAp Bytes

~TI

~d:

Arithmetic
Shift Left

JuMP

even or odd byte

Arithmetic
Shift Right
Byte

JMP

110

1514

~I~

Arithmetic
Shift Right

0001DD

even or odd byte

~:I
DE
15

0 I

: «0

°I

8·7

C.3.4

Operate Instructions

Op-Code

MNEMONIC

Operation

000000

HALT

The computer stops all
functions.

000001

WAIT

The computer stops and
waits for an interrupt.

Stands for

C-5

Status Word
Condition Codes
Z
C
N
V

PAL-llA ASSEMBLY LANGUAGE AND ASSEMBLER
000002

RTI

ReTurn
from
Interrupt

The PC and PS are popped
off the SP stack:
( (SP) ) -+
(SP) +2 .....
«SP» .....
(SP) +2 .....

PC
SP
PS
SP

RTI is also used to return from a trap.
000005

C.3.5

RESET

Trap Instructions

Returns all I/O devices
to power-on state.

Op or Op E

Where

0~E~377 8

Status Word
Condition Codes
C
N
Z
V

Op-Code

MNEMON;IC

*000003

(none)

(breakpointTrap to location 14. This *
is used to call ODT.
trap)

*000004

IOT

Input/Out- Trap to location 20. This
put Trap
is used to call IOX.

104000104377

EMT

EMulator
Trap

Trap to location 30. This *
is used to call system programs.

104400
104777

TRAP

Trap to location 34. This *
is used to call any routine
desired by the programmer.

Stands for

Operation

*Op (only)
CONDITION CODE OPERATES
Op-Code

MNEMONIC

Stands for

000241

CLC

CLear Carry bit in PS.

000261

SEC

SEt Carry bit.

000242

CLV

CLear oVerflow bit.

000262

SEV

SEt oVerflow bit.

000244

CLZ

CLear Zero bit.

000264

SEZ

SEt Zero bit.

000250

CLN

CLear Negative bit.

000270

SEN

SEt Negative bit.

000254

CNZ

CLear Negative and Zero bits.

000257

CCC

Clear all Condition Codes.

000277

SCC

Set all Condition Codes.
C-6

PAL-IIA ASSEMBLY LANGUAGE AND ASSEMBLER
C.3.6

Op-Code

Branch Instructions Op E where-128 l0 S. (E-.-2)/2s.l27 l0

MNEMONIC

Condition to be met i f
branch is to occur

Stands for

0004XX

BRanch always

OOlOXX

BNE

Branch i f Not Equal
(to zero)

Z=O

0014XX

BEQ

Branch i f EQual
zero)

Z=l

0020XX

BGE

Branch if Greater than
or Equal (to zero)

N(DV=O

0024XX

BLT

Branch if Less Than
(zero)

N(DV=l

0030XX

BGT

Branch if Greater Than
(zero)

(N(DV) =0

0034XX

BLE

Branch i f Less than or
Equal (to zero)

(N(DV) =1

1000XX

BPL

Branch i f PLus

N=O

1004XX

BMI

Branch i f MInus

N=l

1010XX

BHI

Branch i f HIgher

1014XX

BLOS

Branch i f LOwer or Same

1020XX

BVC

Branch i f oVerflow Clear

V=O

1024XX

BVS

Branch i f oVerflow Set

V=l

1030XX

BCC (or Branch if Carry Clear
BHIS)
(or Branch i f HIgher or
Same)

C=O

1034XX

BCS (or Branch i f Carry Set (or
BLO)
Branch i f LOwer)

C=l

C.3.7

(to

Subroutine Call Op ER, A

Op-Code MNEMONIC

Stands for

004RDD

Jump to SubRoutine

JSR

Operation
Push register
stack,put
the
register:

on
PC

the
in

SP
the

temporary
DE (TEMP) - a
storage
register
to
internal
processor.
(SP) -2+SP
(REG) -+- (SP)
(PC) -+- REG
(TEMP) -+- PC
C-7

PAL-IIA ASSEMBLY LANGUAGE AND ASSEMBLER
C.3.8

Subroutine Return Op ER

Op-Code MNEMONIC

Stands for

00020R

ReTurn from Subroutine

C.4

RTS

Operation
Put register contents into
and pop old contents from
stack into register.

PC
SP

ASSEMBLER DIRECTIVES

Op-Code MNEMONIC

Stands for

Operation

.EOT

End Of Tape

Indicates the physical end
the source input medium

. EVEN

EVEN

Ensures that
the
assembly
location counter is even by
adding I if it is odd

.END mEND
(m optional)

. WORD
E,E, ..
E, E, ...

Indicates the
physical
and
logical end of the program and
optionally specifies the entry
point (m)

WORD

Generates words of data

(the void operator)

Generates words of data

.BYTE
BYTE
E,E, .•.
.ASCII
ASCII
Ixxx • .• xl

Generates bytes of data
Generates
7-bit
for
the
character
enclosed by delimiters

Ascii
text

C.S

ERROR CODES

Error Code
A

Meaning
~ddressing

error.

An address within

the

instruction

incorrect.
B

Bounding error.
Instructions or word
assembled at an odd address in memory.

data

Doubly-defined symbol referenced. Reference was made to a
symbol which is defined more than once.

Illegal character detected. Illegal characters which are
also non-printing are replaced by a? on the listing.

Line buffer overflow.
are ignored.

Multiple definition of a label. A label was encountered
which was equivalent
(in the first six characters) to a
previously encountered label.

~umber

Ehase error. A label's definition or
one pass to another.

Extra characters (more

are

than

being

7210 )

containing an 8 or 9 has a decimal point missing.

C-8

value

varies

from

PAL-llA ASSEMBLY LANGUAGE AND ASSEMBLER

C.6

Q-

Questionable syntax. There are missing arguments or the
instruction scan was not completed, or a carriage return
was not followed by a line feed or form feed.

Symbol-table overflow. When the. quantity of user-defined
symbols exceeds the allocated space available in the
user's symbol table, the assembler outputs the current
source line with the S error code, then returns to the
command string interpreter to await the next command
string to be typed.

Truncation error. A number was too big for the allotted
number of bits;
the leftmost bits were truncated. T
error does not occur for the result of an expression.

Undefined symbol. An undefined symbol was encountered
during the evaluation of an expression. Relative to the
expression, the undefined symbol is assigned a value of
zero.

INITIAL OPERATING PROCEDURES

Loading:

Use Absolute Loader (see Chapter 6). Make sure that the
start address of the absolute loader is in the switches
when the assembler is loaded.

Storage Re- PAL-llA exists in 4K and 8K versions.
quirements:
Starting
Initial
Dialogue:

Immediately upon loading, PAL-llA will be in
initiate dialogue.
Printout

control

and

Inquiry

What is the input
symbolic tape?

What is the output device of the
tape?

~inary

What is the
:histing?

the

What is the output device of the symbol !able?

Each of these questions may
characters:

output

answered

Character

device

one

Answer Indicated

~eletype

:how-speed reader or punch

~igh-speed

line Printer (8K version only)

keyboard

reader or punch

C-9

the

Source
object

assembly

following

PAL-IIA ASSEMBLY LANGUAGE AND ASSEMBLER
Each of these answers may be followed by other
options:
Option Typed

characters

indicating

Function to be Performed

on pass 1

on pass 2

on pass 3

errors to be listed on the Teletype
on the same pass (meaningful or *B
or *L only)

Each answer is terminated by typing the RETURN key.
answer will delete the function.

A RETURN alone as

Dialogue during assembly:
Printout

Response

EOF ?

Place next tape in reader and type RETURN.
A
.END statement may be forced by typing E
followed by RETURN.

END ?

Start next pass by placing
reader and typing RETURN.

EOM ?

If listing on HSP or LPT,
paper and type RETURN.
start assembly again.

Restarting:

Type CTRL/P. The
started again.

C-IO

initial

first

tape

replenish tape or
If binary on HSP,
dialogue

will

APPENDIX D
TEXT EDITOR, ED-II

D.l

INPUT/OUTPUT COMMANDS
R

Reads a page of text from input device, and appends it to
the contents (if any) of the page buffer. Dot is moved to
the beginning of the page and Marked.
(See Band M below.)

Opens the input device when the user
input with a new tape in the reader.

wishes

continue

ARGUMENTS

~n}
o

Lists the character
string

(n)

beginning at Dot and ending
with nth line feed character.

(-n)

beginning with 1st character
following the (n+l)th previous
line feed and terminating at
Dot.

(0)

beginning with 1st character
of current line and ending at
Dot.

(@)

bounded by Dot and the Marked
location (see M).

(/)

beginning at Dot and ending
with the last character in the
page.

~n}
o

Punches the character
string

/
F

Outputs a Form Feed character
tape.

Punches four inches of Trailer (blank tape) n times.

Punches contents of the page buffer (followed by a trailer
if a form feed is present), deletes the contents of the
buffer, and reads the next page into the page buffer.
It
does this n times. At completion, Dot and Mark are located
at the beginning of the page buffer.

Lists the entire line containing Dot (i.e.,
line feed to next line feed or form feed).

Same as -IL. If Dot is located at the beginning of a line,
this simply lists the line preceding the current line.

~iststhe line following the current line.

D-l

and

four

inches

from

blank

TEXT EDITOR, ED-II
D.2

POINTER-POSITIONING COMMANDS
B

Moves Dot to the beginning of the page.

Moves Dot to the end of the page.

Marks the current position of Dot for later reference in a
command using the argument @. Certain commands implicitly
move Mark.
(n)
(-n)
(0)
(@)
(/)

forward past n characters
backward past n characters
to the beginning of the current line
to the Marked location
to the end of the page

(n)

-n

(-n)

forward past n ends-of-lin~s
to first characte~ following the (n+l)th
previous end-of-line
to the beginning of current line
to the Marked location
to the end of the page

Moves Dot:

D.3

SEARCH COMMANDS
nG
XXXX

Gets (searches for) the nth occurrence of the specified
character string on the current page.
Dot is set
immediately after the last character in the found text,
and the characters from the beginning of the line to Dot
are
listed on the teleprinter.
If the search is
unsuccessful, Dot will be at the end of the buffer and a
? will be printed out.

Searches the wHole file for ,the next occurrence, of the
specified character string. Combines G and N commands.
If search is not successful on current page, it continues
on Next page. Dot is set immediately after the last
charact~r
in the found text and the characters from the
beginning of the line to Dot are listed
on
the
teleprinter. If the Search object is not found, Dot will
be at the end of the buffer and a? will be printed out.
In such a case, all text scanned is copied to the output
tape.

XXXX

D.4

(0)
(@)
(/)

COMMANDS TO MODIFY THE TEXT
Character-Oriented
nD
nC
XXXX

DeleteS} the following
Changes n characters

-nD
-nC
XXXX

DeleteS} the previous
Changes n characters

Line-Oriented
Kills
}
eXchanges

the character string
beginning at Dot
and ending at the
nth end-of-line.

-nK Kills
}
-nX eXchanges
XXXX

the character strinq
beginning with th
first character following the (n+l)th
previous end-of-line
and ending at Dot.

nX
XXXX

D-2

TEXT EDITOR, ED-ll

the current line up
to Dot.

Kills
eXchanges

}

the character string
beginning at Dot and
ending at a previously Marked location.

Kills
/K
eXchanges
/X
XXXX

}

the character
string beginning at Dot and
ending with the
last character
of the page.

DeleteS} the character
Changes string beginning at Dot and
ending with the
last character
of the page.

Inserts the specified text.
LINE FEED terminates Text Mode and
causes execution of' the command. Dot is set to the location
immediately following the last character inserted.
If text was
inserted before the position of Mark,
ED~11
performs an M
command.

XXXX

SYMBOLS
Dot

Location following the most recent character operated
upon.

Holding down the CTRL key
(not the t
key)
combination with another keyboard character.

RETURN

If in command mode, it executes the current command;
goes into Text Mode if required.
If in Text Mode, it
terminates the current line, enters a carriage return
and line feed into the buffer and stays in text mode.
At all times causes the carriage to move to the
beginning of a new line.
(RETURN ,is often symbolized
as .,..) ) .

CTRL/FORM

D.6

}

/D
/C

xxxx

D.5

Kills
eXchanges

D,eletes } the current line OK
Changes up to Dot
OX
XXXX
Deletes } The character
@K
Changes string begin@X
ning at Dot and XXXX
ending at a previously Marked
location.

OC
XXXX
@D
@C
XXXX

(Typing the LINE FEED key)
unless the first character
executes the current command.

Terminates Text Mode
typed in Text Mode;

A Form feed which terminates,
page of the user's text.

and

thus

defines,

GROUPING OF COMMANDS
No Arguments
V

<
>
B
E
F
H
I
M
0

(Verify:
Lists current line)
(Lists previous line)
(Lists next line)
(Begin)
(End)
(Form feed)
(wHole)
(Insert)
(Mark)
(Open)
(Read)

Argument n only
G
N
T

D-3

(Get)
(Next)
(Trailer)

All Arguments
A
C
D
J
K
L
P
X

(n,-n,O,@,/)

(Advance)
(Change)
(Delete)
(Jump)
(Kill )
(List)
(Punch)
(eXchange)

TEXT EDITOR, ED-ll

D.7

Requiring
Text Mode

Line
Oriented

Character
Oriented

C
G
H
I
X

A
K
L
P
X

J
D

(Jump)
(Delete)

(Change)

(Change)
(Get)
(wHole)
(Insert)
eXchange)

(Advance)
(Kill )
(List)
(Punch)
eXchange)

OPERATING PROCEDURES

D.7.1

Loading: Use Absolute Binary Loader

D.7.2

Storage Requirements: ED-II uses all of core.

D.7.3

Starting: Immediately upon loading, ED-II will be in control.

D.7.4

Initial Dialogue:

Program Types

(see Chapter 5).

User Response

L-.J
H-.J

(if LSR is to be used for source input)
(if HSR is to be used for source input)

L...J
H-.J

(if LSP is to be used for edited output)
(if HSP is to be used for edited output)

I f the output device is the
(HSP) ,
Editor enters
high-speed punch
command mode to accept input. Otherwise the sequence continues with:

LSP OFF?

~(when

LSP is off)

Upon input of ~ from the keyboard, Editor enters command mode and
ready to accept input.

D.7.5

Restarting:
Type CTRL/P twice, initiating the normal
initial dialogue. The text to be edited
should be loaded (or reloaded)
at this
time.

D-4

APPENDIX E
DEBUGGING OBJECT PROGRAMS ON-LINE, ODT-ll AND ODT-llX

E.I

SUMMARY OF CONTENTS

ODT indicates readiness to accept commands by typing
location by printing its contents.
l.

ODT-ll
n/

opens word n

reopens last word opened

RETURN key

closes open location

opens next location

opens previous location

opens relatively addressed word

$n/

opens general register n (0-7)

n;G

goes to word n and starts execution

niB

sets breakpoint at word n

removes breakpoint

$B/

opens breakpoint status word

* or by opening a

proceeds from breakpoint, stops again on next
encounter

niP

proceeds from breakpoint, stops again on nth
encounter

$M/

opens mask for word search

n;W

searches for words which match n in bits specified
in $M

n;E

searches for words which address word n

n/ (contents) miO

calculates offsets from n to m

$S/

opens location containing user program's status
register

$P/

opens location containing ODT's priority level

E-1

DEBUGGING OBJECT PROGRAMS ON-LINE, ODT-II AND ODT-IIX
NOTE
If a word is currently
open,
new
contents for the word may be typed
followed by
any
of
the
commands
RETURN,1-, t, -+- or. The open word will be
modified and closed before the
new
command is executed.
2.

ODT-llX

In addition to the commands
version has the foll,:owing:

the

regular

version,

the

extended

opens byte

reopens last byte opened

opens the absolutely addressed word

opens the word to which the branch refers

opens next location of previous sequence

n;rB

(r between 0 and 7) sets breakpoint r at word n

;rB

removes breakpoint r

removes all breakpoints

$B/

opens breakpoint 0 status word.
Successive LINE
FEEDs open words for other breakpoints and singleinstruction mode.

inS

enables Single-instruction mode
value
and is not significant)

niP

in single-instruction mode, Proceeds with program
run
for next n instructions before reentering ODT
(if n is missing, it is assumed to be 1)

disables Single-instruction mode

E-2

can

have

any

APPENDIX F
LOADING AND DUMPING CORE MEMORY

F.l

THE BOOTSTRAP LOADER

This appendix pertains only to systems with a Switch Register.

F.l.l

Loading The Bootstrap Loader

The Bootstrap Loader should be toggled into the
bank.
xx7744
xx7746
xx7750
xx7752
xx7754
xx7756
xx7760
xx7762
xx7764
xx7766
xx7770
xx 7772
xx7774
xx7776

highest

core

memory

016701
000026
012702
000352
005211
105711.
100376
116162
000002
xx7400
005267
177756
000765
yyyyyy

xx represents the highest available memory bank.
For example, the
first location of the loader would be one of the following, depending
on memory size, and xx in all subsequent locations would be the same
as the first.
Location

Memory Bank

Memory Size

017744
037744
057744
077744
117744
137744
157744

0
1
2
3

4K
8K
12K
16K
20K
24K
28K

5
6

The contents of location xx7776 (yyyyyy)
in the Instruction column
above should contain the device status register address of the paper
tape reader to be used when loading the bootstrap formatted tapes
specified as follows:
Teletype Paper Tape Reader -- 177560
High-speed Paper Tape Reader -- 177550

F-l

LOADING AND DUMPING CORE MEMORY

11-0068

Figure F-l

Loading and Verifying the Bootstrap Loader

F-2

LOADING AND DUMPING CORE MEMORY

------I

see Fig ure E-1

L..----r-----I

_________ Code 351 must be
over reader sensors

------I

see Figure 5-5

'------.----'

11-0067

Figure F-2

F.2

Loading Bootstrap Tapes into Core

THE ABSOLUTE LOADER
1.

Loading the Absolute Loader
The Bootstrap Loader is used to load the Absolute Loader into
core.
(See Figure F-2.)
The Absolute Loader occupies
locations xx7474 through xx7743, and its starting address is
xx7500.

Loading with the Absolute Loader
When using the Absolute Loader, there are three types of
loads available: normal, relocated to specific address, and
continued relocation.

Optional switch register settings for the three
listed below.

F-3

types

loads

are

LOADING AND DUMPING CORE MEMORY

F.3

Type of load

Switch Register
Bits 1-14
Bit 0

Normal

(ignored)

Relocated - continue loading
where left off

Relocated - load in specified
area of core

nnnnn
(speci1ied address)

CORE MEMORY DUMPS

The two dump programs are
DUMPTT, which dumps the octal representation of the contents
of all or specified portions of core onto the teleprinter,
low-speed or high-speed punch, or line printer.
DUMPAB, which dumps the absolute binary code of the contents
of specified portions of core onto the low-speed (Teletype)
or high-speed punch.
Both d~mps are supplied on punched paper tape in bootstrap and
absolute binary formats.
The following figure summari.zes loading and
using the Absolute binary tapes.

F-4

LOADING AND DUMPING CORE MEMORY

ABS
LOADER
IN CORE

)-~N~O~____.,

HSR = 177550
LSR = 177560
xx IS HIGHEST
CORE MEMORY
BANK

(This is necessary only
if using a reader different
from that used by the
bootstrap loader.)

TAPE
READS IN

YES

CHECKYES
SUM ERROR )----'--<>-{

Figure F-3

Loading with the Absolute Loader

F-5

LOADING AND DUMPING CORE MEMORY

LSR = 177560
HSR = 177550
xx IS HIGHEST CORE
MEMORY BANK

SPECIFY
READER IN
xx7766

BOOT

! SEE FIGURE F-1 !

TOGGLE IN
BOOT LOADER

ABS
! SEE FIGURE F-2!

LOADER
IN CORE

YES

! SEE FIGURE F-21- - - -'-

YES

LOAD
DUMP TAPE

___ -1

SETSRTO
TRANSFER
ADDRESS

PRESS
LOADADDR
AND START

TTY OR LSP

SETSR TO
177564

OUTPUT
DEVICE FOR
DUMP

HSP

SETSRTO
177554

SETSR TO
177514
TTY

OR TTY

PRESS
PUNCH ON

PRESS
CONTINUE

Figure F-4

LOAD
ABS LOADER

Dumping Using DUMPAB or DUMPTT

F-6

SEE FIGURE F-3!

LOADING AND DUMPING CORE MEMORY

Figure F-4 (continued).

Dumping Using DUMPAB or DUMPTT

F-7

APPENDIX G
INPUT/OUTPUT PROGRAMMING, lOX

G.l

'INSTRUCTION SUMMARY
l.

Format
lOT
. WORD
.BYTE
. WORD

(an address)
(a command code, a slot number of a device)
;READR AND WRITR ONLY
(done address)

Command Codes:
INIT
RESET
RSTRT
WAITR
SEEK
READ
WRITE
READR
WRITR

G.2

1
2
3
4
5
11

12
13
14

PROGRAM FLOW SUMMARY
1.

Set up buffer header:
Location

BUFFER
HEADER

Contents

Buffer and
Buffer+l

Maximum number of data bytes (unsigned
integer)

Buffer+2

Mode of data (byte)

Buffer+3
Buffer+4 and
Buffer+5

Status of data (byte)
Number of data bytes involved in transfer (unsigned integer)

Buffer+6

Actual data begins here.

Mode Byte Format
Mode Byte Format
Bits

NO ECHO

EC.HO

Bits

UNFORMATTED

BINARY

FORMATTED

ASCII

INPUT/OUTPUT PROGRAMMING, lOX
Coding Mode Byte
Formatted ASCII

Formatted Binary

Unformatted ASCII

Unformatted Binary

(or 200 to suppress echo)

(or 202 to suppress echo)

Status Byte Format
Status Byte Format
6

1=
DONE

1=
EOM

1=
EOF

SEE CODES

NON-FATAL ERRORS

Coding Non-Fatal Errors
checksum error (formatted binary)
truncation of a long line
an improper mode

2
3

4
2.

Assign devices to slots in Device Assignment Table:
(RESET and INIT commands)
Slot numbers are in the range 0 to 7.
Device Codes:
KBD
TTY
LSR

G.3

1
2

LSP
HSR
HSP

LPT

4
5
6

Use a data transfer command to initiate the transfer.

FATAL ERRORS

Fatal errors result in a jump to 40 with RO set to the error code.
Errors 1-5 cause
R1 is set to the value of the PC for error code O.
Rl to be set to an lOT argument or to the instruction following the
arguments.
Reason

Fatal Error Code

Illegal Memory Reference, SP overflow, illegal
instruction

Illegal command

Slot out of range

Device out of range

Slot not inited

Illegal data mode
G-2

APPENDIX H
SUMMARY OF FLOATING POINT MATH PACKAGE, FPMP-II

This appendix lists all the global entry points of FPMP-II and
provides a brief description of the purposes of each. Sections H.l
and H.2 are for reference when it is desired to call FPMP-II routines
directly
(i.e., without the use of the TRAP handler).
Entry names
preceded by an octal number can be referenced via the TRAP handler.
The number is the "routine number" referred to in the FPMP-II manual.
If the number is enclosed in parentheses, the routine cannot be
accessed by the present TRAP handler, but has been assigned a number
for future use.
For a more detailed explanation of the Floating Point
Math Package, refer to the FPMP-II User's Manual DEC-II-NFPMA-A-D.
Examples

the calling conventions are:

POLISH MODE:
JSR R4,$POLSH
$subrl
$subr2

jenter Polish mode
;call desired subroutines

$subrn
• WORD

;call last subroutine desired
;leave Polish mode

.+2

JSRR:

XX:

JSR RS,subr
BR
XX
• WORD
argl
. WORD
arg2

'jcall desired subroutine

• WORD

;last argument
ireturn point

argn

;subroutine argument address

---------~--~----~--------~----~---------------------- ---------------

H-1

SUMMARY OF FLOATING POINT MATH PACKAGE, FPMP-ll
JPC:
push args onto stack
JSR PC,subr

H.l

OTS ROUTINES

Thesecare the routines taken from the FORTRAN operating
The codes used in the £ollowingtable ar~:

time

system.

Routine is included in the standard single precision (2-word)
package.
D = Routine is included in the standard double precision (4-word)
package.
SD = Routine is included· in both standard packages.
Octal codes shown in parenthe.ses are not yet implemented.

#: OF
ARGU

MODE

Polish

The
double
precision
add
routine.
Adds the top stack
item (4-words) to the second
item
(4-words) and leaves the
four word sum in their place.

Polish

The
single
precision
add
routine.
Same as $ADD except
it uses 2 word numbers.

AINT

J5RR

Returns sign of argument *
integer
=
greatest
real
absolute value of the argument
in RO;Rl.

ALOG

J5RR

Calculates natural logarithm
of its single argument and
returns a two word result in
RO,Rl.

ALOGIO

J5RR

Same
as
ALOG,
except
calculates base-IO logarithm.

ATAN

J5RR

Returns th~ arctangent of
argument in RO,Rl.

ATAN2

(43)

J5RR

Returns ARCTAN (ARGI/ARG2)
RO,Rl.

$CMD

Polish

Compares top 4 word items on
the stack, flushes the two
items;
and
returns
the
following
condition
codes:
4(SP)
@SP
N=l,Z=O
4(SP)
@SP
N=O,Z=l
4(SP)
@SP
N=O,Z=O

NAME

OCTAL
CODE

PKG

$ADD

$ADR

H-2

DESCRIPTION

its
in

SUMMARY OF FLOATING POINT MATH PACKAGE,

FPMP~11

$CMR.

Polish

Same as $CMD except it is
2 word arguments.

cos

J5RR

Single
DCOS.

precision

version

DATAN

J5RR

Double
ATAN.

precision

version

DATAN2

(45)

J5RR

Double
ATAN2.

precision

version

DBLE

(34)

J5RR

Returns in RO-R3 the double
precision equivalent of the
single precision (two word)
argument.

$DCI

(57)

. JPC

for

ASCII to double conversion.
Calling sequence:
Push address of start of
ASCII field. Push length
of ASCII field in bytes.
Push format scale D (from
W.D) position of assumed
decimal point (see FORTRAN
manual) .
Push P format
scale
(see
FORTRAN
manual). JSR PC,$DCI.
Returns 4 word result
of stack.

top

$DCO

(61)

JPC

Double precision
to
A~CII
conversion. Calling sequence:
Push address of start of
ASCII field. Push length
in bytes of ASCII field (W
part ofW.D) Push D pait
of W.D position of decimal
point).
Push P scale.
Push 4 word value to be
converted,
lowest order
word first. JSR PC,$DCO.

DCOS

J5RR

Calculates the cosine of its
double precision argument and
returns the double precision
result in RO-R3.

DEXP

··J5RR

Calculates the exponential of
its double precision argument,
and
returns
the
double
precision result in'RO-R3.

$DI

(11)

$DINT

(76)

Polish

Converts
double
precision
number on the top of the stack
to integer. Leaves result on
stack.

Polish

OTS internal function to find
the integer part of a double
precision number.

H-3

SUMMARY OF FLOATING POINT MATH PACKAGE, FPMP-ll
DLOG

J5RR

Double precision
version of ALOG.

word)

DLOGIO

J5RR

Double precision
version of ALOGIO.

word)

$DR

(6 )

Polish

Replaces the. double precision
item at the top of the stack
with its two word, rounded
form.

DSIN

J5RR

Calculates the sine of its
double
prec1s10n argo
and
returns the double precision
result in RO-R3.

DSQRT

J5RR

Calculates the square root of
its double ptecision argo and
returns the double precision
result in RO-R3.

$DVD

Polish

The double precision division
routine.
Divides the second
4-word item on the stack by
the top item and leaves the
quotient in their place.

$DVI

(24)

Polish

The integer division routine.
and
Calculates
2(SP)/@SP
returns the integer quotient
on the top of the stack.

$DVR

Polish

The single precision division
routine.
Same as $DVD, but
point
for 2 word floating
numbers.

$ECO

(62)

JPC

ASCII
Single precision
to
according
to E
conversion
format. Same calling sequence
as $DCO except that a 2-word
value is to be converted.

EXP

J5RR

Single
DEXP.
RO,Rl.

$FCALL

$FCO

(64)

FLOAT

(32)

$GCO

(63)

precision
Returns

version
result

of
in

Internal OTS routine.
5

JPC

Same as $ECO except
format conversion.

J5RR

Returns in
equivalent
argument.

JPC

Same as $ECO except
format conversion.

H-4

uses

RO-Rl, the real
of
its integer
uses

SUMMARY OF FLOATING POINT MATH PACKAGE, FPMP-ll
$ICI

(65)

JPC

ASCII to integer conversion
calling sequence:
Push address of start of
ASCII field.
Push length
in bytes of ASCII field.
JSR PC,$ICI
Returns with integer result on
top of stack.

$ICO

(67)

JPC

Integer to ASCII conversion.
Calling sequence:
Push address
of
ASCII
field.
Push length in
bytes of
ASCII
field.
Push integer value to be
converted. JSR PC,$ICO
Error will return with C bit
set on.
Ro-R3 destroyed.

IDINT

(31)

J5RR

Returns sign of arg * greatest
integer <= arg in RD. Arg
is double precision.

$ID

( 5)

Polish

Convert full word argument on
the top of the stack to double
precision and return result as
top 4-words of stack.

IFIX

(35)

J5RR

Returns the
truncated
and
fixed real argument in RD.

INT

(30)

J5RR

Same as IDINT
precision args.

$INTR

(27)

Polish

Same function as AINT, but
called in Polish mode with
argument and returns result on
the stack.

$IR

( 4)

Polish

Convert full word argument on
the top of the stack to single
precision and return result as
top 2-words of stack.

$MLD

Polish

precision
multiply.
Double
Replaces the top two doubles
stack
with
their
on the
product.

$MLI

(20)

Polish

Integer multiply.
Replaces
the top 2 integers on the
stack with their full word
product.

$MLR

Polish

precision
multiply.
Single
Replaces the top two singles
on the
stack
with
their
product.

$NGD

( 3)

Polish

Negate
number
stack.

H-5

the
on

for

single

double precision
the top of the

SUMMARY OF

FLOA~ING

POINT MATH PACKAGE, FPMP-ll

$NGI

(1)

Polish

Negate the integer on the
of the stack.

$NGR

(2)

Polish

Negate
number
stack.

$OCI

(66)

JPC

ASCII to octal
conversion.
Same call as $ICI.

$OCO

(70)

JPC

Octal to ASCII
conversion.
Same call as $ICO.

$POLSH

$POPR3

the
on

top

single precision
the top of the

Called whenever it is desired
to
enter Polish mode from
normal in-line code.
It must
be called via a JSR R4,$POLSH.
Polish

Internal
routine
to
2-words from the stack
place them into RO,Rl.

pop
~nd

$POPR4

Polish

Internal
routine
to
4-words from the stack
place them in RO-R3.

$:POPR5

Polish

Internal
routine
to
pop
4-words from the stack and
place them in registers RO-R3.

$PSHRI

Polish

Internal routine to push the
contents of RO onto the stack.

$PSHR2

Polish

Same as $PSHRI.

$PSHR3

Polish

Push RO,Rl onto stack.

$PSHR4

Polish

Push RO-R3 onto stack.

$PSHR5

Polish

Same as $PSHR4.

JPC

ASCII to
single
precision
Same
conversion.
calling
sequence as $DCI.
Returns
2-word result on top of stack.

Polish

Converts the single precision
number on the top of the stack
to double preclslon format.
Leaves result on stack.

pop
and

$RCI

(60)

$RD

(7)

$RI

(10 )

Polish

Converts
single
precision
number on the top of the stack
to integer. Leaves result on
stack.

$SBD

Polish

The double precision subtract
routine. Subtracts the double
precision number on the top of
the
stack from the second
double precision number on the
stack and leaves the result on
the top of the stack in their
place.

H-6

SUMMARY OF FLOATING POINT MATH .PACKAGE, FPMP-ll
$SBR

SIN

SNGL

(33)

SQRT

TANH

H.2

Polish

Same as $SBD
precision.

J5RR

Single
DSIN.

J5RR

Rounds
double
precision
argument to single precision.
Returns result in RO,RI.

J5RR

Single
DSQRT.

J5RR

Single precision
hyperbolic
function.
Returns
tangent
(EXP(2*ARG)-I/(EXP(2*ARG)+I)
in RO,RI.

precision

for

single

version

OCTAL
CODE

PKG

should

not

Internal error handler.

$ERRA

Similar to $ERR.

$LDR

Load FLAC, single precision.

$LDD

Load FLAC, double precision.

$STR

Store FLAC, single precision.

$STD

Store FLAC, double precision.

The TRAP handler routines and tables.

TRAPH

DESCRIPTION

$ERR

H.3

NON-OTS ROUTINES

These routines are written especially for FPMP-II and
called directly by the user.
NAME

but

ROUTINES ACCESSED VIA TRAP HANDLER

The following is a table of the FPMP-II routines which can be accessed
via TRAPH,
the trap handler.
Each routine name (entry point) is
preceded by its TRAP code number to be used to access it, and followed
by a brief description of its operation when called via the TRAP
handler.
Those entries which are preceded by an asterisk (*)
perform
operations only on the FLAC, and address no operands.
For example, a
TRAP call to the single precision square root routine can be coded as
follows:

TRAP

H-1

SUMMARY OF FLOATING POINT MATH PACKAGE, FPMP-ll
The net effect of the above TRAP instruction is to replace the
contents of the FLAC with its square root and then set the condition
codes to reflect the result. Note that since the FLAC is implicitly
addressed in this instruction, the TRAP call supplies no other
address. For such a TRAP call, the addressing mode bits (bits 6 and 7
of the TRAP instruction) are ignored.
All entries not marked by an asterisk require an operand when called.
The operand is addressed in one of the four addressing modes explained
in section 3.1.1. of the FPMP-ll User's Manual. The addressing mode
is specified in bit 6-7 of the TRAP instruction.
("Operand" is the contents of
call. )

the

location

addressed

the

TRAP

OCTAL
CODE

NAME

DESCRIPTION

$ADD

Adds
Double precision addition routine.
Assumes 4-word
operand to the FLAC.
operand.

$ADR

Adds
Single precision addition routine.
Assumes 2-word
operand to the FLAC.
operand.

AINT

Replaces contents of the FLAC by its
part.
SIGN (FLAC)
integer
* greatest
2-word
integer
Assumes
<= IFLACI
argument in FLAC.

the FLAC
Assumes

by its
2-word

calculates

base-IO

ALOG

Replaces contents of
logarithm.
natural
argument in FLAC.

ALOGIO

Same as ALOG, except
log.

ATAN

Replaces contents of
Assumes
arctangent.
FLAC.

$CMD

Compares operand to the contents of the
FLAC, and returns the following condition
codes.

the FLAC by its
2-word argument in

FLACoperand, N=O,Z=O
Assumes 4-word operands.

*
*
*

$CMR

Same as $CMD, but for 2-word operands.

COS

Same as DCOS, but for 2-word argument.

DATAN

Same as ATAN, but for 4-word argument.

DEXP

Replaces the contents of the FLAC by its
exponential.
Assumes 4-word argument in
the FLAC.

H-8

SUMMARY OF FLOATING POINT MATH PACKAGE, FPMP-Il

DLOG

Same as ALOG, but for 4-word argument.

DLOGIO

Same as ALOGIO, but for 4-word argument.

DCOS

Replaces the contents of the FLAC by its
cosine.
Assumes 4-word argument in the
FLAC.

DSIN

Same as DCOS, but calculates sine instead
of cosine.

DSQRT

Replaces the contents of the FLAC by its
square root.
Assumes 4-word argument in
the FLAC.

$DVD

Double
precision
division
routine.
Divides the FLAC by the operand and
stores the result in the FLAC.
Assumes
4-word operands.

$DVR

Same as $DVD, but for 2-word operands.

EXP

Same as DEXP, but for 2-word argument.

$LDD

Same as $LDR, but assumes 4-word operand.

$LDR

Replaces the contents of the FLAC by
operand. Assumes 2-word operand.

MLD

Double precision multiplication routine.
Multiplies the contents of the FLAC by
the operand and stores the result in the
FLAC.
Assumes 4-word operands.

$MLR

Same as $MLD, but for 2-word operands.

$SBD

The double precision sUbtraction routine.
Subtracts the operand from the contents
of the FLAC.
Assumes a 4-word operand.

$SBR

Same as $SBD, but for 2-word operand.

SIN

Same as DSIN, but for 2-word argument.

SQRT

Same as DSQRT, but for 2-word argument.

$STR

Stores the contents of the FLAC into the
operand location.
The contents of the
FLAC are unchanged.

$STD

Same as $STR, but assumes 4-word
location.

TANH

Replaces the contents of the FLAC by its
hyperbolic
tangent.
Assumes
2-word
argument.

H-9

the

operand

APPENDIX I
TAPE DUPLICATION

Duplication of paper tapes can be accomplished via low- or high-speed
I/O devices by toggling (as with the Bootstrap Loader) the following
program directly into memory through the Switch Register.
(Refer to
Section 6.1.1 in Chapter 6 if necessary, for toggling procedure.)
1.

Turn on appropriate device switches and place tape in desired
reader.

Set ENABLE/HALT switch to HALT.

Set Switch Register to the desired starting address and press
LOAD ADDR.

Set Switch Register to each value listed in the CONTENTS
column below, lifting the DEP switch after each setting.
(Addresses are automatically incremented.) The desired input
device
(either Low- or High-Speed Reader) and output device
(Low- or High-Speed Punch) are- specIfied in the last two
words.
ADDRESS

CONTENTS

016700
000024
016701
000022
005210
105710
100376
105711
100376
022021
111011
000764
177560 (LSR) or 177550 (HSR)
177564 (LSP) or 177554 (HSP)

4
6

10
12
14
16
20
22
24
26
30
32
5.

Set Switch Register to starting address specified in 3
and press LOAD ADDR.

Set ENABLE/HALT switch to ENABLE.

Press START switch.

I-I

above

TAPE DUPLICATION
NOTE
This program is recommended as a simple
way of duplicating the system tapes.
However, for extensive tape duplication,
the program shown in section 7.8 is
recommended.

I-2

APPENDIX J
ASSEMBLY AND LINKING INSTRUCTIONS

J.I

J.I.I

SYSTEMS WITHOUT SWITCH REGISTERS

IOX/IOXLPT

IOX/IOXLPT is provided in both source and relocatable object form.
Unless modifications are made to the source it is not necessary to
assemble the source tapes. The object tape may be linked with the
user's object tapes to produce an absolute tape (.LDA).

J.I.I.I Assembling lOX - lOX consists of three source tapes (-PAl to
-PA3) . These tapes are assembled together in sequence with PAL-lIS.

J.I.I.2 Assembling IOXLPT - IOXLPT consists of two source tapes (-PAl
to PA2). These tapes are assembled together in sequence with PAL-lIS.

J.I.I.3 Linking lOX and IOXLPT - IOX and lOXLPT
are
linked
LINK-lIS with the user's object tapes to produce an absolute tape.

J.I.2

ODTIIX

ODTllX is provided in both source and relocatable object form.
Unless
modifications are made to the source, it is not necessary to assemble
the source tape. The object tape may be linked with the user's object
tapes to produce an absolute tape.

J.l.2.1 Assembling ODTIlX - ODTIIX consists of one source tape (-PAL)
which is terminated with the following:
.EOT
form feed
.END O.ODT
When PAL-IlS indicates that it has encountered the .EDT,
so that it will process the .END statement.

J-l

type

return

ASSEMBLY AND LINKING INSTRUCTIONS
J.1.2.2 Linking ODTllX - ODTllX is linked with user object tapes. It
is self starting and should be the first object tape input to LINK-lIS
so that it will be the program started by the Absolute Loader when the
program is loaded.

J.1. 3

ED-II

The ED-II source file is available only in RT-ll format on a flexible
diskette.
The RT-ll MACRO assembler is required to assemble ED-II.
The RT-ll linker (LINK) is us~d to produce the absolute tape.

J.1.3.l Assembling ED-II - The RT-ll commands to assemble
as follows:

ED~ll

are

.R MACRO
*EDITll=DXl:EDITll

J.1. 3.2 Linking ED-II - The RT-ll
follows:

commands

link

ED-ll

are

.R LINK
*PP:EDITll/L=EDITll

J.1. 4

PAL-llS

The PAL-lIS source file is available only in RT-ll format on a
flexible diskette. The RT-ll MACRO assembler is required to assemble
PAL-lIS. The RTll linker (LINK) or LINK-lIS may be used to produce
the absolute tape.

are
J.l.4.l Assembling PAL-lIS - There are three sources
which
assembled separately for PAL-lIS.
One of these, the symbol table
source, is available in three versions:
8K, 12K, and 16K. The RT-ll
commands to assemble PAL-lIS source files are as follows:
.R MACRO
*RELMEM=DXl:RELMEM.PAL
*PSYM08=DXl:PSYM08.PAL
*PSYM12=DXl:PSYM12.PAL
*PSYM16=DXl:PSYM16.PAL
*PALllS=DXl:PALllS.PAL

Clear Memory Program
8K Symbol Table
12K Symbol Table
16K Symbol Table
Assembler

In addition to the above, IOXLPT is used by PAL-lIS.
The IOXLPT
source is also available in RT-ll format on a flexible diskette. The
commands to assemble IOXLPT are:
.R MACRO
*IOXLPT=DXl:IOXLPT.PAL

J-2

ASSEMBLY AND LINKING INSTRUCTIONS
J.I.4.2 Linking PAL-lIS - PAL-lIS may be linked with LINK-lIS or the
RT-II linker, LINK. The PAL-lIS tape actually contains two programs:
RELMEM and PAL-lIS. RELMEM precedes PAL-lIS on the tape.
Using LINK-lIS, link PAL-lIS as follows:
1.

Link RELMEM as a separate program and do not remove the
from the punch when finished.

tape

Link PALIIS.OBJ, IOXLPT.OBJ, and one of the symbol table
object tapes (PSYM08.0BJ, PSYMI2.0BJ, or PSYMI6.0BJ) in that
order. The symbol table tape is selected depending on the
size of the memory of the computer on which the program is to
be executed.
If the target computer has 8K words of memory
then PSYM08.0BJ,
if 12K then PSYMI2.0BJ, and if 16K then
PSYMI6.0BJ. Specify a top address of 57460 for 12K and 77460
for 16K.
DO not link PAL-lIS to run above 16K. The size of the symbol
table is fixed, and there is no need to re-link at a higher
address even on large systems.

Using RT-II LINK, link PAL-lIS as follows:
1.

Link RELMEM as a separate program as shown
.R LINK
*RELMEM/L=RELMEM

Link 8K, 12K, and 16K versions of PAL-lIS
.R UN',
*PAL08/L/B:204=PALllS~IOXLPT~PSYM08
*PAL12/L/B:204~PALllS,IOXLPT,PSYM12

*PAL :l6/L/B: 204~-::PAI... l:l. S, I OXI...PT, PSYM16
3.

Use RT-II PIP to punch the tapes. Remember not to remove the
tape from the punch after punching RELMEM .

• R PIP
*PP:=REI...MEM.I...DA/B
*PP:=PAI...08.I...DA/B
remove 8K PALIIS.LDA from punch

*PP:=REI...MEM.I...DA/B
*PP=::PAL:L2.I...DA/B
remove 12K PALIIS.LDA from punch

*PP:=REI...MEM.LDA/B
*PP:~PAL:l.6.I...DA/B

J.1. 5

LINK-llS

The LINK-lIS source file is available only in RT-II format on a
flexible diskette. The RT-II MACRO assembler is required to assemble
LINK-lIS.
LINK-lIS is composed of two components:
LINK-lIS proper
and IOXLPT.
See Section N.l.4.1 for instructions on how to assemble
IOXLPT using RT-II.

J-3

ASSEMBLY AND LINKING INSTRUCTIONS
J.l.5.l
follow:

Assembling LINK-lIS - The RT-ll commands to assemble LINK-lIS

.R MACRO
*LINKll=DX1:LINKll

J.l.5.2 Linking LINK-lIS - LINK-lIS may be linked with LINK-lIS or
the RT-ll linker, LINK. There are two object tapes which ate linked
together to produce LINK-lIS: LINKll.OBJ and IOXLPT.OBJ.
Using LINK-lIS to link LINK-lIS, link the following two tapes in
order:
LINKll.OBJ and IOXLPT.
If versions are desired for systems
with more than 8K, specify a top address of 57460 for 12K and 77460
for 16K.
Using RT-ll LINK to link LINK-lIS is a two step process because of a
difference in philosophy. An initial link is required which produces
a link map so that the size of LINKllS can be determined.
A final
link is then made with the information obtained in the initial link
used to compute the bottom address.
The initial link is executed as follows:

.R LINK
*,TT:=LINKll,IOXLPT
The value displayed for "HIGH LIMIT" is used to compute the bottom
address for the final link. Assume for an example that the following
was displayed:

HIGH LIMIT

015572

Select 37460, 57460, or 77460 depending on whether an 8K, 12K, or 16K
top address is desired. The bottom address is computed as follows:
B = T - 4 + 1000
Where:

B = bottom address
T
top address
H
high limit

Example:

B
37460-15572+1000
B = 22666

Using the figures in the example above,
system would be executed as follows:

the

final

link

for

.R LINK
*PP:/B:22666/L,TT:=LINKll,IOXLPT
As a check, examine the link map produced and verify
limit matches the one used in the calculations above.
the high limit value must be 37460.

J-4

that the high
In the example,

ASSEMBLY AND LINKING INSTRUCTIONS
J.2

SYSTEMS WITH SWITCH REGISTERS

J.2.1

Assembling PAL-IIA

The following procedures are for assembling the PAL-II Assembler
source tapes.
An 8K version of the PAL-IIA (V007A) Assembler is
required, thus also requiring at least an 8K PDP-II system.
The Assembler consists of two programs. The first program, on tape 1,
is a memory clear program and is very short (DEC-II-UPLAA-A-PAl). The
second program is the Assembler proper, and consists of eleven ASCII
tapes (DEC-II-UPLAA-A-PA2-PA12). They are assembled as follows:
1.

Generate a sufficient amount of blank leader tape.

Assemble
the
memory
clear
program
source
tape
(DEC-ll-UPLAA-A-PAl)
and assign the binary output to the
high-speed punch.
For example, PAL-IIA's initial dialogue to
specify the 2-pass assembly would be:
*S
*B
*L
*T

HIE
(PAl assembly - 1st pass)

END?
000000 ERRORS
C

(PAl assembly - 2nd pass)
(No errors - Do not remove
the binary tape from the punch.)

Assemble the rest of the Assembler's
PA12) in numerical sequence.

source

tapes

(PA2

Assign the binary output to the high-speed punch.
For
example, the initial dialogue should be answered as follows:
*S H
*B HIE
*L
*T
EOF ?
EOF ?
EOF ?
EOF ?
EOF ?
EOF ?
EOF ?
EOF ?
EOF ?
EOF ?
EOF ?
MAXCL13
END ?
EOF ?
EOF ?
EOF ?
EOF ?
EOF ?
EOF ?
EOF ?
EOF ?
EOF ?
EOF ?
EOF ?

(Enter tape PA2 for 1st pass)
(End of tape PA2, enter PA3)
(End of tape PA3, enter PA4)
(End of tape PA4, enter PAS)
(End of tape PAS, enter PA6)
(End of tape PA6, enter PA7)
(End of tape PA7, enter PA8)
(End of tape PA8, enter PA9)
(End of tape PA9, enter PAID)
(End of tape PAID, enter PAll)
(End of tape PAll, enter PA12)
(End of first pass)
****** SIMBC = ******
(Enter tape PA2 for 2nd pass)
(End of tape PA2, enter PA3)
(End of tape PA3, enter PM)
(End of tape PA4, enter PAS)
(End of tape PAS, enter PA6)
(End of tape PA6, enter PA7)
(End of tape PA7, enter PA8)
(End of tape PA8, enter PA9)
(End of tape PA9, enter PAID)
(End of tape PAID, enter PAll)
(End of tape PAll, enter PAI2)
J-S

ASSEMBLY AND·· LINKING INSTRUCTIONS
000000 ERRORS

(End of 2nd pass)

*S
Note that at the end of the first pass there are two undefined
symbols:
MAXCl3 and SIMBC. These undefined symbols are resolved so
that there are no errors reported during the second pass.
Be sure that there is sufficient blank trailer tape on the
output tape before removing the assembled tape from the punch.

binary

Normally, using high-speed paper tape input and output, this process
requires about 45 minutes.
If a symbol table and listing are
requested, there will be about 750 symbols and about 4500 lines of
listing.

J.2.2

Assembling ED-II

ED-II consists of five source tapes (PAl to PAS) which
together in sequence with 8K PAL-IIA.

are

assembled

In subsequent discussion, reference to ODT applies to both
ODT is supplied on both source and absolute binary tapes.

versions.

J.2.3

ODT-II/ODT-IIX

If the program being debugged requires storage where the version of
ODT being used is normally loaded, it is necessary to reassemble ODT
after changing the starting location.
The source tape of ODT is in three segments, each separated
next by blank tape. The first segment contains:
.=n
.EOT

from

the

(standard location setting statement)

where n=13026 for ODT-II or n=12054 for ODT-IIX. This statement tells
the Assembler to start assembling at address n. To relocate ODT to
another starting address, substitute for segment one a source tape
consisting of:
. (n is the new load address for ODT)

.=n
.EOT

The .EOT statement tells the Assembler that this is the end of the
segment but not the end of the program -- the Assembler will stop and
wait for another tape to be placed in the reader.
The second segment of tape contains the
segment is also terminated with .EOT.

ODT

source

program.

This

The third segment of the tape consists of the statement:
.END

O.ODT

where .END means "end of program" and O.ODT
address of the program (see Section 6.2.3).

represents

the

starting

When relocating ODT, the first segment of the source tape must be
changed to reflect the desired load address. The third segment may be
changed to .END without a start address. The latter will cause the
Loader to halt upon completion of loading.
J-6

ASSEMBLY AND LINKING INSTRUCTIONS
The segmentation allows the following assembly forms:
1.

Assemble alone but at a new address. A new segment one
be generated and assembled with segments two and three.

must

Assemble immediately after the user's program to be debugged.
Assemble the tape of the user's program (ending with .EOT)
followed by ODT's segment two and either segment three or a
new segment three.

Assemble inside the program to be debugged.
Assemble the
first part of the user program (ending with .EOT) followed by
ODT's second segment followed by the second part of the user
program.

When setting locations before assembling, it must be noted that
immediately preceding ODT a minimum internal stack of 40 8 bytes is
required for the ODT~ll and 116 8 bytes is required· for ODT-IIX.
Additional room must be allocated for subroutine calls and possible
interrupts while ODT is in control. Twelve bytes maximum will be used
by ODT proper for subroutine calls and interrupts, giving a minimum
safe stack space of 528 bytes for ODT-ll or 1308 bytes for ODT-IIX.
Once a new binary tape of ODT has been assembled, load it using the
Absolute Loader as explained in Section 6.2.2. Normally, the program
to be debugged is loaded before ODT, since ODT will automatically be
in control immediately after loading, unless the third segment of
ODT's source tape was altered before assembly. As soon as the tape is
read in, ODT will print an * on the Teletype to indicate that it is
ready for a command.

J.2.4

Assembling IOX/IOXLPT

In subsequent discussion, reference to lOX applies to both
lOX is supplied on both source and absolute binary tapes.

versions.

If there is more than 4K of core available and it is desired to load
lOX
(or IOXLPT)
in other than the normal location, lOX must be
reassembled.
The code
.=15100

.EDT
appears at the beginning of the first lOX tape (PAl) and contains the
starting address.
Create a new tape containing the new starting
address desired; be sure to allow enough room for 63410 words for
lOX, 72510 for IOXLPT. For example,
.=25100

.EDT
Use PAL-IIA to assemble rox and substitute the new section of
the first part of the old tape (PAl). After the new section
insert the lOX tape in the reader so the read head is past
starting address and .EOT and type the RETURN key to read in
of the tape.

tape for
is read,
the old
the rest

Now read in the second tape (PA2). An EOF? message is output at the
end of the second tape. Type the RETURN key and the END? message is
printed. Put the tapes through for the second pass of the assembler.
J-7

ASSEMBLY AND LINKING INSTRUCTIONS
lOX (IOXLPT) can also be assembled with a user program if desired.
The .=15100 and .EOT lines must be deleted before lOX is assembled
with a user program.
lOX can be assembled into the program wherever desired but if it is
the first tape read by the assembler, remove it from the reader before
typing the RETURN key (after the EOF? message of the second tape.
lOX and IOXLPT have a .END code which would cause the assembly pass to
end when read). Assembling a user program and lOX together eliminates
the need to read in lOX each time the program is run.

J.2.5

Assembling and Linking PAL-lIS

PAL-lIS consists of two independent programs. The first program is a
memory clear program. The second is the assembler. All programs are
available as ASCII source tapes, object modules and as a load module.
The memory clear program, MEMCLR, (DEC-II-UPLSA-A-PAl) consists of one
ASCII tape.
This program should never need to be assembled. The
object module may be used when constructing a new load module of
PAL-lIS.
The assembler consists of three program modules which are assembled
separately and then linked together. The first is the main program
called PAL-lIS. It consists of 13 ASCII tapes (DEC-UPLSA-A-PA2-PA14).
The second module is the symbol table, PALSYM, which consists of 2
ASCII tapes
(DEC-II-UPLSA-A-PA15-PA16).
The
third
is
IOXLPT
consisting of 2 ASCII tapes (DEC-II-UPLSA-A-PA17-PA18). Also included
is PALSYM, specially created for 12K and 16K, consisting of one tape
each (DEC-II-UPLSA-A-PA19-PA20).
If changes are
assembled by
with the other
these programs
Program
PAL-lIS
PALSYM
IOXLPT

made in any of these modules, that module must be
PAL-lIS (V003A) and the new object module can be linked
object modules. It should be noted that assembly of
will result in:

Pages of Listing (Decimal)

Number of Symbols (Decimal)

160
11
29

756
32
191

Also note that there will be two undefined symbols listed at the end
of pass 1. These are forward references on direct assignments which
get defined properly in pass 2.
An example of the PAL-lIS assembly follows:
PAL-lIS
V003A
*S H
*B H
*L P
*T P/2
END ?
000000 ERRORS

(first pass on PAl)
(2nd pass on PAl)
(End of tape tl assembly)
(Remove tape from punch)

J-8

ASSEMBLY AND LINKING INSTRUCTIONS
PAL-llS
V003A
*S H
*B H
*L P
*T P/2
EOF ?
EOF ?
EOF ?
EOF ?
EOF ?
EOF ?
- EOF ?
EOF ?
EOF ?
EOF ?
EOF ?
EOF ?
BINCNT=****** SIMBC=******
END ?
EOF ?
EOF ?
EOF ?
EOF ?
EOF ?
EOF ?
EOF ?
EOF ?
EOF ?
EOF ?
EOF ?
EOF ?
000000 ERRORS

(Insert
(End of
(End of
(End of
(End of
(End of
(End of
(End of
(End of
(End of
(End of
(End of
(End of
(End of
(Insert
(End of
(End of
(End of
(End of
(End of
(End of
(End of
(End of
(End of
(End of
-(End of
(End of
(End of
(Remove

PAL-lIS
V003A
*S H
*B H
*L P
*T P/2
EOF ?
END ?
EOF ?
000000 ERRORS

PA2 for 1st pass)
PA2, insert PA3)
PA3, insert PA4)
PA4, insert PAS)
PAS, insert PA6)
PA6, insert PA7)
PA7, insert PAS)
PAS, insert PA9)
PA9, insert PAlO)
PAlO, insert PAli)
PAll, insert PA12)
PA12, insert PAl3)
PA13, insertPA14)
PA14 and 1st pass)
PA2 for 2nd pass)
PA2, insert PA3)
PA3, insert PA4) PA4, insert PAS)
PAS, insert PA6)
PA6, - insert PA7)
PA7, insert PAS)
PAS, insert PA9)
PA9, insert PAlO)
PAlO, insert PAll)
PAll, insert PA12)
PA12, insert PAl3)
PA13, insert PAl4)
PA14 and 2nd pass)
tape from punch)

(1st pass on PAIS)
(End of PAIS, insert PA16)
(End of PA16, insert PAIS for 2nd pass)
(End of PAIS, insert PA16)
(End of 2nd pass)
(Remove tape from punch)

PAL-llS
V003A
*S H
*B H
*L P
*T P/2
EOF ?
END ?
EOF ?
000000 ERRORS

(1st pass on PA17)
(End of PA17, insert PAIS)
(End of PAIS, insert PA17 for 2nd pass)
(End of PA17, insert. PAIS)
(End of 2nd pass)
(Remove tape from punch)

PAL-llS
V003A
*S H
*B H
*L P
*T P/2
END ?
000000 ERRORS

(Pass 1
(Pass 2
(End of
(Remove

J-9

on PA20)
on PA20)
pass 2)
tape from punch)

ASSEMBLY AND LINKING INSTRUCTIONS
The final load module is constructed by LINK-lIS.
First the· memory
clear program object module is processed by the linker and the
resulting load module is left in the punch while the PAL-lIS, PALSYM,
and IOXLPT object modules are linked to create a second load module.
The resulting tape ~ontains two load modules.
The first clears memory
and then jumps to the absolute loader to load the second.
In order to take advantage of core sizes larger than 8K, PALSYM,
the
symbol table, specially created for 12K core and 16K core, and the
object modules are included with the assembler.
To link for 12K
(or
16K), simply sUbstitute the appropriate object tape for PALSYM (use
DEC-II-UPLSA-A-PR5 for 12K or DEC-II-UPLSA-A-PR6 for 16K)
specify a
top address to LINK-lIS ~f 57460 for 12K (77460 for 16K) and link as
described in the preceding paragraph.
Do not reI ink PAL-lIS to run above 16K. The size of the symbol table
is fixed, and there is no ne.ed to re-link at a higher address even on
large systems.
The supplied tapes are identified as follows:
Contents

Library Code
DEC-II-UPLSA-A-PAI

Tape 1 of 20

DEC-II-UPLSA-A-PA2

DEC-II-UPLSA-A-PA8
DEC-II-UPLSA-A-PA9
DEC-II-UPLSA-A-PAIO
DEC-II-UPLSA-A-PAII
DEC-II-UPLSA-A-PA12
DEC-II-UPLSA-A-PA13
DEC-II-UPLSA-A-PA14

Tape
Tape
Tape
Tape
Tape
Tape
Tape
Tape
Tape
Tape
Tape
Tape
Tape

DEC-II-UPLSA-A-PA15
DEC-11-UPLSA-A-PA16

Tape 15 of 20 }
Tape 16 of 20

One
Assembly

DEC11-UPLSA-A-PA17
DEC-II-UPLSA-A-PA18
DEC-II-UPLSA-A-PA19
DEC-II-UPLSA-A-PA20

Tape
Tape
Tape
Tape

17
18
19
20

One
Assembly
One Assembly
One Assembly

DEC-II-UPLSA-A-PRI
DEC-II-UPLSA-A-PR2
DEC-II-UPLSA-A-PR3
DEC-II-UPLSA-A-PR4
DEC-II-UPLSA-A-PR5

Tape
Tape
Tape
Tape
Tape

1
2
3
4
5

DEC-II-UPLSA-A-PR6

Tape 6 of 6

DEC-II-UPLSA-A~PA3

DEC-II-UPLSA-A-PA4
DEC-II-UPLSA-A-PA5
DEC-II-UPLSA-A-PA6
DEC-Il~UPLSA-A-PA7

DEC-II-UPLSA-A~PL

2 of 20
3 of 20
4 of 20
5 of 20
6 of 20
7 of 20
8 of 20
9 of 20
10 of 20
11 of 20
12 of 20
13 of 20
14 of 20

of
of
of
of
of
of
of
of
of

20 }
20
20
20
6
6
6
6
6

One
Assembly

RELMEM
(Memory Clear Program)
PAL-lIS (Main Program)

One
Assembly

PALSYM (Symbol Table) for 8K

IOXLPT
PALSYM (Symbol Table) for 12K
PALSYM (Symbol Table) for 16K
RELMEM Object Module
PAL-lIS Object Module
PALSYM Object Module for 8K
IOXLPT Object Module
PALSYM Object Module for 12K
assembler
PALSYM Object Module for 16K
Assembler
PAL-lIS Load Module l

This tape is the concatenation of a link of the RELMEM object module
followed by a link of the PAL-lIS, PALSYM for 8K, and IOXLPT object
modules.
J-IO

ASSEMBLY AND LINKING INSTRUCTIONS
J.2.6

Assembling And Linking LINK-lIS

LINK-lIS is available as an absolute load module (for an SK machine),
as two object modules
(for reI inking) and as several ASCII source
tapes. There is one object module fo~ the Linker and one for IOXLPT.
The supplied object modules may be reI inked (using the supplied load
module) to load into any size machine larger than SK.
However, the
resulting Linker will still assume a top of memory corresponding to an
SK machine (this can be overridden in· the command string options).
The assumed top of memory and reserved Absolute Loader space may be
changed by editing the first linker ASCII tape with ED-II.
The
parameters to be changed are HGHMEM (high memory address +1 (always
even)) and ALODSZ (Absolute Loader size (always even)).
The source
tapes for the Linker may then be assembled with PAL-lIS and the new
object module can then replace the supplied Linker object module.
The tapes are identified as follows:
Library Code
DEC-II-ULKSA-A-PAI
DEC-II-ULKSA-A-PA2
DEC-II-ULKSA-A-PA3
DEC-II-ULKSA-A-PA4

Tape
Tape
Tape
Tape

DEC-II-ULKSA-A-PA5
DEC-II-ULKSA-A-PA6

Tape 5 of
Tape 6 of

DEC-II-ULKSA-A-PRI
DEC-II-ULKSA-A-PR2

Tape 1 of 2
Tape 2 of 2

1
2
3
4

of
of
of
of

~1
~

LINK-lIS (Main Program)
One
Assembly

One
} Assembly

IOXLPT
LINK-lIS Object Module
IOXLPT Object Module

DEC-II-ULKSA-A-PL

LINK-lIS Load Module

J-II

APPENDIX K
STANDARD PDP-II ABBREVIATIONS

Abbreviation
ABS
A/D
ADC
ADRS
ASCII
ASL
ASR
B
BAR
BBSY
BCC
BCS
BEQ
BG
BGE
BGT
BHI
BHIS
BIC
BIS
BIT
BLE
BLOS
BLT
BMI
BNE
BPL
BR
BRD
BRX
BSP
BSR

Definition
absolute
analog-to-digital
add carry
address
American Standard Code
for Information Interchange
arithmetic shift left
arithmetic shift right
automatic send/receive

BSY
BVC
BVS

byte
bus address register
bus busy
branch if carry clear
branch if carry set
branch if equal
bus grant
branch if greater or equal
branch if greater than
branch if higher
branch if higher or same
bit clear
bit set
bit test
branch if less or equal
branch if lower or same
branch if less than
branch if minus
branch if not equal
branch if plus
branch
bus register data
bus request
back space
bus shift register
back space record
busy
branch if overflow clear
branch if overflow set

CBR
CLC
CLK
CLN
CLR
CLV

console bus request
clear carry
clock
clear negative
clear
clear overflow

K-l

STANDARD PDP-II ABBREVIATIONS
CLZ
CMP
CNPR
CNTL
COM
COND
CONS
CaNT
CP
CSR

clear zero
compare
console nonprocessor request
control
complement
condition
console
contents
continue
central processor
control and status register

DEL
DEP
DEPF
DIV
DMA
DSEL
DST
DSX

data
digital-to-analog
device address register
data in
data in, pause
data out
data out, byte
data buffer register
decoder
destination effective address
decrement
Digital Equipment Corp.
delay
deposit
deposit flag
divide
direct memory access
device select
destination
display, X-deflection register

EMT
ENB
EOF
EOM
ERR
EX
EXAM
EXAMF
EXEC
EXR

emulator trap
enable
end-of-file
end-of-medium
error
external
examine
examine flag
execute
external reset

FCTN
FILa
FLG

flag (part of signal name)
function
first in,last out
flag

GEN

generator

INDIVR
INC

integer divide routine
increment
increase
increment flag
indicator
inhibit
ini tialize
instruction
interrupt
interrupt flag
input/output
input/output trap
input/output executive routine

D/A
DAR
DATI
DATIP
DATa
DATOB
DBR
DCDR
DE
DEC

INCF
IND
INH
INIT
INST
INTR
INTRF
I/O
lOT
lOX

K-2

STANDARD PDP-ll ABBREVIATIONS
IR
IRD
ISR

instruction register
instruction register decoder
instruction shift register '

JMP
JSR

jump
jump to subroutine

LIFO
LKS
LOC
LP
LSB
LSBY
LSD

last in,first out
line time clock status register
location
line printer
least significant bit
least significant byte
least significant digit

MA
MAR
MBR
MEM
ML
MOV
MSB
MSBY
MSD
MSEL
MSYN

memory address
memory address register
memory buffer register
memory
memory location
move
most significant bit
most significant byte
most significant digit
memory select
master sync

ND
NEG
NOR
NPG
NPR
NPRF
NS

negative driver
negate
normalize
nonprocessor grant
nonprocessor request
nonprocessor request flag
negative switch

ODT
OP

octal debugging technique
operate
operation
operator
operand

OPR
PA
PAL
PB
PC
PD
PDP
PERIF
PGM
PP
PPB
PPS
PR
PRB
PROC
PRS
PS
PTR
PTS
PUN

parity available
program assembly lanquage
parity bit
program counter
positive driver
programmed data processor
peripheral
program
paper tape punch
paper tape punch buffer tegister
paper tape punch status register
paper tape reader
paper tape reader buffer
register
processor
paper tape reader status
register
processor status
positive switch
priority transfer
paper tape software system
punch

K-3

STANDARD PDP-II ABBREVIATIONS
RD
RDR
REG
REL
RES
ROL
ROM
ROR
R/S
RTI
RTS
R/W
R!WSR

read
reader
register
release
reset
rotate left
read-only memory
rotate right
rotate shift
return from interrupt
return from subroutine
read/write
read/write shift register

single
selection acknowledge

SACK
SBC
SC
SE
SEC
SEL
SEN
SEV
SEX
SEZ
SI
SP
SR
SRC
SSYN
ST
STPM
STR
SUB
SVC
SWAB

SUBTRACT CARRY

s.ingle cycle
source effective address
set carry
select
set negative
set overflow
sign extend
set zero
single instruction
stack pointer
spare
switch register
source
slave sync
start
set trap marker
strobe
subtract
service
swap byte

TEMP
TK
TKB
TKS
TP
TPS
TRT
TSC
TST

trap address
track address
temporary
teletype keyboard
teletype keyboard buffer register
teletype keyboard status register
teletype printer
teletype printer status register
trace trap
timing state control
test

UTR

user trap

VEC

vector

WC
WCR

word count
word count register

XDR
XRCG
XWCG

X-line driver
X-line read control group
X-line write control group

YDR
YRCG
YWCG

Y-line driver
Y-line read control group
Y-line write control group

K-4

APPENDIX L
CONVERSION TABLES

L.l

OCTAL-DECIMAL INTEGER CONVERSIONS

0000
to
0777
(Octal)

Octal

0000
0008
0016
0024
0032
0040
0048
0056

0001
0009
0017
0025
0033
0041
0049
0057

0002
0010
0018
0026
0034
0042
0050
0058

0003
0011
0019
0027
0035
0043
0051
0059

0004
0012
0020
0028
0036
0044
0052
0060

0005
0013
0021
0029
0037
0045
0053
0061

0006
0014
0022
0030
0038
0046
0054
0062

0007
0015
0023
0031
0039
0047
0055
0063

0400
0410
0420
0430
0440
0450
0460
0470

0257
0265
0273
0281
0289
0297
0305
0313

0258
0266
0274
0282
0290
029a
0306
0314

0259
0267
0275
0283
0291
0299
0307
0315

0260
0268
0276
0284
OB2
0300
0308
0316

0261
0269
0277
0285
0293
0301
0309
0317

0262
0270
0278
0286
0294
0302
0310
0318

0263
0271
0279
0287
0295
0303
0311
0319

0064
0072
0080
0088
10096
0104
0112
0120

0065
0073
0061
0089
0097
0105
0113
0121

0066
0074
0082
0090
0098
0106
0114
0122

0067
0075
0083
0091
0099
0107
0115
0123

0068
0076
0084
0092
0100
0108
0116
0124

0069
0077
0085
0093
0101
0109
0117
0125

0070
0078
0086
0094
0102
0110
0118
0126

0071
0079
0087
0095
0103
0111
0119
0127

0500 0320 0321
0510 0328 0329
05~0 0336 0337
0530 0344 0345
0540 0352 0353
0550 0360 0361
0560 0368 0369
0570 0376 0377

0322
0330
0338
0346
0354
0362
0310
0378

0323
0331
0339
0317
0355
0363
0371
0379

0324
0332
0340
0348
0356
0364
0372
0380

0325
0333
0141
0349
0357
0365
0373
0381

0326
0334
0342
0350
0358
0366
0374
0382

on7
0335
0343
0351
0359

0128
0136
0144
0152
0160

0129
0137
0145
0153
0161
elsa 0169
0176 0177
0184 0185

0130
0138
0146
0154
0162
0170
0178
0186

0131
0139
0147
0155
0163
0171
0179
0187

0132
0140
0148
0156
0164
0172
0180
0188

0133
0141
0149
0157
0165
0173
0181
0189

0134
0142
0150
0158
0166
0174
0182
0190

0135
0143
0151
0159
0167
0175
0183
0191

0385
0393
0401
0409
0417
0425
0433
0441

0386
0394
0402
0410
0418
0426
0434
OH2

0387
0395
0403
0411
0419
0427
0435
OH3

0388 0389 0390
0396 0397 0398
040~ 0405 0406
0412 0413 04 I 4
0420 0421 0422
0428 0429 0430
0436 0437 0438
0444 0445 0446

0391
0399
0407 I
0415 ,
0423
0431
0439
0447

0193
0201
0209
0217
0225
0233
0240 0241
1
0370 0248 02fe

0194
0202
0210
0218
0226
0234
0242
0250

0195
0203
0211
0219
0227
0235
0243
0251

0196
0204
0212
0220
0228
0236
0244
0252

0197
0205
0213
0221
0229
0237
0245
0253

0198
0206
0214
0222
0230
0238
0246
0254

0449
0457
0465
0473
0481
0489
0497
0505

0450 0451 0452 0453 0454
0458 0459 0460 0.61 0462
0466 0467 0468 0469 0470
0414 0~75 0476 0471 0478
0~82 0483 0484 0485 0486
0490 0491 0492 0493 0494
0498 0499 0500 05dl· 0502
0510
0506 0507 0508 nS09
. ---

0455
0463
0471
0479 1
0487

0000
0010
0000
0020
to
0030
0511
(Decimal ) 0040
0050
0060
0070

Decimal

10000 - 4096
20000 - 8192
30000 - 12288
40000-163a4
50000 - 20480
60000 -·24576
70000 - 28672

0100
0110
0120
0130
0140
0150
0180
0170
0200
0210
0220
0230
0240
0250
0260
0270
0300
OliO
0320
10330

0192
0200
0208
0216
i 0340
1 0224
03~0 0232

i~36~
1000
to

1777

(Octal)

03~7

0375
03B3

i
i

~~~i I

05.!..!J

--,
7

i
I

0518 0519 11400 0768 0769
0526 0527
1410 10776 0777
0534 0535
1420 0784 0785
~542 0543
1430 0792 0793
0550 0551
1443 08JO 0801
0558 0559
1450 0808 080~
0566 0567
1460 0816 0817
0574 0575
1470 0824 0825

071Q
0778
0786
0794
0802
0810
0818
0826

0771
0719
0787
0795
0803
0811
0819
0827

0712
0780
0788
0796
0804
0812
0820
0828

0713
0781
0789
0797
0805
0813
0821
0829

0774
0782
0790
0798
0808
0814
0822
0030

0581
0589
0591
0605
0813
0621
0629
0637

0582
0590
0598
0606
0614
0622
0630
0638

0583
0591
0599
0607
0615
OE2l
0631
0639

1500,0832 0833
0841
1520 08~8 0849
1530; 0856 0857
1540 I 0864 0865
1550' 0872 0873
1560 1 0880 0881
1570 0888 0889

0834
0842
0850
0858
0866
0874
0882
0890

0835'
0843
0851
0859
0367
0875
0883
0891

0836
0844
0852
0860
0868
0875
0864
0092

0837
0845
0853
0861
0869
0877
0885
oe93

0838
0846
0854
0852
0870
0878
0886
089~'

0878
0887
0095

0644
0652
0660
0668
0616
0684
0692
0700

0645
0653
0661
0669
0677
0685
0693
0701

0646
0654
0662
0670
0678
0686
0694
0702

0647
0655
0663
0671
0679
0607
0695
0703

1600 0896
1610 0904
1320 0912
1630 0920
1640 0928
1650,0936
1660 10944
1670 10952

0897
0905
0913
0921
0929
0937
0945
0953

0898
0906
0914
0922
0930
0938
0946
0954

0899
0907
0915
0923
0931
0939
0947
0955

0900
0908
0916
0924
0932
0940
0948
0956

0901
0909
0917
0925
0933
0941
0949
0957

0902
0910
0918
0926
0934
0942
0950
0958

0903 [
09111
0919
0927
0935
0943!
0951!
0959 j

0708
0716
0724
0732
0740
0748
0756
0764

0109
0717
0725
0733
0141
0740
0757
0765

0710
0718
0726
0734
0742
0750
0758
0766

0711
0719
0727
073S
0743
0751
075'
0781

1100 0960 0961 0962 0963
1710 i0968 0969 0970 0971
1720,0976 0977 0978 0979
0985 0986 0987
1740 0992 0993 0994 0995
1750 1000 1001 1002 1003
ITGOlooe 1009 1010 1011
1770 IOU 1017 IOU 1019

0964
0972
0980
0988
0996
1004
IOU
1020

0512
0520
0528
0536
05H
0552
0560
0568

0513
0521
0529
0537
0545
0553
U561
0569

0514
0522
0530
0538
0546
055.
0562
0570

0515
0523
0531
0539
0547
0555
0563
0571

0515
0524
0532
0540
0548
0556
0564
0512

0517
0525
0533
0541
0549
0557
1)565
0573

1100
1110
1120
1130
1140
1150
1160
1170

0576
0584
0592
0600
0608
0616
0624
0632

0571
0585
059.3
0601
0609
0617
0625
0633

0578
0586
0594
0602
0610
0618
0626
0634

0579
0587
059S
0603
0611
0619
0627
0635

0580
0588
0596
0604
0612
0620
0628
0636

1200
1210
1220
1230
1240
1250
1260
1270

0640
0648
0656
0664
0672
0680
0688
0696

0641
0649
0657
0665
0673
0681
0689
0691

0642
0650
0658
0666
0674
0682
0690
0698

0643
0651
0659
0667
0675
0683
0691
0699

1300
IJIO
1320
1330
IHO
1350
1360
1370

0704
0712
0720
0728
0736
0744
0752
0760

0705
0713
0721
0729
0737
0745
0753
07GI

0706
0714
0722
0730
0738
0746
·0754
0762

0707
07 I 5
0723
073!
0739
0747
0755
07&3

0512
to
1023

·0384
0600 0392
0610
0620 0400
0630 0408
0640 0416
0650 0424
0660 I 0432
0670,0440
1
0199 0700' 0448
0207 0710 0456
0215 0720,0464
0223 0730! 0472
0231 0740 0480·
0239 0750 0488
0247 0750 0496
02~ 0770 0504
1

0
1000
1010
1020
10301
(Decimal) 1040
1050
1060
1070

0
0256
0264
0272
0280
0288
0296
0304
0312

1510108~0

0775
0783
1
0791
0799,
0807i
0815i
0823 1

0831
0839 1
0847 1
0855 1

::::1
1

"'"11"

L-l

0965 0966 0967
0973 0974 0915 1
Ogel 0982

0989
0997
100;
1013
1021

0990 0991
0983
009a 0999 1
1006
1014 1007
lOIS
IO~1

CONVERSION TABLES
OCTAL-DECIMAL INTEGER CONVERSIONS (Continued)

2000
to
2777

I0
1024
to
1535

1091
1099
!I07
1115
1:23
1131
1139
1147

1092
1100
1108
1116
1124
il3Z
1140
1148

1093
1101
1109
1117
1125
1133
1141

1094 1095
1102 1103
1110 1111
1118 1119
1126 1127
1134 1135
1142 1143
1149 1150 1151

2500 1344 1345
2510 1352 1353
2520 1360 1361
2530 1368 1369
2540 1376 1377
2550 1384 1365
2560 1392 1393
2570 1400 1401

1153
1161
1169
1177
IIB5
1193
1201
1209

1154
1162
1170
117a
1186

1155
1163
1171
1179
118'
1194 \195
1202 1203
1210 1211

1156
1164
1112
1180
1188
1196
1204
1212

liS? 1158 1159
IUS 1166 1167
1173 1174 i175
1181 1182 1183
JlB9 1190 USI
1197 1196 1199
1205 1206 1207
1213 1214 1215

2600
2610
2620
2630
2640
265C1
2660
2670

1408
1416
1424
1432
1440
1448
1456
1464

1409
1417
1425
1433
1441
1449
1457
1465

1221
1229
1237
1245
1253
1261
1269
1277

2700
2710
2720
2730
2740
2750
2760
2770

1472
1480
1488
1496
1504
1512
1520
1528

1489
1497
1505
1513
1521
1529

1216
1224
1232
1240
1246
12511
1264
1212

1217 1218 1219 1220
1225 1226 1227 1228
1233 1234 1235 1236
1241 1242 1243 1244
1249 1250 1251 1252
1257 1258 1259 1260
1265 1266 1267' 1268
un 1274 1275 1276

~
3777
(Octal)

1349
1357
1365
1373
1381
1:.189
1397
1405

1090
1096
1106
1114
1122
1130
1138
lUG

1088
1096
2120 1104
2130 1112
2140 1120

1222
1230
1238
1241
1254

1348
J 356
1364
1372
1380
1388
1396
1404

1089
1097
1105
1113
1121
1129
215011128
2160 1136 1137
2170 1144 1145

1048
1056
1064
1072
1080

lOll

nos

1029
1037
1045
1053
1061
1069
1077
1085

1030
1038
1046
B054
1062
1070
1076
1086

1400 1280 1281 \282 1283 12e'4

1028
1036
1044
1052
1060
1068
1076
1084

2300
2310
2320
2330
2340
2350
2360
2370

1027
1035
1043
1051
1059
1067
1075
1083

nOOllU2
2210 1160
2220 1168
2230 1176
2240 1184
22~0 1192
2260 1200
2210 1208

3000

1026
1034
1042
1050
1058
1066
1074
1002

(Decimal) 2030!
2040
2050
2060
2070
Octal Decimal
10000 - 4096
2100
20000· 8192
2110
30000 - 12288

40000 - 16384
50000 - 20480

1025
1033
1041
1049
1051
1065
1073
1081

2001l!

1032
201°1
202Q 1040

(Octal)

60000 - 24576
70000 - 28672

I02~

1039 2410 1288 1289 1290 1291 1292 1293
1041 2420 1296 1297 1298 1299 1100 1301
1055 2430 1304 1305 1306 1307 B308 1309
1063 2440 1312 1313 UI4 1315 1316 1311
1071 2450 1320 1321 132~ 1323 1324 1325
1079 2460 1328 1329 1330 1331 1332 1333
1081 2470 1336 1337 1338 1339 1340 1341

1223
1231
1239

1247
1255
1262 1263
1270 1271
1278 1279

1346 1347
135~ 1355
1332 1363
1370 1371
1378 1379
1386 1387
1394 1395
1402 1403

-7

1286 1287
1294 1295
1302 1303
1310 1311
1318 1319
1326 1327
1334 1335
1342 1343

1350
1358
1366
i3H
1302

1351
1359
1367
IJ1S
138J

1390 1391

1398 1399
140e 1407

1410 1411 1412 1413 1414 1415
1418 1419 1420 1421 1422 1423
1426 1427 1428 1429 1430 1431

1436
1444
1452
1460
1468

1437
1445
1453
1461
1469

1438 1'139
1446 1447
1454 IUS
1462 1463
1470 1471

1473 1474 un 1476
1481 1432 1483 1484
1490 1491 1492
1498 1499 1500
1506 1507 1508
1514 1515 1516
1522 1523 1524
1530 1531 1532

1477
1485
1493
1501
1509
1517
1525
1533

1494
1502
1510
1518
1526
1534

1434
1442
1450
1458
1466

1435
1443
1451
1459
1467

1478 147.
1486 1417
1495
1503
1511
1519
1527
1535
7

1537
1545
1553
1561
156S
1577
i585.
1593

1538
1546
1554
1562
1570
1578
1586
1594

1539
1547
1555
1563
1571
1579
1587
1595

1540
1548
1556
1564
1572
1580
1588
1596

1541
1549
1557
1565
1573
1581
1589
1597

1542
1550
1558
1566
1574
1582
1590
1598

1543
1551
1559
1567
1575
1583
1591
1599

3400
3410
3420
3430
3440
3450
3460
3470

1792
1800
1808
1816
1824
1832
1840
1848

1601
1609
1617
1$25
1633
1641
1649
1657

1602
1610
1610
1626
1634
1642
1650
J65a

1603
1611
1619
1627
1635
1643
1651
1659

1604
1612
1620
1626
1636
1644
1652
1660

t605
1613
1621
1629
1637
1645
1653
1661

1601
1615
1623
1631
1639
16~6 1647
16.4 1655
1662 1663

3500
3510
3520
3530
3540
3550
3560
357D

1856
1064
1872
1880
1888
1896

3200 1664 1665
3210 1672 1673
322011680 1681
3230 1688 1689
3240 1696 1697
3250 1104 1705
326011712 1713
3270 1720 1721

1666
1674
1682
1690
U9a
1706
1714
1722

1667
1675
1683
1691
1699
1707
1115
1723

1668
1676
1684
1692
1700
1708
1116
1724

1669
1677
1685
1693
1701
11091717
1725

1670
1678
1686
169 ,

~
2039

_1~~~'12~ 2M7

CONVERSION TABLES
OCTAL-DECIMAL INTEGER CONVERSIONS (Continued)

4400 2304 2305
4410 2312 2313
4420 2320 2321
4430,2,328 2329
4440 2336 2331
445012344 2345
4460 2352 2353
4470 2360 2361

2301
2314
2322
2330
2338
2346
2354
2362

2307
2315
2323
2331
2339
2347
2355
:i363

2308
2311
2324
2332
2340
2348
235!
2364

not
2317
232$
2333
2341
2349
2357
2365

2310 ,t;1\
2311,\13.'
2321 '~327
2334 335
2342 1343
1350 2351
2351 235.
nil 2317

2119
Zl27
2,135
2143
2151
2159
Zl67
,2175

1
4500 2368
4510 2376
4520 \2384
4530,2392
45401 2400
4550i 2408
4560 2416
457012424

2369
2371
2385
2393
2401
2409
2417
2425

2370
2378
2386
2394
2402
2410
24\8
2426

2371
2379
2387
2395
2403
2411
2419
2427

2372
2380
2388
2396
2404
2412
H2O
2428

2373
2311
2389
2397
2405
2413
2421
2429

2374
2312
2390
239.
2401
2414
2422
2430

2375
2383
2311
2.319
2407
2415
2423
2431

2180
2188
2196
2204
2212
2220
2228
2236

2181 2182 2183
2119 2190 2191
2111~ 2198 2199
2205 2206 2201
2213 2214 2215
2221 2222 ·2223
2229 2230 2231
2231 2238 2239

4600 2432
4610 2440
4620 2448
4630 2456
4640' 2464
4650: 2472
4660,24,80
4670 \2488

2433
2441
2449
2457
2465
2473
2481
2489

2H4
2442
2450
2458
2466
2474
2482
2490

'2435
2443
2451
2459
246?
2475
2483
2491

2436
2444
2452
2460
2468
2476
2184
2U2

2437
2445
2453
2461
2469
2477
24,85
2493

2438
2446
2454
2462
2470
2471
2485
2494

243.
2447
2455
2413
2471
2479
2417
2495

2244
2252
'2260
22S8
2216
2284
2292
2300

2245
2253
2261
2269
2277
2285.
2293
2301

4700 2496
471012504
4720,2512
413012520
474012528
4150 2536
4760: 2SH
i4770! 2552

2497
2.505
2513
2511
2529
2537
2545
2553

2.498
2506
2514
2522
2530
2538
2546
2554

2499
2507
2515
2523
2531
2539
2547
2555

2500
2508
2516
2524
2532
2540
2548

2501
2509
2517
2525
2533
2541
2549
2~..s6 2557

2502
2510
2518
2UI
2534
2542
2550
2558

2503
2511
2519
2527
2535
2543
2551
2559

2050 2051
2058 2059
2066 2067
2074 ,2075
2082 2083
2090 2091
2098 2099
2106 2107

2052
2010
2068
2076
2084
2092
2100
2108

2053
2011
2069
2071
2085
2093
ZlOI
2109

2054
2012
2070
207'
2086
2094
2102
2110

2055
2013
2071
2079
2017
2095
2103
2111

2\13
2121
2129
2137
2145
2153
2161
2169

2114
%122
2130
2138
2146
2154
2162
2170

2115
2123
2131
2139
2147
2155
2163
2171

2116
2124
2132
2140
2148
2156
2164
2172

2117
2125
2133
2141
2149
2157
2165

2118
2126
2134
2142
2150
2158
2166
2174

4200
4210
4220
4230
4240
4250
4260
421p

2176 2171
218~ 2185
2192 2\93
2200 2201
2208 2209
2216 2217
2224 2225
2232 2233

2178
2186
2\94
2202
2210
2218
2226
22H

2179
2187
2\95
2203
2211
2219
2227
2235

4300
4310
4320
4330
4340
4350
4360
4370

2240
2248
2256
2254
2212
2280
2288
2296

2241
2'49
2257
2265
2273
2281
2289
2297

2242
2250
2258
2266
2274
2282
2290
2298

2243
2251
2259
2267
2275
Ull
2291
2299

4000
4010
4020
(Octal) (Decimal ) 4030
4040
4050
Octal Decimal
4060
10000· 4096
4070

2048
2056
2064
2072
2080
2018
2et6
2104

2057
2065
21173
2081
2089
2091
2105

4100
4110
4120
4130
4140
4150
4160
4170

2112
2120
2128
2136
2144
2152
2160
2168

~I~
4777
2559

20000·, 8192
30000 . 12288
40000 . 16384
50000 . 20480
60000 . 24576
70000 . 28672

20~9

217~

2246
2254
2262
2270
2278
2286
2294
~302

2247
22S5
2263
2271
2279
2287
2295
2303,

i
i

2560
2568
257&
2584
2592
2&00
2608
2616

2561
256.
2577
2585
25"
21101
2109
2617

2562
2570
2571
25841
2594
2102
2etO
2618

2563
2571
'2579
2517
25.5
2S03
2811
2619

2564
2572
2580
2511
25,.
2804
2&12
2620

2565
2573
2511
251"
2597
2105
Ut3
:1821

540012816
2~66
258!1 541012824
2574 2575
5420 2832
~582
2590 2591
2583 5430 2840
5440
2848
2599
25'8
1
2101 2&07 5450 2856
2114 2615 5460 2864
2622 2823 5470 2872

2817
2825
2833
2841
2849
2857
2865
2873

2818 2819 2820 281.1
2826 2827 2828 2829
2834 2835 2836 2837
28~2 2843 2844 2845
2850 2851 2852 2853
2858 2859 2860 2861
2866 2861 2868 2869
2874 2875 ~876 2877

510012624
5110 2632
5120 2640
5\30 2648
5140 2656
5150 2664
5180 2672
5170 2680

2625
2633
2641
2649
2657
2665
2673
2681

2626
2634
2642
2650
2658
2666
2674
2682

2627
2635
2643
2651
2659
2667
2615
2683

2628
2636
2644
2652
2660
2668
2676
2684

2629
2637
2645
2653
2661
2669
2677
2685

2630
2638
2646
2654
2682
2670
2618
2686

2631
2639
2647
2655
2663
2671
2679
2687

5500 2880
551012888
5520 2896
5530 2904
5540 29\2
555012920
5560 2928
557012936

2881
2889
2897
2905
29\3
2921
2929
2937

2882
2890
2898
2906
29\4
2922
2930
2938

2883
289\
2899
2907
29\5
2923
2931
'2939

2884
2891
2900
2908
2916
2924
2932
2940

5200 12688
5210 2696
522012704
5230 1 271 2
5240,2720
5250! 2728
5260 2736
5270 2744

2689
2697
2705
2713
2721
2729
2137
2745

2690
2698
2706
2714
2722
2730
2738
2746

2691
2699
2707
2715
2723
2731
2139
2747

2692
2700
2108
27\6
2724
2732
2740
2748

2693
2701
2709
2717
2725
2733
2741
2749

2694
2702
2710
2718
2726
2734
2742
2750

2695
2703
2711
27\9
2727
2135
2143
2751

5600 12944
'5610 i2952
562012960
5630 2968
564012976
5650 2S8~
5660 2992
,5670 3000

2945
2953
2961
2969
29n
2985
2993
3001

2946
2954
2962
2970
2978
2986
2994
3002

2947
2955
2983
297:
2979
2987
2995
3003

5300
5310
5320
5330
5340
5=50
5380
5370

2753
2761
2759
2777
2785
2793
2801
2809

275~

2762
2770
2778
2716
2794
2102
'21t 0

2755
2763
2771
2779
2787
2795
2803
2ItI

2756
2764
2772
2780
2788
2796
2804
2112

2757
2765
2773
2781
2789
2797
2105
2113

2758
2766
2774
2782
2790
2791
2106
2814

2759
2767
2775
2783
2791
2799
2807
2115

3009
3017
3025
3033
3041
3049
3057
3015

3010
3018
3026
3034
3042
3050
3051
301S

3011
3019
3027
3035
3043
3051
3059

2152
2160
2768
2171
2784
2792
2100
2101

L-3

~700

5710
5720
5730
5740
5750
5760
5770

3008
301'6
3024
3032
3040
3048
305S
3014

301~

5000
5010
!iOOO
2560
to
to
5020
5777
3071
(Octal) (Decimal ) 5030
5040
5050
5060
5070

•

2822 2121
2130
2831
2846
2154
2162
2170
2871

2131
213.
:147
2151
2..'
2nl
217"

2885
2893
2901
2909
2917
2925
2933
2941

2886
2894
2902
2910
2918
29U
2934
2942

21.7
2895
2903
2911
21119
2927
2935
2943

2948
2956
2964
2972
2980
29!8
2996
3004

2949
2957
2965
2973
2981
2989
2997
3005

2950
2958
2966
2974
2982
2990
2998
3006

2951
2959
2967
2975
21113
2911
2919
3007

3012
3020
3028
3036
3044
3052
3010
3068

3013
3021
3029
3037
3045
3053
3061
3069

3014
3022
3030
3038
3046
3054
3012
3070

3015
3023
3031
3039
3047
3055
3013
3071

CONVERSION TABLES
OCTAL-DECIMAL INTEGER CONVERSIONS (Concluded)

8000 3072
8010 3080
6020 3088
6777
3583
I)
(Octal) (Decima 6030 3096
8040 3104
e050 3112
6060,3120
Octal Decimal
607013128
10000· 4096
1
20000· 8192
6100 :3136
30000·12288
6110.3144
40000 . 16384
6120.3152
50000 . 20480
G130 i 3160
60000 . 24576
6140131S8
70000 . 28672
6150·3176
6160 3184
GJ7D 3192

6000

3072

3073
3081
3089
3097
3105
3113
3121
3129

3074
3082
3090
3098
3108
3114
3122
3130

3075
3083
3091
3099
3107
3115
3123
3131

3076
3084
3092
3100
3108
3116
3124
3132

3077
3085
3093
3101
3109
3117
3125
3133

3078
3086
3094
3102
3110
3118
3126
3134

3079
3087
3095
3103
3111
3119
3127
3135

3137
3145
31,3
316i
3169
3177
3185
3193

3138
3146
3154
3162
3170
3178
3186
3:194

3139
3147
3155
3163
:'171
31H
3187
3195

3140
3148
3156
3164
3172
3180
3188
3196

3141
3149
3157
3165
3173
3181
3189
3197

3142
3150
3158
3166
3174
3182
3190
3198

J143
JI51
3159
3167
3175
3183
3191
JI99

64~! me

-7

3333
3341
3349
3357
3365
3373
3381
3389

3334
3342
3350
3358
3366
3374
·3382
3390

3335
3343
3351
3359
3367
3375
3383
3391

3397
3405
3413
3421
3429
34~6 3437
3444 3445
3452 3453

3398
3406
3414
3422
3430
3438
3446
3454

3399
3407
3415
3423
3431
3439
3447
H55

3329
3337
334.5
3353
3361
3369
3377
64701.3384 3385

.3330
3338
3346
3354
3362
3370
3378
3386

3331
3339
3347
3355
3363
3371
3379
3387

3332
3340
3348
3356
3364
3372
3380
3388

6500 3392 3393
3401
3409
6530: 3416 3417
.6540 1 3424 3425
6550,3432 3433
6560 3440 3441
6570 3448 3449

3394
J402
3410
3418
3426
3434
3442
3450

3395
3403
3411
3419
3427
3435
3443
3451

3396
3404
3412
3420
3428

641°1 3336
6420
3344
6430 3352
6440 3360
645013368
6460 3376

I:~;~: ~:~~

3461 3462 3463
15200 3200 3201 3202 3203 3204 3205 3206 3207 1 6600 I 3456 3457 3458 3459 3460
3210 3208 3209 3210 3211 3212 3213 3214 3215 16610; 3464 3465 3466 3467 3468 3469 3470 3471

J216
3224
3232
3240
3248
3256

3!l17
3225
3233
3241
3249
3257

3218
3226
3234
3242
3250
3258

3219
3227
3235
3243
3251
3259

3220
3228
3236
3244
3252
3160

3221
3229
3237
3245
3253
3261

3222
3230
3238
3246
3254
3262

3223 6620 3472 3473 3474 3475
3231 ·6630: 3480 3481 3482 3483
3239 6640· 3488 3489 3490 3491
3247 6650 I 3496 3497 3498 3499
3255 6660! 3504 3505 3506 3507
3263 6670i 3512 3513 3514 3515

3476
3484
3492
3500
3508
3516

3477
3485
3493
3501
3509
3517

3478
3486
3494
3502
3510
3518

3479
3487
3495
3503
3511
3519

6300,3264
8310 : 3272
832013280
6330 3288
5340 : 3296
6350 3304
6360,3312
6370 I 3J20

3265
3273
328 !
3289
3297
3305
3313
3321

3266 3267
3274 3275
3282 3283
3290 3291
3298 3299
3306 3307
3314 3315
3322 3323

3268
3276
3284
3292
3300
3308
3316
3324

3269
3277
3285
3293
3301
3309
3317
3325

3270
3278
3286
3294
3302
3310
3318
3326

3271
3279
3287
3295
3303
3311
3319
3327

3588
359.
3604
3612
3620
3628
3636
3644

3589
3597
3605
3613
3621
3629
3537
3645

3590
3598
3608
3614
3622
3630
3638
3646

3591
3599
3607
3615
3623
3631
3639
3647

8220
6230
6240
6250
G260
6270

7000

7777
(Octal)

3584
10
4095
(DEcimal )

7000 I
7010
7020
7030
7040
7050
7060
7070

3584
3592
3600
3608
3616
3624
3632
3640

3585
3593
3601
3609

7100
7110
7120
7130
7140
7150
7100
7170

3648
3656
3664
3672
36@0
3698
3ege
3704

36~9 3650 3651 3652
3657 3658 3659 3650
3685 3666 3667 3568
3673 3674 3675 3676
3881 3682 3683 3664

3587
3595
3603
3611
~617 3618 3619
3625 3626 3627
3633 3634 3635
3641 3642 3643
3586
3594
3602
361U

6700 I 3520
16710 13528
16720 I 3536
16730 3544
,6740 3552
1615013560
,6760: 3568
[!770: 3576

3eu

3751
3759

3787
3775

3793 3799
3106 3807
3814 3815
Jan 3823
3UO ~31
3i3e 31139

L-4

3845

3846
3854
3862
3870
3878
3886
3894
3902

3847
3855
3663
3871
3879
3807
3395
3903·

3906
3914
3922
3930
393S
3946
3954
3962

3907
3915
3923
3931
3939
3947
3955
3963

3908
3915
3924
3932

3909
3917
3925
3933
39~0 3941
3949 3949
3956 3957
3964 3965

3910
3918
3926
3934

3968
3976
3904
3992
7MO 4000
7650 4008
7660 4016
7670 402~

3969
3977
3985
3993
4001

3970
3978
3986
3994
4002
~OO9 4010
4017 4018

3971
3979
3987
3995
4003
4011
4019
4027

3972
3980
3988
3996

3973
3981
3989
3997
400~ 4005
4012 4013
4020 4021
4028 ~O29

39H
3982
3990
3998
4006
4014
1022
4030

4037
4045
4053
4061
4069
~O7~ ~076 4071
4083 4034 4085
4091 4092 4083

4038
4046
4054
4062
4070
4078
40GG

3782 3783
3790 3791

3743

3527
3535
3543
3551
3559
3567
3575
3583

3905
3913
3921
3929
3937
3945
3953
3961

3778 3717 3778 3779 3780 3781
nB4 3785 ~718 3797 3768 3789
3792 3793 379~ 3795 3796 3797
~IOO 3MI 3802 3903 3804 3eos
310S 3809 lUO 3@1I 30U J313
73$~
lBn 1918 3819 3320 3821
7~0 3134 3a25 3926 3927 382S 3929
'370 3632 3033 Jil34 JUS 3838 3S37

7300
7310
7320
7330
7:140

3742
3750
3758
3766
3714

3526
3534
3542
3550
3558
3566
3574
3582

3904
3912
3920
3928
3936
3944
3952
3960

7600
7610
7620
7630

3726 3727
3734 3735

3525
3533
3541
3549
3557
3565
3573
3581·

3857
3865
3873
3881
3889
3897

3'18 3719

3715
3723
3731
3139
374?

3524
35J2
3540
3548
3556
3564
3572
3580

3842 3843 3844
3852
3858 3859 3860
3866 3867 3863
3874 3875 3876
3882 3883 3884
3890 3891 3892
3899 3899 3900

3718 3717
3724 3725
3732 3733
3HO 3741
3748 3749
375~ 3756 3757
3763 3764 3785
3771 3772 3773

3714
3723
3730
3738
3746
3754
3782
3770

3523
3531
3539
3547
3555
3563
3571
3579

3841

7500
7510
7520
7530
7540
7550
7560
7570

7200 3712 3713
7210 3720 3721
1220 3721\ ~39
7230 373. 3737
7240 3744 ~745
1250 37~2 3753
7aeo 3780 376\
7270 3108 3769

3522
3530
3538
3546
3554
3562
3570
3578

7400 38~0
7110 3848
1
7420 3856
1
7430 3864
7440 3872
7450 3880
7460 30B8
7470 3896

3653 3654 3655
3663
3671
3679
3687
3695
leu
3703
3711

36S1 3662
3569 3670
3677 3678
3685 3688
3694
3689 3690 3691 33Q2
3701 3702
3700
3691 3698 3699
370~ 3706 3707 3700 3709 3710

3521
3529
3531
3545
3553
3561
3569
3577

7700
'7710
7720
7730
7740
7no
7760
7770

3849 3850 3851

402~

~026

4032 4033
~040 4041
4048 4049
4056 4057
4064 4085
4072 4073
4080 4081
~08e 4089

4034
4042
4050
4058
4066
4074
4082
4090

4035
4043
Q051
4059
4067

4036
4044
4052
4060
4066

3853
3861
3869
3877
3885
3893
3901

3911
3919
3927
3935
39~2 394JI
3950 3951
3958 3959
3966 3967

~084

3975
3983
3991
3999
4007
4015
4023
4031

403S
4047
4055
4063
4071
4079
4087
4005

CONVERSION TABLES
L.2

POWERS OF TWO

-2

1.0
0.5
0.25
0.125
0.062
0.031 25
0.015 625
0.007 812
0.003 906
0.001 953
0.000 976
0.000 488
0.000 244
0.000 122
0.000 061
0.000 030
0.000 015
0.000 007
0.000 003
0.000 001
0.000 000
0.000 000
0.000 000
0.000 000
0·.000 000
0.000 000
0.000 000
0.000 . 000
0.000 000
0.000 000
0.000 000
0.000 000
0.000 000
0.000 000
0.000 000
0.000 000
0.000 000
0.000 000
0.000 000
0.000 000
0.000 000
0.000 000
0.000 000
0.000 000
0.000 000
0.000 000
0.000 000
0.000 000
0.000 000
0.000 000
O.OQO 000
0.000 000
0.000000
0.000 000
0.000 000
0.000 000
0.000 000
0.000 000
0.000 000
0.000 000
0.000 000

5
25
125
562
281
140
070
035
517
258
629
814
907
953
476
238
119
059
029
014
007
003
001
000
000
000
000

oo
Source Exif Data:
File Type : PDF
File Type Extension : pdf
MIME Type : application/pdf
PDF Version : 1.6
Linearized : Yes
XMP Toolkit : Adobe XMP Core 4.2.1-c041 52.342996, 2008/05/07-21:37:19
Create Date : 2010:06:18 11:07:20+10:00
Modify Date : 2017:08:04 11:29:56-07:00
Metadata Date : 2017:08:04 11:29:56-07:00
Format : application/pdf
Document ID : uuid:126e6315-a50d-42c4-8560-77cf276bd5da
Instance ID : uuid:14ac9556-44d2-f34a-b0cd-e2d776ea6f49
Producer : Adobe Acrobat 9.0 Paper Capture Plug-in
Page Layout : SinglePage
Page Count : 280

EXIF Metadata provided by EXIF.tools

DEC 11 XPTSA B D PDP Paper Tape Software Handbook

DEC-11-XPTSA-B-D PDP-11 Paper Tape Software Handbook DEC-11-XPTSA-B-D PDP-11 Paper Tape Software Handbook

Navigation menu

Versions of this User Manual:

Views

Navigation