Digital PDFs

AA-5075A-TC

August 1977

182 pages

Original

66MB

view download

OCR Version

30MB

view download

Document:	PDP-11 MACRO-11 Language Reference Manual
Order Number:	AA-5075A-TC
Revision:	0
Pages:	182
Original Filename:

OCR Text

Order No. AA-5075A-TC

August 1977

To order additional copies of this document, contact the Software Distribution

Center, Digital Equipment Corporation, Maynard, Massachusetts 01754.

digital equipment corporation - maynard. mossachuwtts

First Printing,

August 1977

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

DECsystem-10
DECtape

MASSBUS
OMNIBUS

DECUS
UNIBUS
COMPUTER LABS

EDUSYSTEM
FLIP CHIP
FOCAL

PHA
RSTS
RSX

PDP

COMTEX
DDT
DECCOMM

DIBOL

INDAC
LAB-8

DECSYSTEM-20

0S/8

TYPESET-8
TYPESET-10
TYPESET-11

11/78-15

CONTENTS

Page

PREFACE

MANUAL OBJECTIVES AND READER ASSUMPTIONS

ASSOCIATED DOCUMENTS

DOCUMENT CONVENTIONS

STRUCTURE OF THE DOCUMENT

STATEMENT FORMAT

Label Field
Operator Field

Operand Field

Comment Field

FORMAT CONTROL

Illegal Characters
Unary and Binary Operators

&
.
L2
&
]

DIRECT ASSIGNMENT STATEMENTS
REGISTER SYMBOLS

LOCAL SYMBOLS

CURRENT LOCATION COUNTER

MACRO~11 SYMBOLS
Permanent Symbols
User-Defined and Macro Symbols

1
2

EXPRESSIONS

ADDRESSING MODES

e
s

CHAPTER
e

RELOCATION AND LINKING

CHAPTER

U
W - O

2
3

NUMBERS
TERMS

Separating and Delimiting Characters

CHARACTER SET

WHWWWWWWWWWwWwWwWw
LI
I | | Pl I |
HHHEOITOUTOEA,
WD
N e

SYMBOLS AND EXPRESSIONS

CHAPTER

i
-

PROGRAMMING IN MACRO-11 ASSEMBLY LANGUAGE

PART II

AUTOINCREMENT MODE

iii

PROGRAMMING STANDARDS
AND CONVENTIONS
1
2

N
i
st

SOURCE PROGRAM FORMAT

CHAPTER

OVERVIEW OF MACRO-11l
Assembly Pass 1
Assembly Pass 2

NN N

1
2

INTRODUCTION TO MACRO-1l1l

MACRO-11 FEATURES

CHAPTER

ix
ix

MNM?&&M
1

PART I

(S0

CONTENTS

AUTODECREMENT

DEFERRED

IMMEDIATE

MODE

ABSOLUTE

MODE

RELATIVE

MODE

RELATIVE

DEFERRED

MODE

INSTRUCTION

USING TRAP

PART

aWN

BB BB WWWWWWWWN
P R
e
&
&
&
®
&
=
3
[]
L
L
&
L]
s & &

DATA

STy
W N
.

b
b

CONTROL

and Unary Control Operators

.FLT2 and .FLT4 - Floating=-Point Storage
Directives
Temporary Numeric Control Operators: “C and
COUNTER CONTROL

DIRECTIVES

.EVEN Directive
.ODD Directive
.BLKB and .BLKW Directives
DIRECTIVES

(SR

.END Directive
.EOT Directive

L]
-

]
»
.
L]

PROGRAM SECTIONING DIRECTIVES

“R

TERMINATING

,DSABL

FACILITIES

PROGRAM BOUNDARIES

AND

Radix Control and Unary Control Operators
.RADIX Directive
Temporary Radix Control Operators:
"D, "0,

LOCATION

LTI

.ENABL

DIRECTIVES

DIRECTIVES:

STORAGE

DIRECTIVES

.NLIST Directives

Numeric Directives

HOOOOOWOOOO~IJOon

and

RADIX AND NUMERIC

B> b
.« o
.

UL b

ASSEMBLER DIRECTIVES

.BYTE Directive
.WORD Directive
ASCII Conversion Characters
.ASCII Directive
.ASCIZ Directive
.RAD50 Directive
Temporary Radix-50 Control Operator:

€

OGO O O O O O

S
s

5
&

=
&
S
s

FUNCTION

&
]
L]
L4

DIRECTIVES

Page Headings
.TITLE Directive
.SBTTL Directive
.IDENT Directive
.PAGE Directive/Page Ejection

and

NN
NAN

MACRO-11

LISTING CONTROL

.LIST

FORMS

ADDRESSING

INSTRUCTIONS

III

GENERAL

&
s
&
s
&

INDEX

BRANCH

MODE

INDEX MODE

SUMMARY OF ADDRESSING

MODE

bbdwWwww

MODE

RSOSSN

DEFERRED

AUTODECREMENT

WO ooNOYU

AUTOINCREMENT

Gttt
Ut

J
o
Q
o

(Cont.)

DIRECTIVE:

+LIMIT

.PSECT Directive
Creating Program Sections
Code or Data Sharing
Memory Allocation Considerations
.ASECT and .CSECT Directives
SYMBOL CONTROL DIRECTIVE:
CONDITIONAL ASSEMBLY

.GLOBL

DIRECTIVES

CONTENTS

(Cont.)

Page

6.10.2

Subconditional Assembly
Block Directives:

6.10.3

.IFF, .IFT, .IFTF
Immediate Conditional Awsambly Directive:

6-43

.IIF

6.10.4

PAL-11R Cmndxtimnal Asm&mbly mmr@etlves

6-46
6-46

CALLING MACROS
IN MACRO DEFINITIGMS AND ‘MACRO CALLS

ARGUMENTS

- «NARG,

AND

.NARG Directive
.NCHR Directive

7-11

.NTYPE Directive
.PRINT DIRECTIVES

.ERROR AND
INDEFINITE

REPEAT

_
BLOCK DIw@CTIqu~

7-13
7-14

.IRP AND

7-15

.IRP Directive
.IRPC Directive
REPEAT

BLOCK

7-15

DIRECTIVE:

MACRO LIBRARY DIRECTIVE:

.REPT, .ENDR
+MCALL

7-16
7-17
7-18

MACRO=-11 CHARACTER SETS

A-1

ASCII

A-1
A-4

CHARACTER SET

RADIX-50 CHARACTER SET
MACRO-11 ASSEMBLY LANGUAGE AND ASSEMBLER

APPENDIX

B-1

SPECIAL CHARACTERS

W N

SUMMARY OF ADDRESS MODE

PERMANENT SYMBOL TABLE

Q00

OP CODES

MACRO-11 DIRECTIVES

(PST)

DIAGNOSTIC ERROR MESSAGE

MACRO-11 ERROR CODES

B-1
B-1
B-2

SYNTAX

ASSEMBLER DIRECTIVES

oDww

DIRECTIVES

i
-

BRE

-NCHR,

BIRECTIVES*

Macro Nesting
Special Characters in Mamxm Arguments
Passing Numeric Arguments as Symbols
Number of Arguments in Macro Calls
Creating Local Symbols Amtmmmtmcmlly
Keyword Arguments
Concatenation of Macro Argumentw
MACRO ATTRIBUTE

¢«

*
.
.
.

NN da

.MACRO Directive
.ENDM Directive
+MEXIT Directive
MACRO Definition Foxmattlmg

«IRPC

APPENDIX

QQQQQ\Q\Q&NQMMHH

I
-

> W

=
e
&
o
&
8
e

DEFINING MACROS

~J]OY
UT o W N =

BPWWWWWWWWNR

NTYPE

APPENDIX

6-41

MACRO DIRECTIVES

APPENDIX

vIF"

+ENDC

Conditional Assembly Block Directives:

NN NN NSNSNSNSNNNNNA
S
L4 L4 »
L
|]
* L
& [
L]
L
[]
s
L
L]

CHAPTER

6.10.1

c-1

Cc-1
c-4
SUMMARY

D-1
D-1

(Cont.)

O
o
O
o

CONTENTS

APPENDIX

.
*

[ ]

a2
*
®
&
®
=
&
&
B
$
a8

bhWNKHE

&
®
®

Using the Standard Symbmllca
Symbols
Global Symbols

Symbol Examples
Program-Local Symbols
Macro Names

PROGRAM MODULES

General Comments on Prmgrams
The Module Preface
Formatting the Module Preface
Modularity

Calling Conventions (IntarmMmdule)
Exiting
Intra-Module Calling Conventions
Success/Failure Indication
Module Checking Routines

WNHE

FORMATTING STANDARDS

Program Flow
Common Exits
Code with Interrupts

Inhlblted

PROGRAM SOURCE FILES
FORBIDDEN INSTRUCTION USAGE

RECOMMENDED CODING PRACTICE

Conditional Branches
PDP-11 VERSION NUMBER STANDARD

Displaying the Version Identifier
Use of the Version Number in the Program
ALLOCATING VIRTUAL MEMORY

GENERAL

HINTS AND

MACRO DEFINITIONS
OPERATIONAL

APPENDIX

4
i
=

FORMAT

COMMENTS
NAMING STANDARDS

OO O
&
-

APPENDIX

mmmmmmmmmmmmmm?mwmmmmmmmmmmmm
BN

LINE

HFOWVLWVWVWYWWVWRONVUTUINERWWWWWNNNDNDNDN
M

"INTRODUCTION

UV
WN

HHMOOOJOAOOAUVIUTUTUIUIUIUTUT
UL OB 8B D D0
B D
DD DD WA

SAMPLE CODING STANDARD

TECHNIQUES

WRITING POSITION

INTRODUCTION

SPACE-SAVING GUIDELINES
AND EXPANSIONS

INDEPENDENT

POSITION

CODE

G-1

INDEPENDENT

CODE

G-1
G=~2

REFERENCE

LISTING

H-1

EXAMPLES
APPENDIX

SAMPLE

ASSEMBLY

AND

CROSS

CONTENTS

(Cont.)

Page

FIGURES
FIGURE

3-1

Assembly Listing Showing Local Symbol Block

3-11

3=-2

Sample Assembly Results

3=-12

6-1

Example of Line Printer Assembly Listing

6-6

6-2

Example of Terminal Assembly Listing

6-7

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

Listing Produced With Listing Control
Directives
|
Assembly Listing Table of Contents
Example of .ENABL and .DSABL Directives
Example of .BLKB and .BLKW Directives

7-1

Example of

.IRP

and

.IRPC Directives

6-9
6-12
6-16
6-30
7-17

TABLES

TABLE

3-1

3-2
3-3
3-4
3=-5
6-1
6=-2

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

Special Characters Used in MACRO-11l

Legal Separating Characters
Legal Argument Delimiters
Legal Unary Operators
Legal Binary Operators
Symbolic Arguments of Listing Control
|
Directives

Symbolic Arguments of Function Control

Directives
Symbolic Arguments of .PSECT Directive
Non-IAS/RSX-11 Program Section Default Values
Legal Condition Tests for Conditional Assembly
Directives
Subconditional Assembly Block Directives

vii

3-1

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

6-14
6-33
6-39
6-41
6-44

PREFACE |

0.1

MANUAL OBJECTIVES AND READER ASSUMPTIONS

The

intent of this manual is to enable users ta davelmp pragrams coded
in the MACRO-11aaaembly language.
No prior knawlaflge of the MACRO-11

Relocatable Assembler is assumed.

Although
the
descxiptimn,
of
the
asgembly l&nguage
is
wholly
self-contained
within
this
manual,
the reader is assumed
to be
familiar with
the
PDP-11
processors
and related
t@f%lflfllQQYr
as
presented
in
the
PDP-ll Processor Handbooks. No attempt is made in
this document to describe the PDP-11 hardware or the functions of
the
various PDP-1l1l instructions.
Since the development of
programs
necessarily involves
1linking
to
create
an
executable
image,
the
reader
1is encouraged
to become
familiar with this process, as
present@d ;in
the
applmgable
system

manual

(see Sectlon 0. 3)

gia

In presenting MACRO-11l, a tutorial bias has been
adopted
to
enlarge
upon
the
reference
material.
This
posture
is reflected in the

examples and the accompanying commentary describing'MACRO-11

language

elements
1in
typical
applications.
Portions of text that are shaded
indicate that a particular MACRO-11 feature is not avmxlabl@ in the 8K

version of MACRO-11l.

0.2

‘

STRUCTURE OF THE DOCUMENT

This

manual contains three parts.
Part I, consisting of two chapters,
briefly
introduces
MACRO-1ll.
Chapter
1
1lists the key features of
MACRO-11l,
and Chapter
2
identifies
the
advantages
of
following
programming
standards
and conventions.
Also described is the format
used in coding MACRO-1ll source programs.
Part

II,

consisting

three

chapters,

presents

general

information

essential
to programming with the MACRO-1ll assembly language.
Chapter
3 describes the symbols, terms, and expressions that form the elements
of
MACRO-1l instructions.
The character
set is listed, and the types
of programming symbols that may be defined
by the user are
discussed.
Chapter
4
describes
the
output
of
MACRO-1ll and presents concepts
essential to the proper relocation
and
1linking
of
object
modules.
Chapter

5 briefly describes how data stored
in memory can be accessed
and manipulated using the addressing modes recognized
by
the
PDP-11

hardware.

Part

III,

<consisting

two

chapters,

describes

the

MACRO-11

directives
that
control
the
processing of source statements during
assembly.
Chapter 6 discusses directives which accomplish generalized
MACRO-11
functions, while Chapter 7 deals with directives used in the
definition and expansion of macros.
Finally,
several
appendixes
are
provided,
supplying
information of interest to the MACRO-1l1l programmer.
Appendix A

lists

the ASCII

and

Radix-50

character

sets

additional

that

may

used
in
MACRO-1ll
programs.
Appendix B lists the special characters
recognized
by
MACRO-1l1l,
summarizes
the
syntax
of
the
various
addressing
modes used in PDP-11 processors, and briefly describes the
MACRO-11l directives
have
been
defined
Appendix C.

in alphabetical order.
The permanent symbols that
for use with MACRO-1ll are listed alphabetically in

The diagnostic error codes produced by MACRO-1ll
to identify
various
types
of
errors detected
during
the
assembly process are listed
alphabetically in Appendix D.
Appendix E
contains a sample coding
standard
that is recommended practice in preparing MACRO-11l programs.
Appendix F discusses several
methods
of
conserving
dynamic memory
space
for
wusers
of
small
systems who may experience difficulty in
assembling MACRO-11l programs.
o
|

Appendix G is a discussion of position independent code

0.3

(PIC).

ASSOCIATED DOCUMENTS

The reader should refer

listed
manual.

below

for

the

applicable

descriptions

documentation directory

documents

associated

with

this

IAS Documentation Directory
'RSX-11D Documentation Directory
‘RSX-11M/RSX-11S Documentation Directory
RT-11

0.4

Documentation Directory

DOCUMENT CONVENTIONS

‘The symbols defined b&lmw‘are used throughout this manual.

Symbol

Definition

[]

“

n
|

Ceee

Brackets

argument

Vertical bars indicate that a single choice must

optional.

made

from a list

Elllpgls

list

indicate

indicates

the

form of

that

|
the

enclosed

arguments.

optional

the

last

)

continuation of an argument

spaaifmed argument.
TM

UPPER-CASE
CHARACTERS

lower—case
characters

(n)

Upper-case characters indicate elements of the language
that must be used exactly as shown.

Lower-case characters indicate elements of the language
that are supplied by the programmer.

In some instances the symbol (n) is used following a
For example, 100(8)
number to indicate the radix.
indicates that 100 is an octal value, while 100(10)
indicates a decimal value.

x1i

PART

INTRODUCTION TO MACRO-11

CHAPTER
1

;nacammll,mfikru?%Sf;w
The MACRO-11 Assembler provides the fallowimg f®mgug@az

swurme'and command'mtring contrai mf“amaambly'fiunctians .

'¢3;<;Err0r llfitmng on camm&nd mmtput fi@vama
4

Alphabetlzed,

symbml tabl@ lm&tlmg,,

formatted

cross-reference listing of &ymbwla

Relocatable object modules

Global symbols for

‘

fiptional
e

linking object modules

7. Cdnditional assembly directives
8.

angram aaafiiqming,diréative$'

9. User-defined maprofi and?mamro libfiafiigé .
10.

Comprehensive system macro library

of listing
ive source | and , commandS stringe icontrol
|
11. Extens
functions.
1.1

OVERVIEW OF MACRO-1l

The fiunctioma anw,mperatlmns m&lavant
MACRO-11 is a 2-pass assembler.
to each assembly pass are described in thefollowing sections.

1.1.1

Assembly Pass 1

The main purpose of assembly pass 1 is to locate and read all required
macros
from libraries;
to build symbol tables and program section
while also performing a rudimentary assembly
tables for the program;
of

each

source

statement.

The first stage of asaambly pass 1 is the initialization of all impure
uses internally for the assembly process.
data areas that MACRO-1l1l
These areas include all dynamic storage areas and buffer areas used as
file

After

storage

regions.

initializing memory areas,

subroutine

which

MACRO-1ll issues a call

transfers a command line into memory.
1-1

system

This command

MACRO-11

line contains
assembly.

the

After

specifications
scanning

the

FEATURES

command

the

files

line

for

initializes the specified output files.
These
determine
if valid output file specifications
command line.
They are
active file space.

then

closed

proper

used

syntax,

during
MACRO-11

files
are
opened
have been passed in

minimize

requirements

to
the
for

As the assembly process begins,
retrieves
source
1lines
from

MACRO-11
initiates
a
routine which
the
input
file.
If no such file is
currently open, as is the case at the beginning of assembly,
MACRO-11
opens
the
next
input
file specified in the command line previously
read
and
begins
to
assemble
the
source
statements.
MACRO-11

determines
as

one

the

word,

length

of each

two words,

instruction and assembles

three

words.

it accordingly

At the end of assembly pass
1,
MACRO-11
described above and writes out information

reopens
the output files
that is to be used later in
linking the object modules.
Such information
as
the object
module
name,
the
program
version
number,
and the global symbol directory
(GSD) entries for each program section are output to the object
file.
After
writing
out
the
GSD
entries for
a
given program section,
MACRO-11l scans through the
symbol
tables
to
find all the
global
symbols that
are bound to that particular program section.
MACRO-11
then writes out GSD records to the
object
file
for
these
symbols.
This
process
continues for each program section, bringing
to a close
assembly pass 1.

1.1.2
As an

Assembly Pass
integral

part

2
of

pass

MACRO-11

simultaneously

writes

the

object
records to the output file and generates the assembly listing,
followed
by
the
symbol
table
listing
for
the
program.
A

cross-reference

listing may

also

generated.

Basically, assembly pass 2 consists of the
same
steps performed
1in
assembly
pass
1,
except
that
all
source
statements
containing
"MACRO-ll-detected errors
are
flagged with
an
error
code
as
the
assembly

1listing

file

created.

The

object

file

that

created

the final consequence of pass
2
contains
all
the
object
records,
together
with relocation records containing information necessary for
subsequent linking of the object file.

The information thus passed enables the global symbols in
the
object
modules
to
be
associated with absolute or virtual memory addresses,
thereby forming an executable body of code.
|
|
The

and

user

may wish
description.

system manual

(see

to become

This
Section

familiar with
information
is
0.3

the

macro object file format
presented in the applicable

Preface).

;;;;;

CH&PTER 2
SOURCE PRQGR&M FOR%AT

2.1

PROGRAMMING STANDARDS AND CONVENTIONS

Assembly level programming deals
directly
with
the
host
hardware.
Hence,
great
care
must
be exercised
1in establishing programming
standards and conventions to enable code written by one
group
to
be

interchanged easily with another group.

‘Standards provide a number of

advantages.
When
applied to
the
pr»gram "“w@valmpment
standards make the programming effort easier to:

process,

Plan

Comprehend
Test

Modify
Convert

Even

though
aspects
of

standards must

applicability.

the

The

program

accommodate
develapm&nt

standarda

common

local Vrmqumr@memtw,

pr&wfiww

all

“have

many
uniwwraal

of l%GITAL’

PDP-11

software
products
are
preaanted in %wpemw%~fi~f~**~ model forusers.
Observance of these standards is beneficial to DIGITAL and its
users,
by simplifying both communications and the continuing task of software
maintenance and enhancement.

2.2

STATEMENT FORMAT

A source program
is composed of a sequence
of source
coding
lines.
Each
1line contains
a
single assembly-language statement.
MACRO-1l
will accept a source line of 132 characters,
but 80 characters is
the
recommended
langth,
because of congtralnts imposed by Iistlng format
and terminal

line

size.

A MACRO-1l1l statement may consist of as many as
four fields.
These
fields
are
identified
by
their
order
of
appearance
within
the
statement and/or by specified separating
characters between
fields.
The general format of a MACRO-1ll statement is:
Label:

The

Operator Operand

label

and comment

;Comment(s)

fields are

optional.

The

operator

and

mperand

fields are
interdependent, i.e.,
when both fields are present in a
source statement, each field is evaluated by thntwll in
the
context
of the other.
“
A statement may contain an operator field and no
operand
field,
but
the
reverse
1is not true.
A statement containing an operand with no

operator does not conform to established MACRO-1ll coding
conventions;
such
a statement is currently interpreted by MACRO-1l1l during assembly
as an implicit .WORD directive (see Section 6.3.2).
2-1

SOURCE PROGRAM FORMAT

MACRO=-1l1l interprets and processes source
program
statements
one
by
one,
generating
one
or
more
binary instructions or data words, or
performing a specified assembly process.
Blank lines, although legal,
have no significance in the source program.
An assembly~language statement must be completed on one
no continuation lines are allowed in MACRO-1l1l.
The

tab character

fields

into

can be

aligned

source program format,
Label

= begins

Operator

= begins

used

the

source

columns

below:

shown

column

statement

source

1line;

format

the

accordance with DIGITAL's standard

in column

Operand(s) = begin(s) in column 17
1Cmmment(s) - begin(s) in column 33

For @mampla, the follmwing statement should be farmatted in. thm source

program
into
specific
assembly listing:

columns,

increasing

REGTST:BIT#MASK,VALUE
; COMPARES BITS

its readability
|
P

IN OPERANDS.

REGTST:

BIT

#MASK ,VALUE

; COMPARES BITS IN OPERANDSaif;~

The

above

(columns)

formatting conventions are

programs

(free-field

coding

in the

not mandatory

in coding

permxasmble)..

MACRlel

‘However,

~recommended that source programs be prepared in accmrdance wmth the&e
conventions for confimstancy and clarity.
;
~
|

2.2.1

Label Field

A label

a means

symbolically

referring

program.

1location

A label

is
a user-defined symbol which is assigned the
value of the
current
location
counter
and
entered
into the user-defined symbol
table. The current location counter is the means
by which MACRO-11

~assigns

memory addresses to the

source program statements
as they are

encountered during the assembly process.
The
address value of the
label
1is
absolute
or
relmcatable. depending on whether the current
‘program section being assembled
is
absolute or relocatable.
(The
concept
of program sections and the attributes that may be . fip&filfl@d
for them are discussed in detail in Section 6.8.)
In the case of an absolute program section, the value of
the
current
location
counter
is likewise absolute, i.e., itsvalue references an
absolute virtual memory address (such as
location
100).
Similarly,

;th&._value_ of

~section isalso

the

current location counter in a relocatableprogram

relocatable;

however, a

relocation bias calculated at

‘link time will be added to the apparent value of the current location
counter
to
establish
execution time.

its

effective

absolute

virtual

address

I1f present, a label always appears as the
first field in a source
%tatemgwt and mumt be tarminated by a amlmnfif Emw;emw
1@, if the
e&

ABCD:

Iuav'

A,B

assigns the value 100(8) to the label ABCD.
Subsequent references to
this label would then yield a value mfi absolute 100(8). 1In this

wmampl@,,mf the lmcatzan cmunt@r value
value of ABC

w&rw sr««*~~
ble, thw final
. represents the r location

,biamof the mmwgmww ,fieatiwm, as a&lcul&te« 'bg thw ;mmk%uilfiwm atlink
time.

&0re than one labwl may agpwar thhmn a ;minqlwIab@l fxalfl*
label

specified is assigned the same address value.

if the current location counter value is 100(8),

the

in the following statement:
ABC:

$DD:

A7.7:

MOV

Each

For example,

multiple

1labels

A,B

t value 10&(3);
uhe
are each assigne
Multiple labels may also appear on successive lines.
statements

ABC:
$DD:
A7.7:

MOV

For‘&x&mpler the

A,B

likewise cause the same current lwcmtxon amumt&x walue to be afiflmqm@d
to all three labels.

ROy

Of the two methods of
assigning multiple 1labels mhwwm
abmvm#, the
second
is preferred, because consistency of field ww&mtmanmnq within
the source program improves readability.

A double colon (::) defines the label as a glmbal m;wywlw

label

can

rmfietencwa

Such a

by imdwpandwnhly“awwwwwle%,flbjeat wwwulawfi

References to this label in other modules will be resolved when the
nodules are linked as a composite executable iwwqu For wxawwla, the
SEateront

ABCD::

MOV

A,B

establishes the label ABCD as a global
symbol.
The distinguishing
attribute of a global symbol is that it can be referenced from within
an object module other than the module in which the symbol is defined
(see Section 6.9).

The 1@@&1 charaat&r@ for defifiingglabalfi are:
A through 2
0 through 9
. (Period)

(Dollar Sign)

NOTE
By convention,
are
reserved
software

should

the dollar sign ($) and period
(.)
for
use in defining DIGITAL system

symbols.

Therefore

these

characters

not be used in defining labels in MACRO-11

fimumaa;praqmaM$,

2=3

SOURCE PROGRAM

FORMAT

source pmgramw all 1abels are termxnatdby a colmn (2 }, whlch is not

considered
part of the label.
It is a mandatory delimiter. An error
code (M) is generated
in
the
assembly
1listing
if
the first
six
characters in two or more labels are the same (see,Appendifo)
A symbol used as a
program.
If the

label must
not
be redefined
within
the
source
symbol
is redefined,
a 1label with a: multiple

definition results, causing MACRO-1l to generate an error code

the

assembly listing

(see Appendix D).

the source program which references a multi-defined label
an
additional diagnostic message;
in this case, an error
'qenerated in the assembly listing (see Appendxx D)
Lok

2.2.2

Operator

(M)

Furthermore, any fitat@ment,xn

results
in
code (D) is

Field

The operator field specifies the
action
to
be
performed.
It
may
consist
of an instruction mnemonic (op code), an assembler directive,
or a macro call.

The operator field follows the label

Chapters

and

7 describe

When the operator is an
specifies
the
machine

these

field

three

types

source

operator

statement.

field

entries.

instruction
mnemonic,
the mnemonic
op code
instruction
to
be generated.
MACRO-1l1l then

continues with the evaluation of the
address(es)
of
the
operand(s)
which follow(s).
When
the
operator
is a directive,
the directive
causes MACRO-11l to
perform
certain
control
actions
or
processing
aperatlona
during
the
assembly
of
the
source
program.
When the
operator is a macro call, MACRO-1ll inserts th@ code generat@d
by
the
macro expansion.
The operator

preceded

field

need

not

be preceded

by a

label;

but

may

one or more labels and followedby one or more operands

and/or a comment. Furthermore, leading and trailing spaces or tabs in
the
operator field have no significance;
such characters serve only
to separate the
operator
field from
the
preceding and
following
fields.
”‘
~
7

An operator

is terminated by a space,

tab,

as in the following examples:
MOV A,B
v

dony

MOV

;THE SPACE TERMINATES THE opERhToR
o

A,B

MOV@A,B

or any non-RAD50 character,

s MOV
.
N

;THE

TAB TERMINATES

THE

OPERATOR MOV
.

JTHE

@ CHARACTER TERMINATES THE

;OPERATOR MOV.

Although the statements above are

all

second
statement
is the
recommended
MACRO-11l coding conventions.
o

2.2.3

equivalent
in function,
form

because

conforms

the
to

Operand Field

When the operator field contains an
instruction mnemonic¢
(op code),
the
operand
field
specifies
those program variables that are to be

2-4

Wi,

SOURCE PROGRAM FORMAT

evaluated/manipulated by the operator.

The operand field may also

used
to
supply ~arguments to MACRO-1ll directives and macro calls,
desarlbad in Chapterfi 6 and 7, respectively.
|

Operands may be expressxong or $ymb011c argum&ntm (within the

be
as

context

of the specxfx&fl operation).
Multipl@ expressions u%flfl in the operand
field of a MACRO-11 statement must be &epar&ted by a aaqu;
multmple

symbolic
arguments
sxmxlarly
used may be delmmit,d” by any legal
separator, i.e., a comma, tab, and/or apaaafi
An 0p&rand should be
preceded by an operator field; if it is not, the statement is treated
by MACRO-11 as an implicit .WORD directive (see Section 6.3.2).

‘When the aperatmr field ammtaxna an op g@fi@, afiamfilat@d mp&zands are
always expressions,

RO,A+2(RL)

MOV
On the other

directive

‘arguments,

as shown in the following stat&;@mt,

hand,

when

the

operator

field contains a MACRO-1l1

or a macro call, associatedoperanfi% arenmrmally &ymballc
as shawn in the fallmw1ng statem&mt. .

~ .MACRO ALPHA ARG1,ARG2

Refer to the descrlptlmn of each MkCRO*lldmramtmvw tm determine
type and number of operands required in 1$$umng the fiixamtlve~

The operand field is terminated by

followed by a comment. For example,

LABEL:

MOV

A,B

semicolon

when

the

‘the
fl

field 'is

in the following statement:

; COMMENT FIELD

the tab between MOV and A terminates the operator

field

and

defines

the
beginning of the operand field;
a comma separates the operands A
and B;
and a semicolon terminates the operand field and
defines
the
beginning
of
the
comment field.
When no comment field follows, the
operand field is terminated by the end of the source line.

2.2.4

Comment Field

The comment field normally begins in column 33 and extends through the
end
of
the
line.
This field is optional and may contain any ASCII
characters
except
null,
RUBOUT,
carriage-return,
line-feed,
vertical-tab
or form-feed.
All
other
characters appearing in the

comment field, even special characters reserved for use
in
MACRO-1ll1,
are
checked only for ASCII legality and then included in the assembly

listing as they appear

in the source text.

All comment fields must begin with the semicolon

character(;).

When

lengthy comments extend beyond the end of the source line (column 80),
the comment may be resumed in a following
1line.
Such
a
1line
must
contain
a leading semicolon, and it is suggested that the body of the
comment be continued in
the
same
columnar
position
in which
the
comment
began.
A
comment
line can also be included as an entirely
separate line within the code body.

Comments do not
affect
However,
comments
are
debugging,

assembly
processing
or
program
execution.
useful in source listings for later analysis,
or documentation purposes.

Horizontal fmrm@ttlng of the source program is controlled by ‘the spaue
and

tab

characters.

These characters have no effect on

process unless they are embedded within a

symbol,

the assembly

number,

ASCII

text string, or unless they are used as the operator field terminator.
Thus, the space and tab characters can be used to prmvad@ ‘an
orderly
mmd readable fimwxww program, as rafl&ctafi by tha fmllmwmng statem@nts‘
LABELammv(S@}%WT&G,PQP VALUE OFF STACK.
No

spaces or

‘statement.

tabs have

Note

been

used

separate

the

fields

this

the dafflaulty in recmgnizing wh&re ane fleld ends

and the next begins.
LABEL:

Phis

MOV

@tat&%fmt

(SP)+,TAG

conforms

; POP

the

VALUE

OFF

STACK.

standard horizontal

formatting

conventions,
i.e.,
the
statement
elements are aeparated
distinct fields and are therefore easily discernible.

intm

four

Page fntmattlng and assembly listing considerations are
discussed
1in
,Chapter
6 in the context of MACRO-1l1l directives that may be specified
accomplish desired formatting operations.
Appamflxx E describes the
coding conventions
used
in
all
DIGITAL
PDP-11 0perat1ng
system
software.

2-6

PART

PROGRAMMING

IN MACRO-11 ASSEMBLY
LANGUAGE

CHAPTER 3

SYMBOLS AND EXPRESSIONS
This chapter describes the components of MACRO-1ll
instructions.
The
character set, the conventions observed in cmmatructmng symbols, and
the use of numbers, operators, terms and expresaxmns are discussed
as
they relate to MACRO-1ll programming.
|

3.1

CHARACTER SET

The following characters are legal in MACRO-1l1 source programs:
1.

The
are

letters A through Z.
Both upperm and lower—-case
letters
acceptable, although, upon input, lmwarficmsa letters are

converted to upper-case

(see Section 6.2,

.ENABL LC).

2. The digits 0 thraugh 9.
3.

The,@h@ractara

(period)

The special charactars listed in Table 3-1.
Table
~

afidi $

characters
are
reserved
for
use
Corporation system program symbols.

(&ollar
as

sign).

Digital

3-1

Special Characters Used
in MACRO-11

Character

Designation

Colon

Label

Double colon

Label terminator;

Equal sign

Function

label

Direct

terminator.

global

assignment

and macro keyword

These

Equipment

defines the

label.

operator;

indicator.

Double equal

Direct assignment operator;

sign

defines the fiymbml as a global
symbol.

Percent

sign

term indicator.

Tab

Item or field terminator.

Space

Item or field terminator.
(Continued on next page)

SYMBOLS AND EXPRESSIONS

Table

3-1

Special Characters

(Cont.)
Used

Character

Designation

Number sign

(

Left

)

Right parenthesis

‘Period

Current location counter

Comma

Operand field separator.

Semicolon

Comment field indicator.

Left angle

Initial argfifient’ar'expression

bracket

1ndlaatmr*

Right

angle
bracket

Terminal

Plus sign

Arithmetic

Function

Immediate expression
indicator.

sign

\Defe;red

parenthesis

Initial

sion

Minus sign

Asterisk
|

addressing
register

| Terminal

in MACRO~11

p
h
indicator.

indicator.

argum&nt or
1nd1cator.
N

expres-

addltxan

operator

autoincrement indicator.

~Arithmetic subtraction operatmr

autodecrement indica-

{0)

Arlthmetlc multlpllcatlon op-

erator.

Slash

Arithmetic division operator.

Ampersand

Logical AND

Exclamation point | Logical

Double

quote

operator.

inclusive OR operator.

Double ASCII

character

indica-

tor.

- Single
SR

3.1.1
Legal

quote

Single ASCII
tor;
or
indicator.

character indicaconcatenation

‘Up arrow or

Universal unary operator or

circumflex

argument

' Backslash

indicator.

Macro call numeric argument

indicator.

Sepa:ating,and Delimiting Characters
separatlng characters

and

legal

argument

below inTables 3- 2 and 3-3 respectively.

delimiters

are

defined

LS
AND EXPRESSIONS
SYMBO
Table 3-2
Legal Separating Characters

One or more

Space

Usage

Definition

Character

ceae
fi. Sp aaa

spaces

alegal separator

between 1n%trumt10n fleldm and

~and/or tabs

between symbolic mrgum&ntg
field.
~within the mperand
Spaces within expreammmms mx@

kiqnmr@d (a&@Sactlmn 3.9).

Comma

separatmr
A~mammarifl“a
arguments
symbolic
between
field.
operand
within
the
used in
expressions
Multiple
be
must
field
- the operand

separated by
a comma.

Table 3-3

Legal Argument Delimiters

Definition

‘Usage

Paired angle brackets

may
be
Paired angle brackets
anywh@r@ in a program to
used

Character

for
an @H@K@&filmn
'
enclose
‘single term.
treatfi.n&~?aw
Pair@& mngf@~braaketa are also
macro
a
to enclose
used
when
particularly
argument,
argument contains separthat
Section
ating characters (see
7.3).

"XeooX

Up=arrow

ator)
where

(unary oper-

construction,
the up-arrow is

by an argufollowed
ment that is bracketed

by any paired printing
characters

This

construction

1is equiva-

lent in function to the paired
angle brackets described above
only
used
generally
1is
and
where the argument itself contains angle brackets.

(x).

1I1 legal Characters

3.1.2

A character is determined to be 111egal far an& mf two r@asmns~
1.

racognmzed MACRO-11
th@
mf
element
is not an
A character of this kind is r&placed in the
'Wlmsting by a question mark, and an error code (I) is pxmnt@d
in the assembly listing (see &ppandix D). ‘The exception to
"this is an embedded null whlch, wh@n G@tected, terminates the
scan of the current line.
A character
charaater

set.

3-3

3.1.3

A legal MACRO-1l1l character is illegal in the context
of
its
usage
within
the
source statement, 1i.e.,
its
syntax is
illegal or questionable.
Such a character
causes
an
error
code (Q) to be printed in the assembly listing.

Unary and Binary Operators

Legal MACRO-ll

unary operators

are

described

Table

3-4.

Unary

operators
are
wused 1in connection
with
single terms (arguments or
operands) to indicate an action to be performed on
that term
during
assembly. A term
preceded
by
a
unary
operator is considered to
contain that operator.
The term so specified
thus
becomes
a
value

which can be used alone
or as an element of

Table
Legal

Unary
Operator

3-4

Unary Operators

Explanation

Plus

an expression.

Example

sign

Effect

Produces

value
-

Minus

sign

-A

,"Upwarrawr;uniVQr”
{

~sal unary operator.
(This

usage. is

| described in detail
‘in S@ctlon

6.4.)

the

positive

of A.

Produces
the
(2's complement)

negative
value of

~c24

| Produces the l's comple-

“D127

Interprets 127 as a

ment value of

24(8).

~034

Interprets 34 as an octal
number.,

"B11000111 | Interprets 11000111 as
binary number.

“"RABC

Unary

operators

can be

involving multiple

used

terms,

adjacent

each

shown below:

Evaluatéfi ABC in Radix-50

other

constructions

e g950 - (Equivalent to =-<"D50>)
“c 012 (Equivalent to "C<"012>)

Legal MACR@wll binary operators are

described

in Tabla‘ 3-5.

~contrast to unary
operators, binary operators sp&clfy actions to be
performed on multiple items or terms within an expression. Table
3-5
~shows the relationships that
can be establmfihed betwe@n-expression
terms

through

the

use

binary

operatmrs.~n

SYMBOLS AND EXPRESSIONS

Table 3-=5
Legal Binary Operators

Binary

Operator

All

Explanation

Example

Addition

A+B

Subtraction

Multiplication

A*B (l6-bit product returned)

Division

A/B

Logical AND

A&B

Logical

A!B

A-B

(l6-bit gquotient returned)

inclusive OR

binary operators have equal

priority.

Items

terms

can

grouped
for
evaluation within an expression by enclosing them within
angle brackets.
Terms so enclosed are evaluated first, and
remaining
operations
are performed from left to right, as shown in the examples
below:

. WORD
- WORD

3.2

1+2%*3
14+<2%*3>

;EQUALS 11(8).
; EQUALS 7(8).

MACRO-11l SYMBOLS

Three types of symbols may be defined for use within
MACRO-1ll
source
programs:
permanent symbols, user-defined symbols, and macro symbols.
MACRO-11 maintains three types of symbol tables:
the Permanent Symbol
Table
(PST),
the User Symbol Table (UST), and the Macro Symbol Table
(MST).
The PST contains all the permanent symbols defined within (and
thus automatically recognized by) MACRO-11 and is part of the MACRO-11l
image.
The UST and MST are
constructed
as
the
source
program
is
assembled.

3.2.1

Permanent Symbols

Permanent symbols consist of the instruction mnemonics
(see
Appendix
C)
and
MACRO-11]
directives
(see
Chapters 6 and 7 and Appendix B).
These symbols are a permanent part of the MACRO-1ll image and need
not
be defined
before
being
used
in
the operator field of a MACRO-1ll
source statement (see Section 2.2.2).

3.2.2

User-Defined and Macro Symbols

User—-defined symbols are

those symbols treated by

the

programmer

labels
(see
Section
2.2.1)
or that are equated to a specific value
through a direct assignment statement (see Section 3.3) or
appear
as

macro
names
or dummy arguments. These symbols are added to the User
Symbol Table as they are encountered during assembly.
Macro
symbols

are

those

symbols used as macro names

these symbols are

added

encountered during

assembly.

the

Macro

3-5

(see Section 7.1).
Symbol

Table

Similarly,
they

are

SYMBOLS AND EXPRESSIONS

User-defined
characters,
character is

and
macro
symbols
dollar
signs
($),
illegal.

can
be
composed
and
periods
(.)

of
alphanumeric
only;
any other

NOTE

The

dollar

sign

($)

and

period

(.)

characters

are

reserved
for
wuse
1in
defining Digital Equipment
Corporation system software symbols.
For example,
READS is a file-processing system macro.
The user

is cautioned not
constructing

in order

following rules govern the

these

symbols

to avoid possible

or
future
Digital
software symbols.

The

employ

user-defined

characters
or

macro

symbols

conflicts with existing

Equipment

«creation

Corporation

system

user-defined

and

macro

symbols:

The

in
2.

The

first character of a symbol must not be a number
the case of local symbols;
see Section 3.5).

first

symbol

six
can

characters

be written with

symbol
more

must

than

six

(except

unique.

legal

characters,

but
the
seventh
and subsequent characters are checked only
for
ASCII
legality
and
are
not
otherwise
evaluated
or
recognized by MACRO-11l.
4.

Spaces, tabs, and illegal characters
within a symbol.
The legal MACRO-11l
in Section 3.1.

must
not
be
embedded
character set is defined

The value of a symbol depends upon its use in
symbol
appears
in the operator field, it may

the
program.
When
a
be any one of the three
symbol types described above i.e., permanent, user-defined, macro.
To
determine the value of an operator-field symbol, MACRO-1l1l searches the
symbol tables in the following order:

Macro Symbol Table

Permanent Symbol

User-Defined Symbol

Table

This search
order
allows redefinition
of
Permanent
Symbol
Table
entries
as
macro symbols.
That is, permanent symbols may be used as
macro symbols.
But the user must keep in mind the sequence
in
which
the
search
for
symbols
1is
performed
in
order to avoid incorrect
interpretation of the symbol's use.
When a symbol appears in
Table is searched first,

the operand field,
the
User-Defined
Symbol
then the Permanent Symbol Table is searched.

Depending on their use in the
have
either
attribute.

local

through

symbols

one

'program,

attribute

user~defined
a global

symbols

(external)

;

Normally, MACRO-11l
their
definition
However,

source

(internal)

treats all user-defined symbols as local, that
is,
is
1limited
to
the
module
in which they appear.

can

explicitly

three methods:

3-6

declared

global

symbols

SYMBOLS AND EXPRESSIONS

Use of the

Use of

.GLOBL directive

the double colon

(see Section 6.9).

(::)

in defining a label

(see Section

2.2.1).

Use of the double equal (==
statement (see Section 3.3).

sign

direct

All symbols within a module
that
remain
undefined
assembly are treated as default global references.

assignment
~

the

end

NOTE

Undefined symbols
at
the
assigned
a
value
of
0
user—-defined symbol
table

end
of
assembly
are
and
placed
into
the
as
wundefined
default

global references.
If the .DSABL GBL directive is
~in
effect,
however,
(see
Section
6.2),
the
automatic
global
reference
default
function of
MACRO-11]

inhibited,

causing

the

statement

containing the undefined symbol to be flagged with
an error code (U) in
the
assembly
1listing
(see

Appendix D).

Global symbols provide linkages between independently-assembled object
modules
within
the
task image.
A global symbol defined as a label,

for example, may serve as an entry-point address to another section of
code within the image.
Such symbols are referenced from other source
modules in order to
transfer
control throughout
execution.
These
global
symbols are resolved at link time, ensuring that the resulting
image is a logically coherent and complete body of code.

3.3

DIRECT ASSIGNMENT STATEMENTS

A direct assignment statement allows you

equate

for a direct assignment statement

is:

symbol

specific
value.
When a direct assignment statement is first used to
define a symbol, that symbol is entered into the
User-Defined
Symbol
Table.
A
symbol
defined
in
this
manner
may
be
redefined in a
subsequent direct assignment statement by assigning a new value to the
previously-defined symbol.

The general

format

symbol=expression
or

symbol==expression

where:

expression
,

= can have only one level of forward
.(see

below).

reference

- cannot contain an undefined global reference.

A direct assignment statement embodying the double equal (==)
shown above, defines the symbol as global (see Section 6.9).

sign,

SYMBOLS AND EXPRESSIONS

The

following examples

illustrate

the

coding

direct

assignment

statements:

A=]1

;THE SYMBOL A IS EQUATED TO THE
: VALUE

B=A-1&MASKLOW

;THE

SYMBOL B

sVALUE

THE

sWHICH

FOLLOWS.

EQUATED TO THE

ENTIRE

EXPRESSION

C:
D=.

MOV

#1 ,ABLE

;THE

SYMBOL

;THE

LABELS

AND

; VALUE

;OF

THAT

THE

MOV

EQUATED

AND

E ARE ASSIGNED

EQUAL

THE

LOCATION

INSTRUCTION.

The last of the three examples above is provided
only
to
illustrate
the performance of MACRO-1l in such situations.
See Section 3.6 for a
description of the period (.) as the current location counter symbol.
The

following

conventions

apply

the

coding

direct

assignment

statements:

An equal sign
symbol
from

(=)
the

or double equal sign (==) must separate the
expression
defining
the
symbol's value.

Spaces
preceding
and/or
following
the
operators,
although permissible, have no
resulting value.
2.

The symbol being assigned in
placed in the label field.

Only

one

symbol

can

a direct

defined

direct
assignment
significance in the
~

assignment

single

direct

statement

assignment

statement.

A direct assignment
comment field.

Only

the

statement

one level of forward
following example:
X=Y

(Illegal

Y=2

(Legal

referencing

forward

may

followed

allowed,

only

shown

reference)

z=1
The

above example would result
in
the
assembly
1listing
on
reference.

in the generation of an error code
(U)
the line containing the illegal forward

Although one
1level
of
forward
referencing
1is
allowed
for
1local
symbols, a global symbol defined in a direct assignment statement must
not
contain
a
forward
reference,
i.e.,
the
global
assignment
expression
must
not itself contain an undefined reference to another
symbol.
Such a forward reference is illegal, causing
an
error
code
(A) to be generated in the assembly listing.

SYMBOLS AND EXPRESSIONS

3.4

The eight general registers of the PDP-1ll1
processor
are
numbered
0
through
7 and can be expressed in the source program in the following
manner:

30
3l
®

37
where % indicates a reference to a register rather
than
a
location.
The
digit
specifying
the
register
can
be
replaced by any legal,
absolute term that can be evaluated during the
first
assembly
pass.
Use standard symbolic names for all register references.
The register definitions listed below are
automatically
assigned
MACRO-11,
i.e.,
these
definitions are the normal default values

remain valid

for

all

references within

RO=%0
R1=%1
R2=%2
R3=%3
R4=%4
R5=%5

source program.

sREGISTER
0 DEFINITION.
;REGISTER 1 DEFINITION.
;REGISTER 2 DEFINITION.
;REGISTER 3 DEFINITION.
;REGISTER 4 DEFINITION.
sREGISTER 5 DEFINITION.

SP=%6
PC=%7
Note

the

by
and

that

unique

registers

;STACK POINTER DEFINITION.
; PROGRAM COUNTER DEFINITION.

and

are

given

special

system functions.

names

because

their

A register symbol may be defined
in
a
direct
assignment
statement
appearing
1in
the
program.
The
defining
expression of a register
symbol must be a legal, absolute value.
Although you can reassign the
standard
register symbols through the use of the .DSABL REG directive
(see Section 6.2), this practice is not recommended.
An
attempt
to
redefine a default register symbol without first specifying the .DSABL
REG

directive

override

the

normal

definitions causes

that

assignment
statement
to
be
flagged
with
an error code (R) in the
assembly
1listing.
The
symbolic
default
names
assigned
to
the
registers,
as
listed
above,
are the conventional names used in all
DIGITAL-supplied

well

advised

PDP-1l

system programs.

these

All non-standard register symbols must
referenced
in
the source program.
A

or greater
listing.
The

than

character

specify a

may

CLR

flagged with

used with

For

example,

in function

an error

any

the

to the

3-9

code

legal

(R)

term

statement

the contents of

this

reason,

vyou

are

be
defined
before
they
are
register expression less than 0

%$3+1

is equivalent

and clears

For

conventions.

the

assembly

expression

SYMBOLS AND EXPRESSIONS

contrast,

the

CLR
clears

3.5

the

statement

contents of

virtual

memory

location

LOCAL SYMBOLS

Local symbols are specially formatted symbols used as labels within
a
block
o0f
coding
that
has
been
delimited as a local symbol block.
Local symbols are of the form n$, where n is a decimal integer from
1
to 65535, inclusive.
Examples of local symbols are:

1$
279
59§
1045%

A local symbol block
is delimited in one of three ways:
1.

The range of a local symbol block usually consists
of
statements
between
two normally-constructed symbolic
(see Figure 3-1).
Note that a statement of the form:

those
labels

ALPHA=expression

is a direct assignment

statement

not
<create
a label
local symbol block.

and

The

symbol

range

local

thus

block

encountering
a
.PSECT,
.CSECT,
source program (see Figure 3-1).
3.

The range of
a
1local
MACRO=-11 directives, as

not

symbol
block
follows:

Starting delimiter:

.(ENABL

Ending

.ENABL

delimiter:

(see Section

does

LSB

delimit

normally

3.3),

but

the

range

terminated

.ASECT directive

1is

does

delimited

(see Section

6.2)

(see Section

6.2)

upon

the

through

LSB

.DSABL

LSB

followed

one

of:

.PSECT
.CSECT
.ASECT

(see Section
(see Section
(see Section

Symbolic

label

6.8.1)
6.8.2)
6.8.2)

Local symbols provide a convenient
means
of
generating
1labels
for
branch
instructions
and
other such references within a local symbol
block.
Using local symbols reduces the possibility
of
symbols
with
multiple
definitions
appearing
within a user program.
In addition,
the use of local symbols differentiates entry-point labels from
local
labels,
since
1local
symbols cannot be referenced from outside their
respective local symbol block.
Thus, local symbols of the
same
name
can
appear. in
other
1local
symbol
blocks without conflict.
Local
symbols do not appear in cross-reference listings.

SYMBOLS AND EXPRESSIONS

Local symbols require less symbol table
space
than
other
types
of
symbols.
Their use is recommended.
When defining local symbols, use
the range from 1$ to 63% first, then the range from
1285
to
655358.
Local symbols
within
the range 64$ through 127%, inclusive, can be
generated automatically as a feature
of MACRO-1ll.
Such local
symbols
are
useful
in
the
expansion
of
macros
during
assembly
and are
described in detail in this context in Section 7.3.5.
Be sure to avoid multiple definitions of local symbols within the same
local
symbol
block.
For example, if the local symbol 10§ is defined

two or more times within the same

1local

symbol

block,

each

symbol

represents
a
different
address
value.
Such a multi-defined symbol
causes an error code (P) to be generated in the assembly listing.
For

examples

local

in a source program,

symbols

121

)

123
124

125 ooveen
0ARPR4

128 Q200NR6
129 Qo012
130

symbol

blocks

they

appear

PROGRAM

INITIALIZATION

CODE

)

126 240200

131

local

3-1.

)

122

127

and

see Figure

o0o00pe

21270¢

oQe2000!

@22709
121374

QReoeer!

005720

LPSECTY

XCTPRGreMOV

181

CLR

XCTPRG,GBL

#IMPURE ,RD

‘

+PSECT

XxCTPAS,GBL

012700

opo0oR!

XCTPASt MOV

HIMPRPAS,RD

134 PPORRE
135 9neas2

P227¢0
121374

PAPARO!

CMp
Bml

BIMPPAT,RD
1%

212700

eeopo@!

136
137 ee0ape

138 2000232

139 200004
140 20206
141 200012

142

Q05020

Q05020
P22702
141374

Figure

3.6

181

3-1

CLR

PSECT

CLR
cMp
Bml

INITIALIZATION

JPASS INITIALIZATION

(RP)

XCTLINIIMQV

eRRRAR!

HIMPURT,RQ
1%

132 ooveed
133 200024

JIMPURE DATA

(RBY)s

CMP
Bml

XCTLIN,GBL

BRIMPLIN,RD

JLINE INITIALIZATION

(RRYe
BIMPLIT,RD
1%

Assembly Listing Showing Local Symbol Block

CURRENT LOCATION COUNTER

The period (.) is the symbol for the current location
counter.
When
used in the operand field of an instruction, it represents the address
of the first word of the instruction, as shown in
the
first
example
below.
When
wused
in the operand field of a MACRO-1ll directive, it
represents the address of the current byte or word, as
shown
in
the
second example below.
A:

(The

#.,R0

MOV

function

the

;THE PERIOD
(.) REFERS TO THE ADDRESS

;OF

symbol

THE MOV INSTRUCTION.

is explained

in Section

5.9.)

SAL=0

.WORD

177535,.+4,SAL

;THE OPERAND .+4
;DIRECTIVE

IN THE

REPRESENTS

;THAT IS STORED AS THE
;OF THREE WORDS DURING

.WORD

A VALUE

SECOND

;ASSEMBLY.
Assume

that

the

current

value

the

location

counter

assembly, MACRO-11l reserves storage in response to the
(see Section
6.3.2),
beginning
with
1location
500.
accompanying
the .WORD directive determine the values

3-11

500.

During

.WORD directive
The
operands
so stored.
The

SYMBOLS AND EXPRESSIONS

value 177535 is thus stored in location 500.
The value represented by
.+4
is
stored in location 502;
this value is derived as the current
value of the location counter (which is now 502),
plus
the
absolute
value
4,
thereby depositing the value 506 in location 502.
Finally,
the value of SAL, previously equated to 0, is
deposited
in
1location
504.
Figure

3-2

illustrates

the

Figure

result

the

example.

LOCATION

CONTENTS

500

177535

502

506

504

3-2

Sample Assembly

Results

At the beginning of each assembly pass, MACRO-1ll resets
the
1location
counter.
Normally, consecutive memory locations are assigned to each
byte of object data generated.
However, the
wvalue
of
the
1location
counter
can
be
changed through a direct assignment statement
of the
following

form:

.=expression

Similar to other MACRO-1ll symbols, the current location counter
(.)
has
an
attribute
of
relocatability associated with it:

symbol
it is

either
absolute
or
relocatable,
depending
on
the
specific
such
attribute
of the current program section.
(A program section and its
attributes are
defined
through
the
wuse
of
the
.PSECT
directive
described
in
Section 6.8.1.) The existing attribute (or mode) of the
current location counter cannot be changed
by
specifying
a
defining
expression having a different attribute.
Furthermore, such a defining expression must not
force
the
1location
counter
1into
another
program section (.PSECT area), even though the
program sections so involved may both be absolute or relocatable.
The
expression
defining
the
1location
counter
value must not contain a
forward reference, i.e., the expression must not contain
a
reference

symbol

that

1is

not

previously

constitute a general assembly error,
the

assembly

Thus,

the

defined.

Such

violations

resulting in an error code

(A)

the attribute

attribute

(or mode)

the

of the current location counter

current

program

section.

takes on

Therefore,

attribute from program section to program section can be changed
through
the
program
sectioning
directives
(.PSECT,
.ASECT,
.CSECT), as described in Section 6.8.
The

following coding

counter:

listing.

illustrates

the

3-12

wuse

the

current

1its

only
and

1location

SYMBOLS AND EXPRESSIONS

.ASECT

.=500

; SET LOCATION COUNTER TO
; ABSOLUTE

FIRST:

MOV

.+10,COUNT

500 (OCTAL).

; THE LABEL "FIRST" HAS THE VALUE
;500 (OCTAL) .

THE

3.+10 EQUALS 510 (OCTAL).

; CONTENTS OF THE LOCATION
;510 (OCTAL) WILL BE DEPOSITED
;IN THE LOCATION "COUNT."

; THE ASSEMBLY LOCATION COUNTER

.=520
|

SECOND: MOV

. , INDEX

;NOW HAS A VALUE OF

; ABSOLUTE 520 (OCTAL)
.

; THE LABEL SECOND HAS THE

; VALUE 520 (OCTAL).
; THE CONTENTS OF LOCATION
3520 (OCTAL), THAT IS, THE BINARY
;CODE FOR THE INSTRUCTION
; ITSELF, WILL BE DEPOSITED
; LOCATION "INDEX."

.=.+20

.PSECT

IN THE

;SET LOCATION COUNTER TO
; RELOCATABLE 20 OF THE
; UNNAMED PROGRAM SECTION.

THIRD:

.WORD

; THE LABEL THIRD HAS THE
;VALUE OF RELOCATABLE 20.

Storage areas may be reserved in the program by advancing the location
counter.
For example, if the current value of the location counter is
1000,

each of the

following statements:

—

=.+40
or

.BLKB 40
or

.BLKW 20

reserves 40(8) bytes of storage space
1in
the
source
program.
The
.BLKB
and .BLKW directives, however, are recommended as the preferred
ways to reserve storage space (see Section 6.5.3).

3.7

NUMBERS

MACRO-11 assumes that all numbers in the

source

program

are

1nterpreted
in octal radix, unless otherwise specified.
An exception
to this is that operands
associated with Floating
Point
Processor
instructions and Floating Point Data directives are treated as decimal
(see Section 6.4.2).
This default
radix
can be
altered with
the

.RADIX directive (see Section 6.4.1.1).
Also, individual numbers can
be designated as decimal, binary, or octal numbers through
temporary
radix control operators (see Section 6.4.1.2).

For every statement in the source program that contains a
digit
that
is
not
in
the <current radix, an error code (N) is generated in the
assembly listing.
However, MACRO-1ll continues with the
scan
of
the
statement
and evaluates each
such
number encountered as a decimal
value.

3-13

SYMBOLS AND EXPRESSIONS

Negative

numbers

must

such

numbers

translates

may

(but

need

not)

preceded

into

two's

preceded

minus

complement

a plus

sign;

form.

the

Numbers

not

assembly

listing.

always

considered

are

absolute

values,

greater

than

error

code

i.e.,

they

are

relocatable.

Single-word floating-point
numbers
may
be
generated
with
operator (see Section 6.4.2.2) and are stored in the following

Sign
Bit

8-bit
Exponent

to the appropriate PDP-11
floating=point number format.

the
°F
format:

7-bit
Mantissa

Refer

3.8

numbers

sign.

A number contalnlnq more than 16 significant bits, i.e.,
177777(8),
is
truncated from the left and flagged with
(T)

MACRO-11

Positive

Processor

Handbook

expression

and

for

details

the

value

TERMS

term

component

may

one

following:

number,

as defined

in Section

3.7,

as defined
follows:

in Section

3.2.

whose

1l6-bit

used.
2.

symbol,

as
a.

are

A period (.) specified in an expression causes
of the current location counter to be used.

A defined
Table

symbol

(UST)

A permanent

substituted
all op codes
d.

Symbols

and

is
its

symbol's

for
and

located
value

basic

the

evaluated

the

User-Defined

value

Symbol

used.
value

the addressing
their values.)

1is

modes.

wused,

with

zero

(Appendix C

lists

An undefined symbol is
assigned
a
value
of
=zero
and
inserted in the User-Defined Symbol Table as an undefined
default global reference.
If the
.DSABL
GBL
directive
(see
Section
6.2)
1is
in
effect, the automatic global
reference default function of MACRO-1l1l is
inhibited,
in
which case, the statement containing the undefined symbol
is flagged
with
an
error code
(U)
in
the
assembly
listing.

A single quote followed by a single
ASCII
character,
or
a
double
quote followed by two ASCII characters.
This type of
expression construction is explained
in
detail
in
Section
6.3.3.

3-14

SYMBOLS AND EXPRESSIONS
4.,

A term may also be an expression enclosed in
angle
brackets
(<>).
Any expression so enclosed is evaluated and reduced to

a single term before the remainder of the expression in which
it appears is evaluated.
Angle brackets, for example, may be
used to alter the left-to-right evaluation of expressions (as
in
A*B+C versus A*<B+C>), or to apply
a unary operator to an
entire expression (as in =<A+B>).

3.9

A unary operator followed by a symbol or number.

EXPRESSIONS

Expressions are
combinations
of
terms
J@lned
together
by
binary
operators
(see Table
3-5) and which reduce to a 1l6-bit expression
value.
The evaluation of an @Xpt&fifil@fi includes the determination of
its
attributes.
A resultant expression value may be any one of four
types (as described later in this
section):
absolute,
relocatable,
external, or complex relocatable.
bt
:
|
|
Expressions
are evaluated
from
left to
right
with no
operator
hierarchy
rules,
except
that
unary
operators take precedence over
binary operators.
A term preceded by a unary operator
1is
considered
to
contain
that
operator.
(Terms
are evaluated, where necessary,
before their use in expressions.)
Multiple unary operators are valid
and are

treated as

follows:

o -\

to:

is equivalent

=<L+L=A>>

A missing term,

expression,

or external symbol

Zero.
A m1351ng or
illegal
analyala, causing an error code
in

the

itself.

assembly

1listing,

For example,

TAG

1is

interpreted

operator terminates
the
expression
(A) or (Q),or both, to be generated

dep@milng on the cant@xt of the expression

the expression:

LA 177777

is evaluated as
!

TAG

because the first non-blank character following the symbol LA is not a
legal
binary operator,
an expression separator (i.e., a comma), or an
operand field terminator (i.e., a semicolon or the end of
the
source
line).
It should be noted that spaces within expressions are ignored.
The value of an external expression is

absolute

part

that

expression.

equal

For

the

example,

wvalue

the

expression

EXTERN+A, where "EXTERN"
is
an
external
symbol,
has
a
value
at
assembly-time
that
is
equal
to the value of the internal symbol A.
This expression, however, when evaluated at link
time
takes
on
the
resolved value of the symbol EXTERN, plus the value of symbol A.
Expressions,
four
types:
relocatable.
l.

when evaluated by MACRO-11l, are determined to be
one
of
absolute, relocatable, external (or global), or complex
The following distinctions are important:

An
expression
1is
absolute
expression
whose
terms
are

3-15

if
its
wvalue
1is
fixed.
An
numbers
and
ASCII conversion

SYMBOLS AND EXPRESSIONS

characters will
expression
or

reduce to an absolute value.
term
minus
a
relocatable

A
relocatable
term, where both

elements being evaluated belong to the same program
section,
are
also
absolute, since such an expression is reduced to a
single term by MACRO-1l upon
completion
of
the
expression

scan.

For

expression

absolute

expression.

Terms

absolute

section will

have

defined

the

where

program

section,

that

contain

absolute

labels

both

TAGl

defined

value.

expression is relocatable if its value is
fixed
relative
to
the
base
address
of
the
program
section in which it
appears, but it will have an offset value added at link time.
Terms
that
contain
labels
defined
1in relocatable program
sections will have a relocatable value;
similarly, a
period
(.)
in a relocatable program section, representing the value
current

An expression

location

counter,

is external

will

also

(or global)

have

relocatable

if it contains a single

global reference (plus or minus an absolute expression value)
that is not defined within the
current
progranm.
Thus,
an

assembly

expression
and must be

is complex
applies:

contains

- It

contains

program

is
only
partially
defined
resolved at link time.

expression

conditions

The

global

relocatable

reference

than

relocatable

terms

belonging

from

introduced,

= An

since each
bias

operation

= An

evaluation

has

link

symbol.

different

for

than

is evaluated

the

than

one

addition

external,

time.

terms of

the

program section.
is

specified

than addition, subtraction,
specified for a relocatab}e

relocatable,

completed

symbol

symbol.

operation other
complementation is

expressions

expression

following

For
example,
1if
the relocatable
associated
with
the
same
program
in an expression construction in the
levels
of
relocation
would
be

effect

other

undefined global

the

following

the

relocatable

reference.

resulting

relocation

any of

global

sections.

value

and

one

level
of
relocation.
symbols TAGl and TAG2
section
are specified
form
TAGl+TAG2,
two

The

TAG2-TAGl,

same

external

the

are

of the
value.

-3.

example,

and TAG2Z

and

complex

negation,

value.
relocatable

CHAPTER 4
RELOCATION AND LINKING

The output of MACRO-1l1l is an object module that must be
processed
or
linked
before
it
can
be loaded and executed.
Essentially, linking
fixes (i.e., makes absolute) the values
of
external
or
relocatable
symbols
in the object module, thus transforming the object module, or
several such object modules, into an executable image.

To allow the value of an expression to be fixed at link time, MACRO-1l
outputs
certain
directives
1in
the object file, together with other
required parameters.
In the case of relocatable
expressions
in
the
object
module, the base of the associated relocatable program section
is added to the
value
of
the
relocatable
expression
provided
MACRO-11l.
1In the case of external expression values, the value of

external term in the expression

(since the

external

symbol

must

by
the

defined
in
one of the other object modules being linked together) is
determined and then added to the
absolute
portion
of
the
external
expression, as provided by MACRO-1l.
|
All instructions that require modification at link time are flagged in
the
assembly
1listing,
as
1illustrated
in
the
example below.
The
apostrophe (')
following
the
octal
expansion
of
the
instruction
indicates
that simple relocation is required;
the letter G indicates
that the value of an external symbol must be
added
to
the
absolute

portion
of
an
expression;
and the letter C indicates that complex
relocation analysis at link time is required in order to fix the value
of

the

expression.

EXAMPLE:

005065
CLR
000000G

EXTERN (R5)

;THE VALUE OF THE SYMBOL "EXTERN"
;ASSEMBLED AS ZERO AND IS
;RESOLVED AT

005065
G

CLR

.
|

EXTERN+6 (R5)

LINK

TIME.

;THE VALUE OF THE SYMBOL "EXTERN"
; IS RESOLVED AT LINK TIME
;AND ADDED TO THE ABSOLUTE
; PORTION (+6) OF THE EXPRESSION.

005065
000040"

CLR

RELOC (R5)

;ASSUMING THAT THE VALUE OF THE
; SYMBOL "RELOC" IS RELOCATABLE
;40, THE RELOCATION BIAS
;WILL BE ADDED TO THIS VALUE.

005065
CLR
000000C

-<EXTERN+RELOC>
(R5) ;THIS EXPRESSION IS COMPLEX
s RELOCATABLE BECAUSE IT REQUIRES
;THE NEGATION OF AN EXPRESSION
; THAT CONTAINS A GLOBAL "EXTERN"
; REFERENCE AND A RELOCATABLE TERM.

For a complete description of object records output by MACRO-11l, refer
to the applicable system manual (see Section 0.3 in the Preface).
4-1

CHAPTER 5

ADDRESSING MODES
The program counter (PC) 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
assemble,
the
action of the program counter must be understood.
key rule

to remember

they
The

is:

"whenever the processor implicitly uses the program
counter
(PC)
to
fetch
a
word from memory, the program counter is

automatically 1ncrem@ntad by 2 after th@ fetch operation
completed."

In the case of 2- or 3-word instructions,
fetch the following words as well.

The
following
symbols are
throughout this chapter:
is any expression,

used

the processor uses the PC to

describing

addressing

in Chapter

modes

R is a register expression, i.e., any expression containing a
term
preceded
by
a percent sign (%) or a symbol previously
equated to such a term, as shown in the examples below:
RO=%0
R1=R0+1
R2=1+%1

as defined

;GENERAL REGISTER 0.
;GENERAL REGISTER 1.
;GENERAL REGISTER 2.

The symbol R may also represent any
of
register definitions (see Section 3.4).

the

normal

default

ER is a register expression or an absolute expression in
range

A is a general

inclusive.

the

addressing

specification

which

produces

6-bit
mode
address
field,
as
described
in
the
PDP-l1l
Processor Handbooks.
The
addressing
specification,
A,
is
described
in
terms of E, R, and ER, as defined above.
Each
addreflsxng specification within this section
is illustrated
using either the single operand instruction CLR or the double

operand

instruction MOV.

5.1

The

itself

(R)

contains the operand to be manipulated by

instruction.

5-1

the

ADDRESSING MODES

- Format

for

Example:

CLR

5.2
The

;CLEARS

the

Format

(R)

contains

the

address

the

operand

manipulated

instruction.
for

(ER)

Examples:

CLR

@R1

CLR
CLR

5.3

sALL THESE INSTRUCTIONS CLEAR

(R1)
(1)

s THE WORD AT THE ADDRESS
s CONTAINED IN REGISTER 1.

AUTOINCREMENT MODE

The contents of the register
being

used

Format

for

the

address

(ER) are

the

incremented

operand

(see Note

immediately

below).

(ER)
+

Examples:

CLR

(RO) +

CLR
CLR

(R4) +
(R2)+

s EACH INSTRUCTION CLEARS
: THE WORD AT THE ADDRESS
s CONTAINED IN THE SPECIFIED
+REGISTER AND INCREMENTS
; THAT REGISTER'S CONTENTS
+BY

TWO.

NOTE

Certain

special

combinations,
not

which

operate

processors,

ifistruction/address

are

exactly

rarely

the

described

same

never

mode

used,

all

PDP-11

below.

the autoincrement mode, both the
JMP
and
JSR
instructions autoincrement the register before its
use on the PDP-11/40, but not on the PDP-11/45
or
11/10.
In

double

operand

instructions

having

the

addressing
form
Rn, (Rn)+
or Rn,-(Rn), where the
source and destination registers are the same, the
source operand is evaluated as the autoincremented
or

autodecremented

value,

originally

but

the

destination

it is used, still contains
effective address.
1In the
executed on the PDP-11/40,

contains

100(8):

after

ADDRESSING MODES

MOV

RO, (RO) +

; THE QUANTITY 102 IS MOVED

MOV

RO,-(RO)

; THE QUANTITY 76 IS MOVED

; TO LOCATION 100.

The use of these forms should

they

avoxded,

- since

are not compatible with the entire family of

PDP-11 processors.

An error code
with
listing

(2)
1is
printed
1in
the
aasembly
not
1is
which
instruction
each

compatible among all members of the PDP-11 family.

5.4

AUTOINCREMENT DEFERRED MODE

the operand.
The register (ER) contains a pointer to the address of
The
contents
of
the
register are incremented after being used as a
pointer.

Format for A:

+
@ (ER)

Example:

CLR

5.5

@ (R3)+

; THE CONTENTS OF REGISTER 3 POINT

:TO THE ADDRESS OF A WORD TO BE
; CLEARED BEFORE THE CONTENTS OF THE
; REGISTER ARE INCREMENTED
BY TWO.

AUTODECREMENT MODE

The contents of the register (ER) are decremented before being used as
the address of the operand (see Note above in Section 5.3).
Format

for A:

- (ER)

Examples:

5.6

CLR

- (RO)

CLR
CLR

- (R3)
- (R2)

: DECREMENT THE CONTENTS OF THE SPECIsFIED REGISTER (0, 3, OR 2) BY TWO
:BEFORE USING ITS CONTENTS
:AS THE ADDRESS
OF THE WORD TO BE
: CLEARED.

AUTODECREMENT DEFERRED MODE

The contents of the register (ER) are decremented before being used as
a pointer to the address of the operand.

Format for A:

@- (ER)

Example:

CLR

@- (R2)

s DECREMENT THE CONTENTS OF

sREGISTER 2 BY TWO BEFORE
sUSING ITS CONTENTS AS A POINTER
;TO THE ADDRESS OF THE WORD TO BE

; CLEARED.
5-3

ADDRESSING MODES

5.7
The

INDEX MODE
value

expression

(E)

stored

the

second

third

word

of
the
instruction.
The
effective
address
of
the
operand
is
calculated as the value of E, plus the contents of register
ER.
The
value E is the offset of the instruction, and the contents of register

form

Format

the

for

base.

E (ER)

Examples:

CLR

X+2 (R1)

MOV

RU, -2 (R3)

;THE EFFECTIVE ADDRESS OF THE WORD
;TO BE CLEARED IS X+2, PLUS THE
; CONTENTS OF REGISTER 1.
; THE EFFECTIVE ADDRESS OF THE
;DESTINATION

;THE

5.8

LOCATION

CONTENTS

-2,

PLUS

INDEX DEFERRED MODE

An expression (E), plus the contents of
a
register
(ER),
vyields
a
pointer
to
the
address of the operand.
As in index mode above, the
value E is the offset of the instruction, and the contents of register
ER

form

Format

the

base.

for A:

@E (ER)

Example:

CLR

@114 (R4)

; IF

;VALUE,

PLUS

CONTAINS

THE

OFFSET

100,

THIS

114,

YIELDS
214.
IF LOCATION 214
; CONTAINS THE ADDRESS 2000, LOCATION
;2000 WOULD BE CLEARED.

;THE

5.9

POINTER

IMMEDIATE MODE

Immediate mode allows the operand itself
second or third word of the instruction.
autoincrement of the PC.
Format

for

(E)

This

mode

be
is

stored
as
the
assembled as an

Examples:

MOV
MOV

The

#100, R0

#X,R0

:MOVE THE VALUE 100 INTO REGISTER O.
;MOVE THE VALUE
;REGISTER 0.

OF SYMBOL X

INTO

number
sign
(#)
in
the
MACRO-11
character
set
has
special
significance
as
an
addressing
mode indicator.
When this character
appears in the
operand
field,
as
shown
above,
it
specifies
the
immediate
addressing
mode,
indicating
to MACRO-1l1l that the operand
itself immediately follows the instruction word.

ADDRESSING MODES

The operation of this mode can be shown

through

MOV #100,R0, which assembles as two words:

Location 20:

2 7 0 0

Location 22:

0 0

Location 24:

Next instruction

the

first

example,

0 O

Note that the source operand (the value 100) is assembled
immediately
following the
instruction word,
i.e., as the second word in the
instruction.
Upon execution of the instruction, the processor fetches
the first word (MOV) and increments the PC by 2 so that it points to
location 22 (which contains the source operand).

After the next fetch and increment cycle, the source operand (100)
moved

into

instruction).

5.10

ABSOLUTE MODE

leaving the PC pointing to location
24 (the

Absolute mode is the
equivalent of
immediate mode deferred.
The
address expression
@#E specifies an absolute address which is stored
as the second or third word of the instruction.
In other
words,
the
value
immediately
following
the
instruction word
is taken as the
absolute address of the operand.
Absolute mode
1is
assembled
as
an
autoincrement deferred of the PC.

Format

for A:

Q#E

Examples:

MOV

@#100,R0

CLR

Q#X

;MOVE THE CONTENTS OF ABSOLUTE

; LOCATION
100 INTO REGISTER RO.

;CLEAR
THE CONTENTS OF THE LOCATION
;WHOSE ADDRESS

;THE SYMBOL X.

IS SPECIFIED BY

The operation of this mode can be shown
through
MOV@#100,R0, which assembles as two words:
Location 20:

3 7 0 0

Location 22:

0 0

0 1

Location 24:

Next instruction

the

first

example,

0 O

Note that the absolute address 100 is assembled immediately
following
the
instruction word,
i.e.,
as the second word in the instruction.
Upon execution of the instruction, the
processor
fetches the
first
word (MOV) and increments the PC by 2 so that it points to location 22
(which contains the absolute address of the
source
operand).
After
the
next
fetch and increment cycle, the contents of absolute address
100 (the source operand) are moved into register
pointing to location 24 (the next instruction).

5.11

leaving

the

RELATIVE MODE

Relative mode is the normal mode for
memory
references within
your
program.
It
is
assembled
as index mode, using the PC as the index
register.

5-5

ADDRESSING MODES
Format

for

Examples:

CLR

100

;CLEAR ABSOLUTE

MOV

RO, Y

sMOVE

; TO

THE

LOCATION

CONTENTS

LOCATION

100

In relative mode, the offset for the address calculation is
assembled
as the
second or third word of the instruction. This value is added
to the contents of the PC (the base register) to yield the address
of
the source operand.

The
MOV

operation of
100,R3, which

Note

that

relative

mode
can
be
as two words:

assembles

Location

20:

Location

22:

Location

24:

the

constant

shown

and

increments

the

statement

instruction
is

assembled

immediately

instruction
word,
1i.e., as the second word
execution of the instruction, the processor
(MOV)

with

the

that

following

in the instruction.
fetches
the
first
it

points

the

Upon
word

location

(containing the value 54).
After the next fetch and increment
cycle,
the
processor
calculates the effective address of the source operand
by taking the contents of location 22 (the offset) and
adding
it
to
the current value of the PC, which now points to location 24 (the next
instruction).
Thus, the source operand address is the result
of
the
calculation
OFFSET+PC = 54+24 = 100(8),
causing
the
contents
of
location 100 to be moved into register 3.

Since MACRO-11l considers the contents of the current location
counter
(.) as the address of the first word of the instruction, an equivalent
index mode statement is shown below:
MOV

100-.-4
(PC) ,R3

This instruction has a relative addressing mode
because
address
1is
calculated
relative to the current value of
counter.
The offset is the distance (in bytes)
between
and the current value of the location counter.

the
the
the

operand
location
operand

5.12 RELATIVE DEFERRED MODE
The relative

deferred mode

similar

operation

the

relative

mode
above,
except that the expression E is used as a pointer to the
address of the operand.
1In other words,
the
operand
following
the
instruction word is added to the contents of the PC to yield a pointer
to the address of the operand.
Format

for

Example:

MOV

@X,R0

;sRELATIVE
sTHE

PC,

s LOCATION

;TO

TO THE
MOVE

CURRENT VALUE

THE

WHOSE

LOCATION

CONTENTS

ADDRESS

INTO

OF
THE

POINTED

ADDRESSING MODES

5.13

SUMMARY OF ADDRESSING FORMS

Each PDP-11 instruction takes at least one word.
listed below do not

increase

the

length of

Form

Operands of the form

instruction.

Meaning

@R or (ER)

(ER) +

Autoincrement mode

@(ER) +

Autoincrement deferred mode

- (ER)

Autodecrement

mode

@- (ER)

Autodecrement

deferred

"
)

Operands

for

occurrence

each

the

following

Form

The syntax
Additional

forms

add

operand

one word

that

Index mode

QE (ER)

Index

Immediate mode

Q#E

Absolute

Relat{;e mode

Relative

the
addressing
discussion
of

applicable PDP-11

the

instruction

deferred

mode

(see Note

deferred

modes
1is
addressing

below)

mode

summarized in
Appendix
modes
1is
provided
in

Processor Handbook.

NOTE

alternate

form

@(ER)

equivalent

form

1is

for

only

the

(ER).

logically,

but

expression

However,

the

not physically

@0 (ER) .

The

addressing
form
Q#E
differs from 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 instruction CLR
@#100 clears absolute location 100,
even
if
the
instruction
1is
moved
from the point at which it
was assembled.
See the description of the
.ENABL
AMA

length
-

form:

Meaning

E (ER)

mode

function
1in
Section
relative
mode
addresses
absolute mode addresses.

6.2,
which causes
to
be
assembled

all

B.
the

ADDRESSING MODES

5.14

BRANCH

INSTRUCTION ADDRESSING

The

branch instructions are l-word instructions.
The high-order
byte
contains the operator, and the low—-order byte contains an 8-bit signed
offset (seven bits, plus sign), which
specifies
the
branch
address
relative
to the current value of the PC.
The hardware calculates the
branch address as follows:
1.

Extends

Multiplies

than
3.

the

sign

the

a word

the

result

offset

through

creating

bits

8-15.

byte

offset

rather

offset.

Adds

the result to the current
effective branch address.

value

the

form

the

MACRO-11 performs the reverse operation to form the word
offset
from
the
specified address.
Remember that when the offset is added to the
current value of the PC, the PC is pointing
to the word following
the
branch
instruction;
hence,
the
factor
-2
in
the
following
calculation:

Word
Since

the

value

Word
In

using

offset
of

the

offset

branch

following error

(E-PC)/2
PC

.+2,

truncated to eight
we

have:

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

instructions,

you

bits.

must

exercise

bits.

care

avoid

the

conditions:

Branching

from one

Branching to a
1location
(global) symbol;
or

Specifying a branch address that is out of range,
i.e.,
the
branch
offset
1is a value that does not lie within the range
-128(10) to +127(10).

The above conditions
assembly listing for

5.15

TUWSING TRAP

program

section

that

1is

cause an error code (A)
the statement in error.

another;

defined

to be

generated

external

the

INSTRUCTIONS

The EMT

and TRAP instructions do not use the
low-order
byte
of
the
instruction
word,
allowing information to be transferred to the trap
handlers in the low-order byte.
If the EMT
or
TRAP
instruction
is
followed
by
an
expression, the value of the expression is stored in
the low-order byte of the word.
However, if the expression is greater
than
377(8),
it 1is truncated to eight bits and an error code (T) is
generated in the assembly listing.

5-8

PART

MACRO-11

III

DIRECTIVES

Chapters 6 and 7 describe
all
the
directives
used
with
MACRO-11.
Directives
are
statements
that
cause
MACRO-1l1l
to perform certain
operations during assembly.
Chapter
6
describes
several
types
of
directives,
including
those
which
control
symbol
interpretation,
listing header material, program sections, data storage
formats,
and
assembly
listings.
Chapter
7 describes those directives concerning
macros,

MACRO-11

macro

arguments,

directives

can

and

repetitive

preceded

coding

sequences.

label

restrictions
associated
with
specific directives)
comment.
A MACRO-1ll directive occupies the operator

statement.

Only

one

directive

can

included

(subject

any

and followed by a
field of a source
any

given

source

line.
The operand field may be occupied by one or
more
operands
left blank;
1legal operands differ with each directive specified.

CHAPTER 6
GENERAL ASSEMBLER DIRECTIVES

This category of directives
1.

Listing

Function control

Data

Radix and

Location counter

Terminators

Program boundaries

Program sectioning

Symbol

includes:

control

storage
numeric

control

10.

Conditional assembly

11.

PAL-11R conditional

assembly.

Each is described in its own section of this chapter.

6.1

LISTING CONTROL DIRECTIVES

Listing control directives control the content, format, and pagination
of
all
1line
printer and teleprinter listing output generated during
assembly.
Facilities also exist for creating object module names
and
other identification information in the listing output.

6.1.1

.LIST and

Listing control
program through
of the form:

.NLIST Directives
options can be specified in the
the .LIST and .NLIST directives.

.LIST

.LIST
.NLIST
.NLIST

arg
arg

text
of
a
MACRO-11
These directives are

GENERAL ASSEMBLER DIRECTIVES

where:

arg

represents one or more of
arguments defined in Table

the
6-1.

optional

symbolic

As indicated above, the listing control directives may be used without
arguments,
1in
which
case
the
1listing directives alter the listing
level count.
The listing level count is initialized to zero.
At each
occurrence
of
a
.LIST
directive,
the
1listing
level
count
1is
incremented;
at each occurrence of an .NLIST directive,
the
1listing
level count is decremented.
When the listing level count is negative,
the listing
1is
suppressed
(unless
the
1line
<contains
an
error).
Conversely,
when
the
1listing
1level count is greater than zero, the
listing is always generated.
Finally, when the
count
1is
zero,
the
line is either listed or suppressed, contingent upon the other listing
controls currently in
effect
for
the
program.
For
example,
the
following
macro definition employs the .LIST and .NLIST directives to

selectively list

portions

the

macro

body

when

the

macro

expanded:

7
7

A-THIS
B-THIS

.MACRO
LTEST
LINE SHOULD LIST

;LIST TEST
; LISTING LEVEL

COUNT

NLIST

; LISTING LEVEL COUNT

-1.

; LISTING LEVEL COUNT

-2.

; LISTING LEVEL COUNT

IS -1.

; LISTING LEVEL COUNT
; LISTING LEVEL COUNT

IS O.
IS BACK TO

LINE

SHOULD

NOT

LIST

SHOULD NOT

LIST

SHOULD

LIST

NLIST
;

C-THIS

;

D-THIS

;

E-THIS

LINE

+LIST
LINE

NOT

.LIST
LINE SHOULD LIST

. ENDM

.LIST

s LIST

LTEST

;
;

A-THIS
E-THIS

LINE
LINE

MACRO

; CALL. THE

SHOULD
SHOULD

LIST
LIST

EXPANSION.

MACRO

; LISTING LEVEL COUNT
; LISTING LEVEL COUNT

IS
IS

0.
BACK TO

An important purpose of the level count is to allow
macro
to
be
listed
selectively
and yet exit with the listing
restored to the value existing prior to the macro call.

expansions
level count
|

When

used with arguments, the listing
directives
do
not
alter
the
listing
1level count;
however, the .LIST and .NLIST directives can be
‘used to override current listing control,
as
shown
1in
the
example
below:
.MACRO

.LIST

:LIST

LINE.

.NLIST
.

:DO NOT LIST REMAINDER OF MACRO
s EXPANSION.

X=.

. ENDM

.NLIST

:DO NOT LIST MACRO EXPANSIONS.

XX
xm'

The symbolic arguments allowed for use with the listing directives are
described
1in
Table
6-1.
These
arguments can be used singly or in
combination with each other.
If multiple arguments are specified in a
listing directive, each argument must be separated by a comma, tab, or
6-2

GENERAL ASSEMBLER DIRECTIVES

space.
control

For any
argument
not
specifically
1included
in
statement, the associated default assumption (List

a
1listing
or No list)

is applicable throughout the source program.
The default
assumptions
the listing control directives also appear in Table 6-1.

for

Symbolic Arguments
Argument

SEQ*

Table 6-1
of Listing Control

Default

List

Directives

Function

Controls

the

sequence

1listing

numbers.

source

MACRO-11

1line

assigns

sequence number 1 to
the
first
source
line
in
a
file,
and
increments
the
sequence

number

the

for

file.

suppressed

each

additional

this

through

line

field

.NLIST

1is
SEQ

directive,
MACRO-1l1]
generates
a
tab,
effectively
allocating
space
for
the
field,
but fills the field with blanks.
Thus,

the inter-positional relationships
of subsequent
fields
in
the
1listing
remain undisturbed.
During the assembly
process,
MACRO-1ll
examines each source
line for possible error conditions.
For
any
line in error, an appropriate error
flag
is
printed
preceding
the
line
sequence
number field (see Appendix D).
MACRO-11
does
not
assign
sequence
numbers for files that have had sequence
numbers assigned by other programs, such
as an editor.

LOC*

List

the
1listing
of
the
current
counter field.
Normally, this
not suppressed.
However, if it
is
suppressed
through
the
.NLIST LOC
directive, MACRO-1l1l does not generate
a
tab,
nor does it allocate space for the
field, as is the case
with
the
source
line
sequence
number
field
(SEQ)
described above.
Thus, the
suppression
of
the
current
1location counter (LOC)
field
effectively
left-justifies
all
subsequent
fields
(while
preserving
inter-positional relationships) to
that
position
otherwise normally occupied by
this field.

BIN*

List

Controls

Controls
location
field is

the

listing of

generated

binary

code.
If
this
field
1is
suppressed
through
an
.NLIST
BIN
directive,
left-justification
of
the
source code
field
occurs
in
the
same
= manner
described above for the current location
counter (LOC) field.

BEX

List

Controls

the

listing

binary

extensions,
i.e.,
the
locations
and
binary contents beyond those
that
will
fit
on the source statement line.
This
is a subset of the BIN argument.
|
SRC*

List

Controls

the

listing of

source

lines.

(Continued on next page)
6-3

GENERAL ASSEMBLER DIRECTIVES

Table 6-1 (Cont.)
Symbolic Arguments of Listing Control

Argument

COM

Default

List

Directives

Function

Controls the listing of comments.
This
is
a
subset
of the SRC argument.
The
.NLIST
COM directive
reduces
1listing

time

and

space

when

comments are

not

desired.
MD

List

Controls
the
definitions and

List

Controls the
repeat range

list

Controls
the
expansions.

MEB

list

Controls the listing of macro
expansion
binary
code.
A
.LIST
MEB
directive
causes
only
those
macro
expansion
statements
that generate binary code to
be listed.
This is a subset of
the
ME

listing
of
macro
repeat range expansions.

listing of macro
expansions.
listing

calls

and

macro

argument.

CND

List

Controls
the
1listing
of
unsatisfied
conditional
c¢oding
and
associated .IF
and
.ENDC
directives
1in
the
source
program.
This
argument
permits
conditional
assemblies
to
be
listed
without
including
unsatisfied
conditional coding.

Controls
the
1listing
of
all
1listing
directives
having
no
arguments, i.e.,
those listing directives that alter
the
listing level count.

TOC

List

Controls the listing
of
the
table
of
contents
during
assembly
pass
1 (see
Section
6.1.4
describing
the
- SBTTL
directive).
This
argument
does
not
affect the printing of the full assembly
listing during assembly pass 2.

SYM

List

Controls the listing of the symbol table
resulting
from
the
assembly
of
the

list

source

TTM

List

program.

Controls the listing output format.
The
default
can
be
set
by
the
system
manager.
If the system manager does not
set a default, it is set to line printer

format.
Figure 6-1 illustrates the 1line
printer
output
format.
Figure
6-2
illustrates
the
teleprinter
output
format.
*
If the .NLIST arguments SEQ, LOC, BIN, and SRC are in effect at the
same
time, i.e., if all four significant fields in the listing are to
be suppressed, the printing of the resulting blank line is inhibited.

6-4

GENERAL ASSEMBLER DIRECTIVES

An example of an assembly
1listing,
as sent
to a
132-column
line
printer,
is
shown
in Figure
6-1.
Note that binary extensions for
statements generating more than one word are formatted horizontally on
the source line.

An example of an assembly listing,

as sent to a

teleprinter

(in

the

Notice that binary extensions for statements generating more than

one

same format as for an 80U-column line printer), is shown in Figure 6-2.

word are printed on subsequent lines.

There is no explicit truncation

of output to 80 characters by the assembler.

Any argument specified in a .LIST/.NLIST directive

listed

code

(A)

other

than

those

Table 6-1 causes the directive to be flagged with an error

in the assembly listing.

The listing control options can also be specified at
assembly
time
through
switches
included in the command string to MACRO-11 (see the
appropriate system manual).
The use of these switches
overrides
all
corresponding 1listing control (.LIST or .NLIST) directives specified
in

the

source program.

v98c9/vgvc1ieee229v99pO02(s//G92ct1t1itl0P0eo8eeLI9BpPpVgE9oRTtTR020Vd 1i31g518Vd01§§1A323IOdv88vWA)]l 8¥)2(yNLOd3IIENS2VNT)8AI32=37R+?0(I@W82)"29+QI914W°3odL2)FIILNBNVSIiNSHCmIV’NNAS(OOIN2LI8LOH4)NTSO00dLH8!I¥HN33OIS43NIIH2NO1LN84S30DSH#J10I88LL3NIL3ON/V7W1¥3II7QWSvQi3v3LSN¥IgVNdI¥1SNI
g6(p2@Z2.i(L1¢e0c0i9cLto2e0€det@couvRe12002 3g48Nlg1e 1(I3@69H4)wdL]lVIS*I'HOW"SOL‘wS39I¢AST¥Y3INHLOILNIVNQYOADVOINVWL[WINOMADNILIAN£AI8NTJ3dN8VIS

“61t2(v¢@6B9gp0c1¢p£¢f22€v2Zm22e2p@92z9eP02pvA52€9EaOp/G9ilLeo9S96scvrcpvYp1Il1ti8l020Re0ovegIu®d@0QQ®e8de2Z©©L6(0Q20@9oVG99/9Cr2L/9L@Tl5102G9F2e@p1¢90f01T0Q029292Q0l011v0020o0@802RL8¢890Al00Y081c2Y0Ee2Poe09Qp2Lg0o0Tee2ant11I§53Nb20gTt1HLvId9Tl-91g1a2WV98wTiWgJ3$aNiE3LdI18SniQIAgvvldNO]isgdv1Wl)gIASW8AIL5oL)l92ydu38l(Bei2L(NT)$4O(rY(SNNEdP2rT@1R0@(eDd'I44‘EHLaNPv8IxX))RvN3n)])V¥M(x0PdLIA0Od9QbS2)eL0ET*33V0vEiST2+8LWL®‘+wQ.D2J)I’SoTQIj"70L®.oA%7*1Q*I2wLN24‘WdM7°4'aNI2E"(2H°ONuRa702D(r*WN(3°2P°*@(dHS)AH0WIoT)yLI‘Od“234¥AOLNnt10)8IIMFN3SW394I2¥3L9(LI8HINA°*4I3¢SPS7VvNEd2QH°tH3T4U‘¥ISN¢1%!J)Y5TIIQDISL04XN3iNvNv1ID4LW3IVSJaNMdLT“@LI2II8gNALEOrO3dvvOV/NTNSA#H4LYNIOInLQS0NAIIDAt(AWwNTAOT3NV0TSJ9INHqSq3!4V7tON43M¥WWS3Iw0IHNLQ2WedNSEL2IT9WgOl9vS1IV3svnI1NNd8SA94MN393B0vIT9NLS3i4SsVETO3ILNH0ty809NSATJOI4V38NdI0LSm3AILO$pN4S4WN8ON0JuIL/0VVXWT7S39Wt7LOIVId84N6a7aSNLESMYvNlvyOSIu)INTsJldvSOW3N34Tud8IS¥Tl¥TdyL4Vvu01ddv¥dVIdNINLS 0f3>MwM<mB=2OoHx)z

L22

o012

LQv5iiIlySs2<6g98£2110Q22NY=9p29g€rcvo3e2EcSLttliWS1e0V3e]@ewdLA40Ng9POVeTRWdp1Ac9Ee6ORt21DTSOPINYONIBTR2YIeS3OMIVW(B/OW PL=7Sg31TN20Nd$§L[8gA1=l8IA62XS83*(W(Q%$Pv21dID3GSI8TN)1(OB8ON3MHV0)V128d3+76180ST9Y'Wd9SI$*%iO12Nd*H#°VIW3(Ox2W8dNs1SV3tO)9H¢D+S!%41L7SI8O4Fx1NON7L3I32vTSIO0HNH90I2¥TLH+YvNO0¥M3L¥GI3LWSIL¥Q2AI3O4N13ILTl1¥34v0dL40SHI4NI¥LuSvd
e

GENERAL ASSEMBLER DIRECTIVES
T
CSITS»=
READ AND

TESY
PARSE

OF CSI1
COMMAND

AND CS12
LINES

211
212
213
214
215
216

201230
201244
201246
po1254

201300

2m1324

217 201326
218 291332
-

219 291336
220 221360
221

2014P4

222 201412

MB707

B9=JUL=74

15159

PAGE

JSBYTL

READ AND PARSE COMMAND LINES

GCMLS
BCC
EXITSS
TYPE
CS1sy
BCC

#GCLBLK
13

166016

suB

C.CMLD*2(RB), (SP)

Q2004

209

2102

MACRO

1023003

GETLNS

181
123064

MOV

e00R20

TYPE
TYPE

Q66060

ADD

‘200@16
200020

JGET LINE VIA GCML
JSKIP IF NO ERROR
JELSE,

EXIT

G,CMLD+2(R2),G,CMLD(R®),»'0
JSEND OUT THE INPUT LINE
#CSIBLK,GCLBLK+G,CMLD*2,GCLBLK*G,CMLD
23
JBRANCH IF NO ERROR DETECTED

C.,FILDe2(RR),~(SP)
|

JIPUT STRING ERROR ADDR IN STK

JCALCULATE LENGTH OF FIRST PARTY
,

C,FILD(RO),C,FILDe2(RB)

JCALC ADDR OF (AST PART OF STRING

162660

C,CMLDe2(RB),(SP),#'S
JSEND OUT FIRST PART OF STRING
C.FILD#2(RQ),C,FILD(RB),n'S
JSEND QUT SECOND PART
;

suB

(SPY+,C,CMLD(RA)

JDEDUCT LENGTH OF FIRST PART

sus

C,FILD(RO),C .CMLD(RD)

JCALC LENGTH OF LAST PARTY

TYPE

C.,FILD#2(RD)Y,C,CMLD(RB), %40
STX,42
GETLN

JSEND SYNTAX ERROR MESSAGE
JTRY FOR MORE

18T

C.CMLD(RD)

JCHECK LENGTH OF LINE

BEG

GETLN

JIF NULL, SKIP BACK FOR NEXT LINE

220002

223 201416

224 701424

225 001450
226 PR1474
227

228 AM1476

229 201502

238 201504

231

Q01512

166060

eaQeR16
eoee0?2

TYPEM
BR

©AR6ESS

@AS5768

281

200002

Q81652
112767

MOVB

e2R0N60

176432

BITR

132760

#'@,EQUBIT

JSEND OUY LAST PART

JASSUME EQUAL SIGN NOT FOUND

#CS,EQU,C,STAT(RA)Y

JCHECK STATUS

200n4R

232 201520

233 v21%22

220001

@142
176416

234 271526

1081

235 231852
236 2A1576

237 22a16@2
238 201620

239 02021622

240 201626
241
242
243
244
245
246

221634
221636
001662
201666
Q21704
201706

247 BA1712

BEQ

185267

OPARSE:

108

INCB

EQUBIT

TYPEM

EOQU, 40

TYPEM
Call

CsIs2

OPT,42
INIT2

+OUTPUT,#SWTBL

JSKIP IF EQUAL SIGN NOT SEEN

JELSE,

INDICATE EQUAL SIGN FOUND

JSEND EGUAL SIGN STATUS MESSAGE
JSEND OQUTPUT SCAN MESSAGE
JINIT LOCNS FOR CSI2 CALL/TEST

JPARSE OUTPUT SPEC

ISKIP ON ERROR

1P344y

BCS

CS2ERR

1327680

BITB

BCS,MOR,C,STAT(R@)

BNE
TYPEM
CALL
CsIs2
BCS
CALL

OPARSE
1IPT,d¢
INITZ
» INPUT, #SWTBL
CS2ERR
EVALUB

JYES, CONTINUE WITH QUTPUT SCAN
JSEND INPUT SCAN MESSAGE
JINIT LOCNS FOR CSI2 CALL/TEST

BNE

IPARSE

PYES, CONTINUE wWITH INPUT SCAN

CaLL

eReV20
PoRRB!

Q201360
IPARSE:

103407

1327680

8178

EvaLUB

JEVALUATE RESULTS OF SEMANTIC PARSE

jADDITIONAL OUTPUT SPECS?

IPARSE INPUT SPEC
)ISKIP ON ERROR

JEVALUATE RESULTS OF SEMANTIC PARSE
#CS,MOR,C,8TAT(R®)
JADDITIONAL INPUT SPECS?

2ee20

248 001720

200001

201360

Figure 6-2

Example of Terminal Assembly Listing

GENERAL ASSEMBLER DIRECTIVES

Figure
©6-3
reflecting
program and

shows
a
1listing,
produced
in
1line
printer
format,
the
use
of the .LIST and .NLIST directives
in the source
the effects such directives have on the
assembly
1listing

output.

6.1.2

Page Headings

MACRO-11l prints
Figure
©6-1
or

each assembly page
in
the
Figure
6-2,
depending on

Table 6-1).
On the first line
following (from left to right):

each

N
W

Date.

Time-of-day.
Page

Assembler

Title of the object module, as
directive (see next section).

.
.

version

format
shown
in
either
the listing mode (see TTM,

page,

MACRO-1ll

established

through

prints

the

.TITLE

identification.

number.

The second line
text
specified
6.1.4).

of each
in the

assembly listing page
contains
the
subtitle
last-encountered .SBTTL directive (see Section

w02

w03

GENERAL ASSEMBLER DIRECTIVES

X398
Qyom* X3g

X38

03¢

Wil

013s
03s

quom’®

JVWiET

x38°'W02 X38'wW02

6-9

LKL

piioee

togae vénoe
gaie 0 9ate0e 21 0 0

£€

14Y

‘NIvVW®

100000
conoee

vioeee 1eeeaeo

(2 8e

roe e
v¢(o uo vid0 0
ot

1eeee0

JYWiST

l1evoe
ecie e

@ WD

O N

0D O

6-10

lege e
9ce or

vs0000

t£odeee "

foveee

NIg

aceave vo0oee100000

9cQ000

14" &4

vcoeae

Zlo e

{

1e29e
veeeee9 500200

bUT3lsSIpeonpoxdUYITMBUTISTTOIUODS®AT309ITd
wan3*

wil

SIWLE ST Vv INIW OD

ISvNImWMTLOt]

ca'ypiogime’*liMs1Y9* J2vw0iE7"

Li1sS1I®N® X38¢

SAvNIWILOHD

Javyyowmi*ST X¢3'8pig2’l

39vd I

*NIVW® vL=N[=60 62161 QHIVW’ JVYWLIET 94y
GENERAL ASSEMBLER DIRECTIVES

{

20vevY

§co0 0

2anbTJg

GENERAL ASSEMBLER DIRECTIVES

6.1.3

.TITLE Dixective

The .TITLE directive is used to assign a name to the object module
as
the
first
entry
in the header of each page in the assembly listing.
The name so assigned is the first six non-blank
characters
following
the
.TITLE directive.
This name should be six Radix-50 characters or
less in length;
any characters beyond the first six are
checked
for
ASCII
1legality,
but
they
are not used as part of the object module
name.
For example, the directive:
.TITLE

PROGRAM TO PERFORM DAILY ACCOUNTING

causes the assembled object module to be named PROGRA.
Note that this
6-character
name
bears no relationship to the filename of the object
module, as specified in the command string to MACRO-11l.
The
name
of
an
object
module
(specified in the .TITLE directive) appears in the
load map produced at link time.
This is also the
module
name
which
the Librarian will recognize.
|
If the .TITLE directive is not specified, MACRO-1l1l assigns the default
name
.MAIN.
to the object module.
If more than one .TITLE directive

is
specified
in
the
encountered establishes

source
program,
the
last
.TITLE
directive
the name for the entire object module.

All spaces and/or tabs up to
the
following the .TITLE directive are
the text string.
If

the

.TITLE directive

first
non-space/non-tab
character
1gn0r&d by MACRO-11 when evaluating

specified

without

object

module

name,

if
the first non-space/non-tab character in the object module name is
not a Radix-50 character, the directive 15 flagged with an error
code
(A)

the

assembly

listing.

Section A.2 of Appendix A contains
a table
of Radix-50

6.1.4
The

characters.

.SBTTL Directive
.SBTTL

directive

1is

wused

produce
a

table

contents

immediately
preceding
the
assembly 1listing and to further identify
each page in the 1istingm
In the latter case, the text following
the
.SBTTL
directive
1is printed as the second line of the header of each
page in the listing, continuing until altered by a
subsequent
.SBTTL
directive in the program.
For example, the directive:
.SBTTL
causes

the

CONDITIONAL ASSEMBLIES

text

CONDITIONAL ASSEMBLIES

to be printed as the
listing.

second

1line

the

header

the

assambly

During assembly pass 1,
a
table
of
contents
is
printed
for
the
assembly
1listing,
containing
the
1line
sequence
number,
the page
number, and the text accompanying each .SBTTL directive.
The
1listing
of
the
table
of
contents
1is
suppressed
whenever
an
.NLIST TOC
directive is encountered in the source program (see
Table
6-1).
An
example of

table

contents

listing

6-11

shown

in Figure

6-4.

GENERAL ASSEMBLER DIRECTIVES

CSITST
TABLE

TEST

2= 55
Je 74
4=153
5-209
6-255
7=345%5

CSI1

AND CS12

MACRO

M@787

0@9=JUL=74

15147

MACRO DEFINITIONS
MESSAGE STRINGS
MISCELLANEOUS ODATA
READ AND PARSE COMMAND LINES
EVALUATE THE SEMANTIC ANALYSIS
SUBROUTINES

Figure

6.1.5

CONTENTS

6-4

Assembly Listing Table

of Contents

LJIDENT Directive

The

.IDENT directive provides an
additional
means
of
labeling
the
object
module produced by MACRO-11l.
1In addition to the name assigned
to the object module with the .TITLE directive (see Section 6.1.3),
a
character
string
up
to
six
Radix-50
characters
can be specified
between paired printing delimiters to label the object module with the
program version number.
This directive takes the following form:

.IDENT
where:
”

/string/

string

represents six or fewer legal Radix-50
characters
which
establish
the
program
identification
or
version number.
This number is
included
1in
the
global symbol directory of the object module;
the
first four characters are printed in the load
map
and librarian listing.

represent delimiting characters.
These delimiters
may
be any paired printing characters, other than
the equal sign (=), the left angle bracket (<), or
the
semicolon
(;),
as
1long
as
the delimiting
character is not contained within the text
string
itself.
If
the
delimiting
characters
do
not
match, or if an illegal
delimiting
character
1is
used,
the
.IDENT
directive
1is
flagged with an
error code (A) in the assembly listing.

An example of
.IDENT

the

.IDENT directive

shown below:

/VO05A/

The character string VO5A is converted to Radix-50 representation
and
included
in
the
global symbol directory of the object module.
This
character string also appears in the load map produced
at
1link
time
and the Librarian directory listings.
When more than one .IDENT directive is encountered in
the
1last
such directive encountered establishes the
which forms part of the object module identification.

6-12

a given program,
character string

GENERAL ASSEMBLER DIRECTIVES

.PAGE Directive/Page Ejection

6.1.6

Page ejection is accomplished in one of four ways:

MACRO-11
After reaching a count of 58 lines in the 1listing,
page
skip over
eject to
a page
automatically performs
formulate
to
and
perforations on 1line printer paper
is not
The page number
into pages.
teleprinter output
changed.

source
the
In addition, the .PAGE directive is used within
in the
points
desired
at
eject
page
a
perform
to
program
The format of this directive is:
listing.
. PAGE

This directive takes no arguments and causes a skip to the
It also causes the
the next page when encountered.
top of
sequence counter
page number to be incremented and the line

listing.

<cleared.

The .PAGE directive does not appear in the

When used within a macro definition, the .PAGE

directive

ignored during the assembly of the macro definition.

Rather,

is
the page eject operation is performed as the macro itself
In this case, the page number is also incremented.
expanded.

is
A page eject is performed when a form-feed character
If the form-feed character appears within a
encountered.

macro definition, a page eject occurs during the assembly of
but not during the expansion of the
the macro definition,
the
wuse of
A page eject resulting from the
macro itself.
the -page number to be
causes
likewise
form-feed character
incremented and the line sequence counter to be cleared.

6.2

Encountering a new source file causes the page number
incremented and the line sequence count to be reset.

FUNCTION DIRECTIVES:

.ENABL AND .DSABL

Several function control options are provided by MACRO-1l through the
These directives are included in a
.DSABL directives.
.ENABL and
functions and
source program to invoke or inhibit certain MACRO-11
These
itself.
process
assembly
the
to
incidental
operations
directives take the following form:
.ENABL
.DSABL

arg

where:

arg
arg

represents one or more of

the

arguments defined in Table 6-2.

optional

symbolic

Specifying any argument in an .ENABL/.DSABL directive other than those
e be flagged with an error
listed 1in Table 6-2 causes that directivto
code

(A)

in the assambly listing.

6-13

GENERAL ASSEMBLER DIRECTIVES

Symbolic Arguments
Argument

Default

ABS

Disable

AMA

Table 6-2
of Function Control

Directives

Function

Enabling this function produces absolute
binary
output
in
FILES-11 format.
To

Disable

convert
format

this
(as

Loader),

use

Enabling

output to Formatted
Binary
required
by
the
Absolute
the

FLX

this

utility.

function

causes

all

relative
addresses (address mode 67) to
be
assembled
as
absolute
addresses
(address
mode
37).
This
function is
useful during
the
debugging
phase
of
program development.

CDR

Disable

Enabling

this
function
causes
source
columns 73 and greater, i.e., to the end
of the line, to be treated as a comment.
The

most

common

use

to
permit
sequence
columns 73-80.
CRF

Enable

Disabling

this

generation
of
This
function

Disable

1in

card

inhibits

the

command

generation

Enabling

this

point
function

truncation;
disabling
this
causes floating-point rounding.

Enabling

this

the

output

specified

LSB

function

feature

<cross-reference
output.
only
has
meaning
if

cross-reference

FPT

this

numbers

function

causes

floating-

causes

instead

of converting it to upper-case.
function
1is
not
enabled,
all
converted to upper-case.

If this
text is

argument

ASCII

MACRO-11

input

This

lower-case

string.

permits

the

enabling

disabling
of
a
local
symbol
block.
Although
a
1local
symbol
block
is
normally
established
by encountering a
new symbolic label or a .PSECT directive
in
the
source
program,
an .ENABL LSB
directive establishes a new local symbol
block
which is not terminated until (1)
another .ENABL LSB
1s
encountered,
or
(2)
another
symbolic
1label
or .PSECT
directive
1is
encountered
following
a
paired .DSABL LSB directive.
Although

the

. ENABL

LSB

directive

permits
a
local
symbol block to cross
.PSECT boundaries, local symbols
cannot
be
defined
in
a program section other
than the one that was in effect when the
block
was
entered.
The basic function
of
this
directive
with
regard
to
.PSECT's
1is
1limited to those instances
(Continued on next page)

GENERAL ASSEMBLER DIRECTIVES

Table
Symbolic - Arguments

Argument

LSB
(Cont.)

6-2

(Cont.)

Function Control

Default

Directives

Function

where it is desirable to leave a program
section
temporarily
to
store
data,
followed
by
a
return
to the original
program
section.
Attempts
to
define
local
symbols
in
an alternate program
section are flagged with an
error
code

Disable

(P)

the

assembly

listing.

example of the .ENABL LSB and
.DSABL
LSB
directives,
as typically used in a
source program, is shown in Figure 6-5.

PNC

Enable

Disabling

output

this

until

function

encountered within
REG

inhibits

binary

an .ENABL PNC statement
the

same

module.

When

specified, the .DSABL REG directive
inhibits
the
normal
MACRO-11
default
register definitions;
if not
disabled,
the
default
definitions
1listed
below
remain in effect.

Enable

RO=%0
R1=%1
R2=%2
R3=%3
R4=%4
R5=%5
SP=%6
PC=%7
The

.ENABL

REG

statement

may

used

the logical complement of the .DSABL REG
directive.
The use of these directives,
however,
is
not
recommended.
For
logical
consistency,
use
the
normal
default
register
definitions
1listed
above.
T
GBL

When

Enable*

the

.ENABL w GBL

directive

specified,
MACRO-1ll
treats
all symbol
references that are undefined at the end
of
assembly
pass
1
as default global
references;
when
the
.DSABL
GBL
directive
1is specified, MACRO-1ll treats
all

The

default

is Disable

such
references
as
undefined
symbols.
In
assembly
pass
2, if the
.DSABL GBL function is still in
effect,
these undefined symbols are flagged with
an
error
code
(U)
in
the
assembly
listing;
otherwise, they continue to be
regarded
by
MACRO-11
as
global
references.

for RT-11

MACRO programs.

GENERAL ASSEMBLER DIRECTIVES

2d’ty

1124

p3e

AOW

03d CELE

BdWd INT

L] 338

39vd

1502
1§0¢

dnW

4PIWSON

038

§1

T8YN3*

I8dS dW3l 2u’ly
1¥'2H

)
£1719

Y-AT TH'EN

N3¥NL 18vsg’
180v

cacge iv2oe

062 262 £62

962

6-16

get cet

tec

842

162

2(2 vi2 L2

cvitce

182

reigee

1egoco
19ceae

civoe Zotese Zivioe

viZcee iévoee

tecage
Zocelo
lecece

GENERAL ASSEMBLER DIRECTIVES

6.3

DATA STORAGE DIRECTIVES

A wide range of
data
and
data
types
can
be
generated with
the
following directives,
ASCII conversion characters, and radix-control
operators:

.BYTE
.WORD
L]

.ASCII

These MACRO-11l facilities are described in the following sections.

6.3.1

.BYTE Directive

The .BYTE directive is used to generate
successive
bytes
data in the object module.
The directive is of the form:

binary

.BYTE

exp

; STORES THE BINARY VALUE OF THE
3;EXPRESSION "EXP" IN THE NEXT BYTE.

.BYTE

expl,exp2,expn

;STORES THE BINARY VALUES OF THE LIST
;OF

EXPRESSIONS

IN SUCCESSIVE BYTES.

A legal expression must reduce to eight bits of
data
or
less.
The
operands of a .BYTE directive are evaluated as word expressions before
being truncated to the low-order eight bits.
The l6-bit value of
the
specified
expression must have a high-order byte (which is truncated)

that is either all zeros (0) or all ones (l).
Each
expression
value
is
stored
in
the
next
byte
of
the
object
module.
Multiple
expressions,
which
must
be
separated by
commas,
are
stored
1in
successive bytes,

SAM=5
.=410

.BYTE

as described below:

"D48,SAM

;sTHE VALUE 060

(OCTAL EQUIVALENT OF 48

;DECIMAL) IS STORED IN LOCATION 410.
s THE VALUE 005 IS STORED IN LOCATION
+411.,

If the highmorder byte of the expression reduces to a value other than
0
or
=1,
the
value
1is
flagged with an error code

truncated
to the low-order eight bits and
(T) in the assembly listing.

The construction "D in the first operand of the .BYTE directive
above
illustrates
the
use
of
a
temporary
radix-control
operator.
The
function of such special
wunary
operators
is
described
1in
Section
6.4.1.2.

At link time, it is likely that a relocatable expression
in

wvalue

having

than

eight bits,

6-17

will

in which case the

result
linker

GENERAL ASSEMBLER DIRECTIVES

issues

truncation diagnostic for the object module in question.
following statements create such a possibility:

example,

For

the

.BYTE

:STORES OCTAL

IN NEXT BYTE.

.BYTE

: RELOCATABLE

VALUE

WILL

PROBABRLY

: CAUSE TRUNCATION

:DIAGNOSTIC.
If

expression

interpreted

following

zero,

the

.BYTE

as described

directive

1is

null,

1is

below:

.=420
.BYTE

Note

that

.BYTE

the

; ZEROS

above

directive.

example,

The

three

ARE

STORED

7422,

AND

423,

four

bytes

commas

the

implicit declaration of four null
values,
other
by a comma.
Hence, four bytes, each
(0), are reserved in the object module.

6.3.2

The

the

BYTES

storage

operand

420,

result

421,

from

the

field represent

an
each
separated from
the
containing a value of zero

.WORD Directive

.WORD directive is used to generate
object module.
The directive is of
.WORD

.WORD

exXp

; STORES

expl,exp2,expn

successive

data

form:

THE

BINARY EQUIVALENT

; EXPRESSION

EXP

;STORES

BINARY

;LIST

words

the

THE

WORD.

EQUIVALENTS

EXPRESSIONS

THE

SUCCESSIVE

: WORDS.

legal expression must result in 16
bits
of
data
or
less.
Each
expression is stored in the next word of the object program. Multiple
expressions must be separated
by
commas
and
stored
in
successive

words,

shown

the

following

example:

SAL=0
.=500

.WORD

177535, .+4,SAL

;STORES THE VALUES 177535, 506, AND
;0 IN WORDS 500, 502, AND 504,
s RESPECTIVELY.

expression

interpreted

following

the

zero,

.WORD

shown

directive

the

contains

following

null

value,

example:

.=500
.WORD

; STORES THE VALUES 0, 5, AND
; LOCATION 500, 502, AND 504,

s RESPECTIVELY.

statement

containing

blank

operator

not recognized by MACRO-1l1l as a macro
MACRO-11 directive, or a semicolon is
an implicit .WORD directive, as shown
.=440
LABEL:

100,LABEL

; STORES
;AND

field,

i.e.,

symbol

that

call, an instruction mnemonic, a
interpreted during
assembly
as
in the example below:

THE

VALUE

THE VALUE

440

100
IN

LOCATION

440

442.

GENERAL ASSEMBLER DIRECTIVES

CAUTION

You

should

not

use

this

technique

generate .WORD directives because it may
not
be
included
in
future
PDP~-11
assemblers.

6.3.3

ASCII

Conversion Characters

The single quote (') and the double quote
(")
characters
are
unary
operators
that
can appear in any MACRO-1l1l expression.
When so used,
these characters cause a 16-bit expression value to be generated.

When the single quoteis used, MACRO-1ll takes the
next
character
in
the
expression and converts it from its 7-bit ASCII value to a 1l6-bit
expression value.
The l6-bit value is then used as an
absolute
term
within the expression.
For example, the statement:
MOV

#'A,RO

results in the fallnwing l6-bit

0O:

expression

value

being

moved

into

00000000/01000001

ImmBinary Value of ASCII A
Thus, in the example above, the expression 'A results in
a
value
101(8).
Note
that
the
high-order
byte 1is always zero (0) in
resulting expression value when the single
quote
unary
operator
used.

The ' character must not
RUBOUT,
1line-feed,
or
(A) is generated
in the

fallmwed

carriage-return,

form—-feed character;
assembly listing.

if it

is,

error

of
the

null,

code

When the double quote is used, MACRO-1ll takes the next two
characters
in
the
expression
and converts
them to a 1l6-bit binary expression
value from their 7-bit ASCII values.
This 1l6-bit value is
then
used
as
an
absolute
term
within
the
expression.
For
example,
the
statement:

MOV

results in the
register O:

#"AB,RO

following

16-bit

expression

value

being

moved

into

0100001001000001
]

1ww&inary Value of ASCII A

——Binary Value

of ASCII

Thus, in the example above, the expression "AB results in a
041101 (8).

6-19

value

GENERAL ASSEMBLER DIRECTIVES

The " character also must not be followed by a carriage-return,
null,
RUBOUT,
line-feed,
or
form—-feed character;
1if it 1is, an error code
(A) is likewise generated in the assembly listing.

The ASCII

6.3.4

character

.ASCII

set

listed

in Section A.l,

Appendix A.

Directive

The .ASCII directive translates character
strings
into
their
7-bit
ASCII equivalents and stores them in the object module.
The format of
the .ASCII directive is as follows:

.ASCII

where:

/string

string

1/.../string

is a
string of
printable
ASCII characters.
All
printable
ASCII
characters
are
legal.
The
vertical-tab,
null,
line-feed,
RUBOUT,
and all
other
non-printable
ASCII
characters,
except
carriage-return
and
form-feed,
are
illegal
characters.
Such
an
illegal
non-printing
character is flagged with an error code (I) in the
assembly
listing.
The
carriage-return
and
form-feed
characters
terminate
the
scan of the
source line.
This premature
termination
of
the
.ASCII
statement
results in the generation of an
error code (A) in the
assembly
1listing,
because
MACRO-11
1is
wunable
matching delimiter at
string.

the

complete

end

the

scan

of the
character

represent delimiting characters.
These delimiters
may
be any paired printing characters, other than
the equal sign (=), the left angle bracket (), or
the
semicolon
(;),
as
1long
as
the delimiting
character is not contained within the text
string
itself.
If
the
delimiting
characters
do
not
match, or if an illegal
delimiting
character
is
used,
the
.ASCII
directive
1is
flagged with an
error code (A) in the assembly listing.

A non-printing character can be expressed in an .ASCII statement
by
enclosing
its equivalent octal value within angle brackets.
set of angle brackets so used
represents
a
single
character.
example, in the following statement:

L,ASCII

only
Each
For

<15>/ABC/<A+2>/DEF/<5><4>

the expressions <15>, <A+2>, <5>, and
<4>
represent
the
values
of
non-printing
characters.
Furthermore, the expressions must reduce to
eight bits of absolute data or less, subject to
the
same
rules
for
generating data as with the .BYTE directive (see Section 6.3.1).
Angle

brackets can be embedded between delimiting
characters
in
the
character
string,
but
angle
brackets
so used do not take on their
usual significance as delimiters
for
non-printing
characters.
For
example, the statement:

.ASCII

/ABC<Kexpression>DEF/

contains a single ASCII character string, and performs
no
evaluation
of the embedded, bracketed expression.
This use of the angle brackets
is shown in the third example of the .ASCII directive below:

6-20

GENERAL ASSEMBLER DIRECTIVES

; STORES THE BINARY REPRESENTATION

.ASCII

/HELLO/

.ASCII

/ABC/<15><12>/DEF/ ;STORES THE BINARY REPRESENTATION

;OF THE LETTERS HELLO IN FIVE
; CONSECUTIVE BYTES.

;OF THE CHARACTERS A,B,C,CARRIAGE
;RETURN,LINE FEED,D,E,F IN EIGHT
; CONSECUTIVE BYTES.

.ASCII

; STORES THE BINARY REPRESENTATION

/A<K15>B/

;OF THE CHARACTERS A, <, 1, 5, >,
;AND B IN SIX CONSECUTIVE BYTES.

The semicolon (;) and equal sign (=) can be used as delimiting
characters in an ASCII string, but care must be exercised in so doing
because of their significance as a comment indicator and assignment
operator, respectively, as illustrated in the examples below:
.ASCII

;sSTORES THE BINARY REPRESENTATION OF

;ABC;/DEF/

A, B, C, D, E, AND F
CHARACTERS
;THE
|
NOT
;IN SIX CONSECUTIVE BYTES;
; RECOMMENDED PRACTICE.

; STORES THE BINARY REPRESENTATIONS OF

.ASCII /ABC/;DEF;

.ASCII

; THE CHARACTERS A, B, AND C IN THREE
; CONSECUTIVE BYTES; THE CHARACTERS D,
;E, F, AND ; ARE TREATED AS A COMMENT.

/ABC/=DEF=

‘

;STORES THE BINARY REPRESENTATION

A, B, C, D, E, AND
;OF THE CHARACTERS
sF IN SIX CONSECUTIVE BYTES; NOT
; RECOMMENDED PRACTICE.

An equal sign is treated as an assignment operator when it appears

the

first

character

following example:

.ASCII

6.3.5

the

=DEF=

ASCII

string,

illustrated by the

; THE DIRECT ASSIGNMENT OPERATION

; .ASCII=DEF IS PERFORMED, AND A Q
; (SYNTAX) ERROR IS GENERATED UPON
; ENCOUNTERING THE SECOND = SIGN.

.ASCIZ Directive

The .ASCIZ directive is equivalent to the .ASCII directive described
above, except that a zero byte is automatically inserted as the final
character of the string. Thus, when a list or text string has been
created with an .ASCIZ directive, a search for the null character in
the last byte can effectively determine the end of the string, as
reflected by the coding below:

CR=15
LF=12

HELLO:

10S$:

.ASCIZ

<CR><LF>/MACRO-11 VO1lA/<CR><LF>

MOV
MOV
MOVB
BNE

#HELLO,R1
# LINBUF ,R2
(R1)+, (R2) +
109

.EVEN

;INTRODUCTORY MESSAGE

; GET ADDRESS OF MESSAGE.
; GET ADDRESS OF OUTPUT BUFFER.
;MOVE A BYTE TO OUTPUT BUFFER.
;IF NOT NULL, MOVE ANOTHER BYTE.

6-21

GENERAL ASSEMBLER DIRECTIVES

The

.ASCIZ directive is subject
legality
and
proper character

for

the

6.3.6
The

.ASCII

directive.

to
the
same
checks
for
character
string construction as described above

.RADS50 Directive

.RAD50

directive

allows

the

user

generate

data

Radix-50

packed
format.
Radix=-50
form
allows three characters to be packed
into sixteen bits (one word);
therefore, any 6-character
symbol
can
be stored in two consecutive words.
The form of the directive is:
.RAD50

where:

/string

string

1/.../string

represents a series of
characters
to
be
packed
(three
characters
per
word).
The
string must
consist of the characters A through Z,
0
through
9,
dollar sign ($), period (.) and space ( ).
An
illegal printing character causes
an
error
flag
(Q) to be printed in the assembly listing.
I1f

fewer

than

three

characters

are

packed,

the
string
1is
packed
1left-justified within
word, and trailing spaces are assumed.
As with
©6.3.4,

the
the

.ASCII directive described in
vertical-tab, null, line-feed,

Section
RUBOUT,

and
all
other
non-printing
characters,
carriage-return
and
form-feed,
are
characters, resulting in an error code (I)

except
illegal
in
the

assembly

the

carriage-return

and form-feed characters result in
(A)
because
these characters end
line,
preventing
MACRO-11]
from
terminating matching delimiter.

listing.

Similarly,

an
error
code
the scan of the
detecting
the

represent delimiting
may

the
the

any

paired

equal sign
semicolon

characters.

printing

These delimiters

characters,

other than
(=), the left angle bracket (<), or
(;),
provided that the delimiting

character is not contained within the text
string
itself.
If
the
delimiting
characters
do
not
match, or if an illegal
delimiting
character
is
used,
the
.RAD50
directive
1is
flagged with an
error code (A) in the assembly listing.
Examples

.RADS50

directives

.RAD50
.RAD50
.RAD50

/ABC/
/AB/
/ABCD/

.RAD50

/ABCDEF/

are

shown

below:

s PACKS ABC INTO ONE WORD.,
s PACKS AB (SPACE) INTO ONE WORD.
; PACKS ABC INTO FIRST WORD AND
;D

(SPACE)

; PACKS ABC
i SECOND

(SPACE)

INTO

WORD.

INTO

SECOND

FIRST WORD,

DEF

WORD.,

INTO

GENERAL ASSEMBLER DIRECTIVES

Each

character

translated

into

indicated in the following table:

its

Radix-50

Character

Radix-50 Octal Equivalent

(space)
A-Z
$
.
(undefined)
0-9

0
1-32
33
34
35
36-47

The Radix-50 equivalents for characters 1

through

combined as follows:

Radix~50 Value =
For

Section

example below:

.RAD50

are

A.2

((1*50)+2)*50+3 = 3223

Appendix
~

equivalents.

Angle brackets
special
codes

(Cl1,C2,C3)

Radix-50 Value of ABC =

((C1l*50)+C2)*50+C3-

example:

Refer

equivalent,

for a
,

table

Radix-50

(<>) must be used
in
the
.RAD50
directive
whenever
are to be inserted in the text string, as - shown in the
PRy

/AB/<35>

; STORES 3255 IN ONE WORD.

CHR1=1
CHR2=2
CHR3=3
»

.RAD50

6.3.7

;EQUIVALENT TO .RADS0 /ABC/.

Temporary Radix-50 Control Operator:

The "R operator specifies that an
argument
1is
to
be
converted
to
Radix-50 format.
This allows up to three characters to be stored in

one word.

The "R operator

is coded as follows:

"Rccce

where ccc represents a maximum of three characters to be converted
to
a 16-bit Radix-50 value.
If more than three characters are specified,

any following the third character are ignored.
specified,

If fewer

following example shows how the "R operator might be used
3-character file type specifier
MOV # "RMAC,FILEXT

The number

than

pack

it is assumed that the trailing characters are blanks.

sign

(#)

(MAC)

into a single 1l6-bit word.

; STORE RAD50 MAC AS FILE EXTENSION
used to

indicate

immediate data,

i.e.,

be
assembled
directly
into
object
code.
"R
specifies
characters
MAC are to be converted to Radix-50.
This
wvalue
stored in

are

The

location FILEXT.

6-23

data

that the
1is
then

GENERAL ASSEMBLER DIRECTIVES

6.4

RADIX AND NUMERIC CONTROL FACILITIES

6.4.1

Radix Control

and Unary Control Operators

The

normal default assumption for numeric values or expression
values
appearing
1in
a
MACRO-1ll source program is octal.
However, numerous
instances may occur where an alternate radix is useful for portions of
a
program
or
for
variables
within
a
given statement.
It may be
useful, for example,
to
declare
a
given
radix
for
applicability
throughout a program or to specify a numeric value or expression value
in a manner that causes it to be interpreted as a
binary,
octal,
or
decimal
value
during
assembly.
In other such instances, it may be
useful to complement
numeric
values
or
expression
values.
These
MACRO-11 facilities are described in the following sections.

NOTE

When

two

unary

modifying
the
same
applied, from right to

operators

appear

term,
the
left, to the

together,

operators
term.

are

6.4.1.1
.RADIX Directive - Numbers used in a MACRO-11l source
program
are initially considered to be octal values;
however, you can declare
any one of the following
radices
for
applicability
throughout
the
source program or within specific portions of the program:
2,

This

accomplished
.RADIX

where:

via

.RADIX directive

the

form:

represents one of
listed above.
If
the octal default

the
three
acceptable
radices
the argument n is not specified,
radix is assumed.

The argument in the
.RADIX
directive
1is
always
interpreted
as
a
decimal
value.
Any
alternate
radix declared in the source program
through the .RADIX directive remains in effect until
altered
by
the
occurrence
of another such directive, i.e., a given radix declaration
is valid
throughout
a
program
until
changed.
For
example,
the
statement:

.RADIX

;BEGINS A SECTION OF

.RADIX
Any

value

other

than

.RADIX
directive
assembly listing.

null,

<causes

;DECIMAL

RADIX.

s REVERTS

TO OCTAL

error

specified

code

(A)

CODE

HAVING A

RADIX.
as

argument

the

be generated

the

In general, macro definitions should not
contain
or
rely
on
radix
settings
established
with
the
.RADIX directive.
Rather, temporary
radix control operators should be
used
within
a
macro
definition.
Where a possible radix conflict exists within a macro definition or in
possible future uses of that code, it is
recommended
that
the
user

6-24

GENERAL ASSEMBLER DIRECTIVES

specify numeric or expression values using the temporary radix control

operators described below.

"D, "0, and "B - Once the
6.4.1.2 Temporary Radix Control Operators:
user has specified a given radix for a section of code or has decided
to use the default octal radix, he may discover a number of cases
where an alternate radix is more convenient or desirable (particularly
The creation of a mask word, for example,
within macro definitions).
might best be accomplished through the use of a binary radix.
MACRO-11 has three unary operators that allow the user to establish an
alternate radix,

"D"number"
O"number"
"B"number"

Thus,

an alternate

localized

as shown below:
("number"
("number"
("number"

radix

requirement

can

is evaluated as a decimal number)
is evaluated as an octal number)
is evaluated as a binary number)

declared

in the source program.

temporarily

meet

Such a declaration can

be made at any time, regardless
of the existence of the default
octal
radix or another specific radix declaration elsewhere in the program.
In other words, the effect of a temporary radix control
operator
is
limited
to the term or expression immediately following the operator.
Any value specified in connection with
a
temporary
radix
control
operator
is evaluated during assembly as a l6-bit entity.
Temporary
radix control declarations can
be
included
1in
the
source
program
anywhere a numeric value is legal.

The expressions below are representative of the methods of

specifying

temporary radix control operators:
"D123
"0 47
"B 00001101
“0<A+13>

Decimal radix
Octal Radix
Binary Radix
Octal Radix

Note that the up-arrow and the
radix
control
operator
may
not
be
separated,
but the
radix control operator and the following term or

expression
can be physically
separated
by
spaces or
tabs
for
legibility or formatting purposes.
A multi-element term or expression
that is to be interpreted in an alternate
radix
should
be
enclosed

within
angle
brackets,
as
shown
radix control expressions above.
The following example also

in the last of the four

illustrates the use of

angle

temporary

brackets

delimit an expression that is to be interpreted
in an alternate radix:

A=10

.RADIX

.WORD

"O<A+10>*10

When the temporary radix expression in the .WORD
evaluated,

directive

above

it effectively yields the following equivalent statement:

.WORD

180.

MACRO-11 also allows a temporary radix change to decimal by specifying
a number, immediately followed by a decimal point (.), as shown below:
100..

1376.
128.

Equivalent to 144(8)

Equivalent to 2540(8)
Equivalent to 200(8)
6-25

GENERAL ASSEMBLER DIRECTIVES

The

above

expression

forms

are

below:

equivalent

function

those

listed

“D100
"D1376
"D128

6.4.2

Numeric Directives and Unary Control Operators

Two

storage directives and two numeric control operators are available
to
simplify
the
use
of
the floating-point hardware on the PDP-11.
These facilities allow
floating-point
data
to
be
<created
in
the
program,
and
numeric
values
to
be
complemented
or
treated
as
floating-point numbers.

A floating-point number is represented by a string of decimal
digits.
The
string
(which
can
be
a single digit in length) may optionally
contain a decimal point, and may be followed by an
optional
exponent
indicator
in
the
form
of the letter E and a signed decimal integer
exponent.
The number may not contain embedded blanks, tabs
or
angle
brackets
and
may not be an expression.
Such a string will result in
one or more errors (A or Q) in the assembly listing.
The

list of numeric representations
below
contains
seven
valid representations of the same floating-point number:

distinct,

3
3.
3.0
3.0E0
3E0

. 3E1
300E-2
As can be
3000E-3,

inferred, the
.03E2, etc.).

list could
A leading

be
plus

extended
indefinitely
sign is optional (e.g.,

(e.g.,
3.0 is

considered to be +3.0).
A leading minus
sign
complements
the
bit.
No other operators are allowed (e.g., 3.0+N is illegal).
All floating-point numbers
format:

64
S

are

EEEEEEEE

evaluated

bits

the

MMM.....MMM

Exponent

(55 bits)
(8 bits)

Sign

MACRO-11l returns a value of the appropriate
of
the
floating-point
directives.
The
truncated or rounded (see Section 6.2).

bit)

size and precision via one
values
returned
may
be

Floating=-point

numbers
are
normally
rounded.
That
floating=-point
number
exceeds the limits of the field
to be stored, the high-order bit of the unretained word
1low-order

the number

are

needed

bit

is
to

following

Mantissa

the

sign

to be

the

retained

stored

express

1its

in a
exact

word,

2-word
value,

6-26

shown

field,
the

1is,
when
in which it
1is
added

below.

For

but more

than

highest

bit

a
is
to

example,

(32)

32
of

bits
the

GENERAL ASSEMBLER DIRECTIVES

unretained field is added to the least
significant
bit
(0)
of
the
retained
field (see illustration below).
The .ENABL FPT directive is
used to enable floating-point
truncation;
.DSABL
FPT
is
wused
to
return

floating-point

rounding

Bit
32

Bit
0

(see Table

Bit
32 31

Retained

6-2).

Bit
0

Untetained

field

Note
that
all
numeric
operands associated with
Floating
Point
Processor
instructions
are
automatically
evaluated as single-word,
decimal,
floating-point
values
wunless
a
temporary
radix
control
operator
1is
specified.
For
example,
to
add (floating) the octal
constant 41040 to the
contents
of
floating
accumulator
=zero,
the
following instruction must be used:

ADDF
where:

#°041040,F0
FU

is assumed

Floatingwpoint

represent

numbers are

applicable PDP-11

mporary

floating accumulator

daacribéd, in greater

Processor Handbook.

Numeric Control Operators:

zero.

detail

"C and

"F -The

unary
operator ;allmwa you
to
&p@cmfy an
argument
that is
complemented
as it
is
evaluated
during assembly.
The
°F

operator allows
you
flamtinquomnt number‘
As with
2

control

the

rad1X'

operator

contrml

( C)

specify

an arqument consisting of
|
,

operators

d@sarlb@d

abmv&,

the

°C

to be
unary

a l-word
|
numeric

can be used anywhere in the source program that

an expression value is legal. Such a construction
is
evaluated
by
MACRO-11
as
a
1l6-bit binary
value before being complemented.
For
example, the following statement:

TAG4:

.WORD

"C151

causes the 1l's cmmpl@ment of the value 151 (octal) to be stored
as
a
l16-bit
wvalue
in the program.
The resulting value expressed in octal
form is 177626(80

GENERAL ASSEMBLER DIRECTIVES
Because the "C construction is a unary operator, the operator and
its
argument
are
regarded as a term.
Thus, more than one unary operator
may
be
applied
to
a
single
term.
For
example,
the
following
construction:

“C°D25

causes the decimal value 25 to be complemented during
assembly.
The
resulting
binary
value,
when
expressed
1in
octal form, reduces to
177746 (octal).

The term created through
operator
thus
becomes
combination
with
other
following construction:

the use
of
the
temporary
numeric
control
an
entity
that
can
be
used
alone
or in
expression
elements.
For
example,
the

“C2+6
is equivalent

in function

to:

<"C2>+6

This expression is evaluated during assembly as the 1l's complement
of
2,
plus
the absolute value of 6.
When these terms are combined, the
resulting
expression
value
generates
a
carry
beyond
the
most
significant bit, leaving 000003(8) as the reduced value.

As shown above, when the temporary numeric control
operator
and
its
argument
are
coded
as
a
term within an expression, angle brackets
should
be
used
as
delimiters
to
ensure
precise
evaluation
and
readability.
|

MACRO-11 also supports a unary
operator
for
numeric
control
which
allows
you
to
specify
an
argument
consisting
of
a
1l-word
floating-point number.
For example, the following statement:
MOV

$#"F3.7,R0

creates a l-word floating-point number at location A+2 containing
value

3.7

formatted as shown below.

BIT 15

Sign

EEEEEEEE

(bit 15)

Exponent

the

MMMMMMM

(bits 14-7)

Mantissa

(bits 6-0)

GENERAL ASSEMBLER DIRECTIVES

6.5

LOCATION COUNTER CONTROL DIRECTIVES

The directives used in controlling the value of the current location
in the object program are
space
storage
in reserving
and
counter
described in the following sections.

MACRO-11
be noted that several
it should
this connection,
In
statements may cause an odd number of bytes to be allocated, as listed
below:

.BYTE directive

.BLKB directive

.ASCII or

.0ODD directive

A direct assignment statement of
the
form
.=.+expression,
which results in the assignment of an odd address value.

.ASCIZ directive

In cases that yield an odd address
value,
the
next word-boundaried
instruction
automatically
forces
the
1location
counter
to an even
in the
value, but that instruction is flagged with an error code (B)

assembly listing.

6.5.1

.EVEN Directive

The .EVEN directive ensures that the current location counter contains
If the current
an even value by addlng 1 if the current value is odd.
Any aperands
location counter is already even, no action is taken.
following an .EVEN directive are flagged with an error code (Q) in the
assembly

The

listing.

.EVEN directive

6.5.2

is used as

follows:

.ASCIZ
.EVEN

/THIS IS A TEST/

.WORD

XYZ

|
;ENSURES THAT THE NEXT STATEMENT WILL
;BEGIN ON A WORD BOUNDARY.

.0ODD Directive

The .0DD directive ensures that the current location counter
contains
If the current
an odd value by adding 1 if the current value is even.
location counter is already odd, no action
1is
taken.
Any
operands
following an .ODD directive are also flagged with an error code (Q) in
the assembly listing.

GENERAL ASSEMBLER DIRECTIVES

6.5.3

.BLKB and

.BLKW Directives

Blocks of storage can be reserved in
the
.BLKB
and
.BLKW
directives.
reserve
blocks.

byte
The

blocks;
similarly, the
.BLKW
two directives are of the form:

. BLKB

exp

. BLKW

exp

where:

the object program
by
The
.BLKB
directive

exp

directive
|

in either of

is not an absolute value,
with an error code (A) in
illustrates

the

caeooo
geeeno

PASStI1

e2e000

use

the

.BLKB

PSECY

IMPUR
,E
D

,BLKW

«PSECT

IMPPAS,D,GBL
2

the
the

and

these directives

.BLKW directives.

JPASS FLAG
INEXT GROUP

MUST

STAY

SYMBOL
¢ ,BLKW

popen4d
22eee4

MODE
s 3

,BLKB

{

)

poavees
poeRes

SECTORss ,BLKB
VALUEES BLKW

i
i
i

JSYMBOL/EXPRESSION
FJEXPRESSION VALUE
JRELOCATION LEVEL

RELLVLSS ,BLKW

evean3

177

PSYMBOL

178

JREPT

HAXXMTW!Q.*SYMBOL)/?)

o BLKW

{

TOGETHER

ACCUMULATOR

JMODE/FLAGS

FLAGSi1t

220010

176

word

statement is flagged
assembly listing.

gepoas

BYTE

TYPE

« ENDR

179

CLCNAMg: BLKW
CLCFGS11,BLKB
CLCSEC11,BLKB
CLCLOCs
1 ,BLKW
g ,BLKW
CLCMAX
T
S8 ,BLKW
CHRPN
SYMBEGI: ,BLKW
ENDFLGES ,BLKW

180 020020

PA0R24
182 200025
183 220a26
184 AGBo3a
188 @30032
186 28va34
187- BBRB36
{88 0200008
181

6-6

directive in
statement:

Example

LOCATION

COUNTER

NAME

)

)
|
JEND OF GROUPED DATaA
JCHARACTER POINTER

JPOINTER

STARY

SYMBOL

)

o PSECT

Figure

The .BLKB
following

JCURRENT

.flp“-"-«“““&}

166
167
168
169
17@
171
{172
173
174
175

6-6

reserves

of
to

represents the specified number of bytes or
words
to
be
reserved
in
the
object
program.
If no
argument is present,
a
default
wvalue
of
1
is
assumed.
These
directives
should
not
be used
without
arguments.
Any
expression
that
is
completely
defined
at
assembly-time
and
that
reduces to an absolute value
1is
legal.
If
the

expression specified

Figure

means
is used

.BLKB

and

source program has

.BLKW Directives

the

same

effect

the

.=.texpression
the
current
which causes the value of the expression to be added
to
The
.BLKB directive, however, is
location
value
the
counter.
of
source
code
1in
which
it
easier to interpret in the context of the
appears and is therefore recommended.

GENERAL ASSEMBLER DIRECTIVES

6.6

TERMINATING DIRECTIVES

6.6.1

.END Directive

The .END directive indicates the logical end of the source input,
takes

the

following

.END

where:

and

form:

exp

represents an optional expression value which,
if
present,
1indicates the program-entry point, i.e.,
the transfer address at which program execution is
to begin.

When MACRO-11 encounters a valid occurrence of the .END directive,
it
terminates the current assembly pass.
Any additional text beyond this
point in the current source file, as
well
as
in
additional
source
files identified in the command line, will be ignored.
When creating an image consisting of several object modules, only
one
object
module may be terminated with an .END exp statement specifying
the starting address.
All other object
modules
must
be
terminated
with
an
.END
statement
without
an address argument;
otherwise,
a
diagnostic message will be
1issued
at
1link
time.
If
no
starting
address
1is
specified
in
any of the object modules, image execution
will begin at location 1 of the image and immediately fault because of
an odd addressing error.
|
|
|
The

.END

statement must

not

used within

macro

expansion

conditional
assembly
block;
if it is so used, it is flagged with an
error code (0) in the assembly listing.
The
.END
statement
may
be
used,
however,
in
an
1immediate
conditional statement (see Section
6.10.2)
.

If the source program input 1is not terminated with an .END
an error

6.6.2

code

(E)

results

the

assembly

listing.

directive,

LJ.EOT Directive

Under RSX-1ll, RT-11l, and IAS
operating
systems,
the
MACRO-11
.EOT
directive 1is ignored and simply treated as a directive without effect,
i.e., as a no-op.

6.7

PROGRAM BOUNDARIES DIRECTIVE:

.LIMIT

It is often desirable to know the upper and
of
the
image.
When the .LIMIT directive
program, MACRO-11 effectively generates the

. BLKW

lower
address
boundaries
is specified in the source
following instruction:

causing two storage words to be reserved in the object module.
Later,
at
1link
time,
the lowest address in the load image is inserted into
the first reserved word, and
the
address
of
the
first
free
word
following the image is inserted into the second reserved word.

During linking,
2-word

boundary.

the size of the

image

is rounded upward to the nearest

GENERAL ASSEMBLER DIRECTIVES

For a discussion of

memory

applicable system manual

6.8

allocation

(see Section 0.3

and

mapping,

refer

in the Preface).

the

PROGRAM SECTIONING DIRECTIVES

The MACRO-11 program sectioning directives are used to

for

program

sections

and

establish

certain

attributes essential to the linking processing.

6.8.1

declare

program

names

section

.PSECT Directive

The
.PSECT
directive
allows
absolute
control
over
the
memory
allocation
of
a program at link time, because any program attributes
established through this directive are passed to the linker.
For example, if you are writing programs for a multi-user environment,
a program
section
containing
pure
code
(instructions
only) or a
program section containing impure code (data and instructions) may
be

explicitly
declared through the .PSECT directive.
Furthermore,
program
sections
may
be
explicitly
declared
as
read-only
qualifying them for

use as protected,

The advantages gained

through

reentrant programs.

sectioning

programs

this

these
code,

manner

therefore
relate
primarily
to control of memory allocation, program
modularity, and more effective partitioning of memory.
Refer
to
the
applicable
system manual
for a discussion of memory allocation (see
Section 0.3
The

in the Preface).

.PSECT directive

.PSECT

where:

formatted as

follows:

name,argl,arg2,...argn

name

’
arqgl,
arg2,...
argn

represents
the
symbolic
name
of
section, as described in Table 6-3.

represents any legal separator

.
space)

the

(comma,

tab

program

and/or

represent one
or
more
of the
1legal
symbolic
. PSECT
the
use with
for
defined
arguments
directive, as described in Table 6-3.
The
slash
separating each pair of symbolic arguments listed
optional
these
that
indicates
table
the
in
arguments are mutually exclusive, i.e., one or the
other, but not both, may be
specified.
Multiple
arguments
must be separated by a legal separating
character.
Any symbolic argument specified in the
.PSECT
directive other than those listed in Table
6-3 will cause that statement to be
flagged
with
an error

code

(A)

in the assembly listing.

GENERAL ASSEMBLER DIRECTIVES
Table

6-3

Symbolic Arguments of .PSECT Directive

Argument

Default

NAME

Blank

Meaning

Establishes the

program

specified

which

section

one

name,

six

argument
this
Radix~-50 characters. If
is omitted, a comma must appear in place

Radix-50
The
name parameter.
of the
is listed in Section A.2
set
character
of Appendix A.

RO/RW

access
of
type
which
Defines
permitted to the program section:
RO=Read-Only Access
RW=Read/Write Access

NOTE

IAS
and
RSX-11lD
for
protection

sections.

not provide
Defines

1/D

hardware
program

RSX-11lM and RT-11

the

such protection.

program

section

(I) or
instructions
either
containing
the
allow
attributes
These
data (D).
differentiate global symbols
to
linker
entry-point
program
are
that
that are
those
from
(I)
instructions
data values

GBL/LCL

set
RO

LCL

Defines

section,

the

link time.

(D).

scope

the

program

as subsequently interpreted at

In building single-segment
nonoverlaid
programs,
the GBL/LCL arguments have no
memory
total
the
because
meaning,
for the program will go into
allocation
the root
segment
of
the
image.
The
GBL/LCL arguments apply only in the case
of

overlays.

If an object

program

module

section,

contains

then

the

local

storage

occur
will
module
allocation for that
segment in which the module
the
within
reference
can
modules
Many
resides.
the
and
section,
program
same
this
is
module
each
for
memory allocation
or overlaid within
concatenated
either
the

segment,

depending on

of the program section

the

(.PSECT)

argument

defining

(continued on next page)

6-33

GENERAL ASSEMBLER DIRECTIVES

Table 6-3 (Cont.)
Symbolic Arguments of .PSECT Directive
Argument

Default

GBL/LCL

LCL

(cont'd)

Meaning

its

allocation

below).

requirements

If an object

global

(see CON/OVR

module

contains

|
program
contributions to
this

section,
the
program
section
segment boundaries,

are collected across
and the allocation of
memory
for
that
section will go into the segment nearest
the root in which the first contribution
to
this program section appeared.
(The
term contribution implies an
allocation
of memory to the program section.)

ABS/REL

REL

Defines the

relocatability attribute

the

section:

program

ABS=Absolute

(non-relocatable).
When
the
ABS
argument is specified, the
program section is regarded at
1link

time
as
an
absolute
module,
requiring
no
relocation.
program
loaded,

address
The

thus
The
section
1is
assembled
and
starting at absolute virtual
0.

location

program

data

sections

must

absolute

fall

the virtual
memory
1limits
segment
containing
the
section;
otherwise,
an

within

of
the
program
error

results
at link time.
For example,
the following code,
although
wvalid
at
during assembly, may generate an
error message
if
wvirtual
1location
100000
is
outside
the
segment's
virtual address space:
.PSECT

ALPHA,ABS

.=.+100000
.WORD

The above coding assembles properly,
but
the
resulting load address may
be outside the respective
segment's

boundaries.
In
such
cases,
the
linker recognizes this as an attempt
to
load
data outside the image and
responds with an error message.

REL=Relocatable.
argument
is

When

the

REL

specified,
the linker
calculates
a
relocation
bias
and
adds
it
to
all
references
to
locations
within
the
program
section, i.e., all references to the

program

section

must

relocation
bias
added
make them absolute.

have

them

(Continued on next page)
6-34

GENERAL ASSEMBLER DIRECTIVES

Table

(Cont.)

Symbollc Arguments of

.PSECT Directive

Argument

Default

Meaning

CON/OVR

CON

Defines the allocation
the program section:

requirements

CON=Concatenated.
All
program
section
contributions are to be concatenated
with other references to
this
same
program
section
in
order
to
determine

allocation

the

total

requirement

program section.

for

memory

this

OVR=0Overlaid.
All
program
section
contributions
are
to
be overlaid.
Thus,
the
total
allocation
requirement
for the program section
is equal to the
1largest
allocation
request
made
by
any
individual
contribution
to
this
program
section.

The only argument in the .PSECT directive that
1is
position-dependent
is NAME.
If
it is omitted, a comma must be used in its place.
For
example,

the directive:

.PSECT

,GBL

shows a .PSECT directive with
a
blank
name
argument
and
the
GBL
argument.
Default
values
(see Table 6-3) are assumed for all other
unspecified arguments.

Once the attributes of a program section are declared through a

.PSECT

directive, MACRO-1l1l assumes that these attributes remain in effect for
all subsequent .PSECT directives of the same name that are encountered

within

the module.

MACRO-11 provides for

The

256(10)

program sections,

One default

One default unnamed relocatable program section

Two-hundred-fifty-four

.PSECT directive

absolute program section

as listed below:

enables

ABS.)

named program sections.

the

Create program sections

Share code and

data

user

to:

(see Section 6.8.1.1)

among

6-35

program

sections

(see

Section

GENERAL ASSEMBLER DIRECTIVES

For each program section
following information:

specified

Program section

name

Contents

current

Maximum location

4.,

Program

the

implied,

location

counter

section

described

value

attributes,

in Table

6-3

MACRO-11l maintains

counter

encountered
i.e.,

the

.PSECT

2zero
of
a
source

The

a given

occurrence

arguments

above.

6.8.1.1
Creating
Program
Sections - MACRO-1l1
assembling
source
statements
at
relocatable
program section, i.e., the first statement
of
always an implied .PSECT directive.
first

the

.PSECT directive

with

automatically

begins
the unnamed
program
is

name

assumes

that
the
<current
location
counter is set at relocatable zero.
The
scope of this directive then extends until
a
directive
declaring
a
different
program
section
1is
specified.
Further occurrences of a
program
section
name
in
subsequent
.PSECT
statements
cause
the
resumption
of
assembly
where
that
section
previously ended.
For
example:
.PSECT
«WORD

; DECLARES UNNAMED RELOCATABLE PROGRAM
;SECTION
ASSEMBLED
AT
RELOCATABLE

« WORD

;sADDRESSES

« WORD

AND

.PSECT

ALPHA

+DECLARES

. WORD

: NAMED ALPHA ASSEMBLED

« WORD

s ADDRESSES

. PSECT
. WORD

s RETURNS
: PROGRAM

;BLY AT

RELOCATABLE

AND

PROGRAM

SECTION

RELOCATABLE

TO UNNAMED RELOCATABLE
SECTION AND CONTINUES ASSEM-

RELOCATABLE ADDRESS

A given program section may be defined
completely
wupon
encountering
its first .PSECT directive.
Thereafter, the section can be referenced
by specifying
its
name
only,
or
by
completely
respecifying
its

attributes.
the

directive:
.PSECT

and

later

example,

requires

a program section can be declared

through

ALPHA,ABS,OVR

referenced

.PSECT

which

For

through

the

equivalent directive:

ALPHA

arguments.

By maintaining separate location counters for
each
program
section,
MACRO-11
allows
the user to write statements that are not physically
contiguous within the program, but that
c¢an
be
1loaded
contiguously
following assembly, as shown in the following example.

GENERAL ASSEMBLER DIRECTIVES

.PSECT
.WORD

SEC1,REL,RO
0

.WORD

ST:

+sADDRESSES 0, 2, AND 4.

. WORD

s START A RELOCATABLE PROGRAM SECTION
+:NAMED SEC1 ASSEMBLED AT RELOCATABLE
s ASSEMBLE CODE AT RELOCATABLE
;ADDRESSES 6 THROUGH 12.

A
B

CLR
CLR

CLR

;START AN ABSOLUTE PROGRAM SECTION

SECA,ABS

.PSECT

;ABSOLUTE ADDRESSES 0 AND 2.
:RESUME RELOCATABLE PROGRAM SECTION

.+2,A
SECl1

«WORD
.PSECT

ASSEMBLE CODE AT RELOCATABLE
+SECl.
sADDRESSES 14 AND 16.

A
ST

INC
BR

ASSEMBLE CODE AT

;:NAMED SECA.

1likewise, all
All labels in an absolute program section are absolute;
location
current
The
relocatable.
are
section
relocatable
a
labels in
counter symbol (.) is also relocatable or absolute when referenced in
a relocatable or absolute program section, respectively.

Any labels appearing on a line

containing

(or

.PSECT

.ASECT

directive are assigned the value of the current location
.CSECT)
counter before the .PSECT (or other) directive takes effect. Thus, if
|

the first statement of a program is:
ALT,REL

.PSECT

the label A is assigned to relocatable address zero of the unnamed (or

blank)

program section.

It is not known during assembly where relocatable program sections
will be loaded, therefore all references between relocatable sections
in a single assembly are translated by MACRO-1ll to references relative
Thus, MACRO-11l provides the
to the base of the referenced section.

linker with the necessary information to resolve the linkages between
Such information is not necessary, however,
various program sections.

when referencing an absolute program section, because all instructions
in an absolute program section are associated with an absolute virtual
|

address.

In the following example,

references

relative

references

into

translated

program section named SEN.

.=.+1000

.PSECT

ENT,ABS

CLR

JMP

.PSECT
MOV

SEN,REL
RO,R1

JMP
Y:

.WORD

the

symbols

and

are

to the base of the relocatable

;ASSEMBLED AS CLR BASE OF

; RELOCATABLE SECTION + 10.
;ASSEMBLED AS JMP BASE OF
; RELOCATABLE SECTION + 6.

;ASSEMBLED AS JMP 1000.

HALT

0
NOTE

in conjunction
constant
In the preceding example, using a
location counter symbol (.) in the form
current
the
with
.=1000 would result

error,

because

constants

are

always absolute and are always associated with the program's
.ASECT

section

(. ABS.).

If the form .=1000 were

incompatibility would be detected.

used,

for a discussion of the current location counter.
6-37

program

See Section 3.6

GENERAL ASSEMBLER DIRECTIVES

6.8.1.2
Code or Data
Sharing - Named
relocatable
program
sections
with
the
arguments GBL and OVR operate in the same manner as FORTRAN
COMMON, i.e., program sections of the same name with the arguments GBL
and
OVR from different assemblies are all loaded at the same location
at link time.
All
other
program
sections,
i.e.,
those
with
the
argument CON, are concatenated.

Note that no conflict
exists
between
internal
symbolic
names
and
program section names, i.e., it is legal to use the same symbolic name
for both purposes.
Considering FORTRAN again, using the same symbolic
name 1is necessary to accommodate the following statement:

COMMON

/X/ A,B,C,X

where the symbol X
the fourth element

6.8.1.3

represents

that

Memory Allocation

generate

error

you to place odd
several
modules

when

the

base

the

program section

and

also

does

not

section.

Considerations

a module

length data at
contain object

ends

- The
odd

assembler
location.

the end of a
module.
code contributions
to

This

allows

However,
when
the same program

section having the concatenate attribute (see Table 6-3),
odd
1length
modules
(except
the
1last) may cause succeeding modules to be linked
starting
at
odd
locations,
thereby
making
the
linked
program
unexecutable.
To
avoid
this
problem,
code
and
data
should
be
separated

from each

other

and

placed

separately

named

program

sections.
This permits the linker to automatically begin each program
section on an even address.
Refer to the applicable system manual for
further
information on memory allocation of tasks (see Section 0.3 in
the Preface).

6.8.2

J.ASECT and

.CSECT Directives

IAS and RSX-1ll assembly-language programs use the
.PSECT
and
.ASECT
directives
exclusively,
since
the .PSECT directive provides all the
capabilities
of
the
.CSECT
directive
defined
for
other
PDP-11
assemblers.
MACRO-11
will accept both .ASECT and .CSECT directives,
but assembles them as though they
were
.PSECT
directives
with
the
default
attributes
1listed
in Table 6-4.
Also, compatibility exists
between other MACRO-11l programs
and
the
IAS/RSX-11
Task
Builders,
since
the
respective
Task
Builders recognize the .ASECT and .CSECT
directives that appear
in
such
programs
and
likewise
assign
the
default values listed in Table 6-4.

GENERAL ASSEMBLER DIRECTIVES

Table 6-4
Program Section Default Values

Non-IAS/RSX-11

Default

Attribute

.ASECT

Name

Access

Type

Scope

Value

.CSECT (named)

ABS.

.CSECT (unnamed)

name

Blank

GBL

LCL

Relocation

ABS

REL

Allocation

OVR

CON

The allowable syntactical forms of the

.ASECT

and

.CSECT

directives

are:

.ASECT
.CSECT

.CSECT
Note

that

the

statement:

.CSECT
is

identical

symbol

JIM

to the

.PSECT

statement:

JIM,GBL,OVR

because the .CSECT default values GBL and
named program section.

6.9

SYMBOL CONTROL DIRECTIVE:

OVR

are

assumed

for

the

.GLOBL

MACRO-11 produces a relocatable
object
module
and
a
1listing
file
containing
the
assembly
listing and symbol table.
The linker joins
separately-assembled object modules into a
single
executable
image.
During
1linking, object modules are relocated as a program function of
the specified base of the module.
The object modules are then
1linked
via
global
symbols, such that a global symbol in one module, defined
either by a global assignment operator (==), a global
label
operator
(::),
or
the .GLOBL directive can be referenced from another module.
Thus,

all symbols which will be referenced by

other

program

must be singled out as global symbols in the defining modules.

modules

The .GLOBL directive is provided to define (and thus
provide
1linkage
to)
symbols
not otherwise defined as global symbols within a module.
For example, if the .DSABL GBL directive is
in
effect
(see
Section
6.2),

.GLOBL

directives

might

.GLOBL A,B,C

is equivalent

==gxpression

(or A
B

£4

==expression
B==expression

(or C

the directive

(or

effect linkage to library routines.
)
)
)

6-39

included

1in a source program to

For a global
to:

symbol

definition,

GENERAL ASSEMBLER DIRECTIVES

Thus,

the

general

.GLOBL

where:

form of

.GLOBL

directive

is:

syml,sym2,...symn

syml,
sym2,...
symn

.GLOBL

the

represent
legal
symbolic
names.
When
multiple
symbols
are specified,
they are separated by any
legal separator (comma, space, and/or tab).

directive

may

also

embody

label

field

and/or

comment

field.
At

the end of assembly pass 1, MACRO-1l1
determines
whether
a
given
global
symbol is defined within the current program module or whether
it is to be treated as
an
external
symbol.
All
internal
symbols

appearing

within

given

they

assembly pass

references.

Refer

program

will

must

Section

enabling/disabling of global

defined

assumed
6.2

for

be
a

the end of

default

global

description

references.

In the example below, A and B are entry-point symbols.
The
symbol
A
has
been explicitly defined as a global symbol by means of the .GLOBL
directive, and the symbol B has been explicitly defined
as
a
global
label
by
means
of the double colon (::).
Since the symbol C is not
defined as a label within the current
assembly,
it
1is
an
external
(global) reference if .ENABL GBL is in effect.
;

DEFINE

SUBROUTINE

;

EXTERNAL

SUBROUTINE

WITH

ENTRY

POINTS

WHICH

CALLS

;

.PSECT

A:
X:

B::

.GLOBL
MOV

;DECLARE

A
@ (R5) +,R0

MOV

#X,R1

JSR

PC,C

RTS

MOV
CLR

(R5)+,R1
R2

External

symbols can

MACRO-11
below:

directive

CLR

EXT

.WORD

EXT

CLR

@EXT

External symbols may
shown below:

;DEFINE
;DEFINE
; CALL

THE UNNAMED PROGRAM SECTION.
A AS A GLOBAL SYMBOL.
ENTRY POINT A.

EXTERNAL

SUBROUTINE

;EXIT.

; DEFINE

appear

also

CLR
.WORD

EXT+A
EXT-2

CLR

@EXT+A (R1)

in the
direct

appear

ENTRY

POINT

operand

field
reference, as

term within

an
shown

instruction
or
in the examples

expression,
~

It should be noted that an undefined external symbol cannot be used in
the evaluation of a direct assignment statement or as an argument in a
conditional assembly directive (see Sections 6.10.1 and 6.10.3).

GENERAL ASSEMBLER DIRECTIVES

CONDITIONAL ASSEMBLY DIRECTIVES

6.10

Conditional assembly directives allow you to include or exclude blocks
of source code during the assembly process, based on the evaluation of
condition

stated

tests

within

the

body

This

program.

the

capability allows several variations of a program to be generated from
the same source code.

Conditional Assembly Block Directives:

6.10.1

.IF, .ENDC

The general form of a conditional assembly block is as follows:
(s) ;START CONDITIONAL ASSEMBLY BLOCK.
cond,argument

.IF

; RANGE OF CONDITIONAL ASSEMBLY BLOCK.

:END OF CONDITIONAL ASSEMBLY BLOCK.

. ENDC

cond

where:

represents a specified condition that must be met
1is to be included in the assembly.
if the block
the
by
tested
that may be
The conditions

conditional

assembly

directives

represents

any

legal

separator

Table

6-5.

and/or
t
(s)
argumen

tab).

are

defined in
space,

(comma,

or
argument (s)
symbolic
the
represent(s)
the specified conditional test.
of
expression(s)
the
thus a function of
These arguments are
specified condition to be tested (see Table 6-5).

either
represents the body of code that is
included in the assembly or excluded, depending

range

upon whether the specified condition is met.
. ENDC

This
terminates the conditional assembly block.
directive must be present to end the conditional
assembly block.

an 1illegal
A condition test other than those listed in Table 6-5,
argument, or-a null argument specified in an .IF directive causes that

line to be flagged with an error code (A) in the assembly listing.
Table

6-5

Legal Condition Tests for Conditional Assembly Directives
Conditions

Positive | Complement

Arguments

Assemble Block If:

Expression

is equal to 0

(or not equal to 0).

than

equal

(or

0).

less

greater

than or

GENERAL ASSEMBLER DIRECTIVES

Legal

Table
Condition Tests for

6-5 (Cont.)
Conditional Assembly Directives

Conditions
Positive | Complement

Arguments

Assemble

Block

Expression

(or greater
to 0).
DF

NDF

IDN

DIF

Symbolic

Symbol

argument

defined).

Macro-type

Argument

argument

non-blank).

Two

Arguments

macro-type

arguments

(or

If:

less

defined

than

than or

blank

are

equal

(or

not

(or

identical

different).

Expression

Same

EQ/NE.

Expression

Same

GT/LT.

NOTE

A macro-type

argument

argument),
angle
brackets

construction

(which

shown
below,
or
denoted

(as described

form of

symbolic

enclosed within
an
up—arrow

with
in Section

7.3.1).

<A,B,C>

“/124/

An example

.IF

conditional

assembly

ALPHA+1

directive

;ASSEMBLE

follows:

BLOCK

ALPHA+1=0.

. ENDC

The

two

operators

conditions,

& and

that

special

allowed

Logical

AND

Logical

inclusive

For

have

they are

example,

the

.IF DF

conditional

SYMl

meaning

within

and

NDF

in grouping symbolic arguments.

operator

assembly

operator

statement:

SYM2

« ENDC

results

and SYM2

the

are

assembly

both defined.

the

conditional

block

the

symbols

SYMI

GENERAL ASSEMBLER DIRECTIVES

Nested

conditional directives

take

the

form:

Conditional Assembly Directive
Conditional Assembly Directive
-

. ENDC
.ENDC

For

example,

the

following

conditional

directives:

SYMl
SYM2

.IF DF
.IF DF
»

. ENDC

.ENDC

can govern whether assembly 1is to occur.
In the example
outermost
condition
1is unsatisfied, no deeper level of
nested conditional statements within the program occurs.
Each

conditional

directive.
assembly
listing.

block

.ENDC
is

MACRO-11 permits
levels.
depth is

6.10.2

assembly

block

flagged with

must

directive

neéting

Any
statement
that
flagged with an error

terminated

with

.(ENDC

encountered

outside

a conditional

error

code

(0)

the

assembly

depth

16(10)

conditional

assembly

1in

attempts
to exceed this nesting
code (0) in the assembly listing.

Subconditional Assembly Block Directives:

Subconditional directives may
blocks to indicate:

above, if the
evaluation of

be placed

within

.IFF, .IFT, .IFTF
conditional

The assembly of an alternate body of
of the block tests false.

The assembly of a non-contiguous
body
of
conditional
assembly block, depending upon
conditional test in entering the block.

The unconditional
assembly
conditional assembly block.

code when

body

level

the

assembly

condition

code
within
the result of

code

within

the
the

The subconditional directives are described in detail
1in
Table
©6-6.
If
a
subconditional directive appears outside a conditional assembly
block, an error code (0) is generated in the assembly listing.

6-43

GENERAL ASSEMBLER DIRECTIVES

Table 6-6
Subconditional Assembly Block Directives

Subconditional

Directive

Function

.IFF

the

condition

tested

upon

entering

the

the code
£false,
1is
block
assembly
conditional
following this directive, and continuing up to the
of a subconditional directive or
occurrence
next

"to the end of the conditional assembly
to be

.IFT

included

in the program.

block,

If
the
condition
tested
upon
entering
the
conditional
assembly
block
1is
true,
the
code
following this directive, and continuing up to the
next
occurrence
of a subconditional directive or
to the end of the conditional assembly
block,
1is
to be

.IFTF

included

in the program.

The code following this directive, and
continuing
of a subconditional
occurrence
next
the
to
up

directive
or
to
the
end
of
the
conditional
block, is to be included in the program,
assembly
tested
<condition
regardless of the result of the
upon entering the conditional assembly block.

The implied argument of a subconditional directive
is
the
condition
test
specified
upon
entering
the
<conditional
assembly
block, as
reflected by the initial directive in the conditional coding
examples
below.
Conditional or subconditional directives in nested conditional
assembly blocks are not evaluated if the previous (or outer) condition
in
the
block
1is
not
satisfied.
Examples 3 and 4 below illustrate
nested
directives
that
are
not
evaluated
because
of
previous
unsatisfied conditional coding.
EXAMPLE

1l:

Assume

.IF DF
.

that

SYM

symbol SYM

is defined.

; TESTS TRUE, SYM IS DEFINED.
; THE FOLLOWING CODE.

.IFF
.

;TESTS FALSE.
;ASSEMBLE THE

. IFT

;TESTS TRUE.

SYM IS DEFINED.
FOLLOWING CODE.

SYM IS DEFINED.
;BLE THE FOLLOWING CODE.

ASSEMBLE

DO NOT

ASSEM-

.IFTF
.

;ASSEMBLE FOLLOWING CODE UNCONDITION;ALLY.

LIFT
.
.

;TESTS TRUE. SYM IS DEFINED. ASSEM;BLE REMAINDER OF CONDITIONAL ASSEM;BLY BLOCK.

. ENDC

GENERAL ASSEMBLER DIRECTIVES
EXAMPLE 2:

Assume that symbol X is defined and

that symbol Y

is not

defined.

.IF DF

.IF DF
.IFF
.

.IFT
;

; TESTS TRUE, SYMBOL X IS DEFINED.

s TESTS FALSE, SYMBOL Y IS NOT DEFINED.
; TESTS TRUE, SYMBOL Y IS NOT DEFINED,
sASSEMBLE THE FOLLOWING CODE.

s TESTS FALSE, SYMBOL Y IS NOT DEFINED.
;DO NOT ASSEMBLE THE FOLLOWING CODE.

.ENDC
.ENDC

EXAMPLE

Assume that symbol A is defined and that symbol B is not
defined.

.1IF DF

MOV

A,R1

.1FF

: TESTS TRUE.
A IS DEFINED.
;ASSEMBLE THE FOLLOWING CODE.
|

+:TESTS FALSE.
s ASSEMBLE

MOV

THE

A IS DEFINED.

DO NOT

FOLLOWING CODE.

R1,R0

.1IF NDF B
.

;sNESTED CONDITIONAL DIRECTIVE

IS NOT

+sEVALUATED.

. ENDC
. ENDC

EXAMPLE

Assume that symbol X is not defined and that symbol Y
defined.

.IF DF

;sTESTS FALSE.

SYMBOL X IS NOT DEFINED.

;DO NOT ASSEMBLE

.IF DF
.

THE FOLLOWING CODE.
s NESTED CONDITIONAL DIRECTIVE IS NOT
: EVALUATED.

. IFF
.

:NESTED SUBCONDITIONAL DIRECTIVE
:NOT EVALUATED.

. IFT
.

+ NESTED SUBCONDITIONAL DIRECTIVE IS
:NOT EVALUATED.

. ENDC
«.ENDC

GENERAL ASSEMBLER DIRECTIVES

©.10.3

Immediate Conditional Assembly Directive:

.IIF

An immediate conditional
assembly
directive
provides
a
means
for
writing a l-line conditional assembly block.
1In using this directive,
no terminating .ENDC statement is regquired, and the
condition
to
be
tested
is
completely
expressed
within
the
1line
c¢ontaining
the
directive.
Immediate conditional assembly directives are of the form:
.IIF

where:

cond,arg,statement

cond

represents
one
of
the
defined
for
<conditional
6-5.

’

represents

and/or

1legal

separator

(comma,

space,

tab).

arg

represents
the
argument
associated
immediate
conditional
directive,
expression,
symbolic
argument,
or
argument, as described in Table 6-5.

’

represents the separator between
the
conditional
argument
and
the
statement
field.
If
the
preceding argument is an expression, then a
comma
must
be
wused;
otherwise, a comma, space, and/or
tab may be used.

statement

For

any

1legal
condition
tests
assembly blocks in Table

example,

represents the specified statement
if the condition is satisfied.

the

.IIF

immediate
DF

conditional

with
the
i.e.,
an
macro-type

be assembled

statement:

FOO,BEQ ALPHA

generates thé code
BEQ

the

ALPHA

symbol

As with

FOO

defined within

the

.IF

directive,

Table

illegal

.IIF

directive

6.10.4

results

condition

argument,

the

error

code

source program.

test
a

other

null

(A)

than

argument

the

those

listed

specified

assembly

in
an

listing.

PAL-11lR Conditional Assembly Directives

In order to
maintain
compatibility
with
programs
developed
under
PAL-11R, the following conditionals remain permissible under MACRO-11,
It is advisable, however, to develop future programs using the
format
for MACRO-11l conditional assembly directives.
Directive

.IFZ
or
.IFNZ or
.IFL
or
.IFG
or
.IFLE
.IFDF
. IFNDF

.IFEQ
.IFNE
.IFLT
.IFGT

Arguments

Assemble

expression
expression
expression
expression
expression
symbolic argument
symbolic argument

expression=0
expression not equal 0
expression<0
expression>0
expression
is < or =0
symbol is defined
symbol is undefined

The

Block

1if

rules governing these directives are the same as for
conditional assembly directives previously described.

6-46

the

MACRO-11

CHAPTER 7
MACRO DIRECTIVES

7.1

DEFINING MACROS

In assembly-language programming, it is often convenient and desirable
to generate a recurring coding sequence by invoking a single statement
within the program.
In order to do this, the desired coding
sequence
is first established with dummy arguments as a macro definition.
Once
a macro has been defined, a single statement calling the macro by name
with
a
1list
of
real
arguments
(replacing the corresponding dummy
arguments in
the macro definition) generates
the
desired
coding
sequence.
This sequence is called the macro expansion.

7.1.1

.MACRO Directive

The first statement of a macro definition must be a .MACRO
This directive takes the

label:
where:

.MACRO

form:

name,

directive.

dummy argument list

label

represents an optional statement label.

name

represents the programmer-—-assigned
symbolic
name
of
the
macro.
This name may be any legal symbol
and may be
used
as
a
label
elsewhere
in
the
program.

represents
any
and/or tab).

dummy

argument
list

represents

legal

a number of

separator

(comma,

legal symbols

space,

(see 3.2.2)

that may
appear anywhere in the body of the macro
definition, even as a label.
These dummy
symbols
can be
used
elsewhere
1in
the
program with no
conflict of definition.
Multiple dummy
arguments
specified
in
this
directive may be separated by
any legal separator.
The detection of a duplicate
or
an
1illegal
symbol
in
a dummy argument list
terminates the scan and causes an error code to be
generated.

A comment may follow the dummy argument list in a

.MACRO directive,

shown below:

.MACRO

ABS

A,B

;DEFINES MACRO ABS WITH TWO ARGUMENTS.

MACRO DIRECTIVES

NOTE

Although it is legal for
.MACRO
directive,
this

label to
appear
on
a
practice is discouraged,

especially
in
the
case
of
nested
macro
definitions,
because
invalid
labels
or
labels
constructed with the concatenation character
will
cause the macro directive to be ignored.
This may
result
1in
improper
termination
of
the
macro
definition.
This
NOTE
also
applied
to
.IRP,
.IRPC,

7.1.2

and

.REPT.

.ENDM Directive

The final
directive

statement
of
of the form:

. ENDM

where:

every

macro

definition

must

.ENDM

name

represents

optional

argument

specifying

the

symbolic name of the macro being terminated by the
directive, as shown in the following example:
.

ENDM

i TERMINATES
sMACRO

. ENDM

ABS

s TERMINATES
~

sMACRO

THE

CURRENT

DEFINITION.

THE

CURRENT

DEFINITION

NAMED

ABS.

If specified, the symbolic name in the .ENDM statement must match
name
specified in the corresponding .MACRO directive.
Otherwise,

the
the

statement is flagged with an error code (A) in
the
assembly
1listing
(see
Appendix
D).
In
either case, the current macro definition is
terminated.
Specifying the macro name in
the .(ENDM
statement
thus
permits
MACRO-11
to
detect
missing
.ENDM
statements
or
improperly-nested macro definitions.
The

.ENDM

contain

directive

may

label,

shown

.MACRO
JSR
.WORD
.

An
is

followed

TYPMSG MESSGE
R5,TYPMSG
MESSGE

ENDM

comment

field,

but

must

not

below:
s TYPE A MESSAGE.

;END

TYPMSG

MACRO.

.ENDM statement encountered by MACRO-11 outside a macro
definition
flagged
with
an
error
code
(0)
in
the assembly listing (see

Appendix D).

NOTES

Labels

.ENDM directives

Illegal labels will
to be bypassed.

cause

are

the

ignored.

directive

MACRO DIRECTIVES

7.1.3

.MEXIT Directive

The .MEXIT directive may be used to terminate a macro expansion before

the

end

the

macro is encountered.

This directive is also legal

It is most useful in
within repeat blocks (see Sections 7.6 and 7.7).
The .MEXIT directive terminates the
the context of nested macros.

current macro as though

.ENDM

directive

had

encountered.

been

.MEXIT directive bypasses the complexities of nested
Using the
conditional directives and alternate assembly paths, as shown in the
following example:

.MACRO

ALTR N,A,B

IF EQ

; START CONDITIONAL ASSEMBLY BLOCK.

+MEXIT
.ENDC

;sTERMINATE MACRO EXPANSION.
;END CONDITIONAL ASSEMBLY BLOCK.

. ENDM

; NORMAL END OF MACRO.

Considering the above macro,

in an assembly where

the

real

argument

(see Table 6-5), the
for the dummy symbol N is equal to zero
conditional block would be assembled, and the macro expansion would be
terminated by the .MEXIT directive. When macros are nested, a .MEXIT
directive causes an exit to the next higher level of macro expansion.
1is

A .MEXIT directive encountered outside a macro definition
with an error code (0) in the assembly listing.

7.1.4

flagged

MACRO Definition Formatting

A form-feed character used within a macro

during

eject

however,

the

assembly

definition

the macro definition.

is not performed when the macro is expanded.

causes

page

A page eject,

Conversely, when the .PAGE directive 1is specified within a macro
definition, it is ignored during the assembly of the macro definition,
but a page eject is performed when that macro is expanded.

7.2

CALLING MACROS

.MACRO
the
A macro definition must be established by means of
(see Section 7.1.1) before the macro can be expanded within
directive
Macro calls are of the general form:
the source program.
label:

where:

name

label

name

real arguments

represents an optional statement label.

represents the name of the macro, as specified

the .MACRO directive

7-3

(see Section 7.1.1).

MACRO DIRECTIVES

real
arguments

represent

symbolic

arguments

the

arguments

specified

dummy

directive.
they
are

When multiple
separated
by

Arguments

the

which
in

arguments are
any
legal

macro

call

replace

the

are

. MACRO

specified,
separator.
treated

character strings whose usage is determined by the
macro
definition.
Note that MACRO-1l1l accepts the
ASCII value of
1lower-case
alphabetic
characters
when .ENABL LC has been specified.
When a macro name is the same as a user label, the appearance
of
the
symbol
in
the
operator field designates the symbol as a macro call;
the appearance of the symbol in the operand field designates it
as
a
label, as shown below:
ABS:

7.3

MOV

(RO),R1

sABS

DEFINED AS A LABEL.

ABS

sABS

CONSIDERED

ABS

#4 ,ENT,LAR

;ABS

IS A MACRO CALL.

ARGUMENTS

BE A LABEL.

IN MACRO DEFINITIONS AND MACRO CALLS

Arguments within a
other arguments by
Section 3.1.1.

macro

any

definition

the

legal

or macro call are
separated
from
separating characters described in

—

Macro definition arguments (dummy) and
macro
call
arguments
(real)
normally
maintain
a
strict
positional
relationship.
That is, the
first real argument in a macro call corresponds with the
first
dummy
argument
in a macro definition.
Only the use of keyword arguments in
a

macro

For

call

can

override

example,

the

following

expansion

this

contain multiple

.MACRO

correspondence

macro

definition

(see

and

Section

its

7.3.6).

associated

arguments:

macro

REN A,B,C

REN

Arguments

ALPHA,BETA,<Cl1l,C2>

which

enclosed
1in
macro call:
REN

causes

themselves

paired

angle

<MOV

contain

brackets,

separating

shown

characters

above.

For

must

example,

the

X,Y>,#44 ,WEV

the entire expression
MOV

X,Y

to replace all occurrences
Real
arguments
within
a

of the
macro

strings

and

are

treated

symbol A in
the
macro
definition.
«call are considered to be character
single entity during the macro expansion.

7-4

««««««

MACRO DIRECTIVES

The up-arrow (~) construction is provided to allow angle brackets to
be passed as part of the argument. This construction, for example,
could have been used in the above macro call, as follows:
~/<MOV X,Y>/,#44,WEV

REN

causing the entire character string

passed

argument.

The following macro call:

#44 ,WEV~ /MOV X,Y/

REN

however, contains only two arguments (#44 and WEV"/MOV X,Y/), Dbecause
the wup-arrow is a unary operator (see Section 3.1.3) and it is not
preceded by an argument separator.

As shown in the examples above, spaces can be wused within bracketed
of such expressions.
the legibility
qument constructions to increase

7.3.1

Macro Nesting

The nesting of macros, where the expansion of one macro includes a
call to another, causes one set of angle brackets in the macro
The
definition to be removed from an argument with each nested call.

depth

nesting

allowed

dependent

upon the amount of dynamic

memory used by the source program being assembled.

To pass an argument containing legal argument delimiters to nested
macros, the argument in the macro definition should be enclosed within
one set of angle brackets for each level of nesting, as shown in the
It should be noted that this extra set of
coding sequence below.
angle brackets for each level of nesting is required 1in the macro
definition,

not in the macro call.

.MACRO

LEVEL2
LEVEL2

LEVEL1 DUM1,DUM2
<DUMI1>
<DUM2>

.ENDM

.MACRO

LEVEL2 DUM3

ADD
MOV

#10,R0
RO, (R1) +

DUM3

. ENDM

A call to the LEVEL1 macro, as shown below, for example:
LEVEL1

<MOV

X,R0>,<MOV

R2,R0>

causes the following macro expansion to occur:
MOV
ADD
MOV
MOV

X,RO
#10,R0
RO, (R1)+
R2,R0

MOV

RO, (R1)+

MACRO DIRECTIVES

When macro definitions are nested, i.e., when a
macro
definition
is
contained
entirely
within the definition of another macro, the inner
definition is not a callable macro until
the
outer
macro
has
been.
called and expanded.
For example, in the following coding:

.MACRO

LV1 A,B

.MACRO

LVZ2

. ENDM
. ENDM

the LV2 macro cannot
been
so0
invoked.
definition
invoked.

7.3.2
An

cannot

be called
Likewise,
called

Special Characters

argument

bracketed

may

include

construction

semicolons,

and expanded until the
LV1
macro
has
any macro defined within the LV2 macro
and
expanded
until
LV2
has
also
been

in Macro Arguments

special

commas.

that

For

.MACRO

PUSH ARG

MOV

ARG,
- (SP)

characters

without

argument does not

example,

the

enclosing

contain

them

spaces,

macro definition:

tabs,

.ENDM
-

PUSH
causes

the

following

MOV

7.3.3

X+3(%2)
code

generated:

X+3(%2),-(SP)

Passing Numeric Arguments

as Symbols

When macro arguments are passed,
an
absolute
symbol
value
passed which is treated by the macro as a numeric string.
An
preceded by the unary operator backslash (\) is treated as
a

value in the current program radix.
The ASCII characters
this value are inserted in the macro expansion, and their
defined
in
the
context
of
the
resulting
code,
as

following

example:

.MACRO

INC

CON

A,\B

CON

-WORD

. ENDM

C=0

INC

TREATED AS

; PROGRAM

. ENDM

A'B:

X,C

numeric
representing
function
is
shown in the

A,B

B=B+1

.MACRO

can
be
argument

A NUMBER

CURRENT

RADIX.

A,B

;A'B

IS DESCRIBED

SECTION

7.3.6.

MACRO DIRECTIVES

The above macro call
X0:

.WORD

Note

this

(INC) would thus expand to:

4
expanded

code

that

the

label

X0:

the

result

the

concatenation
of
two real arguments.
The single quote (') character
in
the
label
A'B:
causes
the
real
arguments
X
and
0
to
be
concatenated
as
they
are
passed during the expansion of the macro.
This type of argument construction is described in further
detail
in
Section 7.3.6.
A

subsequent

call

the

same

macro

would

generate

the

following

code:

and so on, for later calls.
The two macro definitions
are
necessary
because
the
symbol associated with dummy argument B (i.e., C) cannot
be updated in the CON macro definition, because its numeric value
has
already
been substituted for its symbolic name, i.e., the character 0
has replaced C in the argument string.
In the CON
macro
definition,
the
number
passed is treated as a string argument.
(Where the value
of the real argument is 0, only a single 0 character is passed to
the
macro expansion.)
Passing

numeric values

listings.

For

this

example,

manner

versions

useful

identifying

programs

source

created

through

conditional assemblies of a single source program
can
be
identified
through
such
coding
as that shown below.
Assume, for example, that
the symbol ID in the macro call (IDT) has been
equated
elsewhere
in
the source program to the value 6.
.MACRO

IDT
SYM
/VO5A'SYM/

.IDENT

.ENDM

;WHERE

above

macro

7.3.4
If

call

would

then expand

numeric value

of Arguments

arguments

fewer

THE

UPDATE

PROGRAM.

to:

/V05a6/

the

Number

definition,

\ID

.IDENT

where

VO5A IS

;VERSION

IDT
The

;ASSUME THAT THE SYMBOL ID TAKES
;ON A UNIQUE 2-DIGIT VALUE.

arguments

code

appear

the

symbol

ID.

in Macro Calls

appear

error

(Q)
1in

the

macro

call

is generated
the

macro

call

than

the

the assembly
than

macro

listing.

the

macro

definition,
missing
arguments
are
assumed
to be null values.
The
conditional directives .IF B and .IF NB (see Table 6-5)
can
be
used
within the macro to detect missing arguments.
The number of arguments
can also be specified using the .NARG directive (Section 7.4.1).
Note
that a macro can be defined with no arguments.

7.3.5

Creating Local Symbols Automatically

A label is often required in an expanded macro.
In
the
conventional
macro
facilities
thus far described, such a label must be explicitly

7-7

MACRO DIRECTIVES
specified as an argument with each macro call.
Be careful in
issuing
subsequent
calls
to
the same macro, to avoid specifying a duplicate
label as a real argument.
This concern c¢an be
eliminated
through
a
feature
of
MACRO-11l
which
<creates a unique symbol where a label is
required

in an expanded macro.

As noted in Section
3.5,
MACRO-11l
<can
automatically
c¢reate
1local
symbols
of the form n$, where n is a decimal integer within the range
64 through 127, inclusive.
Such local symbols are created by MACRO-1l1
in numerical order,

as shown below:

64$
65%
-

1265
1278

This automatic facility is invoked on
each call
of
a
macro
whose
definition contains a dummy argument preceded by the guestion mark (?)
character, as shown in the macro definition below:
.MACRO

ALPHA,

TST
BEQ

A
B

ADD

#5,A

A,?B

; CONTAINS DUMMY ARGUMENT B PRECEDED BY
s+QUESTION MARK.

B:
. ENDM

A local symbol is generated automatically by MACRO-11l only when a real
argument
of
the
macro
<call
is either null or missing, as shown in
Example 1 below,
defined above.

which

reflects

the

expansion

the

ALPHA

nmacro

If the real argument is specified in the macro call, however, MACRO-1l1
inhibits
the
generation
of
a
local symbol
and
normal
argument
replacement occurs,

EXAMPLE

1l:

as shown

in Example

Generate a Local Symbol

ALPHA

TST

BEQ
ADD

for

2 below.

the Missing Argument:

s SECOND ARGUMENT IS MISSING.

643
$#5,R1

; LOCAL SYMBOL IS GENERATED.

64S:

EXAMPLE

Do Not Generate

ALPHA

R2,XYZ

TST

BEQ

ADD

XYZ

#5,R2

a Local Symbol:

;SECOND ARGUMENT XYZ IS SPECIFIED.

; NORMAL ARGUMENT REPLACEMENT OCCURS.

XYZ:

Automatically-generated local symbols
16 (10)

arguments of

a macro definition.

are

restricted

the

first

Note that
automatically-created
1local
symbols
resulting
from
the
expansion
of a macro, as described above, do not in any way influence
local
symbol
block
boundaries.
In
other
words,
such

“automatically=-created
1local
block in their own right.

symbols

do not establish

local

symbol

MACRO DIRECTIVES

However, when a macro has several arguments
earmarked
for
automatic
local
symbol
generation,
substituting a specific label for one such
argument introduces a risk that assembly errors will result.
This
is
because
MACRO-1l1
constructs
its
argument
substitution list at the
point of macro invocation.
Therefore, the appearance
of
any
label,
the
.ENABL
LSB
directive,
or
the
.PSECT
directive, in the macro
expansion will create a new local
symbol
block.
This
could
leave
local
symbol
references
1in
the
previous
block
and
the
symbol
definitions in the new one, resulting in error codes in
the
assembly
listing
(see
Appendix D).
Furthermore, a subsequent macro expansion
that generates local symbols in the new block may duplicate one of the
symbols
in question, resulting in an additional error code (P) in the
assembly

7.3.6

listing.

Keyword Arguments

Macros may be defined with and/or invoked with
keyword argument has the following form:

keyword

arguments.

name=string
where

The

name

represents

the

dummy

string

represents

the

real

keyword

argument may

not

argument,

symbolic argument.

contain

unless properly delimited as described
When

keyword

definition,

argument

the

appears

embedded

section

the dummy

specified string becomes

argument

separators

7.3.

argument

the default

list

macro

real argument at

macro call.

When a keyword argument appears in the real argument list of
a
macro
call,
the
specified
string
becomes the real argument for the dummy
argument that exactly matches the specified name, whether or
not
the
dummy
argument
was
defined
with
a keyword.
If a match fails, the
entire argument specification is treated as the next
positional
real
argument.
A
keyword argument may be specified anywhere in the dummy
argument list of a macro definition and
is
part
of
the
positional
ordering
of
argument.
On the other hand, a keyword argument may be
specified anywhere in the real argument list of a macro call but
does
not affect the positional correspondence
of the remaining arguments.
1

2
3

5
)
7
8
9
10
11
12
13
14
15
l6

LLIST

;

DEFINE A MACRO HAVING KEYWORDS

.MACRO
.WORD
. WORD
. WORD
. ENDM

H
:
:

NOW INVOKE

IN DUMMY ARGUMENT LIST

TEST CONTRL=1l,BLOCK,ADDRES=TEMP
CONTRL
BLOCK
ADDRES

SEVERAL TIMES

MACRO DIRECTIVES

18
19

20
21

000000
000000
000002
000004

000O0OOG
000000G
00000O0G

TEST
. WORD
.WORD
. WORD

A,B,C
A
B
C

000006
000006
000010
000012

000040
000030
000020

TEST
.WORD
.WORD
.WORD

ADDRES=20,BLOCK=30,CONTRL=40
40
30
20

TEST
.WORD

BLOCK=>5
1

. WORD
. WORD

5
TEMP

TEST
. WORD

CONTRL=5,ADDRES=VARIAB
5

-.WORD
. WORD

VARIAB

000014
000014

000016
000020

000001
000005
000000G

24
25

000022
000022
000024
000026

000005
000000
000000G

000030
000030
000032
000034

000001
000000
000000G

.WORD
.WORD
.WORD

000036
000036
000040
000042

000001
000000
o00000O0C

TEST
.WORD
.WORD
.WORD

000001

.END

TEST

1
TEMP

28
29
30

7.3.7

ADDRES=JACK!JILL
1
JACK!JILL

Concatenation of Macro Arguments

The apostrophe or single quote
character
(')
delimiting
character
in
macro
definitions.

precedes and/or

follows a dummy argument

removed,
and
the
substitution
of
point.
For example, in the following

A'B:

when

.MACRO

.ASCI1IZ
.BYTE

it
XY:

expanded,
.ASCIZ
.BYTE

macro

definition

the real argument occurs
statements:

1is

that

DEF A,B,C

/C/
''"A,''B

the macro DEF

DEF

operates
as
a
legal
A
single
quote that

is called

through

the

statement:

X,Y,<MACRO-11>
as

follows:

/MACRO-11/
'X,'Y

expanding
the
first
1line,
the
scan
for
the
first
argument
terminates
wupon
finding
the
first ' character.
Since A is a dummy
argument, the ' is removed.
The scan then resumes with B;
B is
also
noted
as
another dummy argument.
The two real arguments X and Y are
then concatenated to form the label XY:.
The third dummy argument
1is
noted
1in
the operand field of the .ASCIZ directive, causing the real
argument MACRO-11l to be substituted in this field.

7-10

MACRO DIRECTIVES

When evaluating the arguments to the .BYTE directive during

expansion

of the second line, the scan begins with the first ' character.
Since
it is neither preceded nor
followed
by
a
dummy
argument,
this
'
character
remains
in
the macro expansion.
The scan then encounters
the second ' character, which is followed by a dummy argument
and
1is
therefore
discarded.
The
scan of
argument
A
1is terminated upon
encountering the comma (,).
The third ' character is neither preceded
nor
followed
by
a
dummy
argument
and
again remains in the macro
expansion.
The fourth (and last) ' character is followed
by
another
dummy
argument
and
is
likewise
discarded.
(Note
that
four
‘'
characters were necessary in the macro definition to
generate
two
'
characters in the macro expansion.)
|

7.4

MACRO ATTRIBUTE DIRECTIVES:

.NARG,

.NCHR,

AND

.NTYPE

Three directives are available in MACRO-11 which
allow
the
wuser
to
determine
certain
attributes
of
macro arguments.
The use of these
directives permits
selective
modifications
of
a
macro
expansion,
depending
on
the
nature
of
the
arguments
being
passed.
These
directives are described separately below.

7.4.1

.NARG Directive

The .NARG directive is used to determine the number
of
arguments
in
the
macro
call currently being expanded.
Hence, the .NARG directive
can appear only within a macro definition;
if it does not,
an
error
code
(0)
1is generated in the assembly listing.
This directive takes
the

form:

label:

where:

.NARG

symbol

label

represents an optional

symbol

represents
any
legal
symbol.
This
symbol
1is
equated
to
the
number of arguments in the macro
call currently being expanded.
If a symbol is not
specified,
the .NARG directive is flagged with an
error code (A) in the assembly listing.

7-11

statement

label.

MACRO DIRECTIVES

example

the

.NARG

directive

follows:

LTITLF NARG

NE LN

JMACRNA

NOPP, NUM

MNARG

SYM

JIF

EQ,SYM

O P~

JMEXTIT

TFF
JRERPT

M
N

NOR

i@
11
12

e ENDM
o ENDC
« ENDM

NOPP

Apegen

prarae

« NARG

SYM

o IF
MEXTITY

EQ,SYM

. IFF

JREPT

NOP

ENDM

s ENDC

{6
17

annaee

APAAA Y

NOPP

JNARG

SYM

JMEXTT

o TFF

P2QARA

REPT

()

NOP

e ENDM
A00QA0Q2

QAag2un

NOP

NOP
NOP
NOP
NOP
NP

AARAA2 . AAR2UR
R
002U
ARRAGU
QCOQA24Q
PARBREe
QQ@a2uR
AAAA1I2
o0agu~
araRr12

«ENDC

{19
20

7.4.2

Q20001

« ENP

J.NCHR Directive

The .NCHR directive, which can appear anywhere in a MACRO-1l1]
program,
is used to determine the number of characters in a specified character
string.
This directive, which is useful in calculating the length
of
macro

arguments,

label:

.NCHR

where:

label

symbol

takes

the

following

form:

symbol,<string>
represents

optional

statement

label.

represents

any
legal
symbol.
This
symbol
1is
the
number
of
characters
in
the
specified character string.
If a
symbol
is
not
specified,
the .NCHR directive is flagged with an
error
code
(A)
1in
the
assembly
1listing
(see
equated

Appendix

D).

7-12

MACRO DIRECTIVES

represents

and/or

{string>

tab).

legal

separator

(comma,

space,

represents a string of printable characters.
The
character
string
need
be
enclosed within angle

brackets
specified

if
the
a
1legal
separator (comma,
space,
and/or
tab).
If
the
delimiting characters
do
not
match
or
if the
ending delimiter cannot be detected because
of
a
syntactical
error
in
the character string (thus
prematurely terminating its evaluation), the .NCHR
directive is flagged with an error code (A) in the
assembly listing.
,

An example of

the

.NCHR directive

N
AL
D O P

NCHR

C(CHAR,MESS

SYM,MESS

JWORD

SYM

«ASCII
«EVEN

/MESS/

o ENDM

ARPARD

MSG1 ¢t

CHAR

HELLO>

PENARS

SYM,HMELLO

pe@roe

repARS

«WORD

earee?

SYM

112

«ASCIT

/HELLOY/

NCHR

tas

earangd

t14

aearns

114

tt?

«EVEN

1e
13
ta

The

JENR

.NTYPE Directive

.NTYPE directive is used to determine the

specified macro argument.

within a macro definition;

with

(")
only
contains

follows:

«MACRD

aRraRrR6

7.4.3

(<>)
or
up-arrows
character
string

«TITLF NCHR

PrAQRY

any

an error code (0)

the form:
label:

J.NTYPE

where:

label

addressing

mode

Hence, the .NTYPE directive can appear only
if

it appears

elsewhere,

in the assembly listing.

1is

flagged

This directive takes

symbol,aexp

represents an optiqmal statement label.
represents
any
legal
symbol.
This
synm bol
is
equated
to
the
6-bit
addressing
mode
of
the
following argument.
If a symbol is not specified,
the .NTYPE directive is flagged with an error code

(A)

in the assembly listing.

represents
any
and/or tab).
7-13

legal

separator

(comma,

space,

MACRO DIRECTIVES

represents any legal address expression,
as
used
with
an opcode.
If no argument is specified, the
result will be zero.

aexp

example of
shown below:

the

use

.NTYPE

directive

.TITLE

NTYPE

«MACRO

JNTYPE

a macro

definition

SAVE, ARG
SYM, ARG

o I1F

MOV

ARG,=(SP)

$REGISTER

o JFF
MOV

#HARG, = (SP)

tNONeREGISTFR

8
9

FR,SYMRTQ

sENDC

i1
12
13

MOQODE

“MODE

JENPM

P207200

0OQAAQAA

TFMP

JWORD

SAVE

15
16

2pPnn2
ARROA Y

ArARR2

21P1Ub

+NTYPE

SYM,R1

o IF

FQR,SYMRT7Q

MAV

R{,=(SP)

tREGISTER

#R1,«(SP)

tNONeREGISTER

. IFF

MAV

« ENNC

18
19 peac0uy

“
SAVFE
PPARET

«MNTYPE

M127U6

TEMP

W IF

ER,SYMR
T

MAY

TFMP,=(8P)

sREGISTER

#TEMP,=(SP)

INON=REGISTFR

MAY

PORARR®

MNDF

MDDE

+ ENDC

el
22

feAa2M 1

o FND

For additional
information
concerning
Chapter 5 and Appendix B, Section B.2.

7.5

MONDE

SYM, TEMP

o« IFF

PAAPRY

MODE

.ERROR AND

addressing

modes,

refer

.PRINT DIRECTIVES

The .ERROR directive is used to output messages to
the
1listing
file
A common use of this directive is to provide
during
assembly pass 2.
a diagnostic announcement of a rejected or erroneous macro call or
to
alert
the
wuser
to
the
existence
of
an illegal set of conditions
specified in a conditional assembly.
If
the
1listing
file
1is
not
specified,
the
.ERROR
messages
are
output
to
the command output
device.
The .ERROR directive takes the form:
label:

.ERROR

expr

;text

MACRO DIRECTIVES

represents an optional statement label.

label

where:

represents an optional expression whose

expr

output

when

the

during assembly.

value

denotes the beginning of the text string.

;

represents the specified message

text

the

.ERROR directive.

Upon encountering an .ERROR directive anywhere in
MACRO-11 outputs a single line containing:

associated
source

The sequence number of the .ERROR directive statement

The value of the current location countér

The value of the expression,

program,

if one is specified

.
.

The source line containing the .ERROR directive.

For example,

with

(P)

An error code

.ERROR directive is encountered

the following directive:

.ERROR

: INVALID MACRO ARGUMENT

causes a line in the following form to be output to the listing file:
Seq.

No.

Loc.

No.

512 005642

Exp.

Text

Value

000076

.ERROR

. INVALID MACRO ARGUMENT

The .PRINT directive is identical in function to the .ERROR directive,
except that it is not flagged with the P error code.
7.6

INDEFINITE REPEAT BLOCK DIRECTIVES:

.IRP AND .IRPC

An indefinite repeat block is a structure that is similar to a macro
essentially a macro definition that has only one dummy
definition;

this
argument. At each expansion of the indefinite repeat range,
dummy argument is replaced with successive elements of a specified
its
An indefinite repeat block directive and
real argument list.
associated repeat range are coded in-line within the source program.
This type of macro definition and expansion does not require calling
the macro by name, as required in the expansion of conventional macros
previously described in this section.

An indefinite repeat block can appear either within or outside another
or repeat block (see
indefinite repeat block,
macro definition,

The rules for specifying indefinite repeat block
Section 7.7).
arguments are the same as for specifying macro arguments (see Section
7.3).

7.6.1

.IRP Directive

The .IRP directive is used to replace a dummy argument with successive
This replacement
in an argument string.
specified
real arguments
process occurs during the expansion of an indefinite repeat block
range.

This directive takes the following form:
7-15

MACRO DIRECTIVES

label:

L.IRP

sym,<argument

(range

indefinite

list>

repeat

block)

. ENDM

where:

label
sym

represents

optional
dummy

statement

argument

label.

that

successively

replaced

with
the
specified
real
arguments
enclosed within the angle brackets.
If
no
dummy
argument
1is
specified,
the
.IRP
directive
is
flagged with an error code
(A)
in
the
assembly
listing.
’

represents

and/or
<argument

list>

any

legal

separator

(comma,

space,

tab).

represents
a
1list
of
real
arguments
enclosed
within
angle
brackets
that is to be used in the
expansion of the indefinite repeat range.
A
real
argument
may
consist
of one or more characters;
multiple arguments must be separated by any
1legal
separator

(comma,

arguments
range

are

space,

specified,

represents

the

for

occurrence

each

block

and/or

no
code

tab).

action

real

taken.

repeated

argument

once

the
list.
The
range
may
contain
other
macro
definitions
and
repeat
ranges.
The
.MEXIT
directive (see Section 7.1.3) is legal within
the
range

. ENDM

indicates

indefinite

the

end

repeat

the

block.

indefinite

repeat

block

range.

example

7.6.2
The

the

use

the

.IRP

directive

shown

Figure

7-1.

.J.IRPC Directive
.IRPC

directive

1is

available

permit

single

character

substitution, rather than argument substitution.
On each iteration of
the indefinite repeat range, the dummy argument is replaced with
each
successive
character in the specified string.
The .IRPC directive is
specified as follows:
label:

L.IRPC

sym,<string>

(range

indefinite

repeat

block)

. ENDM

where:

label

represents

optional

7-16

statement

label.

MACRO DIRECTIVES

represents a dummy argument that

sym

successively

arguments
no dummy

real
specified
the
with
replaced
If
enclosed within the angle brackets.

the .IRPC directive is
argument is specified,
in the assembly
(A)
flagged with an error code
listing.

represents any
and/or tab).

’

legal

separator

(comma,

space,

represents a list of characters enclosed within
angle brackets to be used in the expansion of the

indefinite repeat

Although

range.

the

angle

required only when the string
brackets are
is
their use
contains separating characters,
recommended for legibility.

range

represents the block of code to be repeated once
for each occurrence of a character in the list.
The range may contain macro definitions and repeat

ranges.

repeat

indicates the end of the indefinite repeat

block

block.

.ENDM

The .MEXIT directive (see Section 7.1.3)

is legal within the range of an indefinite
|

range.

An example of the use of the .IRPC directive is shown in Figure 7-1.
LTITLE

«LIST

IRPTST

3
4

5
6

,Iflp

200002
2080P04
AAPN1ID
A20014
PRA20
V20024

Q16720
16722
©B16720
Q16720
016727
016722

. 1RPC

APAN3E
220234

P2004Q

200044
POPN50
Vi P0QAS4
15

112744y
112741

112741

112741
(12741
112741

#ix,=(R1)

MOvVB

MA,~(R1)
#B,=(RY{)
#C,=(R1)

JENDM
MOVEB
MOVB

0QAROQ006
QPQGQQG
QQPR0POG

MOVB
MOVvB
MOVB

QQOQQ0G

X,<ABCDEF>

MOVB

QP0QR06
Q000206

#D,=(R1)
BE,~(R1)
#F,»(R1)

JEND

pegeey’

Figure 7-1

7.7

MOV
MOV

AA, (RD)e
BR, (RA)+
CCy (R
DD, (RA)«
EE, (RDB)+
FF,(RQ)»

X,(RO)e

L ENDM
MOV
MOV
MOV
MOV

QAABARAG
0ROPROG
200200206
nNpRR226
QPAAMAG
ARYO0CAG

x,tlA,BEl,fiCaD&.EE,FFb

MOV

Example of .IRP and .IRPC Directives

REPEAT BLOCK DIRECTIVE:

.REPT, .ENDR

of
fula block of code a number
to duplicate
It is sometimes use

times

This duplication of code |is
with other source code.
in-line
accomplished by creating a repeat block, wusing a directive in the
form:

7-17

MACRO DIRECTIVES

label:

.REPT

exp

(range

repeat block)

. ENDM

where:

label

represents

exp

represents

optional

any

statement

1legal

label.

expression

whose

controls

value

the number of times the block of code is
to be assembled
within
the
program.
When
the
expression
value
1is
1less
than or equal to zero
(0), the repeat block is not assembled.
If
this
expression

statement

assembly
range

1is

not

absolute

flagged with

listing.

represents

the

block

value,

an error

code

the

(A)

.REPT

repeated

the

number
of
times
determined
by
the
specified
expression value.
The repeat
block
may
contain

macro
other

definitions,
indefinite
repeat blocks,
repeat
blocks.
The
.MEXIT
directive
legal within the range of a repeat block.

. ENDM

indicates
the end of the repeat block range.
The
terminating
statement
in a repeat
block can be
either an .ENDM directive or an .ENDR directive.

or
.ENDR

7.8

MACRO LIBRARY DIRECTIVE:

The

.MCALL

and/or

or
1is

directive

allows

user-defined

macro

.MCALL

you

indicate

definitions

advance

that

are

those

system

required

the

assembly of the source program.
The .MCALL directive
allows
vyou
to
specify
the names of all system or user macro definitions not defined
within the source program but
which
are
required
to
assemble
the
program.
The .MCALL directive must appear before the first occurrence
of a call to any externallywdeflned macro.
The .MCALL directive is of
the form:
.MCALL

where:

argl,arg2,...argn

argl,
arg2,...

represent
definitions

argn

program.

The

separated

and/or
The

.MCALL
user-defined

the
symbolic
required in the
symbolic

any

tab).

directive

and

thus
provides
system macro libraries

legal

names
of
assembly of

macro

names

separator

the
means
to
during assembly.

the
the

macro
source

may

(comma,

space,

access

both

The /ML sw;tch under RSX-1ll and the /LIBRARY qualifier under
IAS
and
RT-11,
specified
in
connection
with
an
input file specification,
1ndlcate to MACRO-11 that the file is a macro llbrary.
When
a
macro
call is encountered in the source program, MACRO-11 first searches the
user

macro

continues

library

the

for

the

search with

named

the

macro

definitions, and,

system macro library.

7-18

necessary,

MACRO DIRECTIVES

In
Any number of such user-supplied macro files may be designated.
cases of multiple 1library files, the search for the named macros
The search continues 1in
begins with the last such file specified.
order until the required macro definitions are found,
reverse
terminating again, if necessary, with a search

the

system

macro

library.

If any named macro is not found upon completion of the search, 1i.e.,
if the macro is not defined, the .MCALL statement is flagged with an
Furthermore, a statement
error code (U) in the assembly 1listing.

elsewhere

the

source

program

which

undefined macro is flagged with an error

attempts to expand such an

code

(O)

1in

the

assembly

listing.

The command strings to MACRO-11, through which file specifications are
supplied, are described 1in detail in the appropriate system manual
(see Section 0.3 in the Preface).

7-19

APPENDIX A

MACRO-11 CHARACTER SETS

A.1

ASCII CHARACTER SET
EVEN
PARITY

7-BIT
OCTAL

CODE

CHARACTER

0
1

000
001

NUL
SOH

002

STX

003

ETX

004

EOT

005

ENQ

0
1
1

006
007
010

ACK
BEL
BS

BIT

0
0

011
012

013

014

HT
LF

VT
FF

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.
also EOM, end of
End of text;
message, CONTROL/C.

End of transmission (END);
off TWX machines, CONTROL/D.
also
(ENQRY) ;
Enquiry

shuts
WRU,

CONTROL/E.

Acknowledge; also RU, CONTROL/F.
Rings the bell. CONTROL/G.
format
FEO,
also
Backspace;

some

backspaces

effector.

machines, CONTROL/H.

Horizontal tab. CONTROL/I.
Line feed or Line space (new line);
to

paper

advances

duplicated by CONTROL/J.

Vertical tab (VTAB).
to

Feed

Form

top

CONTROL/L.

(PAGE).

return

line,

CONTROL/K.

page

to beginning of

015

Carriage

016

shift out;

changes ribbon color to

017

shift in;

changes ribbon color

020

DLE

Data link escape.

021

DC1l

line;

duplicated by CONTROL/M.

red.

CONTROL/N.

black.

CONTROL/O.

Device

control

023

DC3

024

DC4

CONTROL/P (DCO).

turns

transmitter (READER) on, CONTROL/Q
(X ON). ©0 022 DC2 Device control
2;

turns

CONTROL/R

punch

or auxiliary on.

(TAPE, AUX ON).

control

Device control 4;

turns

Device

transmitter (READER)
off, CONTROL/S (X OFF).

auxiliary

OFF).

off.

turns

punch

CONTROL/T

oOr

(AUX

MACRO-11

CHARACTER SETS

EVEN

7-BIT

PARITY

OCTAL

BIT

CODE

CHARACTER

025

NAK

REMARKS
Negative

ERROR.
SYN

also

Synchronous

file

ETB

End

transmission

LEM,

logical

Cancel

(CANCL).

End

medium.

SUB

Substitute.

ESC

Escape.

File

Group

Record

Unit

Space.

=@ 00N

| =~

block;

end

CONTROL/W.

CAN

also

medium.

CONTROL/X.
CONTROL/Y.

CONTROL/Z.

CONTROL/SHIFT/K.

separator.

CONTROL/SHIFT/L.

separator.

Accent

ERR,

(SYNC)
.

CONTROL/V.

HIQPMBUOQOWIP®
VIV I A o OOLNOUID WNHON -

026

acknowledge;

CONTROL/U.

acute

CONTROL/SHIFT/M.

CONTROL/SHIFT/N.
CONTROL/SHIFT/O.

apostrophe.

7-BIT
OCTAL

BIT

CODE

CHARACTER

112
113
114
115
116
117
120
121
122
123
124
125
126
127
130
131

135
136
137
140

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
175

170
177

-~ 4

134

TN
X ECSC AN RQUOIERAURITAQMMD QWO T

132

Yt S

EVEN
PARITY

N KX E<SCHNODOWOZERHRG

MACRO-11 CHARACTER SETS

DEL

REMARKS

SHIFT/K.
SHIFT/L.
SHIFT/M.
*

* %

Accent

grave.

This code generated by ALTMODE.
This code generated by
PREFIX
(if present).
DELETE,

RUBOUT.

Appears as # or ~ on some machines.

** + Appears as

_ on some machines.

A-3

key

MACRO-11 CHARACTER SETS

A.2

RADIX-50

CHARACTER SET

Character
space

ASCII Octal Equivalent
40
|

Radix-50 Equi?alent
0

A-Z

101-132

1-32

unused

0-9
The

60-71

maximum Radix-50 value

1is,

36-47

thus,

47*%50% +47*50+47=174777
The following table provides a convenient means of translating between
the
ASCII
character
set and its Radix-50 equivalents.
For example,
given the ASCII string X2B, the Radix-50 equivalent is (arithmetic
1is
performed in octal):
X=113000
2=002400
B=000002
X2B=115402

SINGLE

CHAR.

OR
FIRST

CHAR.

SECOND

THIRD

CHARACTER

Space
A
B
C
D
E
F
G
H
I
J
K
L
M
N
0
P
Q
R
S
T
U

000000
000001
000002
000003
000004
000005
000006
000007
000010
000011
000012
000013
000014
000015
000016
000017
000020
000021
000022
000023
000024
000025

X
Y

000030
000031

Space
A
B
C
D
E
F
G
H
I
J
K
L
M
N
O
P
Q
R
S
T
U

000000
003100
006200
011300
014400
017500
022600
025700
031000
034100
037200
042300
045400
050500
053600
056700
062000
065100
070200
073300
076400
101500

Space
A
B
C
D
E
F
G
H
I
J
K
L
M
N
0
P
Q
R
S
T
U

X
Y

113000
116100

X
Y

\Y
W
Z

104600
107700
121200

124300

'
W
Z

000000
000050
000120
000170
000240
000310
000360
000430
000500
000550
000620
000670
000740
001010
001060
001130
001200
001250
001320
001370
001440
001510

- 001560
001630

\'
W

002020

001700
001750

002070

A-4

000026
000027
000032

000033

MACRO-11 CHARACTER SETS

SINGLE

CHAR.

OR
FIRST CHAR.

.
Unused
0
1
2
3
4
5
6
7
8
9

127400
132500
135600
140700
144000
147100
152200
155300
160400

163500
166600
171700

SECOND
CHARACTER

.
Unused
0
1
2
3
4
5
6
7
8
9

002140
002210
002260
002330
002400
002450
002520
002570
002640

002710
002760
003030

THIRD

CHARACTER

.
Unused
0
1
2
3
4
5
6
7
8
9

000034
000035
000036
000037
000040
000041
000042
000043
000044
000045
000046
000047

APPENDIX B

MACRO-11 ASSEMBLY LANGUAGE AND ASSEMBLER DIRECTIVES

B.1

SPECIAL CHARACTERS

Function

@“ ..

Character

(comma)

2 e @

space

IV A

(apostraphe)V

vertical

B.2

tab

Label terminator
Direct assignment operator
Register term indicator
Item terminator or field terminator
- Item terminator or field terminator
Immediate expression indicator
Deferred addressing indicator
Initial register indicator
Terminal register indicator
Operand field separator
Comment field indicator
Arithmetic addition operator or auto
increment indicator
Arithmetic subtraction operator or auto
decrement indicator
Arithmetic multiplication operator
Arithmetic division operator
Logical AND operator
Logical OR operator
Double ASCII character indicator
Single ASCII character indicator or
concatenation indicator
Assembly location counter
Initial argument indicator
Terminal argument indicator
Universal unary operator or argument
indicator
Macro call numeric argument indicator
Source line terminator

SUMMARY OF ADDRESS MODE SYNTAX

Address mode syntax is
expressed
1in
the
summary
below
using
the
following
n
symbols:
is
an integer between 0 and 7 representing a
register number;
R is a register expression;
E
1is
an
expression;

and ER is either
to 7.

a register

expression or

an expression

in the

range

MACRO-11 ASSEMBLY LANGUAGE AND ASSEMBLER DIRECTIVES

Address
Mode

- Format

Address
Mode

Name

Number

@R or

(ER) +

Autoincrement
o

The contents of
the
register
specified
as
(ER)
are
incremented after
being
used
as the address of the operand.

@(ER)+
|

Autoincrement
Deferred

The register specified as
contains the
pointer
to
address
of
the operand;

(ER)

deferred

Meaning
Register
operand.

dress

after

contains

the

the ad-

operand.

(ER)

1is

the
:

(ER)
the
the

incremented

use.

- (ER)

Autodecrement

The contents of

the

@- (ER)

Autodecrement

The

contents of

the register

Index

én

The

expression

| E (ER)

Deferred

specified
as
(ER)
are
decremented
before being used
as the pointer to the
address
of the operand.

@E (ER)

Index Deferred

specified
as
(ER)
are
decremented before being
used
as the address of the operand.

plus

the

contents
of
the
register
specified as
(ER),
form
the
address of the operand.

The
expression
E,
plus
the
contents
of
the
register
specified
as
(ER),
vyield
a
pointer
to the address of the
operand.

Immediate

Q#E

B.3

The
expression
operand itself.

~Absolute

The

Relative

The address of the operand
E,
relative
to
the instruction,
follows the instruction.

Relative

Deferred

expression

address

the

' The address

the

operand.

of the operand is

pointed to by E whose address,

‘relative
to
the instruction,
follows the instruction.

ASSEMBLER DIRECTIVES

The MACRO-11 assembler directives
table.

For

detailed

are

description

B-2

summarized
of

each

the

directive,

following
the

table

MACRO-11

ASSEMBLY
LANGUAGE AND ASSEMBLER DIRECTIVES
AS

contains references to the appropriate sections in
manual.

Section

Form

3.3
3.6

body

the
|

Reference

the

Operation

A
single
followed
by

quote
(apostrophe)
one
ASCII
character

generates a word which contains

the

7-bit ASCII

the

representation

character in the low-order byte and
zero in the high-order byte.
This
character
is
also
used
as
a
concatenation
indicator
in
the
expansion
of
macro arguments (see
Section 7.3.6).
"

6.3.3

| A
double
gquote
followed by
two
| ASCII
characters
generates a word
which
contains
the
7-bit
ASCII
representation
of
the
two

~characters.
The first character is
stored
in the low-order byte;
the
- second character is stored
in
the
high-order byte.

"Bn

6.4.1.2
- 5

- "Cexpr

6.4.2.2

6.4.1.2

"Fn

radix

the
wvalue
n
binary number.

“Dn

| Temporary

the
expression's value
complemented.

| Temporary

radix

wvalue

causes

treated

Temporary numeric control;
the

causes

to be ones-

control;
be

causes

treated as a

Temporary numeric control;

the
wvalue
n
to
decimal number.

6.4.2.2

control;

causes

treated as

sixteen-bit floating-point number.
“On

6.4.1.2
'

Temporary
the
value
octal

“"Reccce
.ASCII /string/

- 6.3.7
6.3.4

radix
control;
causes
n
to
be treated as an

number.

Convert ccc to Radix-50 form.
Generates
a
block
of
data
| containing
the ASCII equivalent of
the character string
(enclosed
in
' delimiting
characters),
one
character per byte.

MACRO-11 ASSEMBLY LANGUAGE AND ASSEMBLER DIRECTIVES

Section
Reference

Form

.ASCIZ

/string/

6.3.5

Operation

Generates

a
block
containing
the ASCII
the character string

delimiting
character per

.ASECT

characters),
byte,
with
the

specified

Begin

the

absolute

resume

6.5.3

Reserves a block of
storage
space
whose length in words is determined
by the specified expression.

.BYTE expl,exp2,..

Generates
each byte

.CSECT

[name]

Begin

.DSABL

arg

Disables the function
the argument.

.ENABL

arg

Enables
(invokes)
the
specified by the argument.

[exp]

Indicates the logical
end
source
program.
The

.END

.ENDM

.ENDR

successive bytes of data;
contains the value of the
corresponding specified expression.
or

resume

named

[name]

unnamed

relocatable
program section.
This
directive
is
provided
for
PDP-11
other
with
compatibility
assemblers.
specified

function

of
the
optional
argument
specifies
the
transfer
address
where program execution is
to begin.

Indicates the end
assembly block.

.ENDC

one
=zero

Reserves a block of
storage
space
whose length in bytes is determined
by the specified expression.

exp

.BLKW exp

byte
terminating
string.

program section.
.BLKB

of
data
equivalent of
(enclosed
1in

Indicates

the

end

conditional

the

current

repeat
block,
indefinite
repeat
block, or
macro
definition.
The
optional
name,
1if
used,
must be
identical to the name specified
1in
the macro definition.
Indicates

the

end

the

current

repeat
block.
This
directive is
provided
for
compatibility
with
other PDP-11 assemblers.
.EOT

Ignored;
indicates
end-of-tape
(which is detected automatically by

the hardware).
compatibility
assemblers.

B-4

is included
for
with
earlier

MACRO-11 ASSEMBLY LANGUAGE AND ASSEMBLER DIRECTIVES

Form

Section
Reference

Operation

.ERROR exp;text

7.5

User-invoked
error
directive;
causes
output
to the listing file
or
the
command
output
device
containing
the optional expression
and the
statement
containing
the
directive.

.EVEN

6.5.1

Ensures that the
current
location
counter contains an even address by
adding 1 if it is odd.

6.9

Defines

.GLOBL

syml,sym2,...

global

symbol(s)

specified

.IDENT /string/

6.1.5

Provides a means
of
1labeling
the
object
module
with
the
program
version number.
The version number
is
the
Radix-50
string appearing
between
the
paired
delimiting
characters.

.IF

©.10.1

Begins a conditional assembly block
of source code which is included in
the assembly
only
if
the
stated
condition
1is
met
with respect to
the argument(s) specified.

6.10.2

Appears only within
a
conditional
assembly
block,
indicating
the
beginning of a section of
code
to
be
assembled if the condition upon
entering the block tests false.

cond,argl

IFF

LIFT

6.10.2

Appears only within
assembly
block,

conditional
the

indicating
section of
code

“

the

symbol (s).

beginning of a
be
assembled if the condition
entering the block tests true.

IFTF

06.10.2

Appears only within

conditional

assembly
block,
indicating
beginning of a section of
<code
be assembled unconditionally.

.IIF cond,arg,
statement

to
upon

the
to

Acts
as
a
1l-line
conditional
assembly block where the
condition
is
tested
for
the
argument
specified.
assembled
tests

B-5

true.

The
only
1f

statement
is
the
condition

MACRO-11 ASSEMBLY LANGUAGE AND ASSEMBLER DIRECTIVES

Section
Reference

Form

.IRP

Indicates
the
beginning
of
an
indefinite repeat
block
in
which
the
symbol
specified
1is replaced
with
successive
elements
of
the
real
argument list enclosed within
angle brackets.

sym,

<argl,arg2,...>

.IRPC

Operation

sym,<string>

Indicates

the
beginning
indefinite
repeat
block
the specified symbol takes

value
of
optionally
brackets.

Task
high
.LIST

successive
characters,
enclosed
within
angle

Reserves two words into

.LIMIT

[arg]

.MACRO name,argl,
arg2,...

6.1.1

of
an
in which
on
the

which

the

Builder
inserts
the low and
addresses of the task image.

Without
an
argument,
the
LLIST
directive
increments
the
1listing
level
count
by
1.
With
an
argument,
this
directive does not
alter the listing level count,
but
formats
the
assembly
listing
according
to
the
argument
specified.
Indicates

the

definition
name
and

having
the

arguments.

start

a macro

the
specified
following
dummy

Specifies the symbolic names of the
user
or
system
macro definitions
required in
the
assembly
of
the
current user program, but which are
not defined within the program.

Causes an

.MEXIT

macro
repeat

.NARG symbol

exit

from

expansion
block.

the

current

indefinite

Can
appear
only
within
a
macro
definition;
equates the specified
symbol to the number
of
arguments

the

macro call

currently being

expanded.

.NCHR

NLIST

symbol,<string>

[arg]

Can appear
program;
specified
characters
Without

directive

anywhere
equates
to
the

the

1in
a
source
the
symbol
number
=
of

specified string.

argument,

decrements

the

.NLIST

listing

level
count
by
1.
With
an
argument, this directive suppresses

that
portion
of
the
specified by the argument.

B-6

listing

MACRO-11 ASSEMBLY LANGUAGE AND ASSEMBLER DIRECTIVES
Section
Reference

Form

.NTYPE

symbol,aexp

7.4.3

Operation

Can
appear
definition;

only
within
a
macro
equates the symbol to

the 6-bit addressing
mode
of
specified address expression.

the

.ODD

6.5.2

Ensures that the
current
1location
counter
contains an odd address by
adding 1 if it is even.

. PAGE

6.1.6

Causes the assembly listing to skip
to the top of the next page,
increment the page count.

.PRINT exp;text

7.5

and

User-invoked
message
directive;
causes
output
to the listing file
or
the
command
output
device
containing
the optional expression
and the
statement
directive.

containing

the

.PSECT name,attl,...
attn

6.8.1

Begin
or resume a named or unnamed
program
section
having
the
specified attributes.

.RADIX n

6.4.1.1

Alters the current program radix to
n,

where n

10.

.RAD50 /string/

6.3.6

Generates
a
block
of
data
containing
the Radix-50 equivalent
of the
character
string
enclosed
within delimiting characters.

.REPT exp

7.7

Begins a repeat block;
causes
the
section
o0of
code
up
to
the next
.ENDM
or
L.ENDR
directive
to
be
repeated
the
number
of
times
specified as exp.

.SBTTL string

6.1.4

Causes the specified string
to
be
printed
as
part
of
the assembly
listing page
header.
The
string
component
of each .SBTTL directive
is
collected
1into
a
table
of
contents
at
the
beginning of the
assembly listing.

.TITLE string

6.1.3

Assigns

the

first

six

Radix-50

characters
in
the
string
as
an
object module name and
causes
the
string
to
appear
on each page of
the assembly listing.

.WORD expl,exp2,..

6.3.2

Generates successive words of data;

each word contains the value of the
corresponding specified expression.

APPENDIX C

PERMANENT SYMBOL TABLE (PST)

The

permanent

symbol

table

(PST)

contains

those

symbols

which

are

automatically
recognized
by MACRO-11l.
These symbols consist of both
op
codes
and
assembler
directives.
The
op
codes
(i.e.,
the
instruction
set)
are
1listed first, followed by the directives which
cause specific actions during assembly.
For a detailed description of
PDP-11 Processor Handbook.

C.l1

the

instruction

set,

see

the

appropriate

OP CODES

OCTAL
MNEMONIC

VALUE

FUNCTIONAL

ADC
ADCB
ADD
ASH
ASHC
ASL
ASLB
ASR
ASRB
BCC
BCS
BEQ
BGE
BGT

005500

Add Carry

105500
060000
072000
073000
006300
106300
006200
106200
103000
103400
001400
002000
003000

Add Carry (Byte)
Add Source To Destination
Shift Arithmetically
Arithmetic Shift Combined
Arithmetic Shift Left
Arithmetic Shift Left (Byte)
Arithmetic Shift Right
Arithmetic Shift Right (Byte)
Branch If Carry Is Clear
Branch If Carry Is Set
Branch If Equal
Branch If Greater Than Or Equal
Branch If Greater Than

BHI
BHIS
BIC

101000
103000
040000

Branch If Higher
Branch If Higher
Bit Clear

BICB
BIS
BISB
BIT
BITB
BLE
BLO
BLOS
BLT

140000
050000
150000
030000
130000
003400
103400
101400
002400

Bit Clear (Byte)
Bit Set
Bit Set (Byte)
Bit Test
Bit Test (Byte)
Branch If Less Than Or Equal
Branch If Lower
Branch If Lower Or Same
Branch If Less Than

NAME

Same

PERMANENT SYMBOL TABLE

(PST)

MNEMONIC

OCTAL
VALUE

BMI
BNE
BPL
BPT
BR
BVC
BVS

100400
001000
100000
000003
000400
102000
102400

Branch If Minus
Branch If Not Equal
Branch If Plus
Breakpoint Trap
Branch Unconditional
Branch If Overflow Is Clear
Branch If Overflow Is Set

CMPB

120000

Compare Source To

COM
COMB

005100
105100

Complement Destination
Complement Destination

DEC
DECB

005300
105300

Decrement Destination
Decrement Destination

DIV
EMT
FADD

071000
104000
075000

Divide
Emulator Trap
Floating Add

FDIV
FMUL
FSUB

075030
075020
075010

HALT

000000

Halt

INC
INCB

005200
105200

Increment Destination
Increment Destination

10T
JMP
JSR
MARK

000004
000100
004000
006400

Input/Output Trap
Jump
»
Jump To Subroutine
Mark

MFPS

106700
~

Instruction Space
Move from PS
(LSI-11)

MOV
MOVB

010000
110000

MTPI

006600

MTPS

106400

‘MUL
NEG
NEGB
NOP
RESET
RETURN
ROL
ROLB
ROR

070000
005400
105400
000240
000005
000207
006100
106100
006000

CALL
CCC
CLC
CLN
CLR
CLRB
CLV
CL%Z
CMP

004700
000257
000241
000250
005000
105000
000242
000244
020000

FUNCTIONAL

NAME

Jump To Subroutine (JSR PC,xxXx)
Clear All Condition Codes
Clear C Condition Code Bit
Clear N Condition Code Bit
Clear Destination
Clear Destination (Byte)
Clear V Condition Code Bit
Clear 7Z Condition Code Bit
Compare Source To
Destination

Destination

(Byte)

MFPI

006500

Floating Divide
Floating Multiply
- Floating Subtract

(Byte)

Move From Previous

Move Source To Destination
Move Source To Destination
(Byte)
Move To Previous
Instruction Space
Move to PS
'
(LSI-11)
Multiply

Negate Destination
Negate Destination

(Byte)

No Operation
Reset External Bus

Return From Subroutine
Rotate Left
Rotate Left

Rotate Right

(Byte)

(RTS PC)

PERMANENT SYMBOL TABLE (PST)
OCTAL
MNEMONIC

VALUE

RORB

106000

RTI

000002

FUNCTIONAL NAME

Rotate Right (Byte)
Return From Interrupt

(Permits a trace

trap)

RTS
RTT

SBC

SBCB

SCC

SEC
SEN
SEV
SEZ
SOB
SUB
SWAB
SXT
TRAP

TST
TSTB
WAIT
XOR

000200
000006

Return From Subroutine

005600
105600

Subtract Carry

000277
000261
000270
000262
000264
077000
160000
000300
006700
104400
005700
105700
000001
074000

Return From Interrupt

(inhibits trace trap)

Subtract Carry (Byte)
Set All Condition Code Bits
Set C Condition Code Bit
Set N Condition Code Bit
Set
V Condition Code Bit
Set Z Condition Code Bit
Subtract One And Branch
Subtract Source From
Destination
Swap Bytes

Sign Extend
Trap

Test Destination
Test Destination

‘
(Byte)

Wait For Interrupt
Exclusive OR

OP CODES FLOATING POINT PROCESSOR ONLY
OCTAL

FUNCTIONAL NAME

MNEMONIC

VALUE

ABSD
ABSF
ADDD
ADDF
CFCC

170600
170600
172000
172000
170000

Make Absolute Double
Make Absolute Floating
Add Double

CLRD

LDCDF

170400
170400
173400
173400
174400
174400
177400

Clear Double
Clear Floating
Compare Double
Compare Floating
Divide Double
Divide Floating
Load And Convert From

LDCFD

177400

Load And Convert From
Floating To Double

LDCID

177000

LDCIF

177000

LDCLD

177000

LDCLF

177000

LDD

172400
176400

CLRF
CMPD
CMPF
DIVD
DIVF

LDEXP

Add Floating

Copy Floating Condition
Codes

Double To Floating

Load And Convert Integer To
Double
Load And Convert Integer To
Fleoating
Load And Convert Long
integer To Double
Load And Convert Long
Integer To Floating
Load Double
Load Exponent

Cc-3

PERMANENT SYMBOL TABLE

(PST)

OCTAL
MNEMONIC

VALUE

FUNCTIONAL NAME

LDF

172400

Load

Floating

LDFPS

170100

Load

FPPs

Program Status

MFPD

106500

Move

From

171400

Space
Multiply

And

Integerize

171400

Double
Multiply

And

Integerize

MODD
MODF

Data

Floating
MTPD

106600

Move

MULD
MULF

171000
171000

Multiply Double
Multiply Floating

NEGD
NEGF

170700

Negate
Negate

170700

Data

Space

Double
Floating

SETD

170011

Set

Double

SETF

170001

Set

Floating

SETI

170002

Set

Integer

SETL

170012

Set

Long

SPL

000230

Set

Priority

STCDF

176000

Store

Mode
Mode

Mode

Integer

And

Mode

Level

Convert

From

Double

C.2

To Floating
And Convert From

STCDI

175400

Store

STCDL

175400

Double To Integer
Store And Convert From

STCFD

176000

Double To Long Integer
Store And Convert From

STCFI

175400

Floating To Double
Store And Convert From

STCFL

175400

Floating To Integer
Store And Convert From

STD

174000

Floating To
Store Double

STEXP

175000

Store

Exponent

STF
STFPS

174000
170200

Store
Store

Floating
FPPs Program

STST

170300

Store

FPPs

SUBD
SUBF

173000
173000

Subtract
Subtract

TSTD

170500

Test

Double

TSTF

170500

Test

Floating

MACRO-11

Long

Integer

Status

Double
Floating

DIRECTIVES

DIRECTIVE

FUNCTIONAL

SIGNIFICANCE

.ASCII

Translates

character

string

to ASCII

.ASCIZ

Translates

character

string

.ASECT

inserts

zero

byte

Begins

absolute

Reserves

specified
. BLKW

Reserves

byte

other

block

.CSECT

word

block

section

PDP-11

(provided

for

assembliers).

accordance

with

wvalue

accordance

with

value

Generates

successive

specified

arguments.

Begins

equivalents;

argument.

specified argument.
.BYTE

equivalents.

character.

program

compatibility with
. BLKB

last

ASCII

relocatable

compatibility with
C-4

byte

data

program

other

section

PDP-11

accordance

(provided

assemblers).

with

for

(PST)

PERMANENT SYMBOL TABLE

FUNCTIONAL SIGNIFICANCE

DIRECTIVE
. DSABL
.ENABL
.END

Disables specified function.
Enables specified function.
Defines logical end of source program.

.ENDC
. ENDM

Defines end of conditional assembly block.
Defines end of macro definition, repeat block,

. ENDR

for
Defines end of current repeat block (provided
compatibility with other PDP-1l1l assemblers).
Define End of Tape condition (ignored).
file or
to 1listing
Outputs diagnostic message

.EOT

. ERROR

indefinite repeat block.

command output device.
Word—aligns the current location counter.

Declares global
Labels

object

attribute

module

for specified

with

specified

symk

program

version number.

Begins conditional assembly block.

Begins

subconditional

assembly

block

(1f

Begins

subconditional

assembly

block

(1f

conditional assembly block test is false).

.IFTF

LIIF

.IRP

conditional assembly block test is true).
Begins
subconditional
assembly
block
(whether
conditional assembly block test is true or false).
Assembles immediate conditional assembly statement
(if specified condition is satisfied).
Begins
indefinite
repeat
block;
replaces

symbol

specified

. IRPC

.LIMIT

.LIST

block:;
repeat
of
value
with

indefinite
Begins
symbol
specified

characters in specified string.
Reserves two
words
of storage
addresses of task image.

Controls

assembly

1listing

listing.

definition.
- MCALL

specified successive real

with

arguments.

Identifies

1level

<count

.NARG
.NCHR

.NLIST
.NTYPE
.ODD

. PAGE

format

and

required

macro

definition(s)

Exit from current macro definition
repeat

high and low

.MACRO Denotes start of macro

assembly.
.MEXIT

for

replaces
successive

for

indefinite

block.

Equates
specified
symbol
to
the
number
of
arguments in the macro expansion.
of
number
the
to
symbol
specified
Equates
characters in the specified character string.
suppresses
and
count
1level
1listing
Controls
specified portions of the assembly listing.
mode
addressing
Equates specified symbols to the
of the specified argument.
Byte—aligns the current location counter.
Advances form to top of next page.

« PRINT
. PSECT

Prints specified message on command output device.
specified
Begins specified program section having

.RADIX
.RADS50

Changes current program radix to specified radix.
equivalents
Radix-50
Generates data block having

+.REPT

Begins repeat block and replicates it according to
the value of the specified expression.

attributes.

of specified character

C-5

string.

PERMANENT SYMBOL TABLE

DIRECTIVE

.SBTTL

.TITLE

MACRO-11l

Prints

specified

assembly

listing

specified

title

the

Prints

SIGNIFICANCE

subtitle

text

page
text

the

second

line

object

module

name

header.
as

in
the
first
1line
of the assembly listing page
header.
Generates successive word data in accordance
with
specified arguments.

. WORD

The

FUNCTIONAL

(PST)

directives

in Appendix B.

listed

above

are

summarized

greater

detail

APPENDIX

DIAGNOSTIC ERROR MESSAGE SUMMARY

D.1

MACRO-11] ERROR CODES

A diagnostic error code is printed as the first character in a source
line which contains an error detected by MACRO-11l. This error code
identifies a syntactical problem or some other type of error condition
detected during the processing of a source line. An example of such a
source line

is shown below:

26 000236

MOV R1,R2,A

010102

The extraneous argument A in the MOV instruction above causes the line
to be flagged with a Q

(syntax)

error.

Meaning

Error Code

Assembly error.

Because many different

types

conditions produce this diagnostic message,
error
all the possible directives which may yield a
general assembly error have been categorized below
to reflect specific classes of error conditions:

" CATEGORY 1l:

ILLEGAL ARGUMENT SPECIFIED.

.RADIX -- A value other than

specified as a new radix.

.LIST/.NLIST -- Other than

argument

directive.

(see Table 6-1)

legally

defined

is specified with the

1legally defined
.ENABL/.DSABL -- Other than a
(see Table 6-2) is specified with the
argument

directive.

.PSECT -- Other than a legally-defined argument
the
specified with
is
6-3)
Table
(see

directive.

defined
legally
a
than
.IF/.IIF -- Other
test (see Table 6-5) or an illegal
conditional
argument expression value is specified with the
directive.

.MACRO -- An illegal or duplicate symbol

in dummy argument list.

found

DIAGNOSTIC ERROR MESSAGE

Error

Code

(Cont'd)

SUMMARY

Meaning

CATEGORY 2:
-TITLE

NULL ARGUMENT OR SYMBOL SPECIFIED.

Program

directive,
following

or
the

character.

-IRP/.IRPC
the

name

-- No
directive.

dummy

.NARG/.NCHAR/.NTYPE
in

the

directive.

IF/.IIF -- No
specified in the

not

first
directive

argument

-- No

3:
UNMATCHED
CONSTRUCTION.

symbol

an illegal
an illegal
directive.
.NCHAR

match,

CATEGORY

ILLEGAL

FORWARD REFERENCE.

This type of error results
possible conditions:
l1.

from either

global
assignment
(symbol==expression)
contains
reference to another symbol.

expression defining

the

current location counter

two

statement

value

contains a

forward

the

forward

reference.

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

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

End directive not found.
When the end-of-file
1is
reached during source input and the .END directive
has not yet been encountered,
MACRO-1l1l
generates
this
error
code,
ends
assembly
pass
1,
and
proceeds with assembly pass 2.

Illegal

character

which
are
also
question mark (?)
is then ignored.
L

Input

line

length.

detected.

Illegal

characters

non-printable
are replaced by a
on the listing.
The
character

is greater

Currently,

than

132(10)

characters

1in

this error condition is caused

only
through
excessive
substitution
of
real
arguments for dummy arguments during the expansion
of a macro.

Multiple definition

1label.

label

was

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

A number contains a

digit

Opcode
error.
Directive
Permissible
nesting
1level

current program radix.
a decimal value.

that

The number

out
depth

not

the

is evaluated as

of
context.
for conditional

assemblies has been exceeded.
Attempt to expand a
macro which was unidentified after .MCALL search.

'DIAGNOSTIC

Error

ERROR MESSAGE SUMMARY

Code

Meaning

Phase

error.

from

one

label's

assembly

definition

pass

another

value

varies

a multiple

definition of a local symbol has occurred within a
local
symbol block. Also, when in a local symbol
block defined by
the
.ENABL
LSB
directive,
an
attempt has occurred to define a local symbol in a
program section
other
than
that
which
was
1in
effect when the block was entered.
A P error code
also appears if an .ERROR directive is assembled.
Q

Questionable
many
scan

syntax.

arguments
was

not

are

Arguments

are

specified,

missing,
the

too

instruction

completed.

error.

invalid

use

reference
to
a
register
has
been
made, or an
attempt has
been
made
to
redefine
a
standard
register
symbol
without first issuing the .DSABL
REG directive.
T

Truncation

error.

bits
1in
a
word,
than 8 significant
.BYTE directive or
U

Undefined

encountered

number

generated

than

or an expression generated more
bits
during
the
wuse
of
the
trap (EMT or TRAP) instruction.

symbol.

undefined

Instruction

error.

The

the

symbol

evaluation

was

an
expression;
such an undefined symbol is
assigned
a
value of zero.
Other possible conditions which
result in
this
error
code
include
unsatisfied
macro
names in the list of .MCALL arguments and a
direct
assigment
(symbol=expression)
statement
which
contains
a
forward
reference to a symbol
whose
definition
also
contains
a
forward
reference;
also,
a
local
symbol may have been
referenced that does
not
exist
in
the
current
local symbol block.
Z

dur ing

instruction

not
compatible
among
all
family.
See Section 5.3 for

members of
details.

flagged

the

1is

PDP-11

APPENDIX E

SAMPLE CODING STANDARD

E.1

INTRODUCTION

Standards
eliminate
wvariability
and
the
requirement
to
make
a
decision.
Much of the difficulty in establishing standards stems from
the notion that they should be optimal. However, to be
successfully
applied,
standards must represent an agreement on certain aspects of
the programming process.

This Appendix contains DIGITAL's PDP-11 Program Coding
Standard.
It
is suggested that this be used as a model to assist users in preparing
standards for their own installations.

E.2

LINE

FORMA’

All source lines shall consist of from one
to
a
maximum of
exghty
characters
{(not
including
the
audit
trail
added by SLIPR (SLP in
RSX-11M) editor.
This program is described in the applicable
RSX-11M
or RSX-11D Utilities Manual or in the IAS Editing Utilities Reference

Manual (see Section 0.3 in the Preface).

Assembly language code lines shall have the following format:
1.

Label Field - if present,

Operation
1 (column
Operand

the label shall start at tab stop 0

field - the operation
9).

field

the

operand

field shall start at

field

shall

start

tab

tab stop

stop

(column 17).

Comments field - the comments field shall start at tab stop 4
(column

33)

and may continue to column 80.

Comment lines that are included in the code body shall be delimited by
a
line <c¢ontaining only a
leading
semicolon.
The comment itself
contains a 1@@&1»@ semac@lnn and starts in column 3.
Indents shall be
fi. étf%h -

If the operand field extends beyond tab
stop
4
(column
33)
mp
leave
a
space
and
start
the
comment.
Comments which apply to an
instruction but reguire continuation should always line
up
with
the
character position which started the comment.

SAMPLE CODING STANDARD

E.3

COMMENTS

Comment

all coding to convey the global role of an instruction, rather
simply a literal translation of the instruction into English.
1In
general this will consist of
a
comment
per
1line
of
code.
If
a

than

particularly
difficult,
obscure,
used, a paragraph
of comments must
code.
Preface
text,
which
variables, etc., will
start

the

automated

start

column

and

is
of

describes
formats,
algorithms,
program-local
be delimited by the character sequence ;+ at the
;-

the

narrative

end;

these

commentary.

delimiters
The

comment

facilitate
itself

will

THE

A LIST

example:

APPLIES

For

text

extraction

or elegant instruction sequence
immediately precede that section

INVERT
OF

ROUTINE

ACCEPTS

RANDOM NUMBERS

AND

THE

KOLMOGOROV ALGORITHM
ALPHABETIZE THEM.

E.4

NAMING

E.4.1

STANDARDS

E.4.1.1
permitted

General Purpose
as
register

Registers - Only
names;
and
may

the
not

purpose:

RU=%0

;REG

R1=%1
R2=%2

;REG
; REG

R3=%3
R4=%4
R5=%5

;REG
;REG
sREG

SP=%0

;STACK

PC=%7

E.4.1.2

Hardware
identically to the

E.4.1.3

Device

following
names
are
be used for any other

0
1
2

3
4
5

POINTER (REG 6)
; PROGRAM COUNTER (REG 7)

Registers - These
hardware definition.

registers

For

must
example, PS

be
named
and SWR.

Registers - These are symbolically
named
identically
hardware
notation.
For example, the control status register
for the RK disk is RKCS.
Only this symbolic name may be used to refer
to this register.
to

the

SAMPLE CODING STANDARD

E.4.2

Processor Priority

Testing or altering the processor priority is done using the

symbols

which are equated to their corresponding priority bit pattern.

E.4.3"

Other

Symbols

Frequently-used

bit

patterns

such

conventional symbolics on an as-needed basis.

E.4.4
The

and

will

made

assembler.

All

Using the Standard Symbolics

other

standards will

defined

standard symbols will appear

within

the

in a file and will be linked prior

to program execution.

E.4.5

Symbols*

E.4.5.1
Global Symbols Global symbols should be easily
recognized
by
their format.
The following standards apply and completely define

symbol standards for PDP-11 Medium/Large software products.

symbol

pos-1

non-glbl-sym

letter | a-num/|
null

glbl-sym

S/.

pos-3

pos-4

pos-5

pos-6

length

a-num/|
null

>=1

a-num/|

null

a-num/|
null

>=3

a-num/|

>=4

null

glbl-offset

letter|

$/.

a-num | a-num/|
null

glbl-bit-ptrn

letter|

a-num

S/.

local-sym

number |

pos-2

a-num/|

>=2

Symbols that are branch targets are
always use the term "symbol".
Number

*** The

use of

software.

the

$ or

range

for

also called

labels,

but we

will

0<number<65535.

global

names

reserved

for

DEC-supplied

SAMPLE CODING STANDARD

where:

a—-num
|
non-glbl-=sym

is an alphanumeric character.
are non—-global symbols.

local-sym

local

glbl-sym
glbl-offset

symbols, as defined by
MACRO-11.
are global symbols (addresses).
are global offsets (absolute

glbl=-bit-ptrn

are global bit patterns.

quantities).

program never

E.4.5.2

Symbol

Non-Global

contains

.GLOBL

statement

without

showing

cause.

Examples

Symbols

AlB

ZXCJ1
INSRT

Global

Address

Symbols

SJIM
. VECTR

S$SEC
Global

Absolute

Offset

Symbols

ASJIM

A$xT
A . ENT

Global

Bit

Pattern Symbols

A1$20

B3.6
JI.M
Local

Symbols

378
2718

E.4.5.3
Program—Local
(.+n)
1s
absolutely
other contexts must be

Symbols - Self-relative
address
arithmetic
forbidden
in
branch instructions;
its use in
avoided if at all possible and practical.

E-4

SAMPLE CODING STANDARD

Target symbols for branches that exist solely for positional reference
will use

local

symbols of

the

form

<num>$:

Use of non-local symbols is restricted, within reason, to those

cases

where
reference
to
the
code
occurs
external
to
the
code.
Local-symbols are formatted such that the numbers proceed sequentially
down

the page

and from page to page.

E.4.5.4
Macro
Names - The last
two
characters
(with
character possibly being null) have special significance.

last character

is a $,

the last,

a character specifying

the
last
The next to
the
mode
of

the macro.

For example,

in the three macro forms in-line,

stack,

and

p-section,

the
1in-line
form
has
no
suffix,
the
stack
has
an <S>,
p-section a <C>.
Thus the RSX Queue I/0 macro can be written

and the
as
any

QIOS

QIOSS
QIOSC
depending on the form required.
the

form of

E.5

E.5.1

the name

is standard.

These are not reserved letters.
'

Only

PROGRAM MODULES

General Comments on Programs

In our software, a program provides a single
distinct
function.
No
limits
exist
on size, but the single function limitation should make
modules larger than 1K a rarity.
Since any
software
may
eventually
exploit
the
wvirtual memory capacity of the 11/40 and 11/45, programs
should make every attempt to maintain a dense reference locus (do
not
promiscuously
branch
over
page
boundaries or over a large absolute
address distance).

All code is read-only.
contains
explanatory
read-write data.

E.5.2

Code and data
areas
are
distinct
and
each
text.
Read-only data should be segregated from

The Module Preface

Each program module in the system shall exist as a separate file.
The
file
name will reflect the name of the module and the file type shall
be of the form 'NNN'.
The 'NNN' signifies
the
edit
number
or
the
version
number.
The version number shall be changed only when a new
base level is created.
Furthermore, if no corrections are made
to
a
file
from
one base level to the next, the version number will not be
changed.
The availability of File Control Services and
File
Control
Primitives
will greatly simplify version number maintenance.
Program
modules
adhere to
a
strict
format.
This
format
adds
to
the
readability
and
understandability
of
the
module.
The
following
sections are included in each module:

E=>5

SAMPLE CODING STANDARD

For

the Code
lfl

Section:

.TITLE

statement

a
module
used.

that

contains

An .IDENT
statement
PDP-11 version number
A

.PSECT

the

statement

module

than

the

one

name

specifying
the
standard appears

that

defines

routine,

the

statement,

(C)

SOFTWARE
ON

ONLY

WITH

NOTICE.

CORPORATION,

version
number.
in section E.10.

program

INCLUSION

PROVIDED

ANY

OTHER PERSON

MAYNARD,

UNDER A

COMPUTER

SOFTWARE,

NOT

OWNERSHIP

The

which

AND MAY

THE

ABOVE

OR ANY OTHER

OTHERWISE

EXCEPT

LICENSE

SYSTEM

FOR USE

THE

SOFTWARE

SHALL

THEREOF,

AVAILABLE

SUCH

SYSTEM AND

TITLE TO

AT ALL

USE

COPIED

COPIES

MADE

FOR
BE

TIMES

AND

REMAIN

DEC.

THE

INFORMATION

WITHOUT

DEC

NOTICE

COMMITMENT

THIS

AND

BY DIGITAL

DOCUMENT

SHOULD

RESPONSIBILITY

ITS

SUPPLIED

version

The PDP-11
E.10.

number

version

and

A brief

CHANGE

CONSTRUED AS A

SOFTWARE

FOR

THE

USE

EQUIPMENT WHICH

DEC.

the

number

The name of the principal
module was first created.

The
name
of
modification.

SUBJECT

EQUIPMENT CORPORATION.

ASSUMES

NOT

RELIABLILITY

Note:

section

MASS.

ONE WHO AGREES TO THESE LICENSE TERMS.

line

may

the disclaimer.

FURNISHED

THE
THIS

and

SINGLE

MAY

The

module.

1976

EQUIPMENT

ONLY

the

subtitles

resides.

THIS

specifies

file.

standard

author

and

described

the

date

section

which

the

each
modifying
author
and
the
date
of
Names
and
modification dates appear one per

chronological

statement

the

order.
function of

the module.

Items 1-8 should appear on the same page.

A list of the definitions of all equated local
symbols used
in
the module.
These definitions appear one per line and in
alphabetical order.
10.

All

by
-

11.

local

macro definitions,

preferably

alphabetical

name.

order

All

local

data.

The

data

should

indicate

a.
b.
c.

Description of each element (type, size, etc.)
Organization (functional, alpha, adjacent, etc.)
Adjacency requirements

SAMPLE CODING STANDARD

12.

A more

detailed definition of

the

function of

13.

A list of the inputs expected by the module.
the
calling
sequence
if
non-standard,
settings, and global data settings.

14,

This

list of the outputs produced as a result of
module.
These
1include
delivered
results,

settings,
visible

15.

list
result
in the

but not

the

side

The

(A1l

code

entering

these

this
condition code

outputs

are

caller.)

all effects (including side effects) produced as
a
of entering this module.
Effects include alterations
state of the system not
explicitly
expected
1in
the

calling sequence,
16.

effects.

includes

condition

E.5.3

the module.

those not

visible

to the caller.

module code.

Formatting

the Module

Preface

Rules:

The

first eight

items

appear

have
explicit
headings.
p-section is being used.
2.

Headings

start

indented

the

same

page

and

at the
left
tab position.

margin¥;

template

for

which

the

not

module

apply

may

preface

will

not

if the blank

descriptive

Items 7-14 will have headings which start at the
preceded and followed by lines containing only a
Items

Item 3 may be omitted

text

1is

left margin,
leading <;>.

omitted.

follows.

FILE-EXAMPL.SO1
EXAMPLE

.IDENT

/01/

.PSECT

KERNEL

.TITLE

DIGITAL

(C)

1976
COPORATION,

MAYNARD,

MASS.

MADE

SYSTEM AND TO ONE

WME
Ul

ANY OTHER COPIES THEREOF,

TO AND OWNERSHIP OF THE

THIS SOFTWARE IS FURNISHED UNDER A LICENSE FOR USE ONLY ON A
SINGLE COMPUTER SYSTEM AND MAY BE COPIED ONLY WITH THE
INCLUSION OF THE ABOVE COPYRIGHT NOTICE.
THIS SOFTWARE, OR

AVAILABLE

MAY NOT BE

TO ANY OTHER PERSON

WHO AGREES

TO THESE

SOFTWARE

FOR USE

LICENSE

ON SUCH

TERMS.

SHALL AT ALL TIMES

TITLE

REMAIN

DEC.

PROVIDED OR OTHERWISE

EXCEPT

THE

INFORMATION

THIS

NOTICE AND
SHOULD NOT BE
EQUIPMENT CORPORATION.

DOCUMENT

SUBJECT TO CHANGE WITHOUT

CONSTRUED AS

A COMMITMENT

DIGITAL

NEF

EQUIPMENT

*The
text

left margin consists of a <;> a
of the heading begins in column

E-7

<space>
3.

then

the heading,

the

SAMPLE CODING STANDARD

ASSUMES NO RESPONSIBILITY FOR THE
SOFTWARE ON
EQUIPMENT WHICH
IS

USE OR RELIABILITY OF
NOT SUPPLIED BY DEC.

DEC
ITS

VERSION

1-JAN-72

JOE PASCUSNIK
BY:

%NS

MODIFIED

DOE

21-JAN-73

RICHARD

12-JUN-73

SPENCER THOMAS
statement

the

module's

function

Brief

SYMBOLS

EQUATED

equated

symbols

List

MACROS

LOCAL

Macros

‘g

Local

DATA

wWE

LOCAL

data

Module

function-details

INPUTS:

Local

inputs

OUTPUTS:

Description

of outputs

EFFECTS:

Description

effects

WME

Description

Begin Module Code

E.5.4

Modularity

No other
success
Software

characteristic has more impact on
the
of
a
system
than
does
modularity.
Engineering's products consists of
the

wultimate
engineering
Modularity for PDP-11
application
of
the

single=function
philosophy
described in section E.5.1,
to a set of calling and return conventions.

E.5.4.1
Calling Conventions
conventions must be observed.

(Inter-Module)

E-8

- The
|

and

following
|

adherence

calling

SAMPLE CODING STANDARD

Transfer

of Control

Macros will exist for call and return.
The actual transfer
will
be
wvia
a JSR PC
instruction.
For register save routines, a
JSR

Rn,SAVE will

be permitted.

The CALL macro
CALL

is:

subr-name

The RETURN macro

is:

RETURN

On entry, a subroutine minimally saves all registers
it
intends
to
alter
except
result
registers.
On exit it restores these
registers.
(State preservation is assumed across calls.)
Argument Passing

Any registers may be used,

but

their use should follow a coherent

pattern.
For
example, if passing three arguments, pass them in
RU, Rl and R2 rather than
RO,
R2,
R5.
Saving
and
restoring
occurs

E.5.4.2

1in one place.

Exiting -

All subroutine exits occur

through a single RETURN

macro.

E.5.4.3

Intra-Module Calling Conventions -

consistency

inter-module

E.5.4.4

favors

sequence.

calling

sequence

Success/Failure Indication -

The

Designer

identical

bit

optional,

will

that

but

of the

wused

return
the
success/failure
indicator,
where
success equals 0, and
failure equals 1.
The argument registers can be used to return values

additional success/failure data.

E.5.4.5
Module
Checking
Routines Modules
are
responsible
for
verifying
the
wvalidity of arguments passed to them.
The design of a
module's calling sequence should aim at minimizing the validity checks
by
minimizing invalid combinations.
Programmers may add test code to
perform additional checks during checkout.
All
code
should
aim
at
discovering
an error as close (in terms of instruction executions) to
its occurrence as possible.

E.6

E.6.1

FORMATTING STANDARDS

Program Flow

Programs will

page,

be organized on

the listing such

that they

even at the cost of an extra branch or jump.
E-9

flow down the

SAMPLE CODING STANDARD

For

example:

PROCESS

v
BBB

AAA

COMMON

shall

appear
TST
BNE

AAA:

BR
BBB:

CMN:

Rather

than:
TST
BNE

AAA:

:
CMN

BBB:

the

listing

as:

SAMPLE CODING STANDARD

E.6.2

Common Exits

A common exit appears as the last code sequence on the listing.
the

flow

chart:

appear on the listing as:

will
PR1l:

N
L

LR
I I

EXIT

PR4:

EXIT

PR3:

EXIT

PR2

EXIT

EXIT:

And

not

PR1:

EXIT:

PR2:

PR3:

ow »
-

EXIT

*®

EXIT

PR4:

»
"

EXIT
E-11

Thus

SAMPLE CODING STANDARD

E.6.3
Code

three

Code with

Interrupts

Inhibited

that is executed with interrupts inhibited, shall
semicolon (;;;) comment delimiter.
For example:

. .ERTZ:

10$:

E.7

; ENABLE

SYSTEM

SUBROUTINES,

INHIBIT
C

INTERRUPTS

#PR7,PS

:3;

BIT

#PR7,+2 (SP)

BEQ

10$

HE A ®

RTT

cene

“ oo

;i
HHH

ceon

“ees

HHH

.o e

HHH

753

PROGRAM SOURCE

flagged

RETURNING

;BY

BIS

E
N
T

FILES

Source creation and maintenance shall be done in base levels.
A
base
level
1is
defined
as
a point at which the program source files have
been
frozen.
From
the
freeze
point
to
the
next
base
level,
corrections
will
not be
made
directly
to
the base level itself.
Rather a file of corrections shall be accumulated for each file in
base
level.
Whenever
an
updated
source
file
1is
desired,
correction file will be applied to the base file.

the
the

The accumulation of corrections shall proceed until a logical breaking
point
has
occurred
(i.e.
a milestone or significant implementation
point has been reached).
At this
time
all
accumulated
corrections
shall
be
applied
to
the
previous
base level to create a new base
level.
Correction files will then be started for the new base level.

E.8

FORBIDDEN

INSTRUCTION

USAGE

The use of instructions or
previous instruction.
For

uses
to

MOV

@PC,Register

BIC

Src,Dst

the
be

bit
a

clear

very

index words
example:

instruction

"neat"

way

as a
to

literals

literal.

save

word

the

This

may

seem

but

what

about

maintaining a program using
this
trick?
To
compound
the
problem, it will not execute properly if I/D space is enabled
on the 11/45.
1In this case @PC is a D bank reference.
2.

The
to

use

the

transfer

MOV

instruction

program

control

instead
to

another

JMP

instruction

location.

For

example:
MOV

#ALPHA,
PC

transfers control to location ALPHA.
Besides
taking
1longer
to
execute
(2.3
microseconds for MOV vs.
1.2 for JMP) the
use of MOV instead of JMP makes it nearly impossible to
pick
up someone else's program and tell where transfers of control

E-12

SAMPLE CODING STANDARD

What if one would like to get a

take place.

Jjump

trace

the execution of a program (a move trace is unheard of)? As
a more general issue, perhaps even other operations such as
ADD
and SUB from PC should be discouraged.
Possibly one or
two words can be saved by using these operations but how many
such occurrences are there?

instructions
The seemingly "neat" use of all single word
where one double-word instruction could be used and would
execute faster and would not consume additional memory.

Consider the following instruction sequence:
CMP

-(R1), (=-R1)

CMP

-(R1),-(R1)

The intent of this instruction sequence is to subtract 8 from
This can be
(not to set condition codes).
register Rl
SUB
time via a
the
1/3
in approximately
accomplished
3.8 microseconds) at no additional cost
instruction (9.4 vs.
Another question here is also, what if Rl
in memory space.
is

o0dd?

SUB

properly and

always

wins

is always faster!

since

will

always execute

RECOMMENDED CODING PRACTICE

E.9

E.9.1

Conditional Branches

is
it
When using the PDP-11 conditional branch instructions,
the
and
signed
the
between
made
be
choice
correct
the
that
imperative
unsigned branches.

SIGNED

BGE
BLT
BGT
BLE

UNSIGNED

BHIS
BLO
BHI
BLOS

(BCC)

(BCS)

BGT) when comparing
A common pitfall is to use a signed branch (e.g.
All goes well until the two addresses have
two memory addresses.
(100000(8))
that is, one of them goes across the 16K
opposite signs;
This type of coding error usually shows itself as a result of
bound.
the
1in size of
a change
re-linking at different addresses and/or
program.

E.10

PDP-11 VERSION NUMBER STANDARD

The PDP-11 Version Number Standard applies to all

modules,

parameter

files, complete programs, and libraries which are written or caused to
It 1is
be written, as part of the PDP-11 Software Development effort.
used to provide unique identification of all released, pre-released,
and

in-house software.

It is limited in that, as currently specified, only six characters of
implementations of the Macro
Future
used.
are
identification
for at least nine
Assembler, linker, and librarian should provide
E-13

SAMPLE CODING STANDARD

characters,
and
will be enhanced

Version

possibly
twelve.
as the need arises.

Identifier

<form>

Used

{version>

<edit>

<patch>

a particular

form of

a module

LINK-1ll.

applicable,

One

the

null

(i.e.,

Used

alphabetic

binary

identify
Two

the

character,

not

release,

decimal

1in

case

wused,

and

used.
or

digits,

generation,

starting

00,

a
and

discretion of
the
project
in
what,
in
their opinion, is a

Used to identify the level to which
a
particular
release, or generation, of a program or module has
been
edited.
An
edit
1is
defined
to
be
an
alteration
to
the
source
form.
Two
decimal
digits, beginning at 01, and incremented with each
edit;
null if no edits.

<patch>

Used

to identify the level to which
a
particular
release, or generation, of a program or module has
been patched.
A patch is defined as an alteration
to
a
binary
form.
One
alphabetic
character,
starting at B, and running sequentially toward
2,
each
time
a set of patches is released;
null if

fields

{patch>

<edit>

interrelated.

are

When

bugs

have

Displaying

visible

the

been

the

<form>

following

released

in-house

field

identifier
=

have

program

separates

not

intended

{version>

null.

then

that

null,

when

Dbecause

been

appear

as:

<patch>,

identified:

frozen version

that 'X' corresponds roughly to
support, and 'V' to company support.

<patch>

<edit>

should

experimental version
test, pre-release, or in-house

Note

dash which

changed,

Identifier

version

Key Letters

1is

fixed.

the Version

output

<key—-letter>

and/or

<edit> must be reset to nulls.
It is
incremented, then <patch> w1ll be re-set

various

where

patches.

and

E.10.1

When

identification,

individual

from

<edit>

a version

then

the

release

support,

used

identifier

format

'Y'

only

version
to

group

<edit>

is displayed

is:

Program

<space><key-letter><form><version>-<edit><patch>
Name

standard

<edit>

These

part

this

where

The

that

identify

incremented at the
order
to
reflect
major change.

The

expected

program,

program.

the

SAMPLE CODING STANDARD

Examples:

PIP X03
LINK VB04-C
MACRO Y05-01

E.10.2

Use of the Version Number

in the Program

All sources must contain the version number in
an
L.IDENT
directive.
For
programs
(or
libraries)
which consist of more than one module,
each individual module will follow this version number standard.
The
version number of the program or library is not necessarily related to
the version numbers of
the
constituent
modules;
it
1is
perfectly
reasonable,
for
example,
that
the
first
version of a new FORTRAN
library, V00, contain an existing SIN routine, say V05-01l.
Parameter files are also required to contain the
.IDENT directive.
Because the assembler records
parameter files must precede the program.

version number in
an
the last .IDENT seen,

Entities which consist of a collection of modules or
programs,
e.g.,
the
FORTRAN
Library, will have an identification module in the first
position.
An
identification
module
exists
solely
to
provide
identification, and normally consists of something like:
;OTS

IDENTIFICATION

.TITLE

FTNLIB

. IDENT /003010/
. END

APPENDIX

ALLOCATING VIRTUAL MEMORY

This

appendix

intended

for

the

MACRO-1l user

who wants

avoid

the

problem
of thrashing, by optimizing the allocation of virtual memory.
Users of smaller
systems,
particularly
those
with
the
8K
subset
version
of
MACRO-1ll,
should
become
thoroughly
familiar
with the
conventions discussed herein.
In this regard,
Appendix
F
addresses
the following topics:
l.

General

hints

and

Macro definitions

Operational

space=saving

and

guidelines

expansions

techniques.

The user is assumed to have pursued a policy of

modular

programming,

as
advised
in
Appendix
E.
In
addition to the obvious advantages
accruing from small, distinct, highly-functional bodies of
code,
one
can
wusually
avoid
the problem of insufficient dynamic memory during
assembly by practicing such a policy.
Other
suggestions
as
to
how
available
memory
can be best utilized are discussed in the following
sections.

F.1l

GENERAL HINTS AND SPACE-SAVING GUIDELINES

Work-file memory is shared by a number of MACRO-1ll's tables,
each
of
which
1is allocated space on demand (64K words of dynamically pageable
storage are
available
to
the
corresponding entry sizes are as
l.

User-defined

symbols

Local

Program sections

Macro

names

Macro

text

Source

symbols

tables

and

their

five words.

four words.

four

six words.
words.

- six words.
scratch

Expression analysis

Object

code

The

- nine words.

files

In addition, several
process, as follows:

assembler).
follows:

pad

generation

tables are

used durlng

five words.
-

five words.

the

assembly

ALLOCATING VIRTUAL MEMORY

Macro

argument processing

.MCALL

argument

processing

The

above information can serve as
storage
requirements
and
for
requirements.

For

example,

the

use

local

encouraged,
since
their
dynamic storage than that
The

wusage

scope of
MACRO-11l

1local

five

words.

a
guide
for
estimating
dynamic
determining
ways
to
reduce
such

symbols

whenever

possible

can

often

blocks
through
Sections 3.5 and

maximized

loops

instances

when

indicator.
In

occur

If possible,

keeping with

the

during

temporarily

the
definition

the

same

conditional

assigned

symbol

symbols

the
.ENABL
6.2).

Since MACRO-1l1l does not support a purge function,
defined,
it
permanently
occupies
its
dynamic
Numerous

1is

internal
representation
requires
required for regular
user-defined

symbols

1local
symbol
directives (see

three words.

so used

extending

LSB/.DSABL

should be

and

count

repeat

offset

re-used.

principle,

special treatment should be
commonly-used
symbols.
Instead
of

the

LSB

once
a
symbol
is
memory
allocation.

assemblies

used

highly

25% less
symbols.

given
simply

appending a prefix file which defines all
possibly-used
symbols
for
each
assembly,
users
are
encouraged
to group symbols into logical
classes.
Each class so grouped can then
become
a
shortened
prefix
file or a macro in a library (see Section F.2 below).
In either case,
selective definition of symbolic assignments is achieved, resulting in
fewer defined (but unreferenced) symbols.
An appropriate example of
standard
symbols.
The

seen
in
the
definition
of
library, for example, supplies
several macros used to define
distinct
classes
of
symbols.
These
groupings and associated macro names are, as follows:

F.2

this idea
1is
system
macro

DRERRS

Directive

return

IOERRS

I/0

FILIOS

- File-related

SPCIO$

- Special

return

status

I/0

codes

codes
function

function

codes

MACRO DEFINITIONS AND EXPANSIONS

far, dynamic storage is used most heavily for the storage of
macro
text.
Upon
macro definition or the issuance of an .MCALL directive,
the entire macro body is stored, including all comments
appearing
in

the
macro
definition.
For
this
reason,
comments
should
not be
included as part of the macro text.
An RSX-1l1l utility program (called
SQZ
for
RSX-1lD
only)
and
a
Librarian
function switch (/SZ) are
available to compress macro
source
text
by
removing
all
trailing
blanks
and tabs, blank lines, and comments.
The system macro library
(RSXMAC.SML)

has

already

been

compressed.

User-supplied

macro

libraries
(.MLB)
and
macro
definition
prefix files should also be
compressed.
For additional information regarding
these
two
utility
tasks, consult the applicable RSX-11lM or RSX-11lD Utilities Manual (see
Section 0.3 in the Preface).
It

often

seems

expedient

each
assembly
to
however, may produce

append

macro

definition

provide
commonly-used macros.
the undesirable allocation
of

F-2

prefix

file

This
practice,
valuable
dynamic

ALLOCATING VIRTUAL MEMORY

storage
for
unnecessary
macros.
This side effect can be avoided by
specifying
that the
prefix
file
containing
the
macros
is
a
user-supplied macro library file (see Table 8-1).
This action imposes
the stipulation that the names of all desired macros must be listed as
arguments in the

.MCALL directive

(see Section 7.8).

Storage for macro text
can be
re-used effectively
by
redefining
certain
types of
macros to null after they have been invoked.
This

practice releases their dynamic memory for the storage of later
macro
text
and also eliminates the overhead and the need for dynamic memory
which would otherwise be required during the subsequent invocation and
expansion of
such
non-redefined macros.
The practice of redefining
macros to null applies mainly -to
those
that
only
define
symbolic
assignments,
as shown in the example below.
The redefinition process
may be

accomplished as
.MACRO

follows:

DEFIN

SYM1 = VALI
SYM2 = VALZ2

;DEFINE SYMBOLIC ASSIGNMENTS.

OFF1 = SYMBOL
OFF2 = OFF1+SIZl
OFF3 = OFF2+SIZ2

;DEFINE SYMBOLIC OFFSETS.

OFFN

= OFFM+SIZIM

.MACRO

DEFIN

;sMACRO NULL REDEFINITION.

. ENDM

.ENDM

DEFIN

Macros exhibiting this redefinition property
should
be
defined
(or
read
via
the
.MCALL
directive)
and invoked before all other macro
definition and/or .MCALL processing.
So doing ensures more
efficient
use

of dynamic memory.

The following system
macros
have
property after once being invoked:
DRERRS

- Directive

IOERRS -

return

the

status

automatic

null

codes

I/0 return status codes

FILIOS - File-related I/0 function codes
SPCIOS$ - Special I/0 function codes

CSIS

- Command String Interpreter codes and offsets

GCMLDS - Get Command Line codes and offsets
BDOFF$

- FCS buffer descriptor

FCSBTS

- FCS bit value codes

FDOFF$

- FCS

file descriptor

FSROFS$

- FCS

file

NBOFFS$S

- FCS

filename block

storage

offsets

block

region

offsets

(FSR)

offsets
F-3

offsets

redefinition

ALLOCATING VIRTUAL MEMORY
F.3

OPERATIONAL TECHNIQUES

When,
still

despite adhering to the guidelines discussed above,
performance
falls
below
expectations,
several additional measures may be

taken

improve

performance.

The first measure involves shifting the burden
of
symbol
definition
from
MACRO-11 to the linker.
1In most cases, the definition of system
I/0 and FCS symbols (and user-defined symbols of the same
nature)
is
not

necessary

defaulted
category

during

the

assembly

process,

to global references
(see
Section
4
of
error
code
A).
The linker

since

such

3.9
and
attempts

symbols

are

Section
D.1,
to resolve all

global
system

references from user-specified
default
libraries
and/or
the
object
library
(SYSLIB).
Furthermore,
by
applying
the
selective search option for object modules consisting only
of
global
symbol
definitions,
the
actual
additional
burden to the linker is

minimal.
A

second way of making more dynamic memory
available
only
one
output
file (either object or listing), as
The additional file descriptor block (FDB)
and
file
(FSR)

required

support

the

second

output

file

are

is
to
produce
opposed to two.
storage
region
allocated

from

available dynamic memory at the start of each assembly.
Furthermore,
the
size of the file storage region allocated is the minimum required
for the second (listing) output file.
For disk files, this is 264 (10)
words,

and

for

direct

line

printer

output,

74(10)

words.

The final way of increasing available dynamic memory is
related
only
to
the
operating environment.
Under RSX-11M, MACRO-11l allocates all
storage between its highest address and the end of
its
partition
as
dynamic
memory.
Consequently,
the amount of working storage can be
increased by installing and running MACRO-11l in a larger partition.
In

IAS

and

RSX-1lD,

the

assembler's

time.

dynamic

memory

fixed

1link

If
a
larger assembler is not available, you may build one
increasing the size of the task's
stack.
This
1is
accomplished
altering the STACK= option in the command file to build MACRO-11.

F-4

by
by

APPENDIX G

WRITING POSITION INDEPENDENT CODE

G.l

INTRODUCTION TO POSITION INDEPENDENT CODE

The output of a MACRO-11 assembly is a relocatable object module.
Task

Builder

binds

one

modules

together

create

The

executable task image.
Once built, a task can generally be loaded and
This is
executed only at the virtual address specified at link time.
because the linker has had to modify some instructions to reflect
the
memory locations in which the program is to run.
Such a body of code
is considered position-dependent (i.e.,
dependent on the virtual
addresses to which it was bound).

All PDP-11 processors offer addressing modes that make it possible
to
write
instructions that are not dependent on the virtual addresses to
which
they are
bound.
A
body
of
such
code
is
termed
loaded and executed at any virtual
position-independent and can be
address.
Position-independent code
can
improve
system efficiency,
both
in wuse of virtual address space and in conservation of physical
memory.

In multiprogramming systems like
IAS,
RSX-1lD and RSX-11M,
it
Iis
important that many tasks be able to share a single physical copy of
common code;
for example a library routine.
To make the optimum use
of
a
task's
virtual
address
space,
shared
code
should be
Code that is not position—-independent can also
position-independent.
be shared,
but it must appear in the same virtual locations in every
This restricts the placement of such code by the Task
task using it.
Builder and can result in the loss of virtual addressing space.
The construction of position-independent code is closely linked to the

remainder of this
The
addressing modes.
proper usage of PDP-1l1
Appendix assumes you are familiar with the addressing modes described
in Chapter

All
addressing modes
involving
only
register
These modes are as follows:
position-independent.
R
(R)
(R) +

@-(R)

When

using

are

deferred autoincrement mode

@(R)+

- (R)

references

autodecrement mode

deferred autodecrement mode

these

addressing

providing
position-independence,
been supplied such that they are
virtual memory location.

modes,

the
not

you

are

guaranteed

contents of the registers have
dependent upon a particular

WRITING POSITION

The

relative

relocatable

addressing
address

These modes

are

modes

1is

INDEPENDENT CODE

are

referenced

position-independent
from

relocatable

follows:

relative

when

instruction.

mode

relative deferred mode

Relative modes are not position—-independent when an
absolute
address
(that
1s
a non-relocatable address) is referenced from a relocatable
instruction.
1In this case, absolute addressing
(i.e.,
@#A)
may
be
employed to make the reference position-independent.
Index modes can be
according to their
X (R)

index

@X (R)

index

If the base, X,
the

either position-independent or
position-dependent,
use in the program.
These modes are as follows:

reference

mode
deferred

is an absolute value (e.g;, a control

is position-independent.

For

block

example:

MOV

2(sp),RO

s POSITION~-INDEPENDENT

MOV

N (SP) ,RO

; POSITION-INDEPENDENT

N=4

If,
however,
X
1is
position-dependent.

CLR
Immediate

a
For

relocatable
example:

ADDR(R1)

mode

its usage.

can be

an absolute

address,

the

offset),
|

reference

; POSITION-DEPENDENT

either

Immediate mode
#N

When

mode

position-independent

references are

immediate

not,

formatted as

according

follows:

mode

code

position-independent.
When
a
relocatable expression defines N,
code is position-dependent.
That is, immediate
mode
references

expression defines

the
are

position-independent only when N

the

value

absolute

the

value.

Absolute

mode addressing is position-independent only in
those
cases
where an absolute virtual location is being referenced.
Absolute mode
addressing references are formatted as follows:

Q#A
An

example

the

For

example:

absolute mode
a

MOV

G.2

position-independent

directive

status

word

@#SDSW,RO

absolute

($DSW)

from a

s RETRIEVE

reference

relocatable

DIRECTIVE

a reference

instruction.

STATUS

EXAMPLES

The RSX-11l library routine, PWRUP, is a FORTRAN callable subroutine to
establish
or
remove
a
user
power failure AST entry point address.
Imbedded within the routine is the actual AST entry point which
saves

all

registers,

restores

The

effects

all registers on
following
examples

call

return,

are

the

user-specified entry point,

and executes an AST exit
excerpts
from this routine.

directive.

The first
has been modified to illustrate position-dependent references,
(see
Figure
G-1).
The
second
example,
Figure
G-2,
1is
the
position-independent version.
example

G-2

WRITING POSITION INDEPENDENT CODE
PWRUP::

CLR

CALL
|

208

BA:

:PUSH (SAVE) ARGUMENT ADDRESSES ONTO

.X.PAA

; STACK

+CLEAR DSW, AND SET R1=R2=SP
sGET OTS IMPURE AREA POINTER
:sGET AST ENTRY POINT ADDRESS
NO POWER
+ IF NONE SPECIFIED, SPECIFY
;sRECOVERY AST SERVICE

.WORD
MOV
MOV
BNE
CLR

1.,SDSwW
$OTSV,R4
(SP)+,R2
108
- (SP)

MOV
MOV

R2,F.PF (R4)
#BA,- (SP)

CALL

109.,2.

s ISSUE DIRECTIVE, EXIT.

.BYTE

.X.EXT

MOV
MOV

RO,-(SP)
R1l,-(SP)

;PUSH (SAVE) RO
;PUSH (SAVE) Rl

10S$:

s ASSUME SUCCESS

- (SP)

208

;
H

:SET AST ENTRY POINT
s PUSH AST SERVICE ADDRESS
3

Position-Dependent Code

Figure G-1
PWRUP: :

CLR

- (SP)

+ ASSUME SUCCESS

MOV
MOV

@#SOTSV,R4
(SP)+,R2

;GET OTS IMPURE AREA POINTER
;GET AST ENTRY POINT ADDRESS

CALL
.WORD

BNE

20$:

:IF NONE SPECIFIED, SPECIFY NO POWER

108

s RECOVERY AST SERVICE

- (SP)

MOV
MOV
ADD

R2,F.PF (R4)
PC,~-(SP)
#BA-., (SP)

"+SET AST ENTRY POINT
; PUSH CURRENT LOCATION
; COMPUTE ACTUAL LOCATION OF AST

CALL

.X.EXT

+ ISSUE DIRECTIVE, EXIT.

208

.BYTE

109.,2.

ACTUAL AST SERVICE ROUTINE:

2)
(78]

.
4)

[0 w6 we =8 we w8 ws s o

; PUSH ARGUMENT ADDRESSES ONTO STACK
:CLEAR DSW, AND SET R1=R2=SP.

.X.PAA
1.,$DSW

CLR
BR

10S:

(SAVE) R2

;PUSH

R2,-(SP)

MOV

SAVE REGISTERS

EFFECT A CALL TO SPECIFIED SUBROUTINE

RESTORE REGISTERS
ISSUE AST EXIT DIRECTIVE

MOV
MOV
MOV

RO, - (SP)
R1,-(SP)
R2,-(SP)
Figure G-2

;PUSH (SAVE) RO
;PUSH (SAVE) Rl
;PUSH (SAVE) R2
Position-Independent Code

The position-dependent version of the subroutine contains

relative

reference to an absolute symbol ($OTSV) and a literal reference to a
Both references are bound by the Task
relocatable symbol (BA).

Builder to fixed memory locations. Therefore, the routine will not
execute properly as part of a resident library if its location in
virtual memory is not the same as the location specified at link time.

WRITING POSITION INDEPENDENT CODE
In the position-independent version, the reference to SOTSV
has
been
changed to an absolute reference.
1In addition, the necessary code has
been added to compute the virtual location of BA based upon
the
value
of the program counter.
In this case, the value is obtained by adding
the value of the program counter to the fixed displacement
between the
current
1location
and
the
specified symbol.
Thus, execution of the
modified routine is not
affected
by
its
1location
in
the
image's
virtual address space.

The

MACRO-11
Assembler
provides
a
way
of
checking
the
position-independence
of
code.
In
an assembly
1listing, MACRO-11
inserts a '
character
following
the
contents
of
any
word which
requires
the linker to perform a relocation operation.
In some cases
this character indicates a position-dependent instruction;
in
other
cases,
it
merely
draws
the user's attention to the use of a symbol
which may or may not be position-independent.
The cases which cause a
' character to be inserted in the assembly listing are as follows:

Absolute mode references are flagged with a '
the
reference
1is
relocatable.
References

when

they

are

example:

MOV

position-independent

(i.e.,

;PIC ONLY

@#ADDR,R1

IF ADDR

character
when
are not flagged
absolute).

For

ABSOLUTE.

Index and index deferred mode references are flagged
with
character when the offset is relocatable.
For example:

MOV

ADDR(R1),R5

;NON-PIC

MOV

@ADDR(R1l),R5

IF ADDR

;NON-PIC

ADDR

RELOCATABLE.
RELOCATABLE.

Relative and relative deferred mode
references
are
flagged
with
a ' character when the address specified is relocatable
with respect to another program section.
For example:

MOV

ADDR1,R1

+NON-PIC

MOV

@ADDRI1,R1

:TO

SECTION

Immediate mode
flagged with a
MOV
MOV

WHEN ADDR1
ANOTHER PROGRAM

references to
' character.

#3,R0
#ADDR,R1

relocatable

reference

is made

(i.e.,

addresses

are

always

sALWAYS POSITION-INDEPENDENT.
;sNON-PIC WHEN ADDR IS RELOCATABLE.

There is one case in which the MACRO-11
potential
position-dependent reference.

instruction

BOUND

absolute

MOV $OTSV,R4

virtual

assembler
does
not
flag
a
This occurs where a relative
location from
a
relocatable

in Figure F-1).

Those references requiring more than simple relocation
at
1link
time
are
also indicated in the assembly listing.
Simple global references
are flagged with the letter G.
Those which
contain
multiple
global
references
or
complex relocation, are flagged with the letter C (see

Section

3.9

positively

in general,

and

Chapter

state which

4).

are

such

are

not position-independent.

cases,

1is

difficult

is safe to apply the guidelines
discussed
earlier
resulting address value produced at link time.

this Appendix to the

However,

APPENDIX H

SAMPLE ASSEMBLY AND CROSS REFERENCE LISTING

LTITLE

+ IDENT

y COPYRIGHT 1976, OIGITAL EQUIPMENY C(CORP,, MAYNARD, MASS,
'
FOR USE
y TMIS SOFTWARE 18 FURNISHED TO PURCHASER UNDER A LICENSE
N
y ON A SINGLE COMPUTER SYSTEM AND CAN BE COPIED (WITH INCLUSIO
EXCEPT

’

gy OF DEC’S COPYRIGHT NOTICE)

9y EQUIPMENY CORPORATION,

16
17

y DEC ASSUMES NO RESPONSIBILITY
y OF 178 SOFTWARE ON EQUIPMENT

12
13

TO CHANGE WITHOUT
y THE INFORMATION IN THIS DOCUMENT IS AS SUBJECT
A COMMITMENT AY DIGITAL
y NOTICE AND SHOULD NOT BE CONSTRUED
?

]

2?7

29
31

t C, MONIA AS«FEB=T76
¢ LOCAL MACROS

t DEFINE CANNED SECTION TaBLE ENTRIES

3y SECTN NAME,FLAGS,SEG

} WHERE1

)

38
39

42
43

$$0m,

$$8s,

oMoy

+888S

51
52

sS
81
s7

¢ PERFORM PRELIMINARY SyYMBOL TABLE PROCESSING

FOR THME USE OR RELIABILITY
WHICH IS NOTY SUPPLIED BY DEC,

y VERSION 25

ONLY FOR USE IN SUCH SYSTEM,

y AS MAY OTHERWISE BE PROVIDED IN WRITING BY DEC,

705/

S
é
7
8

P3IPRE

NAMESSECTION NAME

FSET IN SEGMENT DESCRIPYOR TO RECEIVE SECTYION ADDRESS,
OFFSET=QF
BLANK IF NONE,

FLAGSESECTION FLAGS BYTE CONTENTS

SEGa*RO0T* IF SECTION IS DEFINED IN ROOT SEGMENT ONLY

JMACRO SECTN
LWORD

RADSA

/NAME/

«BYTE

FLAGS

JBYTE
« IFF

)

JWORD

JIF 8

«IF NB
JWORD
« IFF

NAME,FLAGS,OFFSET,SEG

SEG

OFFSET
OFFSET

H-1

LoT1)L1 2v e e n2eboe

20pve

v900BY

‘

2fP020

i =4x2»3S43F°L0ONYA%0%Q‘Q%108d

¢ vd01 vivg

JON3’

43384’

WON3®

q0uom*®

9nt2o0 v0ae entLee

3B$1$I13L39WNI44OA80N3JSIA80NNuWTH¥ALS8%I)TSAIMMN¥X¥XyTII1I1l888e°°N°'**OI4‘IM ¢m 01M SM |M ‘1M g2t3WLNOILI3S03¢N¢o1THIL4vI0VOS3nIn8uNQ3i0wILVHWyTdY3EAMISSNOaAD14TAN0H13OMYi88NSIWVOL3IdALIILS84SNNIA9NI0JQv3WAS3Y19VZ3S4v3THQyN4VIiA0LvL9VVSWvHIAU1N4HIV¥NIYuSNDININ;3WIOW39S3S
7OSQoI8wNN49IILINMdSWDI3TsT$§IL170V8vSV8Y443IV¥O SN4nLVEIQmsdN®3LOSDQLISSQVYyvNNSaWNIS¥LlhSS$V01M Ny3YSA31OI38N3HIJ5O4EN0Q0DAN1AUSQddVAN133WivWTG73IL4738SN0VV1G7IL4$I1SNI130304¥IW8d09WSQA3AvNTSV1Q364AH¥98030804Fl0LQSHi1HI3d0SNNI3ID¥H3LIS3$YN83OuQ0EAN°wBSNL3(SdNW$AA3BLN$T4‘I928S¥3HvQ07LI0l)I4YNdMI¥Od3I0L3NH0LW
IwATSI¥VNvW»FNI3YHLyuEJYsEdNLIN9IIQA3vO®OsBaNLWHOnIlSI%8sAY4%N0)IDSsIHNNAIOS480I3V9Ln3IAS3w4QS0YVNQINvI4i0v3AdI8WNDA3IL0dWNS3N3ISLNWI843D3S03)JSlH0d8ILHAO¥d802I18¥3l3IWd0SINN3LIUSSd3Q¥0(IWNLIS

”t!wNWNGOLVWODNOOD2I©=2LS37QI91vV3YSAH4®03LS8SI/NT0IQYv30NN7dL1OVY$IHSI/SA0)3HW1$L4SI1M30u3Q3dNSAv0¢ON4W3I0w¥3IeNINWIN1¥A4JE0OdNWIeTDI"VN¢IAWNNV13YJWW¥3dA1sV3OT(HA0N1dIIL¥NIOSY30¢
08 ="3197qyom*®12% )
SAMPLE ASSEMBLY AND CROSS REFERENCE LISTING

mamam%

‘

“““”h““““”““

n908eo

ISiNaNI
“““*““““““““

Ls2sel

g1t

”3AHWLndSNOILI3SQ3LVNOIS3QSYL108,3uYQ30N7ININI3HLNIVW
T3M¥7YNNNIAT0IW0NlINOLIXML30VN8IwASSIANIAQI
s
e
T08WASANINI¥04AVIN3A0T0HINOIIILNOY

3INOONWI M8

Y
A
0
'
Q
0
S
A
Q
S
O
Mg’

13384’

Q*LF3ANOI1V89FY3VAI7M8YV34OWLM3H
L B

13WNLSd

]

nail
IR

sel

151
e

QONOJ3S 4vyW 40 ARYN

Q0w ses

AN3IW3T3 qugm*®

L5t

951
A

d0ae s gouves
ote se 9ie en

ent ol
e

$9v4

YN3LOS8IWLNI3IMS 4LIANvS0YHNIHd

quom*

]

‘

]

ilag’

€51

est

Ni1J3S

quom*®

@BSAVY® /YNS/

F3WNLSA

90 0 0 oloden

0 0 2L

RNCh
gxig’

2eeoee

$¥3SI348SNQV0YvL

L
A
W
N
G
m
y
u
1
g
*
M¥1E
*
0o0eRY

12384’

921

mu18"’
M8’
mue’
Mu18°*
Mye°

SAMPLE ASSEMBLY AND CROSS REFERENCE LISTING

ESHY N
W

951

§§1

st
e

091 191

991

Ni3J3S

IWNLIS

0950 0

patete nEnioe 21050 nenioe J0eoee p1esa1 giLetlt Talele pasele
ense e 85 0 0 nsS0 0
94i50 8 na9ve 9690 8

Le2

962

1341

fos nos

1711

21t

91t
e

181

2% et nat 92¢% 141

an3*

¥17v3 By'SH

AQOW

L*TINV¥038HSIAOWN9I3S9

s34Lv013I1u08d298T3y

1331

+(ty)

Sy‘dy

TWASNI

139$6380Qv40IN3WH3S401dI¥I830

¢O138N40

41
1

V0T108WAS201dIN3830ONYANIFNdA3NiIN3NOITL0I83WIASSLSNNIW37388VLdNi3S
Mvd

12¢

tses

1sa¢é

MENSL‘HNEMEH

gn90ee

diWd

(0y)

§0¢2

AOwW

W
W

rig

AOKW

)
u‘sSy
AONW 0u’SH
(L) By
03¢ 'I8AND (By)

i18

1331

AOW Ce

038

$A9dvO7D4

eIt

niSO0 9IslL l 8473 ‘s3usSIm (0¥

9
v
i
g
N
O
I
1
3
S
S
v
I
N
I
Q
I
S
3
u
=
%
S
v
Y
L
1
3
9
S
3
¥
Q
v
4
0
v
I
0
7
1
H
0
1
l
d
I
¥
J
3
8
3
Q
INENL3 WNLSde 141384030ON041¥335d4Q0¥a04L073188 dNNIi3H80ldI¥IS3Q
Y
Ta'WAS13

¥0L0d1LIv¥NI4$0O37M0 4LO08N1v374ON74AOWdIV0Dv3YSNM H¥AvO3NT3IT U9dIANY3IVQIHSE3NI¥GT N3$O8¥I3L08ILJ03QSv

90¢

Qes

038

AOW8181 (By)=

$o%

R
P
Y
’
Z
*
N
A
S
T
I
9nse0e enisel +++(((BQ0yyy)))'''+++(((T%8YtyHu))) 10 3 IN3WO3S CAIND
795020

AOW

SAMPLE ASSEMBLY
AND CROSS REFERENCE
LISTING

MWWW@,Q

720 0 esetle l1o0e0

2590 0
i

]

(131 ffs

eWiy,

SAMPLE ASSEMBLY AND CROSS REFERENCE LISTING

i1g=2

Toc=2n
$92=2»

oo oxix

- Y

WAS1IT

NQONW

JYdid

JdSi¥

WASNI]

ind183 WN TS3 AM8I83

WNLJS
vds$
4448

WASINI

L2TYYIR
AWKNG

4843

EYT)

189848 T3u8AS LNvos AN DS

H91438
HTNISGMHSAS

§91le2 12 2

Y TY YL

]

PL2=2

2900200

T08HAS JSANI

ni2=2

nnodes

n2eode

n9e gs

29de=¢

dil=2»

WMEAS

9ig=2

d92=2

pLe=2 192=2

991=
2
9912
n9le?
n9l=2
BL1=2
9822
X9

691=2 Log=2

1fg=2
252=2
n9le2
192=2
nie=2 95d=2

6912

f6d=2

elé=2
n9le2n

B1s=2

9L2=2

342 ToA

6éi=2

L91=2 §91=2 BLi=2

SAMPLE ASSEMBLY AND CROSS REFERENCE LISTING

02i=2C» gGC=2» n é=¢ 292=2 g82=2 LoC=2» Bifelx 0ig=gs Lo2=2s 192
gLd=2» ndi=2 61d=2 682

nig=2

612=2»

068KAS

§52=2»

29¢2=¢ARLT

$3IN34343Y

Li2g=2

6n2e=2 eli=¢ glg=2% §2 =2%

£y

HILSION I0GWASS0¥IFINIYIL3Y

602=2»

ésa=2 gLa=2

922=2
0ig=2

AYLTE. 6hd=¢ L)CYRLT L08=2 91g=2 eni=2»

992=2

BIndelsi90222=2» 182=2%Lne=2»Bhde=2» 922=2» 192=2%02i=2»
hhdeee»§c=2s Thig=2YL LI TEned=2YL
=2» 625 ned=d» 962>
Pide2s

13A

H-6

SAMPLE ASSEMBLY AND CROSS REFERENCE LISTING

one=2 Itfe2sTig=2

Ene=e

902=2

1e2=2

612=2

8L2=2

9Si=1in

NuNl3Y

FSOIMNIUOYI¥JV3IWY

Uei=2n §91=2» RiT=2n

13 80

¥17vd s¥oY 3 UAVS YA YSN123S 808§

SAMPLE ASSEMBLY AND CROSS REFERENCE LISTING

ned=2d

192 981«2 9 1=2

22i=2

=2 991

2922
2ig=2

dg=2 Lei=2

121=2
§9i=2

8912 L91=2 S91=2 891e2

12 =2 enée=¢

g21=2
921i=2

L91=2

Bi12=2
ped=2

n9l=2
L91=2
8Li=2
=2 991
ULl=2

2s51=2
691=2

02e=2
édé=2

it1e=2 9 12 fif=2 691=2
aNg®

Bnd=2¢

pl2=2

6912 =2 BL1=2 Qe2=2 2 d=2 gne=2

JaN3*

I8l gisl

éed=¢ gne=2

n9le

MFIVY

62g=2 pL2=2

n9le2

a1°

2ie=2
db2=2
Ind=2
6622
1st=2

ngd=2

8L1=2
991=2 n9leg 0L1=2

geg=¢

neleg
LT IR

90e=2

292=2

2g=2 Scli=2 9L1=2

gxle’
myie*

dWld

S0é=¢

LS2=2

212=2
222=2
d2é=2
on2=2

1g=2

si¥
3¢

INI
dnl
sl

gsle ile gl e 078 ane

413

ing=2

slg=2
05e=2

n9d=¢

£02=2 LE2=2 2td=2 282e2 nege=2 25d=2 502=2

aqyv 338 01e
$36

T08WAS

$43IN3IUI

2i =2

L9¢=2

hidee
162=2

9
5
d
=
¢
2
1L2=2

ININVWYEd T08WAS 376YL S08I 3IN3HI43Y

glg=2

1g2=2
l1gg=e
B12=2
ande2
Iig=2

2og=2 0e =2 2i2=2

§2é=2 n é=2

B12=2 2 é=2 On2=2

SB2=2
8022
B1e=2
822+2
ane=2

192-2

1ge=2

l1og=2

4340

dadid=2 92 =2

vede2
8ld=2

eBi=¢

niged

9=28

g02=2 ene =e

INDEX

Absolute addresses, 6-14
Absolute binary output, 6-14
Absolute expression, 3-16
Absolute

mode,

5-5,

Absolute

module,

5-7

6-34

Absolute program section,
Address boundaries, 6-31
Address
Address

mode syntax,
modes, 5-1

6-37

B-1

Addressing forms, summary, 5-7
Allocating byte data, 6-17
Allocating dynamic memory, F-1
Allocating word data, 6-18
Allocation requirements, 6-35
Alternate radix, 6-25

operator,

Backslash,

6-25

3-1

Binary operators, 3-15
Blank lines, 2-2
Blocks of storage, reserving,
6-30
.BLKB directive, 6-30
.BLKW directive, 6-30
Branch instruction addressing,
5-8
.BYTE directive, 6-17

Ampersand,

3-1
Angle brackets, 3-3, 3-15,
6-4, 6-25, 6-28, 7-4 to
7-5, 7-16 to 7-17
Apostrophe, 7-10
Argument substitution, 7-16

"C operator, 6-27
Calling conventions, E-8
Calling macros, 7-3
Changing default radix, 3-13
Changing value of location

Arithmetic addition operator
or autoincrement indicator
3-1
Arithmetic division operator

Character set, 3-1
Character substitution, 7-16
Code and data separation, 6-38
Code or data sharing, 6-38
Coding standard, E-1

3-1

Arithmetic subtraction
operator or autodecrement
indicator, 3-1
ASCII character set, A-1
ASCII conversion, 3-14
ASCII conversion characters
6-19

.ASCII directive, 6-20
.ASCIZ directive, 6-21
.ASECT directive, 6-38
Assembler directives, 6-1,
B-1,

Assembler

B-2

version,

language, B-1
listing, 2-6
Assembly pass 1, 1-1
Asterisk, 3-1
At sign, 3-1

Attribute of the current
location counter, 3-12
Autodecrement deferred
mode, 5-3, 5-7
Autodecrement mode, 5-7

Comment field
Complementing
argument,

E-2

2-5
indicator,
an

3-1

6-27

Complex relocatable
expression, 3-16
Complex relocation, 4-1
Concatenated, 6-35
Concatenation of macro
arguments,

Conditional

7-10

assembly

block,

assembly
directive, 6-41, 6-42
Conditional branches, E-13
Continuation lines, 2-2
Creating local symbols
automatically, 7-7
Creating program sections,
6-36
Cross~reference

(CREF)r

listing

H-5

.CSECT directive, 6-17
Current location counter,

Autoincrement deferred

5-2,

3-1
Comma, 3-1
Comment, 6-14,
Comment field,

6-41
Conditional

6-8

Assembly
Assembly

mode, 5-3, 5-7
Autoincrement mode,

3-12

Colon,

Arithmetic multiplication
operator,

counter,

3-11,

5-7

Index-1

3-14,

5-6,

6-29

2-2,

"D operator, 6- 25
Data storage directives, 6-17
Date, 6-8
Default object module name,

(CONT.)

Finding number of characters
in strings, 7-12

Floating-point
data,

6-26

number,

6-11

summary,

truncation,

.FLT2

directive,

usage,

operator,

3-1

Direct assignment statements,
3-7
Directives, 2-5
Double ASCII character
indicator, 3-1
Double colon, 3-1, 3-7
Double equal sign, 3“""1 r 3"""7
Double quote, 3-1, 3-14, 6 -19
.DSABL directive, 3 “71 3“9
6-13 to 6-1 5, 6 =27
Duplication of code, 7-17

6-27

E-12

Formatting standards, E-9
Forward referencing, 3-8
Function control switches,
Function

directives,

8-6

6-13

General purpose registers,
General registers, 3-9
Global
label, 6-40
references,

E-2

6-15

6-40
symbol directory, 1-2
.GLOBL directive, 3-7, 6-39
GSD, 1-2

5-8

.ENABL directive,
6-15,

5-7,

6-27

3-15

Exclamation point, 3-1
Executable task image,
Exiting, E-9

Expressions, 3-14,

4-1

3-15

External
3-16

expression,

External

symbols,

3-15,

6 -40

Externally-defined

macro,

“F operator,

6-27

7-18

Hardware registers, E-2
Horizontal formatting, 2-6

.IF directive, 6-41
.IFF directive, 6-43, 6-44
LIFT directive, 6-43
.IFTF directive, 6-43
.IIF directive, 6-46
Illegal characters, 3-3
Immediate conditional
assembly, 6-46
Immediate expression
indicator, 3-1
Immediate mode, 5-4, 5-7
Immediate mode deferred, 5-5
Implicit .WORD directive, 2-5,
6-18
Indefinite repeat block
directives, 7-15
Index deferred mode, 5-4, 5-7
Index

3- 14,

address

arguments,

mode
7-13

2-3,

6-1 3,

.END directive, 6-3 1
.ENDC directive r6“’ 41
.ENDM directive 17""' 2
End of the source input, 6-31
.ENDR directive
Entry-point instructions, 6-33
.EOT directive, 6-3 1
Equal sign, 3-1
Error codes, D- 1
.ERROR directive, 7 -14
Evaluation of expressions,

Finding

6-27

Form-feed, 6-13, 7-3
Format control, 2-6

symbol,

EMT,

6-14,

.FLT4 directive 6-27
Forbidden instruction

D- 1

assignment

6-28

number specification, 6-27
rounding, 6-15, 6-27
storage directives, 6-27

Default register
definitions, 6 - 15
Deferred addressing
indicator, 3-1
Defining macros, 7-1
Device registers, E-2
Diagnostic, 7-14
Diagnostic error message
Direct

INDEX

macro

mode,

5-4,

5-7

—

Initial argument or expression
indicator, 3-1
Initial

Index-2

indicator,

3-1

INDEX

Instruction set, C-1
.IRP directive, 7-15
.IRPC directive, 7-15, 7-16
Item or field terminator, 3-1

Keyword arguments,

7-4,

7-9

(CONT.)

Macro

call

arguments,

7-4

Macro call numeric argument
indicator, 3-1
Macro definition, 7-1, 7-15
Macro definition arguments,
7-4
Macro definition formatting,
7-3

Macro definition termination,
7-2

Macro definitions and
expansions, F-2
Macro directives, 7-1
Macro expansion termination,

Label field, 2-2
Label terminator, 3-1
Left angle bracket, 3-1
Left parenthesis, 3-1
.LIMIT directive, 6-31
Line format, E-1
Linker, 2-2, 4-1
Linking, 4-1, 4-39
.LIST directive, 6-1
Listing conditional
assemblies, 6-4
Listing control
directives, 6-1
Listing level count, 6-2

7-3

Listing of binary extensions,
6-3
Listing of
Listing of
binary
Listing of
Listing of

comments, 6-4
generated
code, 6-3
macro class, 6-4
macro definitions,

6-4

Listing of macro expansion
binary
Listing of

code, 6-4
repeat range

6-4

symbol block, 6-14
symbol block delimiters,

3-10

Local symbols, 3-6, 3-10,
Location counter, 6-36
Location counter control
directives, 6-29
Logical AND operator,

library directive,

Macro

name,

7-1,

Macro

names,

E-5

3-11

3-1,

6-42

Logical inclusive OR
operator, 3-1, 6-42
Lower-case ASCII, 6-14

Macro nesting, 7-5
Macro symbol table, 3-5
MACRO-11 character sets, A-1
MACRO-11 directives, 5-9, C-4
MACRO-11 symbols, 3-5
.MCALL directive, 7-18
Memory allocation, 6-32, 6-33,
6-38

Memory allocation and mapping,
6-32
.MEXIT directive, 7-18
Minus sign, 3-1
Modularity, E-8
Module checking routines, E-9
.Module preface,

E-5

symbols, 3-11
Multiple labels, 2-4

Naming standards, E-2
.NARG directive, 7-11
.NCHR directive, 7-11,
Negative numbers, 3-14

Macro arguments, 7-6
Macro attritube directives,
Macro

call,

2-5,

7-3,

7-5

7-12

Nested conditional directives,
6-43
Nested macros, 7-3, 7-5
.
.NLIST directive, 6-1, 6-11
.NTYPE directive, 7-11, 7-13
Number of arguments in
macro calls, 7-7, 7-11
Number sign, 3-1, 5-4
Numbers, 3-13
Numeric control, 6-24
Numeric control operators,
6-26, 6-27

Numeric directives,

7-11

7-18

7-4

Multi-defined label, 2-4
Multiple definitions of local

expansions, 6-4
Listing of source lines, 6-3
Listing of the current
location counter, 6-3
Listing of the symbol table,
Local
Local

Macro

"0 operator, 6-25
Object module, 4-1

Object module name,
Index-3

6-26

6-11

INDEX

Octal radix, 3-13
.ODD directive, 6-29

Op codes, 2-4, C-1
Operand field, 2-4
Operand field separator,
Operator field, 2-4
Order of symbol table

3-1

search,

3-6
symbols, E-3

Other

(CONT.)

Relative
Relative
Relative

mode, 5-1, 5-7
standards, E-2
symbols, 3-9
term indicator, 3-1
addresses, 6-14
addressing mode, 5-6
deferred mode, 5-6,

5-7
Relative

mode,

Overlaid,

6-35

Relocatability,

Overlays,

6-33

Relocatable

5-5,

5-7

6-34
expressions,

3-16,

4-1

.PAGE directive,
Page eject, 7-3

6-13

Relocatable

module,

Relocatable

program

6-37
Relocation,

4-1

bias,

Page
Page
Page

ejection, 6-13
formatting, 2-6

Relocation

headings,

.REPT directive,

Page

number,

PAL-11R

6-8

conditional

6-46
Passing numeric
as

symbols,

Percent

sign,

Permanent

Repeat

assembly,

arguments

table,

Program boundaries
6-31

7-17

7-17
storage, 6-30
storage space,

Reserving
Reserving
6-29
Right parenthesis,

3-13,

3-1

3-5,

Program

counter,

3-9,

modules,

E-5

Program
Program

section

.SBTTL
Scope

G-1

directive,

Program

5-1

access,

6-33

name, 6-33
sections, 3-12, 6-32

Program source files, E-12
Program-local symbols, E-4

Programming standards and
conventions, 2-1
.PSECT directive, 3-12, 6-32,
6—-35

directive,
of

the

6-8, 6-11
program section,

6-33
|
Semicolon, 3-1
Sending messages to listing
file, 7-14
Separating and delimiting
characters, 3-2
Single ASCII character
indicator, 3-1
Single quote, 3-1, 3-14, 6-19,
7-10
Slash, 3-1
|
Source line sequence numbers,
6-3
Space,

3-1

Special
Special

characters, B-1
characters in macro

arguments,

Stack
operator,

6-23
directive,

.RAD50
3-13, 6-24
Radix control, 6-24
Radix control operators, 6-25
Radix-50 character set, A-4

Radix-50

6-34

3-1

symbol

Plus sign, 3-1
w
Position independent code,
.PRINT directive, 7-14
Processor priority, E-3

2-2,

directive,

6-46

C-1

Program

block

6-34
sections,

control

operator,

6-23
Radix-50

data, 6-22
.RADIX directive, 3-13,
Read-only access, 6-33
Read/write access, 6-33
Register

deferred

expression,

mode,

5-1

5-2

3-9
format, 2-1

Statement
Storing Radix-50 data, 6-23
Subconditional assembly, 6-43
subtitle, 6-8
Success/failure indication,
E-9

Symbol
Symbol

Symbol
6-24

7-6

pointer,

control directive,
examples, E-4
table listing, 1-2

6-39

Symbolic arguments of listing
control directives, 6-3, 6-4
Symbols, E-3
Symbols and expressions, 3-1
System macro libraries, 7-18

Index-4

INDEX

Tab, 3-1
Tab character,
1
|

2-2

Table of contents, 6-4, 6-11

Teleprinter mode, 6-4
Terminal argument or
expression indicator, 3-1
Terminal register indicator,
|
3-1
Terminating directives, 6-31
Terms, 3-14
Time-of-day, 6-8
.TITLE directive, 6-11
Title of the object module,
6-8
Translating to ASCII,
Translating

(CONT.)

Unary operator ordering,
Unary operators, 3-15

Unconditional assembly,

6-28

6-43

Undefined symbols, 3-7, 3-14
Universal unary operator
or argument indicator, 3-1
Up arrow or circumflex, 3-1
Up—-arrow, 3-3
Up-arrow (°) construction, 7-5
User symbol table, 3-5
|
User-defined and macro
symbols, 3-5
User-defined macro libraries,
7-18
Using the standard symbolics,

to Radix-50,

6-22

Trap

instructions,

Unary and
3-4

5-8

Version number, 6-12
Version number standard,

E-13

binary operators,

Unary control,

6-24

.WORD directive,

Index=-5

3-11,

6-18

PDP-11 MACRO-11
Language

Reference

Manual

AA-5075A~TC

READER'S

This form is for document comments only.
DIGITAL will
use comments submitted on this form at the company's
discretion.
Problems with software should be reported
on a Software Performance Report (SPR) form.
If you
require a written reply and are eligible to receive
one under SPR service, submit your comments on an SPR
form.

Did you

find errors

Did you

find

Please make

there

in this manual?

this manual
suggestions

so,

understandable,
for

usable,

improvement.

sufficient documentation

specify by page.

associated

and well-organized?

system programs

required for use of the software described in this manual?
what material is missing and where should it be placed?

Please

indicate

000000

Please cut alon

this line.

——

NOTE:

COMMENTS

the

Assembly

type

language

Higher-level

user/reader

that you most nearly

If not,

represent.

programmer

language programmer

Occasional programmer

(experienced)

Uaer with little programming experience
Student programmer

‘Non-programmer interested in computer concepts and capabilities

‘Name

Date

Organization

(Stre@t‘
| City___

State

Zip Code
or

Country

Fold Here

Do Not Tear - Fold Here and Staple

FIRST CLASS

PERMIT NO. 33

MAYNARD, MASS.
BUSINESS REPLY MAIL

NO POSTAGE STAMP NECESSARY IF MAILED IN THE UNITED STATES

Postage will be paid by:

Software Documentation
146 Main Street

Maynard,

ML5-5/E39

Massachusetts

01754