Digital PDFs

MISC-6841C4E2

October 1981

226 pages

Original

97MB

Document:	Technical Summary
Order Number:	MISC-6841C4E2
Revision:
Pages:	226
Original Filename:

OCR Text

TECHNICAL SUMMARY

~n~nomn

TECHNICAL SUMMARY

~D~DD~D

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.
Digital Equipment Corporation makes no representation that the interconnection of its products in the manner described herein will not infringe on existing
or future patent rights , nor do the descriptions contained herein imply the
granting of license to make, use, or sell equipment constructed in accordance
with its description .
The software described in this document is furnished under a license for use
only on a single computer system and can be copied only with the inclusion of
DIGIT Al 's copyright notice. This software , or any other copies thereof, may not
be provided or otherwise made available to any other person except for use on
such system and to one who agrees to these license terms. Title to and ownership of this software shall at all times remain in Digital Equipment Corporation.
Digital Equipment Corporation assumes no responsibility for the use or reliability of its software on equipment that is not supplied by DIGITAL.
DEC , DECnet, DECsystem-10 , DECSYSTEM-20, DECtape ,
DECUS , DECwriter, DIBOL, Digital logo, IAS , MASS BUS , OMNIBUS ,
PDP, PDT , RSTS , RSX, SBI , UNIBUS , VAX, VMS , VT
are trademarks of
Digital Equipment Corporation .
Copyrigh ~ 1980 Digital Equipment Corporation

Maynard , Massachusetts

Contents
INTRODUCTION

4 THE VAX PROCESSORS
INTRODUCTION . . .

2 THE SYSTEM
INTRODUCTION ...
COMPONENTS
Processor
Virtual Memory Operating System
Peripherals
PERFORMANCE . . . ..... . . . . .

VAX ARCHI TECTURE
.... . ..... . .... . . 2-1

.... .. ..... . . . . .. . 2-3

RELIABILITY
.. .. .. .. . .
.. .. .. . . ..
.. . 2-3
.... . 2-3
Data Integrity
System Availability ..... . ... . .. . ... . . . .. ... ... ... . .... . 2-4
System Recovery . . . . . . . . .
. . 2-4

USER PROGRAMMING ENVIRONMENT
.. . . . . . . . .. 4-4
Process Virtual Address Space Structure ... . .... ........ 4-4
General Registers . . . . . . . . .
. ............ 4-5
Addressing Modes ......... .. .. . . .. .. .. . . . . . .... ...... 4-5
Addressing Modes Table ...................... . ....... 4-6
Program Counter . . . . . . . . . . . . . . . . . . . . . . . .
. .. . . 4-6
The Stack Pointer, Argument Pointer and Frame Pointer .. 4-7
. . .4-8
Processor Status Word . . . . . . . . . . . . . . . . . . . . . . . .
Handling Exceptions ..... .. . ........ ....... ... .. ...... 4-8

FLEXIBILITY ... . . .
. ... . . . . .. 2-4
Flexi blity in the Operating Environment ..
. .... . ... 2-5
Flexibility in Programming Interfaces .. . . . ... ..... ...... 2-5
Programmer Productivity ................... . .... . ..... 2-5
Extending the System .................. . ..... . . ...... 2-5
. ... . . ... ..... . . 2-5
PDP-11 Compatibility .. . . . . . . . . . . .

3 THE USERS

NATIVE INSTRUCTION SET . ..... .. . .................. 4-8
Instruction Set Summary Table .... .. . . . . .... .. ....... .4-9
Integer and Floating Point Instructions .... . ........... 4-11
Packed Decimal Instructions
.... ...... . . . ......... 4-11
Edit Instruction ... ... . .. . .. .. . . ..... . .. .... . ........ . 4-11
Character String Instructions ............. . .. .. .. . ..... 4-12
.. .. ....... ........ . . ......... 4-12
The Index Instruction
Variable-Length Bit Field Instructions .......... . . . ... .. 4-12
Queue Instructions .... .. ........ .. ........ . . .. . . . . . . 4-13
. . . . . . .4-13
Address Manipulation Instructions . . . . . . . .
General Register Manipulation Instructions
....... 4-13
Branch , Jump and Case Instructions . . . . . .
. ...... 4-13
Subroutine Branch , Jump , and Return Instructions ... . . . 4-14
Procedure Call and Return Instructions ................ 4-15
Miscellaneous Special-Purpose Instructions ... . . . . ..... 4-15

.... 3-1
. ... 3-1
. . 3-1
. . . ... . .... . 3-2
. .. 3-3
. ......... 3-4
. .. ... 3-5

. ... .. ... . . . .. . . .. 3-5
THE SYSTEM PROGRAMMER .
Job and Process Structure ............................ 3-5
Multiprogramming Environment ..... . .. .. .. ... . . .. . .. . . 3-6
............. 3-6
Program Development
THE SYSTEM MANAGER . .... . . . ... . . ... ... ... ... .. . .. 3-6
User Authorization . . . . . .
. . . ................. . .... 3-6
Privileges . . . . . . . . . . . . . . .
. ... . ... . .... . . . . . . .. 3-7
...... 3-7
Resource Quotas and Limits
.... . ... 3-7
Privileges Summary Table
Resource Accounting Statistics
.... 3-8
Performance Analysis Statistics ...... ... ... ... ... .... . . 3-8
Display Utility Program .
. . . . . . . . . . . . . . . . ...... 3-8

COMPATIBILITY MODE ............................ . .4-15
PDP-11 Program Environment . .. .. ... . . . ... ... . . . . ... 4-15
.. . .. .. .. .. .... 4-16
PDP-11 Instruction Set .. .. .. .. .
PROCESSING CONCEPTS FOR SYSTEM
PROGRAMM ING . . . . . . . ................ . . .. .. . .... 4-16
Context Switching .
. ............................ 4-16
Priority Dispatch ing ........................ . • .. ...... 4-16
Virtual Address ing and Virtual Memory . .. ..... . . ...... 4-17

............. . .... 3-8
THE SYSTEM OPERATOR
Spooling and Queue Control
................ ....... . 3-8
Batch Processing
.... . . . . . . ... .. . .. ... ... .. 3-9
Online Software Maintenance
...................... 3-9
System Recovery
.. .. .... . ........ 3-9
Error Logging and Reporting ... . .. . . ... . . . . ... .. . .. ... 3-9
Online Diagnostics ........ . ... . ... . . .. .. . . . . ...... .. 3- 10
Remote Diagnosis . .
. ....... . .... 3-10
THE USER ENVIRONMENT TEST PACKAGE

.. .... . ........ ......... 4-1

PROCESSING CONCEPTS FOR USER PROGRAMMING .. 4-1
Process Virtual Address Space . . . . . . . . . . . . .
. . 4-1
Instruction Sets . . . . . . .
. .... . . .... . .. ... ... . . . . 4-1
Registers and Addressing Modes . . . . . . . . . . . . . . . .
. . 4-1
Data Types ................. ....... . .. . . ..... .. ... . . . 4-1
Data Type Table
.............................. 4-2
Stacks , Subroutines, and Procedures .. .. . . .... .... ... . 4-4
Condition Codes ....... .. .. . . . ........ .. ...... . . ..... . 4-4
Exceptions . . .
. ............................. 4-4

. . .. . .......... 2-1
.... 2-1
........ . ..... 2-1
. . . ... ......... 2-2

THE APPLICATION PROGRAMMER
The Command Language . . . . . . . . . . . . . . . .
Command Procedures
DCL Command Language Summary Table
RUN Command .
Programming Languages ..
Record Management Services .

.... . .. .. ... ...... . .. . .. .... 4-1
..

SYSTEM PROGRAMMING ENVIRONMENT ........ .... 4-17
Processor Status Longword ........ ........ . . ........ 4-17
Processor Access Modes ............................ 4-17
Protected and Privileged Instructions .................. 4-18
Memory Management
.. . ............... ........ 4-18
Virtual to Physical Page Mapping ..... . .. . .. . ... . . . . . . . 4-20
Exception and Interrupt Vectors ......... . ... . ... . .... 4-22
Interrupt Priority Levels . ... ... . ....... . . . . . . ......... 4-22
110 Space and 1/0 Processing . . . .. .... ..... . .. . ...... 4-23
.. . . .. . .. . . .... .... . .. ..... ...... . 4-23
Process Context

.. 3-10

APPLICATIONS EXAMPLES
........ ... 3-10
Commercial System Example .... . . .. . . .............. 3-10
Real-Time Flight Simulation Example ... . ...... . ..... . 3-13
. . .. . . .. 3-13
Design Considerations . . . . . . . . . . .
iii

Interprocess Communication Facilities ... . . . .....•.... 6-12
Common Event Flags ...... . ........... . ............. 6- 13
Mailboxes .... . . . . . . ... . . .. .. .................. . . . . . 6- 13
Shared Areas of Memory . . . . .
. .. . ......... . ... 6-14
Interprocessor Communication Facility ......... .. . ... . 6-14

CONSOLE .............. . . . . . .............. . ........ 4-25
THE VAX-11 / 780 PROCESSOR

.... . .... ... .... . .... . 4-27

INTRODUCTION ..... . .... . ....... ... .. . ... . . . . . .... 4-27
VAX- 11 / 780 PROCESSOR COMPONENTS . .. . . . . . .... 4-28
VAX-11 / 780 Console ..... .. . . . . .. . . . ... . . . . . . . . . . .. . 4-28
VAX- 11 / 780 Memory Interconnect . .. . .. . . . . . . . . .. ... . 4-28

MEMORY MANAGEMENT ....... . . . . . . . ... . ........ . . 6-14
Mapping Processes into Memory ... . . . .. . . . . . . . ....... 6-14
Process Virtual Memory and Work ing Set . . .... . ....... 6-15
Paging ..................... . ........ . . . . . . . . . ... . ... 6-15
Virtual Memory Programming .... . ..... . . . .. . . ....... 6-17
VAX/VMS Memory Management Services . ....... .. ... . 6- 17

VAX-11 / 780 MAIN MEMORY AND CACHE SYSTEMS .... 4-29
Main Memory . . . . . . . . .... . . . . . . ............. . ........ 4-29
Memory Cache . ................................. . ... 4-29
Address Translation Buffer ................ . ... . . . ..... 4-29
Instruction Buffer . . . . . . . ..... .. . .. ....... . . . ... . . . . . . 4-29

PROCESS SCHEDULING .............. . ... . ... .... .. 6-1 8
System Events and Process States .... . . . . . . . . . . . ..... 6-1 8
Priority: Real-Time and Normal Processes . . . . .. ... . . . .. 6-1 8
Scheduling Real-Time Processes .............. . . . ..... 6-1 8
Scheduling Normal Processes ..... . .. . ... . .... . . . .... 6- 19
Swapping and the Balance Set ............... .. .. . ... . 6-19
VAX/VMS Process Control Services .............. . . . . 6-19

1/ 0 CONTROLLER INTERFACES .. . ............. . ..... 4-29
VAX- 11 MASS BUS Interface . . ..... . .. . .. . ............ 4-29
VAX-11 1780 UNIBUS Interface . . . ... . ......... . . . . . . .. 4-30
Data Throughput .............................. . . . ... 4-30

VAX- 11 1780 FLOATING POINT ACCELERATOR ........ 4-30
THE VAX-11 / 750 PROCESSOR

... . ....... . . . . ....... 4-31

1/ 0 PROCESSING .................. . ... . .. .... .. .. . 6-19
Programming Interfaces . . . ..... . .. . . .
. . . ......... 6-20
Ancillary Control Processes .. . . . ... . . . . . . . ........... 6-20
1/ 0 Processing Interfaces Table ........ . ............. 6-21
Device Drivers ... . . .. . ... . . . . . . . . . . ... . .. . . . .... . .. . 6-21
1/ 0 Request Processing . . . . . . . . . ... . . . . . . ... . . . . .... . 6-21

INTRODUCTION ... . ........... .. .. . . . .. ... . .. . ..... 4-31
VAX-111750 PROCESSOR COMPONENTS ...•.. .. . . . . 4-31
VAX- 111750 Console ... .. . .. . .. . . . .. .. . ... . . .. ...... 4-31
VAX- 111750 Main Memory ...... . .... . . .. . . ........... 4-32
VAX-111750 Cache Systems . . . . . . . . . . . . . . . . . ......... 4-32
Peripheral Controller Interfaces ........ . ............. 4-32
VAX- 11 MASS BUS Interface .......... . ............... 4-33
VAX- 11/750 UNIBUS Interface . . . . . . . ... . . . . . . . . . . . . .. 4-33

COMPATIBILITY MODE OPERATING ENVIRONMENT .. 6-22
User Programm ing Considerations . .. .. . . . . . ... . ... .. . 6-22
File System and Data Management .. . . . . . . . .. .. ....... 6-23
Command Languages ........... .. . . . .. .. . . . . . . . . . ... 6-23

5 THE PERIPHERALS

7 THELANGUAGES

COMPONENTS . . . . . . . .. . . ... . . . . . . . ..... . . . . . . . . . . .. 5-1

INTRODUCTION . . . . . .. . . .. . .................... . ..... 7-1

MASS STORAGE PERIPHERALS ..... . ... . . ........... 5- 1
Disk Device Table .. . ..... . . . .. . . . ..... . ... . . .. .. .. . . . 5- 1
Disks .. .. . ..... . . ...... . .. .. ...... . . . . . .. .. .... . ..... 5- 2
Magnetic Tape ... . ........ . .... . .. • .......... . . . .. . ... 5-2

VAX COMMON LANGUAGE ENVIRONMENT ...... . . . ... 7-1
Symbolic Debugger Interface .......... . .... . ... . . . .... 7-1
Symbolic Traceback Facil ity . . .......................... 7-1
Common Run Time Library ... . ... . .. .. ... ... ..... . .... 7-1
VAX Calling Standard ...... . ....... .. ... . . . ... .. ..... 7-1
Exception Handling ... .. .......... . ... .. .... .. ..... . . 7-1
VAX- 11 RMS .............. . ..... . ............ ... . ... 7-1

UNIT RECORD PERIPHERALS ......... . . . .......... . . 5-3
LP11 Line Printers .. . . . . . . . . . ... . .. . . . . . . . . . .... . ..... 5-3
LA 11 Line Printer .... . ..... . .. . . ... . . . ... . . . . ..•..... 5-3
CR11 Card Reader . . . .. . . . . . . .. . . . . . . . . . . . . . .......... 5-3

VAX- 11 FORTRAN ... . . . . . .... . . . . . . . . . . . . . ... . . . ..... 7-2
Introduction ............. .. . . .. . ... . . . ....•.. . .. . . .... 7-2
File Manipulation .......... . ... . . . ...... . ......... . .... 7-2
Simplified 1/ 0 Formats .............. . . . . . . . . . . . . . .. . .. 7-2
Character Data Type . ......... . . . ... . . . ............. . . 7-2
Language Extensions to FORTRAN - 77 Table ... . .. . .. . . . 7-3
FORTRAN-77 Features Table ......... . ... . . . . . ... . .... 7-4
Source Program Libraries ... . ... . . ... . . .............. 7-5
Calling External Functions and Procedures . . . . . . ........ 7-5
Shareable Programs ... . . . .. ... . . . ............. . . . . . .. 7-5
Diagnostic Messages ......... . . . . . . . . . ... . .. . ......... 7-5
Compiler Operation and Optim izations .................. 7-5
Debugging Facilities ................. . .... . . . . . . . ..... 7-7
Cond itional Compilation of Statements ... . .. . ... . ....... 7-7
Symbolic Traceback .................. . . . . . .. .. ....... 7-7

TERMINALS AND INTERFACES .................. ... . . . 5-3
LA 120 Hard Copy Terminal ......... . ..... . ...... . . . . . . 5-4
LA36 Hard Copy Terminal. . . . . . . . . . . ......... . . . . . . . ... 5-4
VT100 Video Terminal . . . . . . . .. .. ... . . . .. . . . .......... 5-4
DZ11 Terminal Line Interface . ...... . ......•... . ... . . . . 5-5
REAL-TIME 1/ 0 DEVICES ...... . ..................... . . 5-5
LPA11 - K . . .. .. . . . . .. . . . . . . . ........... . ....... . ... . . 5-5
DR11 - B . . . .. . . .•..... . ..... . .•.....•........... . . . . . . 5-5
DR780
. . ............. . ...... . . . . . . . . . . . . . ........ 5-5
INTERPROCESSOR COMMUNICATIONS LINK ........ . . 5-6
DMC11 ......... ...... . . . . . . .............. . . . . . .. . ... 5-6
MA780 ............ . ..... . . ... .. . . . . . . . ... . . . . . ... .. . 5-6
CONSOLE STORAGE DEVICES ... . ......... . .. .. . .. . . . 5-6
RXO 1 Floppy Disk Cartridge . . ...... . . . . ..... . . . ........ 5- 7
TU58 Tape Cartridge . . . .................... . . . ... .. . . . 5- 7

. .. ... . . . . .......... . . . . ... . ..... 7-7
VAX - 11 COBOL . .
Introduction ...................... .. . . . .. ........ . .... 7-7
General Characteristics . . . . . . .. ................. . . . .... 7-8
Structured Programming . . . ..... . ....... . . .. ... . . . . . . . 7-8
Data Types ...... . ... . . . .. . . . . . . . ....... . ............ 7-9
Files and Records . . . . . . . .. . . . ......... . ... . ... . . . . . . . 7-9
SORT/ MERGE Facility ................. . ... . . . ... . .... 7-1 0

6 THE OPERATING SYSTEM
INTRODUCTION ..... .. .. . . .. .. . . . . . . . . . ... . ....... .. . 6-1
COMPONENTS AND SERVICES . . ........... . .......... 6-1
PROCESSING CONCEPTS . . .............•... . . . . . . . . . 6- 2
Programs and Processes . . . . . . . ......... ...... . .. . . . . . 6-2
Resource Allocation ... . . . . . . . . ... .. . . ... . .. . .. .. . .... 6-4
Privileges .......... . . . .. . . . . . . . ... . .... . . . .... . . . . ... 6-4
Protection .. .. . . ............. . .......... . . . . .. ........ 6-4

Symbolic Characters Facility ......... . . . . ... .... .. .... 7-1 0
CALL Facility ....................... . . . . . . ... . . . . . . .. 7-11
Source Library Facility ............. . . . ... . . . . . . .. . . ... 7-11
Shareable Programs . . . . . . ... . ....•.... . •....•.... . . 7-11
Debugging COBOL Programs . .... .. .. . . ... . . . . ...... 7-11
Source Translator Utility . . . . . . . . .. . ... . . . .. . . ... . ..... 7- 12
Source Program Formats . . ... . ...................... 7- 13
Additional Features ... . .......... . . . . . . . . . . . ......... 7- 13
SampleVAX-11 COBOL Code ...... . . . ... . . .. ... . .... 7- 13

USER PROCESS ENVIRONMENT . . . ..... .............. 6-5
Virtual Address Space Allocation . . .. .... .... . . . .. . .... 6-5
System Services . . . . . . . .................... . . .. .. . .... 6-5
System Services Table .. ....... .. . .... .... . . .......... 6-7
110 System Services . . . . . . ...... . ..... . ........... . . . . 6-10
Local Event Flags . . .. ...................... . . . . . ... .. 6-11
Asynchronous System Traps . . . . . . . ... . ............... 6-11
Exception Conditions and Condition Handlers . . . ... .... 6-11

VAX-11 BASIC . . . . ... . . . . . ................... . ...... 7- 14
Introduction . . . . . . . . .......... . ... . . . ...... .. . .. . . .. 7-14
General Characteristics . . .. . .. .. ..... . . . ... . .... . .... 7- 15
Structured Programming .... . . . ........... . ......... 7-15

INTERPROCESS COMMUNICATION AND CONTROL. ... 6-12
Process Control Services .. ....... . ... . . . . . .......... 6-12
iv

Data Types ........ . . .... . . .......•... . .. . . . . . .... . .. 7-15
Data Types Table ... . . .. ... . . . .... . .... . ...... . .... . . 7-15
Declarations ... .... .... .. ................... . .. . .... 7-15
Files and Records .. . . . .... . . .. . .. .. . . .... . ........... 7-17
Symbolic Characters .... . ........... • .. . .. . ... . .. ... 7-17
CALL Facility .. . ..... . ..... . . . ...... .. ...... . ... .. ... 7-18
Shareable Programs
.. . .. . ........ . .......... . . 7-18
Developing BASIC Programs . . ... • .. .. .... . ....... .. .. 7-18
The LOAD Command .. ... ..... . . .... .. . . .. ... . ...... 7-18
Error Handling .. . ....... . . .. .. .. .............• . ..... 7-18
.. . ...... .. ..... . ... . .. .. 7-18
Migration to VAX/VMS
Performance . . ... . . .... .. . . ... ........ .. . .. ... ... . . . 7-20
Additional Functions . . .... .. ..... . ........ ... .. • . .. . 7-20

8 PROGRAM DEVELOPMENT AND
SUPPORT FACILITIES
INTRODUCTION . . . . ............ . . . ....... . ........... 8-1
TEXT EDITORS . . . . . .
File Names and File Types

SOS EDITOR .... . . .... .. .. .. ..... . . . .... . . .. .. . ... .. 8-1
Initiating and Terminating SOS ....... . .. . .... ... .. .... 8-1
SOS Examples . . . . . . . . . . . . .
. .... . .. . .... . . . ..... 8-2
. .... .. . . ............. . ... .. 8-2
EDT EDITOR . . . . . . .
What EDT Does .. .. .... .. . . ..... .. .... . . . . . .. .. ...... 8-2
EDT SPECIAL FEATURES ......... . ......... . ... . .... 8-2
Editing with a Window .... . .. . .... . ........... .. ...... 8-2
Start-up File ...... . ..... .. .. . .. . .. . ....... . .... . . . ... . 8-2
HELP Facilities . . .
. .. .. . . .. ... .... . . .... . . . ..... 8-2
The Keypad .. . ....... .. ...... . . .. .....•. . . . ........ . . 8-2
Redefining Keypad Keys .. .. . .... . . . ...... ... .. . . . .... 8-3
The SET and SHOW Commands .... .... .... . . . ......... 8-3
Journal Processing ........ .. ........ . . . ......... . ..... 8-3
The EDT CAI Program .. .. .. . .. ... .. .. . ....... . . ...... 8-3
EDT Modes of Operation .. . .. ...... . . ..... ... ......... 8-3

VAX-11 PL/I

........ .. ..... .. . 7-21
Introduction . . .
. ... . ... . ....... . .. .. . .. .... .. .. 7-21
The G (General-Purpose) Subset . .... . . . .... . . . •. ..... 7-21
Program Structure .. . ....... . ......... .. .. . . ...... .. 7-22
Program Control ................ . .... . . . .... . . . .. . .. 7-22
Storage Control .... . .............. . . .. . . ... .. .... . ... 7-22
Input/Output ... . ........... . .......... . ... . ... . .. . .. 7-22
Attributes and Pictures . . . .. .... . ....... . . . ... ....... 7-22
Built-in Functions and Pseudovariables ...... .. ..... . .. 7-22
.. . ..... . . .. ..... ........... 7-23
Expressions
VAX-11 Extensions to the G Subset Standard .... . ..... 7-23
Procedure-Calling and Condition-Handling Extensions .. 7-23
Support of VAX-11 Record Management Services . ..... 7-23
Miscellaneous Extensions and Deviations ...... . . . .... . 7-23
Full PL/I Features Supported ..... . .... . .. . ..... .. ..... 7-23
Implementation-Defined Values and Features .... . . . ... 7-24
VAX-11 PL/I Programming Example ........... ... .... 7-24

SLP EDITOR ... . .. . ....... ... . .. ...•. .... ..... . . .. . . . . 8-3
Initiating and Terminating SLP ....... .... .. .. . ..... . . ... 8-3
SLP Input and Output Files ....... . . .. . . ......... . ..... 8-3
Correction Input File
. ... . . . ... . .. . ........ .. ..... 8-3
SLP Output File .. .. ........ . .. .. . . ....... . .. . . . ..... . 8-4
LINKER .... .... ........... . .. . ....... .. . . ......... ... 8-4
The LINK Command ......... .. . . .. . ......... . .. .. .... 8-4
Virtual Memory Allocation ... . . ........ . . . ..... . ........ 8-4
.............. 8-4
Resolution of Symbol ic References
Image Initialization ........................... . . . .. . ... 8-4
Overview of Linker Interface to Memory Management . . .. 8-4
Linker Input ............ ... . . .. . ..... .......... . . ..... 8-4
Object Module Files .... .. .. .. ............ . .. ... . . .... 8-5
Object Module Libraries ........ .... . .. . . ..... . .. ..... 8-5
Shareable Image Files ................. . . ... . .. .. . . . .. 8-5
Shareable Image Symbol Tables . . .............. . .. . .... 8-5
Linker Output
.. . .. . ................ . .... . .... ..... . 8-5

. . .. .. . .... . .. 7-26
VAX- 11 PASCAL . . . . . . . . .
Introduction . . . . .
. .......... .. ... .. ... . . . .. . . 7-26
Sample VAX-11 Pascal Code .......................... 7-27
Compiler Listing Format ... .
. ... . ........... . .... 7-27
Source Code Listing. . . . . . . .
. ........ . . . . .. . ... 7-27
Mach ine Code Listing .......... .. . .. . ............ .. . . 7-29
Cross-Reference Listing .............. .. . . ...... .. . ... 7-29
VAX-11 BLISS-32 ........... . . . ......... . ...... .. .. . . 7-29
Introduction . . . . . . . . . . . . . . .
. .. . .. . ...... 7-29
Features of BLISS-32 ........... . . .. . ...... . ..... ... 7-29
VAX-11 Machine-Specific Function Table ..... ......... 7-30
The VAX-11 BLISS-32 Compiler
.................. 7-31
Library and Require Files . . .. . . .. ....... . . ... . .. . .... 7-31
Macros ................ .. ......................... 7-31
Debugging . . .
. . ... . . ..... . ............... . .. .. 7-31
Transportability Features .. ... .. . ...... .. ... . ........ 7-31
VAX-11 BLISS-32 Sample Program ..... . . . . . . . .. . .. ... 7-32
VAX-11 CORAL66

COMMON RUN TIME PROCEDURE LIBRARY ............ 8-5
Resource Allocation Group (LIB$) ............... . .. . . . . 8-5
Signal and Condition Handling ................ . . .. . . . . 8-6
General Utility (LIB$) .... ....... . ..... . . .. . .. .. ... . .... 8-6
Mathematical Functions (MTH$) ............ . ..... . ..... 8-6
Language-Independent Support (OTS$) .......... • ..... 8-6
Language-Specific Support (FOR$, BAS$) ....... . ...... 8-6
String Processing (STA$) . ..... .. .... . ....... ... .. ... . . 8-6
System Procedures .... . ................ . ....... . . . . . 8-6
Compiled-Code Support Procedures . . ... ... . . .. . . . . .. 8-6
Error Processing Procedures ..... . ... .. . ......... . ... . 8-6

........ . .. . ...... ... .......... . .. 7-32

VAX-11MACRO ........ . . .. . . .... . .. . . . .. . ... .. . . .. 7-33
Symbols and Symbol Definitions .... .. . . ........ . ..... 7-33
Directives ... . ........ . .. . . . ...... ... .. . .... ....... . 7-33
Listing Control Directives ..... . .... .. . . ...... .. . . .... 7-34
Conditional Assembly Directives ....... .. .... . ... . ... . 7-34
Macro Definitions and Repeat Blocks .......... .. .... . . 7-34
Macro Calls and Structured Macro Libraries ........... . 7-34
Program Sectioning ... . .... . . .. ................ . . ... . 7-34

VAX-11 SYMBOLIC DEBUGGER ........... . . .. . . . . . .. .. 8-6
DEBUG Commands . . .. . .. .. . . ..... . ....... . . . .... . . . 8-7
THE LIBRARIAN UTILITY ..... • ... . .. . . .. . . . . .. . ....... 8-7
Librarian Routines .. . . .. . ........ . .... .. . . . .. . . . . . . ... 8-7
DCL LIBRARY Command .......................... . ... 8-7
COMMAND LANGUAGE PROCEDURES . . .. . . .......... 8-8
Passing Parameters to Command Procedures . ... .. . ... 8-8
Log ical Commands ....... . . . ........... . . . ........ . . 8-8
Lexical Functions .. .. ................... .. . . .. ... .. . . 8-8
Command Procedure Example ... . .. ......... ......... 8-8

PDP-11 BASIC- PLUS-2/V AX . .. . ................. . .... 7-34
Program Format . . . .. . . . . ...... .. ......... . ... . . .... 7-35
Long Variable and Function Names ............. . . . . .. . 7-35
Powerful File 1/ 0 . . .. . ......... . ......... . . . . ..... .. . 7-35
Powerful String Handling ........ . ... . ... . ...... .... . 7-35
Virtual Arrays . ....... .. ... . .... .. .................. . . 7-35
PRINT USING Output Formats .. ...... . ...... . . .. ..... 7-35
Subprograms and the CALL Statement ... . ........ . . .. 7-35
COMMON Statement .. . ...... . . .... .... . ........ .... 7-35
Debugging Statements ..... . . . ..... . ....... . ... . . . .. 7-35
PDP-11 FORTRAN IV IV AX to RSX

. ..................... .. .. 8-1
... .. . ... . . . . ........ . ...... 8-1

DIFFERENCES UTILITY .......... ................ ...... 8-9
VAX-11 RUNOFF ....... .. . .. . .. . . .. . . ..... . ........ . 8-10
Filling and Justifying . . . .. .... . .......... . . .. . . ...... 8-10
Page Formatting ........... ........ . . .. ............. 8-10
Title Formatting . ... . ........ .. . .. ... .. . ... . . . ........ 8-10
Subject-Matter Formatting ..... . . .. . .. .... . .. .. ....... 8-10
Index and Table of Contents . ........... .............. 8-10
Miscellaneous Formatting .. ..................... . . ... 8-10

. . .......... . ....... 7-36

MACR0-11 . . ..... ...... ... . ... .. ............. . . . . . .. 7-36
Symbols and Symbol Defin itions ............... . •..... 7-36
Directives . ..... ........ ............ ...... . ......... 7-36

9 DATA MANAGEMENT SERVICES
INTRODUCTION .. . .. . ......... ........ . .... . ......... 9-1
FILE MANAGEMENT ... . .. . . .. . . . .. . . ..... . ... . ..... .. 9-1
File Directories and Directory Structures ...... . ... ... ... 9-1
File Specifications ... . ......... . .. . .. . . . ... .. . . .... . .. 9-1

10 DATA COMMUNICATIONS FACILITIES

Logical File Naming .. .. ...... . . . .. ... . .... .. ... ... . . . 9-3
File Management . .. ...... . . . .. . ...... . . . ......... . . . 9-4
RECORD MANAGEMENT SERVICES

INTRODUCTION ........................ . ........... 10-1

.. ... . . . . ...... ... 9-4

RMS FILE ORGANIZATIONS ..... . . . . .. ..•...... . . . . . . 9-5
Sequential File Organization . .. . . . . ...... . ... . . .. •.... 9-5
Relative File Organization .. . . ....... .. . . . .. ... .. . . ..... 9-5
Indexed File Organization ..... ... . . .......... .. . . ...... 9-5

DIGITAL NETWORK ARCHITECTURE . . . . ... . . ... . . . . .. 10-2
User Layer .. . . . . . . . . . .................... . ....... . .. 10-2
Network Services Layer .. . . . . . . . . . . . . . ... . . . . . . . . . . .. 10-2
Data Link Layer .. . .. . ...... . ..... . ................... 10-2
Pysical Link Layer . . ... . . . .. .. . . . . . . . . . . . . . . . . . . . . . . .. 10-2

RMS RECORD ACCESS MODES .. . .. . . .... . . . . . . ..... 9-6
Sequential Record Access Mode ..... . . . . ... .......... 9-6
Random Record Access Mode . ........ . . . . . .......... 9- 7
Record 's File Address (RFA) Record Access Mode . .. . . . .. 9-7
Dynamic Access ... . . .. ........ . .. . . . ........... . . . . .. 9-7

DECNET-VAX FEATURES . . .......................... 10-2
File Handl ing Using a Terminal . . . . . . . . . . ... . ... . . . .... 10- 2
File Handling Using Record Management Services . . . .. . 10-3
Network Command Terminal .................... . ..... 10-3
Intertask Communications ............. . ......... . . . .. 10-3

FILE AND RECORD ATTRIBUTES ........ . . ... . . ....... 9-7
Record Formats . . . .. . ......... . . . . . . . ........... . .... 9-8
Key Definitions or Indexed Files .. . ... .. . . . . .... . ....... 9-8

DIGITAL COMMAND LANGUAGE (DCL) FILE HANDLING 10-3
RECORD MANAGEMENT SERVICES FILE HANDLING

PROGRAM OPERATIONS ON RMS FILES . . ........ . . . . 9-9
File Processing . ......... . ....... . ......... . . . . . . . . . . 9-9
Record 1/ 0 Processing .. . . ......... . . . . .. ............. 9-9
Block 1/ 0 Processing ...... . . .. . ... ........ . ....... . . 9-10
RMS RUN TIME ENVIRONMENT . .. . . . ............... . 9-11
Run Time File Processing ...... . . . ........... • . . . .... 9-11
Run Time Record Processing ... . . . . . . . . ............. 9-11
RMS Record Locking ............ ....... ............. 9-12
LANGUAGE UTILITIES

MACRO TRANSPARENT INTERTASK
COMMUNICATION ...... .. . .. . . .. . .................. 10-4
Creating a Logical Link Between Tasks . ............... 10-4
Sending and Receiving Messages ...... .. ...... . . . ... 10-6
Disconnecting the Logical Link . . . . ........... . .. . ..... 10-6

. . .. . . . .. . . . . . .. . ....•........ 9-12

DATATRIEVE ......
. . .. . . . . ...... . . .... .. .. .. ..... 9- 12
DA TA TRIEVE Inquiry Facility ................. . ... . . . . . 9-12
DATATRIEVE Report Writer Facility ...... .. . ... . ..... . . 9-12
Basic Commands ............. . .. . . . . . .. . . ... . . ... . . . 9-12
Terminology . . .. .. .. .. . ...... . .... . ................. 9-13
Keywords . .. .... ...... . .... . .. . ..... ....... . . . . . . . . 9-13
Additional DATATRIEVE Features . . . . ..... . . . . . . ... . . 9-13
VAX-11 SORT/MERGE

. 10-4

SAMPLE VAX-11 FORTRAN INTERTASK
COMMUNICATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1 0-4
Creating a Log ical Link Between Tasks ..... . ... . . . . . . . 10-4
Send ing and Receiving Messages . . . . . . . .
. ... 10-4
Disconnecting the Logical Link ..... . ....... . . . . . . . . . . . 10-4
VAX-11 FORTRAN Intertask Communication Example .. 10-4

MACRO CALLS ........... . ........

. ... 10-6

MACRO NONTRANSPARENT INTERTASK
COMMUNICATION .. . .. . . . . . ........... . . .. .. . . . .... 10-6
Task Messages ... .. . . ...... . ..... . . ... ......... . .. . . 10-7
PROTOCOL EMULATORS (INTERNETS) . . . . . . ........ 10-7
VAX-112780/ 3780 Protocol Emulator .......... . . .. .... 10-7
MUX200/VAX Multiterminal Emulator ... . . . .... . .. . ... . 10-7

. . ...... .................. .... 9-14

VAX-11 SORT/ MERGE FEATURES .................... 9- 15
SORT / MERGE as a Set of Callable Subroutines . ... ... . 9-16
Filel / Olnterface . ..... . .. . . . . . . . .. ... . .............. 9-16
Record 1/ 0 Interface .......... . .. . . . . . . . ... . ........ 9-16
Programming Considerations . . . . . . . . . . . . . . . . . . . . .. 9-16

APPENDIX A
APPENDIX B

SORT / MERGE PERFORMANCE FEATURES .. .. . . . . .... 9-16

APPENDIXC

VAX-11 FORMS MANAGEMENT SYSTEM (FMS) .. . . .... 9-16
Using Forms in an Application ... . . . . . . . . . ....... . .... 9-16
Developing Applications with VAX-11 FMS .... .... .. . . .. 9-17
Maintaining VAX-11 FMS Applications ...... .. . . .. .... 9-17

GLOSSARY
INDEX

Introduction

The VAX Technical Summary introduces the characteristic features
and capabilities of the VAX system to computer analysts and system
programmers. Application programmers, and system managers and
operators may also use this summary as a tool to become familiar with
the components, services, and operations of the VAX system .

The intent of this summary is to familiar ize the reader with
the features and capabilities of the VAX (Virtual Address
Extension) system . It serves as a detailed technical introduction to the growing family of VAX processors, the
powerful and unique Virtual Memory operating System,
VAX/VMS , and the increasing number of supported peripherals . In particular, this edition of the VAX Technical
Summary introduces several significant system enhancements:
• the newest member of the VAX family of processors, the
VAX-11 / 750
• version 2.0 of the VAX/VMS operating system
• the addition of major new languages including VAX-11
FORTRAN , VAX-11 COBOL, VAX-11 BASIC , VAX-11
PL/I , VAX-11 PASCAL, VAX-11 CORAL 66, and VAX-11
BLISS-32
Although the VAX Technical Summary contains useful information for the application programmer, the system
manager, and the computer operator, the level of technical
detail makes the book particularly appropriate for the system programmer and/or computer system analyst.
As the reader proceeds through the technical summary,
the term " VAX architecture" or simply "architecture" may
seem confusing . VAX architecture is the collection of attributes that all family members have in common that assures software compatibility. For example, the architecture includes the instruction set, the addressing modes,
data types , etc. Examples of attributes not included in the
arch itecture are processor internal bus structure, implementation-specific privileged reg isters , execution
speed , etc. As new processors are added to the VAX family, a significant part of the engineering effort will be dedicated to preserving this software compatibility. This will
assure that programs written for today's VAX computers
will execute on future VAX systems.

in both sections are closely related ; for example, the respective discussions of memory management, 1/0 processing , and the compatibility mode environment should be
read together to gain a full appreciation for the system 's
design . These sections should prove beneficial to systems
programmers or analysts already familiar with an assembly language or software executive.
Perhaps one of the most important features of the VAX
system is that its programmers do not have to know assembly language to use the system effectively. Both the
hardware and software contain many features that promote efficient and productive high-level language programming . High - level language programmers should find
the Users, Languages , Data Management Services, and
Network Services sections, of particular interest. The beginning of the Operating System section is also helpful because it introduces some of the VAX software terminology
and concepts.

The VAX Technical Summary is designed to be read in either a selective or sequential manner. The reader might
start with section 1 and continue through the book , or first
glance through the table of contents to locate those topics
of most interest. As an added convenience, an abstract
can be found prefacing each section of the summary.
Many of the system 's concepts and features are repeated
throughout the text in various contexts . Appendix A contains a collection of most frequently used mnemonics and
their definitions.

Also of interest to high level language programmers is the
Program Development section which describes the basics
of creating , ed iting , debugging , and executing a program
written in any of the VAX high level languages. This section
also contains an introduction to some of the advanced features of the command language.

The following paragraphs introduce the VAX Technical
Summary.
The System section presents a comprehensive overview of
VAX system features . This section serves as an introduction for the reader presently familiar with computer industry terminology, identifying VAX characteristics and introducing the system features described in detail throughout
the remainder of the summary.
The Users section contains a description of the command
language and its features. It also introduces many of the
aspects of the system that support applications programming, system management, and operator control.

If the reader is uncertain about a particular term or phrase,
its definition can probably be found in the glossary at the
end of the summary. The glossary does not generally contain standard computer-related terms , but it does contain
most of the terms found throughout the VAX
documentation that have special meanings in the context
of the VAX system . The glossary is followed by a list of
mnemonics that may be encountered in the text. The mnemonics list is particularly useful during the system familiarization stage.

The VAX Family of Processors and Operating System
sections provide an in-depth discussion of the system 's
characteristics and capabilities. The concepts developed

For additional literature describing VAX features, capabilities and applications please contact the nearest DIGITAL
sales office.

1-1

2
The

System

The VAX system provides the performance, reliability, and programming features often found only in much larger systems. The VAX family
of processors have a 32-bit architecture based on the PDP- 11 fam ily of
16-bit minicomputers . While using addressing modes and stack structures similar to those of the PDP-11 , VAX provides 32- bit addressing
for a 4 gigabyte program address space, and 32- bit arithmetic and data
paths for processing speed and accuracy.
The processor's variable length instruction set and variety of data
types, including decimal and character string , promote high b it efficiency. The processor hardware and instruction set specifically implement many high-level language constructs and operating system functions.
VAX is a multiuser system for both program development and application system execution. It is a priority-scheduled, event-driven system :
the assigned priority and activities of a process in the system determine
the level of service needed. Real-time processes receive serv ice
according to their priority and ability to execute , wh ile the system manages CPU time and memory residency allocation for normal executing
processes.
VAX is a highly reliable system . Built-in protection mechanisms in both
the hardware and software ensure data integrity and system availability. On-line diagnostics and error detecting and logging ver ify system
integrity. Many hardware and software features provide rapid diagnosis and automatic recovery should power, hardware, or software fail.
The system is both flexible and extendible. The virtual memory operating system enables the programmer to write large programs that can
execute in both small and large memory configurations without
requiring the programmer to define overlays or later modify the program to take advantage of additional memory. The command language
enables users to modify or extend their command repertoire easily,
and allows applications to present their own command interface to
users.

• Network Services-includes the DECnet-VAX network
software and the DMC11 interprocessor communications link.

INTRODUCTION
VAX is a high performance multiprogramming computer
system ideally su ited for a wide variety of applications including real time, batch, time sharing, commercial, data
processing , and program development. The system combines a 32-bit architecture, efficient memory management,
and a virtual memory operating system to provide essentially unlimited program address space.

Processor
Architecturally , a VAX processor provides 32-bit addressing, sixteen 32-bit general registers, and 32 interrupt priority levels. The instruction set operates on integer, floating point, character and packed decimal strings , and bit
fields. The instruction set supports nine addressing
modes.

The processor's instruction set includes floating point,
packed decimal arithmetic, and character string instructions. Many of the instructions are direct counterparts for
high-level language statements. The software system supports programming languages that take advantage of
these instructions to produce extremely efficient code.

The processor includes an efficient memory cache resulting in greatly reduced memory access time.
Error Correcting Code (ECG) MOS memory is connected
to the memory interconnect via a memory controller . Each
memory controller includes a request buffer that substantially increases overall system throughput and eliminates
the need for interleaving in most applications.

The VAX/VMS virtual memory operating system provides
a multiuser, multilanguage programming environment on
the VAX hardware. The floating point instructions, efficient
scheduler , and optional VAX-11 FORTRAN language are
ideal for real-time and scientific computational environments. The optional VAX-11 COBOL language, data management services, and large capacity peripherals make
the system equally suited to commercial applications .
VAX/VMS supports many other optional high level languages suited for other applications. The system management facilities, command language, and program development tools provide the resources for efficient program applications development and execut ion . Spooling and
extensive job control capabilities support batch processing.

The processor uses two standard clocks-a programmable real-time clock used by the operating system and
by diagnostics, and a time-of-year clock used for system
operations . The time-of-year clock includes battery backup for automatic system restart operations.
The console is the operator 's interface to the central
processor. Via the console terminal, the operator can execute diagnostics, install new software, examine and deposit data in memory locations or processor registers , halt
the processor, step through instruction streams, and boot
the operating system .

The processor executes variable length instructions in native mode, and non-privileged PDP-11 instructions in compatibility mode. Native mode includes the PDP-11 data
types, and uses addressing modes and instructions similar
to those of the PDP-11 . The software supports compatible
languages and file and record formats .

Virtual Memory Operating System
VAX/VMS is a multiuser, multifunction virtual memory operating system that supports multiple languages, an easy
to use interactive command interface, and program development tools. The VAX/VMS operating system is designed
for many applications, including scientific/real-time, computational , data processing , transaction processing, and
batch .

COMPONENTS

The operating system performs process-oriented paging,
which allows execution of programs that may be larger
than the physical memory allocated to them. Paging is
handled automatically by the system, freeing the user from
any need to structure the program. In the VAX/VMS operating system , a process pages only against itself-thus individual processes cannot significantly degrade the performance of other processes.

The major components of a VAX system are:
• Processor-includes the basic central processing unit
implementing the VAX architecture. The specific implementations of the VAX processors will be treated in
greater detai l in Section 4, The VAX Processors.
• Operating System-includes a virtual memory manager, swapper, system services , device drivers, file system , record management services, command language,
and operator's and system manager's tools.

The memory management facilities provided by the operating system can be controlled by the user. Any program
can prevent pages from being paged out of its working set.
With sufficient privilege, it can prevent the entire working
set from being swapped out , to optimize program
performance for real-time or interactive environments.
Sharing and protection are provided for individual 512byte pages . The processor's memory management includes four hierarch ical processor access modes that are
used by the operating system to provide read/write page
protection between user software and system software.
The access modes from most to least privileged are kernel , executive , supervisor, and user.

• Languages-includes the native mode languages VAX11 MACRO and optionally, VAX-11 FORTRAN, VAX-11
COBOL, VAX-11 BASIC , VAX-11 PL/I , VAX-11 PASCAL,
VAX-11 BLISS-32 , and VAX-11 CORAL 66 . Also supported in compatibility mode are PDP-11 BASIC-PLUS2/V AX , PDP-11 FORTRAN IV/VAX to RSX , and
MACR0-11 . Development tools for both native and
compatibility mode programs include editors , linkers, librarians, and debuggers.
• Peripherals-includes a range of small- and large-capacity disk drives, magnetic tape systems, hard copy
and video term inals, line printers, card readers, and
real-time 1/0 devices.

VAX/VMS schedules CPU time and memory residency on
a pre-emptive prior ity basis. Thus, real-time processes do
2-1

not have to compete with lower priority processes for
scheduling services. Scheduling rotates among processes
of the same priority. The scheduler adjusts the priorities of
processes assigned to one of the low 16 priorities to i mprove responsiveness and to overlap 1/0 and computation . Real-time processes can be placed in one of the top
16 scheduling priorities , in which case the scheduler does
not alter their priority. Their priorities can be altered by the
system manager or an appropriately privileged user.

SIC-PLUS-2/VAX,PDP-11 FORTRAN IV/VAX to RS X, and
MACR0-11 lang:.iage processors produce compatibility
mode code.
The VAX/VMS operating system provides a file and record
management facility that allows the user to create , access ,
and maintain data files and records within the fi les with full
protection . The record management services handle sequential , relative , and multi-key indexed file organ izations,
sequential and random record access , and f ixed and variable-length records. Indexed files with sequential and random record access are available to compatibility mod e
programs , such as those written in PDP-11 BASIC- PLUS2/VAX.

Interprocess communication is provided through shared
files and shared address space, event flags , and mailboxes wh ich are record-oriented virtual devices.
VAX/VMS provides system management facilities . A system manager and a system operator are given the tools
necessary to control the system configuration and the operations of system users for maximum efficiency.

The VAX/VMS operating system supports the Files-11 OnDisk Structure Level 2 (ODS-2) , which provides the facilities for file creation , extension , and deletion with ownerspecified protections and multilevel d irectories . On-D isk
Structure Level 2 is upwardly compatible with Leve l 1, the
file system currently available under the PDP-11 IAS and
RSX-11 operating systems. Both native and compatibility
mode programs can access both Level 1 and Level 2 volume structures.

The VAX/VMS command language is easy to learn and
use, and is suitable for both the interactive and batch environments. Extensive batch facilities under VAX/VMS include job control, multistream spooled input and output ,
operator control, conditional command branch ing and accounting .

DECnet-VAX networ king capabilities are availab le as an
option for point-to-point interprocess communicati on , file
access, and file transfer , and include down-line com mand
file and RSX-11 S system image loading. The Network
Command Terminal facility allows users on one system to
log into another VAX system on the network . The Mail facility allows electronic mail to be addressed to users on
any VAX node in the network.

Command procedures are supported by the command
languages as an easy method of executing strings of frequently used sequences of commands, or creating new
commands from the existing command set. Command
procedures accept parameters and can include extensive
control flow.
VAX/VMS provides a program development capability
that includes editors, language processors, and a symbolic debugger. The VAX-11 FORTRAN, VAX-11 COBOL,
VAX-11 BASIC, VAX-11 PL/I , VAX-11 PASCAL, VAX-11
BLISS-32, VAX-11 CORAL 66, and VAX-11 MACRO language processors produce native code . The PDP-11 BA-

Peripherals
Medium capacity disks, unit record devices, termi nal s, the
interprocessor commun ications links, and user-spec ific
devices are UNIBUS peripherals. The UNIBUS adaptor(s)

"'.

. f4 ¥

mll VAUro

~~£.~ ~

•

·U· .... -~-:

111111111111111111111111111111111111111111
111111111111111111111111111111111111111111

(

2-2

provides the hardware pathways for data and control information to move between the UNIBUS and the memory
interconnect.

• small constant short literals

High-performance MASSBUS mass storage peripherals
are connected to the memory interconnect via a buffered
MASSBUS adaptor. The MASS BUS adaptors provide the
hardware pathways for data and control information to
move rapidly between a MASSBUS peripheral controller
and the memory interconnect.

• the Add Compare and Branch instruction for DO and
FOR loop control

Furthermore, many instructions correspond d irectly to
high-level language constructs:

• the Case instruction for computed GO TO statements
• the 3-operand arithmetic instructions for statements
such as " A = B+C"
• the Index instruction for computing index values, including subscript range checking

PERFORMANCE
Many features of VAX ensure that the system provides
real-time , computation , and transaction processing applications with the processing speed and throughput they
need .

• the Edit instruction for output formatting
Much of the processor architecture also ensures that the
operating system incurs minimal overhead for real-time
multiprogramm ing . For example, the operating system
uses:

The processor provides 64-bit, 32-bit, 16-bit, and 8-bit
arithmetic, instruction prefetch, and an address translation
buffer.

• the context-switching instructions and queue instructions to schedule processes

An optional high-performance floating point accelerator
(FPA ) can be added to the VAX-11 /780 system. The FPA is
an independent processor executing in parallel with the
base CPU. The FPA takes advantage of the CPU 's instruction buffer to prefetch instructions and memory cache to
access main memory. Once the CPU has the required data, the FPA overrides the normal execution flow of the
standard floating point microcode and forces use of its
own code. Then , while the FPA is executing , the CPU can
be perform ing other operations in parallel. The FPA execu tes standard floating point instructions with substantial
performance improvement. This execution is architecturally transparent to the programmer. In addition , the FPA
enhances the performance of a number of additional instructions including :

• the asynchronous system trap (AST) delivery mechanism to speed returns from system service calls
Careful design , coding , and performance measurement
ensure that the flow within the system is as rapid as possible.

RELIABILITY
Built-in reliability features for both hardware and software
provide data integrity, increased up-time , and fast system
recovery from power, hardware, or software failures . Several of the VAX rel iab ility features are discussed in the following paragraphs .

• extended multiply and integerize (EMOD)

Data Integrity
VAX provides memory access protection both between
and with in processes. Each process has its own independent virtual address space which can be mapped to private pages or shared pages . A process cannot access any
other process' private pages. The VAX/VMS operating
system uses the four processor access modes to read
and/or write protect individual pages within a process.
Protection of shared pages of memory, files , and interprocess communicat ion facilities such as mailboxes and
event flags is based on file ownership and application
group identification .

• polynom ial evaluation (POLY) , (F _f loating and
D_floating formats for both instructions)
• all floating to integer and integer to floating conversions
• 8- and 16-bit integer multiply (MULB and MULW)
• extended multiply (EMUL)
• 32-bit integer multiply (MULL)
The EMOD instruction is used for fast, accurate range reduction of mathematical function arguments. The POLY
instruction is used extensively in the evaluation of mathematical functions such as sine and cosine (the VAX/VMS
mathematics library makes use of the POLY instruction to
sign ificantly reduce the execution time of its subroutines).
The MULL instruction is often used in matrix manipulation
subscript calculations.

The VAX/VMS file system detects bad blocks dynamically
and prevents re-use once the files to which they are allocated are deleted .
Integral fault detection hardware includes:

The VAX processor architecture is spec ifically designed to
support high-level language programm ing . Because the
instruction set is extremely bit efficient, compilation of high
level language user programs is also very efficient. Among
the features of the processo r that serve to reduce program
size and increase execution speed are the :

• memory error correcting code that detects all double-bit
errors and corrects all single-bit errors

• variable length instruction format

• peripheral write-verify checking that checks all input
and output disk and tape operations and ensures data
reliability

• disk error correcting code that detects all errors up to
11 bits and corrects errors in a single burst of 11 bits
• extensive parity check ing

• floating point, packed decimal, and character string data types

• track offset retry hardware that enables the operating
system to recover from many disk transfer errors

• indexed , short displacement, and program counter relative addressing modes
2-3

. ,:.:;

"--

System Availability
The VAX/VMS operating system allows the VAX system to
continue running even though some of its hardware components have failed. The system automatically determines
the presence of peripherals on the processor at bootstrap
time. If the usual system device is unavailable, the system
can be bootstrapped from any disk. If memory units are
defective, memory is configured so that defective modules
are not referenced . Software spooling allows output to be
generated even if the normal output devices are not available. Also, devices can be added on line.

system automatically performs machine checks and internal software consistency checks during system operation .
If the checks fail , VAX/VMS performs a system dump and
reboots the system if the operator has set the system for
auto-restart.
Memory battery backup can be used to preserve the con tents of memory during a power outage. A special memory
configuration register indicates to the recovery software
whether data in memory was lost. Following a power failure, the recovery software restarts all possible 1/0 in progress before the failure occurred. Programs can use the
VAX/VMS power-fail asynchronous system trap facil ity to
initiate user-specific power fail recovery processing . If
memory battery backup is used, the time-of-year clock allows the recovery software to calculate elapsed time of the
outage.

The system operator can perform software maintenance
activities without bringing the system down for stand-alone
use. The operator can perform disk backup for both full
volume backup/restore and single file backup/restore
concurrent with normal activities .
The VAX/VMS operating system supports on-line peripherals diagnostics. VAX/VMS performs on-line error logging of CPU errors, memory errors, peripheral errors, and
software failures. The operator or field service engineer
can examine and analyze the error log file while the system
is in operation.

VAX remote diagnosis allows DIGITAL to run diagnostics,
examine memory locations , and diagnose hard ware/software problems from a remote diagnosis center.
The engineer who goes to the site is prepared in advance
to correct any problems that may have occurred .

FLEXIBILITY

System Recovery
Automatic system restart facilities bring up the system
without operator intervention after a system failure caused
by a power interruption, a machine check hardware malfunction, or a fatal software error. The VAX/VMS operating

VAX is a system that is easy to use because it is both flexible and easy to extend. Several of the ways in which VAX
provides the user with flexible operating and programming
environments are introduced below.

2-4

• Interactive editors with CAI startup-VAX/VMS supports several interactive and batch text editors,
including SOS, SLP, and now the DIGITAL standard editor EDT. The system features a computer-aided instruction course to introduce the user to the power and flexibility of EDT.

Flexibility in the Operating Environment
Virtual memory gives the user the ability to write and execute arbitrarily large programs without concern for
addressing limitations. The paging and swapping algorithms allow more programs to execute than the available
physical memory would allow if all programs had to be totally resident.

• Interactive symbolic debugger-The interactive symbolic debugger provides a fast and efficient method by
which the user can trace program errors. The debugger
offers the user such features as; support of various native languages, support of many data types, and its interactive symbolic operation, i.e., the user can refer to
program locations using those symbols created within
the program.

Both paging and swapping are transparent to the user,
and therefore allow the system to be extended without reprogramming . The system's physical memory configuration can change without requiring program redesign or relinking . Programmers never have to structure their programs, although they can, at their option, to achieve maximum efficiency and performance for a given program .
They can control working set size, lock pages in the working set or memory, and lock an entire working set in memory. In addition, the system manager can control the amount of time a process is guaranteed memory residency
once it is swapped in .

• FMS interactive screen format generation-The Forms
Management System contains an interactive editor
which allows the application programmer to create
and/or modify screen formats.
• DATATRIEVE-DATATRIEVE software provides fast
and convenient access to data within a file or files. This
query/report writing system provides the user with either video or hardcopy output.

The VAX/VMS scheduler recognizes 32 scheduling priorities. A program can modify its priority during execution .
Real-time processes execute at one of the high 16 priority
levels, and normal processes (including system processes) execute at one of the low 16 priority levels . The scheduler may temporarily increase the priority of a normal
process to increase its response to 1/0 events or system
events (but it can never lower the priority of real-time process).

• HELP facility-The HELP facility provides the user with
on-line instructions pertaining to selected system operations.
Extending the System
The VAX/VMS command language can be extended with
user-defined commands through the use of command
procedures. A command procedure is a set of commands,
data, or other command procedures processed in sequence. The user can invoke command procedures by a
single command that can include parameters for the procedure, such as file names or values for symbols. Command procedures can execute programs, transfer control
within the command procedure conditionally or unconditionally, request input from the user, and manipulate
numeric and string symbols.

Batch and printer output processing are completely flexible. The operator controls the number of batch jobs that
can run concurrently. The operator defines the number of
spooler queues. There can be multiple print queues: a
generic queue for jobs that can be output on any printer,
and several queues for jobs that are designated for a specific printer.
Batch jobs can be submitted to batch streams from the interactive environment using a terminal command, from
another batch job, or by any program using a system call .
Submitted batch jobs are queued, and a time can be specified after which a batch job should be executed .

VAX/VMS uses a standard procedure call interface supported by the processor's call instructions. The calling
program and called procedure can be written in different
languages. This contributes to the writing of error-free,
modular, and maintainable software that can be shared by
many programs. The standard procedure call interface is
particularly useful to systems programmers who want to
add their own shareable libraries and library procedures
to the VAX/VMS Common Run Time Procedure Library.

Flexibility in Programming Interfaces
The 1/0 programming facilities can be as device-independent or device-specific as desired. The record management services support high-level programming languages
by providing transparent record access and also enable
the programmer to request specific record management
services or system services to control file allocation, record blocking, or record accessing directly. Programmers
can also use the system services to access file-structured
devices or non-file structured devices if they wish to use
their own record processing or volume structuring techniques.

PDP-11 Compatibility
Users who already know the PDP-11 will find the native
VAX-11 instruction set and programming characteristics
easy to learn when developing new applications for the
VAX system. The PDP-11 and the VAX systems have almost identical FORTRAN, BASIC, and COBOL languages.
Users who have programmed in any of these languages on
the PDP-11 will need to spend very little time learning the
VAX system .

Access to network facilities is device-independent, but a
user who so desires can exert control over operations to
obtain error reports or notification of broken connections
(interrupt messages, inbound connections) . System access protection applies to all network access.

VAX offers many PDP-11 compatibility features :
• the VAX processor can execute a subset of PDP-11 16bit instructions in compatibility mode

Programmer Productivity
In addition to the system's reliability and performance features described above, VAX offers many tools to aid programmer productivity.

• the VAX/VMS operating system provides functionally
equivalent system services for many RSX-11 M execu-

2-5

tive directives

record processing methods that are upwardly co mpatible with RMS-11 record management services

• the VAX/VMS high-level language compilers accept
source languages that are upwardly compatible with the
same PDP-11 compilers

• the VAX/VMS operating system provides an RSX-11
MCA command language interpreter
• the DECnet-VAX package supports RSX-11S system
image down-line loading

• the VAX/VMS file system can read and write disk volumes and magnetic tapes written under RSX-11 and IAS
operating systems

These features make VAX an ideal host system to PDP-11
systems in a distributed processing environment.

• the VAX/VMS record management services provide

2-6

3
The
Users

The VAX system is designed to execute many different kinds of jobs
concurrently. Jobs consist of one or more processes, each of which
can be executing a program image that interacts with on-l ine users ,
controls peripheral equipment, and communicates with other jobs in
the same system or in remote computer systems. Jobs include :
• customer-written interactive, batch, and real-time applications
• interactive and batch program development jobs
• system management and control jobs
Typically, VAX users interact with application or system jobs via an online terminal , or benefit from production batch or real-time jobs . To aid
in the development of interactive, batch, and real-time applications ,
and manage and control system resources , VAX enables:

• The application programmer to write, compile, and test programs interactively or in batch mode, taking advantage of source code, object
code, and program image libraries.
• The system programmer to design application systems that require a
high degree of job and process interaction , data sharing, response
time, and system and device independence .
• The system manager to authorize users, limit resource usage, and
grant or restrict privileges individually.
• The system operator to monitor operations , service user requests ,
and control batch production .
Users can directly control the operation of VAX through the operating
system 's command language. In general, the command language is
used by programmers to develop application software, by operators to
monitor the system , and by system managers to assign user privileges .
Application programmers may also employ the command language to
execute their application programs explicitly. The command language
may be easily extended to provide custom-tailored commands defined
by the user. Customer-written application programs can prov ide their
own command interfaces for people using the system . Transaction
processing applications may require several terminals to be slave terminals, meaning that they are tied to particular application programs
that handle requests entered by the user.
The system manager can assign access user names and passwords to
users who log on the system at a command terminal, and determine
their privileges for obtaining services and limits for using resources .
Users who access the system through an application terminal interface
have the resources and privileges granted to the application programs
run on their behalf. An application program itself determines who can
request its services.

THE APPLICATION PROGRAMMER

The command interpreter displays the dollar sign to
prompt for a command , and the dollar sign underscore to
prompt for a missing parameter.

The application programmer has four basic tools for requesting services of the system :
• command interpreter

Command Procedures
To eliminate the need for typing frequently repeated sequences of commands , users can create command procedures . A command procedure is a file containing complete command lines (including the $ prompt character) .
The user can request the command interpreter to read and
process the command lines in a command procedure file
just as if they were being typed successively at the terminal. To execute a command procedure, the user simply
precedes the name of the command procedure file with an
"at" sign( @):

• programm ing languages
• programmed file and record management services
• programmed system services
The application programmer gains access to the system
through the command interpreter. The command interpreter enables the programmer to create, compile, and execute programs written in any of several programming languages. The record management services are available
through any programming language to provide device-independent data processing . The system services , although primarily of interest to systems programmers, are
also available to the application programmer for requesting special services of the operating system .

$ @procedure-file
Figure 3-1 is a simple example of how the user might create, interactively, a new command called EXECUTE. The
user has previously written a VAX-11 PASCAL program
named AVE, and now wishes to compile, link, and execute
this program. To the user, EXECUTE appears to be a command like any other command in the DCL command repertoire. The first step is to create the command procedure file EXECUTE.COM. Following this, the user enters
the PASCAL, LINK , and RUN DCL commands as input to
the command file .

Command Language
The command interpreter is interactive, comprehensive,
easy to use, and extremely flexible. It enables the user to
log onto the system , manipulate files , develop and test
programs , and obtain system information . Furthermore, it
enables users to extend or redefine their command repertoire as well as write command procedures easily. The
command language includes:

The blue dollar signs in Figure 3-1 represent DCL system
prompts, the remainder of the text is user supplied.

• a set of English commands that provide the basic com mand reperto ire
• a set of control characters that provide special functions
such as erase command line, interrupt the program currently executing , etc.

$ CREATE EXECUTE.COM
$PASCAL AVE
$LINKAVE
$RUN AVE
(CONTROL Z)
$

• a set of special operators and commands that can be
used to automate command streams and extend the
command repertoire

Table 3-1 lists the basic set of English commands accepted by the command interpreter. The command interpreter
is easy to use because its commands can be abbreviated ,
because it prompts for necessary arguments , and because it assumes standard or user-selected default values
for command parameters and qualifiers.

Figure 3-1
A Simple Command Procedure

A command line normally consists of a command verb followed by one or more parameters that identify the object
of the operation (a file, for example) or qualify how the operation is to be performed. If the interactive user enters an
incomplete command, the command interpreter prompts
for any necessary parameters. For example , the COPY
command, which creates a copy of an existing file, accepts
a total of two file specifications: one for the file to be created and one for the file to be copied. The file specifications
identify the exact location and name of the files . The COPY
command can be entered in any of several ways:

Now, to execute the command procedure file , EXECUTE.COM , the user can type:
$ @EXECUTE
or the user can create a symbol to execute the command
file . The user may equate a unique series of letters to the
command procedure file . For instance:
$EXE:= @EXECUTE
Now , to execute the command procedure file,
EXECUTE.COM , the user need only type the symbol EXE:
$EXE

$copy
$_FROM : file-name-1
$ TO : file-name-2

The user could just as easily have created a symbol called
GO to execute the command procedure file EXECUTE.COM . In this instance:
$ GO :=@EXECUTE

$copy file-name-1
$_TO : file-name-2

Now, to execute the command file, EXECUTE.COM, the
user can enter GO in response to the DCL prompt($):
$GO

$copy file-name-1 file-name-2
3-1

Table 3-1
DCL Command Language Summary
BATCH AND COMMAND PROCEDURE
SPECIFIC CONTROL *
Places a given batch command file or comSUBMIT
mand procedure in a batch queue for processing .

GENERAL SESSION INFORMATION
AND CONTROL
The user initiates an interactive session with
LOGIN
the system by typing CTRL/C , CTRL/Y or by
pressing the carriage return on a term inal not
currently in use. The system then prompts for
username and password, and validates them .
HELP

Displays information to assist the user in selecting the proper command qualifiers.

SHOW

Displays any of the following information : current day , date, and time; current default device and directory name; status of devices in
the system ; logical device name assignments;
current characteristics and status of specified
mag tape device; name, number, and status
of local network node and lists available remote nodes; default characteristics of system
printer; status of current process ; current file
protection to be applied to all new files created during terminal session or batch job ; current status of entries in printer / batch job
queues; current disk quota; current VAX-11
RMS default multi block and multibuffer
counts; status of currently executing image in
process; current value of a local or global
symbol ; displays a list of processes and status
information in system ; current characteristics
of a specified terminal ; logical name translations; display of working set quota and limit
assigned to current process.

SET

ASSIGN

Defines default translation mode for cards
read into system card reader; controls whether command interpreter receives control
when CTRL/Y is pressed ; changes the user's
default device name or directory name; defines default characteristics for specific magtape device; determines whether command
interpreter performs error checking following
execution of commands in command procedures; changes execution characteristics of
currently executing process; changes a file's
protection ; changes current status or attributes of a file queued for printing or for batch
job execution; defines default values for multiblock and multibuffer counts used by VAX-11
RMS; changes characteristics of a specified
terminal ; controls whether or not command
lines in command procedures are displayed
at terminal or printed in batch job log ; redefines default working set size for the current
process.

$PASSWORD

Specifies the password associated with the
user name specified on a JOB card for a batch
job.

$JOB

Indicates the beginn ing of a batch com mand
file and provides job control information (such
as time limit) .

$INQUIRE

Requests interactive assignment of a value to
a symbol and assigns the symbol a nam e.

$GOTO

Transfers flow of co ntrol to a given label ed
line .

$ON

Transfers flow of control to a given labeled
line if an error of a given severity or greater is
encountered at any time during comma nd
procedure processing .

$IF

Transfers flow of control to a given labeled
line if the result of a logical compar ison of
symbol ic values is true.

$EOD

Signifies the end of data in the input stream
following a $DATA command .

$EXIT

Term inates the command procedure.

$EOJ

Marks the end of a batch job submitted
through the system card reader . EOJ performs the same functions as the LOGO UT
command.

DECK

Marks the beginn ing of an input stream for a
command or program .

•command names preceded by a$ are meaningful only in a batch command file or command procedure . All other commands listed in this table can either be issued interactively or used in a batch com mand file or
command procedure.
VOLUME AND DEVICE RESOURCE CONTROL

Assigns a logical name to a given character
string (equivalence name) and stores the the
pair of names in a process, group , or system
logical name table. Generally used to create a
logical name for a device.

MOUNT

Requests the operator to make a volum e
available to the user and optionally associates
a logical name with the volume or volume set.

INITIALIZE

Writes a directory file and other volum e structuring information on a disk or magnetic ta pe
volume to prepare it for use.

DISMOUNT

Requests the operator to break the log ical association of this device with the user's job.

ALLOCATE

Obtains exclusive ownership of device and
enables the user to assign a logical name to
the device .

DEALLOCATE

Releases allocated devices.

FILE MANIPULATION
DIRECTORY

Reports information (size, protection , ownership, creation time, etc.) on a given fil e or set
of files .

DEFINE

Creates a logical name equivalence. (Same as
ASSIGN except for syntax.)

CREATE

DEASSIGN

Breaks the correspondence between a logical
name and its equivalence name (see ASSIGN
and ALLOCATE) , or deletes a symbol (see
DEFINE) .

Creates a new file from data subsequently entered in the input stream (user at term inal or
batch stream). Creates a directory file on a
volume.

EDIT

MGR

Signifies that the given command or following
command lines are to be interpreted by the
RSX-11 command interpreter.

Opens a text file and accepts comman ds to insert, delete or mod ify data in the file .

DELETE

Deletes one or more files from a mass storage
disk volume .

LOGOUT

Terminates an interactive session and releases all resources allocated to the user.

DELETE/
ENTRY

Deletes one or more entries from a printer or
batch job queue.

REQUEST

Displays a message at a system operator's
terminal and optionally requests a reply .

DELETE/
SYMBOL

Deletes one or more symbol definitio n from a
local symbol table or from the global symbol
table.

3-2

Table 3-1 (con '!)
DCL Command Language Summary

PURGE

Deletes all but the latest version of a given file
or files , optionally keeping the latest two or
more versions.

COBOL

Compiles given COBOL language source
modules using VAX-11 COBOL compiler , producing an object module.

RENAME

Changes the name of one or more existing
files.

BASIC

Compiles given BASIC language source modules , producing an object module.

COP Y

Copies the contents of a file or files, creating
another file or files.

APPEND

Concatenates the contents of sequential fi les
to a given output file , or creates a new output
file from the concatenated contents of given
sequential files.

DIFFERENCES

Compares two files and reports the differences between the two.

SORT

Creates a file by rearranging the records in a
given file based on the contents of key fields
within the records .

OPEN

Opens a file for reading or wr iting at the command level.

BLISS

Invokes the VAX-11 BLISS-32 compiler .

PL/I

Invokes the VAX-11 PL/I compiler to compile
one or more source programs .

PASCAL

Invokes the VAX-11 PASCAL compiler to
compile one or more source programs.

CORAL

Invokes the VAX-11CORAL66 compiler to
compile one or more CORAL source programs.

LIBRARY

Creates, deletes, or maintains libraries of object modules, shareable images, or macro
source modules.

LINK

Links modules to produce images.

RUN

Executes a program image in the current
process context, or creates a detached process and executes a program image in that
process context.

DEBUG

Starts interactive debugging session after interrupting program image execution by typ ing
a Control C or Control Y.

EXAMINE

Displays the contents of a location in virtual
memory.

Closes a file that was opened for input or output with the open command and deassigns
the logical name specified when the file was
opened.

READ

Reads a single record from a specified input
file and equates the record to a specified symbol name.

WRITE

Writes a record to a specified output file.

Sends the contents of a given file or files to a
spooled output device such as a line printer.

DEPOSIT

Replaces the contents of a location in virtual
memory with the given data.

TYPE

Displays the contents of a given file or files on
the device identified by the logical name
SYS$0UTPUT : (default generally the user's
terminal) .

CONTINUE

Resumes execution of a program interrupted
by typing a Control C or Control Y.

STOP

Terminates the program currently interrupted
by a Control C or Control Y.

SUBMIT

Enters a command procedure in the batch job
queue.

SYNCHRONIZE

Places the process executing a command
procedure in a wait state until a specified
batch job completes execution.

WAIT

Places the current process in a wait state until
a specified period of time has elapsed .

CANCEL

Cancels scheduled wakeup requests for a
specified process. This includes wakeups
scheduled with the run command and with the
schedule wakeup ($SCHDWK) system service.

DUMP

Produces a printed listing of the contents of a
file , ignoring any print formatting characters
that may appear in the records .

UNLOCK

Permits access to a file that was improperly
closed .

ANALYZE /
OBJECT

Provides a description of the contents of an
object file or an executable image file.

PROGRAM DEVELOPMENT AND
EXECUTION CONTROL
MACRO

Assembles given assembly lang uage source
modules, producing an object module .

FORTRAN

Invokes the VAX-11 FORTRAN compiler to
compile one or more source programs.

In addition to executing command procedures at a terminal , an interactive user can also submit batch jobs. Batch
jobs execute under control of the system operator and
leave the user's terminal free to continue interactive or
command procedure processing. A batch job can be submitted as a deck of cards or as a batch command file. A
batch command file is identical to a command procedure
file, except that a batch command file submitted as a deck
of cards begins with a $JOB card that provides job control
information.

and accepts a series of up to eight user-defined parameters 'P1' through 'PS'. DCL can be used to completely define and control a user environment tailored to a specific
application .
The power and flexibility of command procedures and the
DCL command language will be treated in greater detail in
the Program Development and Support Facilities section .
RUN Command
The RUN command includes several qualifiers (DELAY,
INTERVAL, and SCHEDULE) which are of particular importance to the real-time programmer.

However, DCL is more than just a string of commands capable of standing alone; it possesses true high level programming language statements such as GOTO, IF, etc.,

Specifying any of the above qualifiers places a process in

3-3

• performing sorting and merging operations at the COBOL source language level

hibernation, a wait state in which the process can be reactivated only when a particular time value occurs. The time
value can be specified in delta time (/DELAY qualifier), in
absolute time (/SCHEDULE qualifier) or at recurrent intervals (/INTERVAL qualifier). When the image completes execution , the process returns to a state of hibernation .

• complete file organization capability including sequential , relative, and indexed 1/0
• structured programming
• support for the full range of data types including packed
decimal and floating point

Programming Languages
The system includes the VAX-11 MACRO assembler for
programming the machine using its native instruction set.
A wide variety of language processors are optionally available to high-level language programmers: VAX-11 FORTRAN , VAX-11 COBOL, VAX-11 BASIC , VAX-11 PL/I ,
VAX-11 PASCAL, VAX-11 CORAL 66, and VAX-11 BLISS32. In addition, VAX/VMS supports several optional language compilers that execute in compatibility mode.
These include PDP-11 BASIC-PLUS-2/VAX, PDP-11 FORTRAN IV IVAX to RSX and MACR0-11. These language
processors, introduced below, are described fully in the
Languages section .

• COBOL programs can call external routines written in
COBOL or other VAX-supported high level languages
• capability of writing shareable code for use by other native mode high level languages
• accepts source programs coded in either ANSI standard format or the shorter easy to read DIGITAL term inal format
VAX-11 BASIC is a native mode, shareable language processing system producing shareable VAX native object
code. The language compiler utilizes VAX floating po int
and character string instructions while supporting a fast
RUN command and immediate mode execution wh ich
makes it well suited for interactive use. VAX-11 BASIC is a
superset of PDP-11 BASIC-PLUS-2, offering the VAX user
major enhancements such as:

The VAX-11 FORTRAN language processor is based on
the American National Standard FORTRAN specification
X3 .9-1977 (commonly referred to as FORTRAN-77). The
VAX- 11 FORTRAN compiler supports this standard at the
full language level. Additionally, however, VAX/VMS provides support for the industry-standard FORTRAN features based on FORTRAN-66 (an option that can be selected at compile time), the previous ANSI standard .

• access to VAX-11 RMS file and record processing
• long variable names (up to 31 alphanumeric characters)
• dynamic string handling
• CALL statement providing interface to common language environment

The VAX-11 FORTRAN compiler produces shareable,
highly optimized VAX-11 native object code. The compiler
takes advantage of the system's large virtual address
space while utilizing the floating point and character string
instructions. FORTRAN 1/0 processing is supported by the
record management services (VAX-11 RMS) . VAX-11
FORTRAN object modules can be linked with assemblerproduced object modules and the system 's run-time library , which is common to all native mode programs, to provide standard library functions . The VAX-11 FORTRAN
language processor offers the programmer such features
as:

• shareable and re-entrant code
VAX-11 PL/I is an extended implementation of the proposed ANSI X3 .74, American National Standard PL/I General Purpose Subset to full PL/I (ANSI X3.53-1976). VAX11 PL/I extensions include some full language features
and VAX/VMS system-specific extensions.
PL/I is a versatile language that is suited to commerc ial ,
scientific, and systems programming applications. Some
of the features of VAX-11 PL/I include:
• block structured language

• Full ANSl-77 FORTRAN language support

• full support for all VAX-11 hardware data types

• Access to ISAM files as well as relative and sequential
files

• powerful 1/0 capabilities including ISAM support

• Access to the VAX/VMS system services and the run
time library procedures

• user control of storage allocation

• FORTRAN program can call external routines written in
other VAX-supported high level languages

• standard VAX-11 CALL interface, including access to
VAX/VMS System Services and the run-time library

• The compiler itself is shareable

• fast, native-mode optimizing compiler

• condition handling

• shareable , position-independent code

The VAX-11 COBOL language processor produces highly
efficient shareable native mode code which utilizes the
system's packed decimal and character instruction set
and extended call facility . The VAX-11 COBOL language is
based on the American National Standard Programming
Language COBOL, X3.23-1974, the industry wide accepted standard for COBOL. Many features of the planned COBOL standard (anticipated in 1981) are also included . The
VAX-11 COBOL language processor offers the programmer such features as :

VAX-11 PASCAL, a re-entrant native mode compiler , is an
extended implementation of the PASCAL language as defined in the PASCAL User Manual and Report (Jensen and
Wirth , 1974). Particularly suited to instructional use, PASCAL is gaining increasing popularity as a general purpose
language. Major features of the VAX-11 PASCAL language
include:
• block structuring via the BEGIN ... END compound statement to allow easy logic flow

• the ability to manipulate data strings via the INSPECT
verb

• data structuring including the ability to declare and use
pointers, records, files and arrays
3-4

execute in the compatibility mode environment.

• predefined procedures and functions to deal with 1/0
handling and data manipulation

MACR0-11 , the PDP- 11 assembly language, is included in
the compatibility mode environment. Programs written in
MACR0-11 can be assembled to produce relocatable
object modules and optional assembly listings.

VAX-11 PASCAL takes advantage of the VAX hardware
floating point and character instructions as well as the
virtual memory capabilities of the VAX/VMS operating
system. Many of the features common to other native languages are available through VAX-11 PASCAL including :

Record Management Services
The record management services (RMS) are a collection
of procedures that extend the programming languages by
providing general purpose file and record handling capabilities. Programmers using RMS include in their programs statements that read, write, find, delete, and update
records within files. Records can be fixed or variable
length .

• separate compilation of modules
• standard CALL interface to routines written in other languages
• access to VAX/VMS system services
The VAX-11 CORAL 66 compiler executes in compatibility
mode and generates native mode object code under
VAX/VMS. The CORAL language, derived from JOVIAL
and ALGOL-60 in 1966, is the standard language prescribed by the British government for military real-time applications and systems implementation . VAX-11 CORAL
66 is essentially a high level block-structured language
whose compiler offers the user many features including:

RMS enables the programmer to choose the file organization and record access method appropriate for the data
processing application. The file organizations and record
access methods are independent of the language in which
they are programmed , although some languages support
file organizations and access methods not provided in others. Every programming language uses RMS to process
files organized to provide sequential, random or multikeyed indexed record accessing .

• several numeric types (byte, long and double)
• generation of re-entrant code at the procedure level
• code optimization

For further information on RMS and the system 's data
management techniques, refer to the section on Data
Management Facilities.

• Engl ish text error messages
• INCLUDE keyword to incorporate CORAL 66 source
code from user-defined files
VAX-11 BLISS-32 is a high-level systems implementation
language for VAX-11, which runs in native mode under
VAX/VMS . The BLISS language is specifically designed
for building language compilers, real-time processors,
utilities, and operating system software. BLISS contains
many of the features of high-level languages (e .g., DO
loops, IF-THEN-ELSE statements, automatic stack, and
mechanisms for defining and calling routines) , but it also
provides the flexibility and access to hardware that one
would expect from an assembly language. VAX-11 BLISS32 can be used as an alternative to assembly language
coding in all except the most machine-dependent systems
programming applications. The VAX-11 BLISS-32 language processor offers the programmer such features as:

THE SYSTEM PROGRAMMER
The system programmer can use this system to design
and build application systems for multiprogramming environments requiring fast response and a high degree of job
interaction and data sharing .
Job and Process Structure
The user program environment consists of a job structure
that can contain many processes. A process is the schedulable entity capable of performing computations in parallel with other processes. It consists of an address space
and an execution state that define the context in which a
program image executes. An executing program is associated with at least one process, but it can be associated
with several processes .

• program execution on architecturally different machines with little or no modification

A multiple process job structure allows one job to execute
more than one program image at the same time . One
process can wait for an event (such as 1/0 completion) to
occur while another process continues its computations.
The processes can communicate in several ways . They
can coordinate their execution synchronously using event
flags or asynchronously using software-simulated interrupts. They can send messages back and forth using virtual record-oriented devices called "mailboxes, " and they
can share code and data on disk and in memory.

• construction of complex expressions in which several
different kinds of operations can be performed in a single program statement
• exploitation of high level language constructs
PDP-11 BASIC-PLUS-2/VAX is an optional language
processing system that includes a compiler and an object
time system . PDP-11 BASIC-PLUS-2 is also available as
an optional language processor for the RSTS/E, RSX11 M, RSX-11M-PLUS, and IAS operating systems . The
PDP-11 BASIC-PLUS-2/VAX compiler produces code
that executes in PDP-11 compatibility mode.

Jobs can be grouped into application subsystems that
share code and data protected from other applications.
The processes within jobs in the same group can coordinate their activities using group interprocess communication facilities such as mailboxes and event flags, as well as
those facilities local to the job. They can access files and
data in memory that are protected from other groups in
the system.

PDP- 11 FORTRAN IV IV AX to RSX is an extended FORTRAN IV processor based on ANSI FORTRAN X3 .9-1966.
It supports mixed mode arithmetic, extended 1/0 facilities
for data formatting, error condition transfer statements, bit
manipulation , library usage, and several debugging facilities. The FORTRAN IV compiler (and its run time system)

3-5

Furthermore , user programs can be written to be
completely device independent through the system service and command language logical naming facilities. All
files and devices can be identified using arbitrarily defined
logical names that can be assigned values at run time.

Multiprogramming Environment
The system supports multiprogrammed applications that
require high performance by providing :
• event driven priority scheduling
• rapid process context switching
• minimum system service call overhead
• processor access mode memory protection
• memory management control

THE SYSTEM MANAGER

The system schedules processes for execution based on
the occurrence of events such as 1/0 completion as well as
time quantum expiration . When scheduled, the context
switching and interrupt processing hardware and software
ensure that processes are activated quickly. Real-time
processes can be assigned high priorities to ensure that
they receive processor time on demand. A process can
schedule its execution at a given time of day or after an interval has elapsed, and an appropriately privileged process can modify its priority during execution .

A job is normally associated with a user known to the system to have certain privileges, quotas, and resources . The
system manager authorizes users, plans data access and
protection , grants privileges, controls resource utilization ,
and analyzes the system's accounting and performance
information.

User Authorization
The system manager controls use of the system primarily
by creating user authorization information . This information is recorded in a specially maintained and protected
file called the user authorization file. The system manag er can create , examine, and update this file at any time.

The system 's memory management hardware and software ensure that paging , swapping , and dynamic memory
allocation are both efficient and transparent to the programmer. Where real-time applications require
performance control, both paging and swapping can be
reduced or eliminated by increasing the amount of memory allocated to a process and by locking a process in memory. Because memory management is transparent, programs can be written and later tuned for performance after
they are tested. The system provides a utility program to
aid system programmers in evaluating the effectiveness of
the memory management system for their processes.

The file contains one entry for each user authorized to access the system . Each entry :
• identifies the user
• supplies defaults
• specifies privileges
• limits resource usage
User identification consists of a unique user name, a password , a default account name, and a user identification
code (UIC). When logging onto the system , people must
always enter their user name and password to gain access
to the system. The password is not displayed on the user
terminal. Privileged users can change the passwords they
are assigned as often as they desire.

Program Development
This system provides the system programmer with tools
that support highly modular program and applications development. By taking advantage of these tools, the programmer can build applications quickly, and easily modify
and extend them later.

This system 's data protection scheme is based on the user
identification code (UIC) that the system manager assigns.
A UIC controls each user's access to the data structures
protected by UICs, which include both files and the interprocess communication facilities such as mailboxes ,
shared areas of memory, and event flags.

The system includes editors, compilers, librarians, linkers,
and debuggers for both the native and compatibility programming environment. All program development utilities
can be used either interactively or in batch mode, including the editors and debuggers . The native symbolic
debugger recognizes a command language similar to the
operating system command language and uses expressions similar to the language in which the program being
debugged was written.

A UIC consists of a group number and a member number.
Every user is assigned a UIC, and every data structure is
assigned both a UIC and a protection code. A protection
code identifies what types of access are available to which
users. There are four types of access (read, write, execute,
and delete), and there are four types of user (owner ,
group, world , and system) . The owner is any user that has
the same UIC as that assigned the data, the group is any
user that has the same group number as that assigned the
data, the world is any user, and the system is any user with
a group number of 1 through 10.

Executable program images can be built using extensive
libraries. In the native programming environment, the programmer can create libraries of assembler macro definitions, of object modules, and of image modules. The
system also includes the common run time procedure library, which provides library functions common to all native programming languages.

Using this protection scheme, a system can have files and
interprocess communication facilities that are available for
access only by users having the same UIC, or for access
only by users in the same group, or for universal access .
Furthermore, since each data structure has its own protection code, it is possible to protect each data structure assigned the same UIC on a different basis. The system UICs
are generally reserved for system users and system pro-

All program interfaces to the operating system and its utilities have uniform calling standards . System programmers
can add new library procedures to the common run time
procedure library and install them on-line without modifying existing programs and utilities, since all arguments are
passed using standard data structures.

3-6

grams and data structures. This arrangement enables a
user to protect a file from access by anyone other than the
owner or group, but still enables the system to access the
file for operations such as backup .
In addition to identifying the user and the set of data structures the user can access, the user authorization file supplies the user with a default file protection , a default directory name, and a default device name. When the user creates a file, the system assigns the default file protection
unless requested otherwise. An owner can modify a file's
protection at any time .

Table 3-2
Privileges Summary

INTERPROCESS CONTROL
• create event flag clusters
• create permanent common event flag clusters
• create temporary mailboxes
• create permanent mailboxes
• create global sections
• suspend, resume, wake, and delete processes within
the same group

Directory names are arbitrary character strings identifying
a directory file. A directory is simply a file containing a list
of file names and other identification information that is
used to find files on a volume. The default directory name
identifies the directory that lists the files the user normally
accesses. The default device name is the name of the device on which the volume containing the files the user normally accesses is mounted.

• suspend , resume, wake, and delete any process
• create detached processes
• create and delete shared memory sections
• map to physical pages

ACCESS TO FILES AND DEVICES

When the user issues a command to the command interpreter that operates on a file, or runs a program that opens
a file, the file system uses the default directory name and
default device name to locate the file unless specifically
requested to use some other directory name or some other device name. The user can change the default directory
and device names for a given session .

• insert logical names in group logical name table
• insert logical names in system logical name table
• allocate spooled devices
• obtain exclusive ownership of a shared device
• override volume protection

For further information on directories and directory structures, refer to the section on Data Management Facilities.

• issue mount requests

Privileges
Each user's authorization file entry contains a list of the
privileges that the user can invoke. They include interprocess communication and control privileges , performance
control privileges , file and device access privileges, and
system operational control privileges. The system manager can grant distinct privileges individually to each user.
Table 3-2 lists some of the privileges.

PERFORMANCE CONTROL
• execute time critical images
• lock process in memory

SYSTEM OPERATION CONTROL
• issue operator commands
• set any privilege bits

Privileges are checked when the user executes program
images. If a user runs an image that attempts to execute a
function requiring a privilege the user is not granted, the
image incurs a privilege violation. For example, diagnostic
programs require the privilege to issue device level diagnostic functions and the privilege to send messages to the
error logger. Users not granted these privileges will receive privilege violations if they attempt to run diagnostics.

• set process priority

In certain cases , however, it is desirable to let a user run an
image that requires privileges the user is not granted . For
example, the login program image requires the privilege to
switch to a more protected processor access mode to set
the user's initial context in a protected area of memory. To
let a user run an image that requires special privileges, the
system enables the system manager to install known Images. When the user runs a known image, the user obtains
the necessary privileges to execute the functions required
by the image, but only for the duration of that image's execution .

• send messages to error logger

PROGRAM EXECUTION
• execute Change Mode to Executive system service
• execute Change Mode to Kernel system service
• bypass file protection
• issue diagnostic functions
• suppress accounting messages
• issue logical and physical 1/0 functions

can use up during an accounting period. The system
manager can assign user quotas for the maximum amount
of CPU time accumulated during a given accounting period, and can limit the amount of dynamic system memory a
job can utilize for buffers. The system manager can set
disk usage quotas via the disk quota utility on a per user,
per volume or volume set basis. VAX/VMS will automatically record usage and enforce the assigned quotas during
file operation. However, each user possessing a private
volume controls the disk quotas on that volume. The limits

Resource Quotas and Limits
The user authorization file also provides the limits on how
many system resources a user can tie up while logged on
the system, and quotas for how much of a resource a user

3-7

• Response Time Histograms-indicate the time it takes
the system to initiate user requests .

imposed by VAX/VMS include the maximum number of:
• outstanding open files

Display Utility Program
The Display Utility Program (DISPLAY) provides a dynamic display of system performance measurement statistics
on a VT100 or VT52 video display terminal. By typing appropriate commands, system users may list information
regarding 1/0 system activity, paging , CPU usage, current
process activity , and other relevant statistics. Figure 3-2
shows a typical screen display.

• CPU time
• outstanding subprocesses created
• pages in a process working set
• pages in system paging files
• outstanding entries in the timer queue
• outstanding system buffered 1/0 requests
• bytes in system buffered 1/0 request
• outstanding direct 1/0 requests

THE SYSTEM OPERATOR

Resource Accounting Statistics
The system maintains an accounting information file for
collecting cumulative resource usage statistics. The system updates the accounting information file with detail records each time a process terminates. The detail statistics
include:

An operator is any user given the privileges by the system
manager to perform operator functions. A system does not
require an operator , but it can have one or several operators , and they can use any terminal to issue commands o r
run programs. Operator functions include:
• system startup and shutdown

• elapsed CPU time

• job control (change process priorities, kill jobs, etc .)

• login (connect) time

• device allocation

• number of volumes mounted
• number of pages printed

• volume mount and dismount request servicing

• largest process virtual size

• on-line disk and magnetic tape volume and file backup

• largest process working set size

• spool and batch queue control

• number of page faults

• software maintenance update installation

• number of system buffered 1/0 requests

• diagnostic execution

• number of direct 1/0 requests

An operator uses the command language to control operations , check system status, and run utility programs .

A detail record identifies the account name, user name,
and user identification code (UIC) to which the statistic applies. The accounting information file can be used to calculate billing information and reporting by account name,
user name, or UIC. Because the system collects all detail
records , system managers can define their own algorithms
for resource usage billing .

A special system program, the Operator Communications
Manager (OPCOM), is the primary operator aid . It collects
and delivers the messages all users and user programs
send to the operators. Any operator can respond to user
requests , and the Operations Communications Manager
will remind operators of any outstanding requests.

Spooling and Queue Control
The operators define the number and kind of input and
output spool queues in the system . The operators can c reate output spool ·queues for any number of devices, in cluding line printers, terminals , or even magnetic tape .
The operators can also create input queues for spool ing
batch input from a card reader.

Performance Analysis Statistics
The system collects statistics on its activities to help system programmers and managers tune the system for maximum performance. The information collected includes:
• System and Job Statistics-indicate the current number
of processes, interactive users, and batch jobs in the
system, the date and time at which the system was booted , and the current date and time.

The operators can assign each queue a priority , merge or
redirect queues to other devices, and modify the queue
set-up at any time. It is possible to have more than one
print queue for the same printer. For example, an operator
can create a generic printer queue that will collect jobs
that can be printed on any of a set of printers, and at the
same time have a print queue for each individual printer. A
user can issue a print request for a generic printer o r a
particular printer , and the operator can override the user's
request.

• Processor Access Mode Usage-indicates how much
time is spent executing at each of the access modes as a
measure of the type of code being executed and the
computational workload .
• Page Fault Activity-indicates how many and what kind
of page faults occurred as a measure of the effectiveness of memory management.
• 1/0 Activity-indicates how many and what kind of 1/0
operations are taking place.

A print job can contain one or more files to be pri nted
together . Print jobs can be submitted by an interactive
user, batch job , or any program . Print jobs are also automatically submitted at the end of a batch job. A print job
can specify the forms type required , the number of copies
of the job, the job priority, and a " hold until given time" request. Each file within a print job has its own copies coun t,

• Network Activity-indicates network workload (current
number of nodes in the network, number of bytes transmitted and received, number of messages transmitted
and received, number of buffers currently in use, number of successful and failing attempts to obtain space
for network buffers).

3-8

FREE LIST: 2016

1/0 SYSTEM RATES
16:17:09

NAME

VALUE

RATE
/ SEC

AVG
RATE

NAME

DIRECT I/ Os

7.30

1.50

BUFFERED I/Os

6.62

MAILBOX WRITES

WINDOW TURNS

MODIFY LIST: 35

VALUE

RATE
/SEC

AVG
RATE

PAGE FAULTS

14.84

1.83

3.24

PAGES READ

0.91

0.11

0.00

READ I/Os

0.45

0.07

0.68

0.14

PAGES WRITTEN

0.00

LOGNAME TRANS

8.90

0.98

WRITE I/Os

0.00

FILE OPENS

0.68

0.07

TOTAL INSWAPS

0.00

Figure 3-2
Display Utility Program (1/0 System Rates)

and each can have these options: double space , inhibit
form feed , print a flag page, label each page , or delete after printing . The operator can choose whether or not to
print burst (job separator) pages, and can put jobs on indefinite hold , modify the priority of a job, or abort a job.

On automatic system recovery after power interruption ,
the system determ ines whether the contents of memory
are still valid , and if so , restarts all possible 1/0 in progress
at the time of the power interruption and continues operations from the point of interruption . If the contents of memory are not valid , either because memory battery backup is
not included in the configuration , or because the power
failure lasted longer than the battery, the system automatically boots itself from disk and executes the start-up command procedures .

Batch Processing
The system supports multiple stream , multiple queue
batch processing . The operators control how many batch
job streams can run concurrently. Batch jobs can be submitted by an interactive user, another batch job, or any
program . When the number of batch jobs submitted
exceeds the number of streams, the remainder of the
batch jobs are held in a batch input queue. As with the
spool queues, the operators can control the batch job
queue. They can change job priority , hold a job until after a
given time , hold a job indefinitely, and kill a job .

Error Logging and Reporting
The error logger is a job that runs continuously. It collects
errors detected by both hardware and software , including:
• device errors

Volume mount commands issued in a batch job can request a generic device, such as any disk , or a specific device , such as disk drive unit 2. The batch job waits until an
operator satisfies the mount request, while other batch
jobs proceed . Operators can find out which job has a given
device.

• interrupt timeouts
• interrupts received from nonexistent devices
• memory, translation buffer, and cache parity errors
In addition , system software sends complete recovery information to the error logger following a power interruption or hardware or software failure.

On-line Software Maintenance
An operator can incorporate maintenance updates to the
software without bringing down the system for stand-alone
use. For example, VAX-11 /780 maintenance patches are
distributed on floppy disk and the operator simply loads
the console floppy disk drive with the maintenance floppy
to update the software on disk. Depending on the nature of
the update, an operator may have to restart the system to
activate the patched modules.

The error logger writes all messages it receives into an
error log file, noting vital system statistics at the time of the
message. The error logger also notes benign events when
they occur, such as when volumes are mounted and dismounted, and period ic time stamps indicating that no entries have occurred for a specified period of time. The error logger can accept messages from the operators at any
time , and from any programs privileged to send messages
to the error logger.

System Recovery
An operator can select manual or automatic system recovery following a power interruption or hardware or software
failure .

The system includes an error report generating program
that converts the information in the log file into a text file
that can be printed for later study.
3-9

• The Compatibility Mode Test Phase-This phase tests
most RSX-11 M utilities running in compatibility mode on
VAX/VMS.

On-line Diagnostics
An operator can run diagnostics to check the operation of
both hardware and software. An operator can run system
exercisers and device verification diagnostics while normal operations proceed. System exercisers test general
purpose software and compare the results with known
answers, reporting any discrepancies to the error logger.

• Termination Phase-In this phase , temporary files are
deleted and other cleanup activities are performed . If
miltiple runs were requested during the initialization
phase, then the UETP is restarted and control is passed
directly to the device test phase .

Operators can run device verification diagnostics either
stand-alone or concurrent with other processes. Diagnostics check the peripheral functions, including disk head
alignment. In addition, fault isolation diagnostics, which
isolate problems to replaceable units, are available for
stand-alone use.

The UETP is invoked via command procedures. The entire
package may be specified by executing the master procedure , UETP.COM , or tests may be executed individually
by specifying particular command procedures as illustrated in Figure 3-3.

Remote Diagnosis
If the system is equipped with the remote diagnosis option,
an operator sets up the system for remote preventive
maintenance or troubleshooting. When a hardware error is
detected or suspected, the operator mounts a diagnostic
disk pack, loads a diagnostic floppy disk in the console,
sets a switch on the processor console, and calls the local
DIGITAL service office. An operator does not need to be
present at the installation once the call is made. The
DIGITAL Diagnostic Center can then connect to the installation, run automated diagnostics, operate the diagnostic
console manually, and check the error log file. If a problem
is found, the Field Service engineer can bring the proper
equipment and replacement modules to make the repairs.

$ RUN UETINITOO
$ RUN UETINIT01
$ RUN UNETPDEV01
$ @UETCOMPOO
$ RUN UETNATV01
$ @UETNATV02
$ @UETNRMSOO
$ RUN UETLOAD01
$ RUN UETTERM01

Initialization Phase
1/ 0 Device Test Phase
Compatibility Mode Test Phase

Native Mode Test Phase
System Load Test Phase
Termination Phase

Figure 3-3
UETP Command Procedures

THE USER ENVIRONMENT TEST PACKAGE
The User Environment Test Package (UETP), consists of a
series of tests designed to demonstrate that the hardware
and software components of a system are in working order. The UETP consists of six phases:

APPLICATIONS EXAMPLES
To illustrate how the multiprogramming capabilities of the
system can be effective in widely diverse applications , Figures 3-4 and 3-5 show two hypothetical application systems: a commercially oriented data processing system
and a real-time flight training simulation system .

• The Initialization Phase-This phase verifies that 1/0 devices are operational via simple read/write operations .
In addition , users are prompted to supply several parameters which define the scope of the test, e.g ., number
of users to be simulated by UETP, amount of information to be displayed at the console, and number of
consecutive runs to be made by the UETP.

Commercial System Example
The commercial system diagram (Figure 3-4) begins with
both a programming group and a data processing operations group. Within the programming group, jobs can be
performing requests for programmers at terminals who
are using the system 's text editors, compilers, and linkers
to write and test programs for both the VAX/VMS and an
RSX-11 M system in the manufacturing department. The
programmers can execute command procedures and
submit batch jobs to automate repetitive development
steps .

• The 1/0 Device Test Phase-In this phase , 1/0 devices
undergo comprehensive testing. Terminals and line
printers generate pages or screens of output containing
header information and a test pattern of ASCII characters . Disks and magnetic tapes are also exercised. Files
are created on the mounted volumes. Data are then
written to the files. The test then checks the written data
for errors and erases the files .

Within the operations group, the system 's operators can
be managing batch and spool queues, backing up disks,
and monitoring performance. They may be down -line
loading tasks into the RSX-11 M system . They may be running an accounting program that interprets and summarizes the accounting statistics collected by the operating
system during the period and sends that data to the business data processing subsystem.

• The Native Mode Phase-This phase includes three
tests, each of which exercises software services provided explicitly for VAX/VMS. The first exercises VAX/VMS
system services; the second exercises native mode utilities such as the Symbolic Debugger and Image File
Patch Utility; and the third exercises VAX-11 RMS.
• The System Load Test Phase- This phase creates a
number of detached processes which simulate the action of a group of users concurrently issuing commands
from terminals; it tests the system's ability to handle
various levels of utilization.

A billing process within the business data processing
group can be suspended until it is activated by a process
(such as the accounting process) that wants to send it bill3-10

DP OPERATIONS

PROGRAMMING

MANUFACTUR ING
& STOCK CONTROL

MANUFAC TURING
RSX-l lM
PROCESS
CONTROL
SYSTEM

Figure 3-4
Commercial Data Processing System

ing information . The accounting process can send to the
billing process the name of an account summary file that it
created , using a permanent mailbox defined for that purpose.

and several subprocesses that perform the actual record
processing functions . As users at the terminals enter their
requests , such as create order record , update order record, etc. , the controll ing process collects the input from a
terminal and sends the request to the appropriate subprocess's mailbox. The subprocess may be hibernating,
having requested the operating system to activate it when

The business data processing group may also run a job
that handles order entry terminals . The job can consist of a
controlling process that handles input from the terminals,
3-11

VAX-11/780 SAMPLE APPLICATION SYSTEM
FLIGHT TRAINING SIMULATION

LIBRARY OF
SCENARIOS

REAL-TIME
INTERFACE

ON-LINE
MAINTENANCE
INPUT

REAL-TIME
GRAPHICS

VAX-11/780
CONSOLE
I

INSTRUCTOR'S STATION

SECONDARY :
USES
I
1---T-----------~
I
I
I
I
I
I
I

I
I
I
I
I
I
I

I
I

I
I
I
I

---------~

/- 4--:t

REAL-TIME
TRAINING INPUTS AND
TRAINING SITUATIONS
\

I
I

\
\
\
\

',
\

BATCH-INTERACTIVE

DElD \\
PROGRAM MODIFICATIONS
AND UPDATES
• EDITOR
• LINKER
• COMPILER
• DEBUGGER

\
\
\
\
\
\
\
\

DESD
PILOT TRAINING RECORDS
SIMULATOR MAINTENANCE
RECORDS
OTHER EDP WORK
Figure 3-5
Real-time Flight Simulation System

3-12

anything is written to its mailbox. Or the controlling process can simply set an event flag to notify a subprocess that
it has received a request for which the subprocess is waiting .
A process in the manufacturing application group may regularly collect orders from the order entry job to create
materials parts lists, or notify stockroom clerks of high-priority orders. The stockroom clerks can keep inventory records up to date by using the shipping and inventory jobs,
because they can collect manufacturing statistics from a
background task in the RSX-11 M process control system
on the manufacturing floor. Once orders are shipped , the
shipping process can notify an accounts receivable process in the business data processing application group,
which in turn can activate the billing process .
Real-Time Flight Simulation Example
To illustrate how VAX's facilities can be as extended to the
real-time environment, Figure 3-5 shows a sample flight
training simulation system . Flight simulation is a particularly good application for the VAX system , since in addition
to fast real-time response, such systems must also be capable of solving large and complex equation systems. In
addition , such systems also require general program development fac ilities such as FORTRAN compilation , assembly, editing , debug , library facilit ies , and fast-access
file management.

The illustrated system shows how the multiprogramming
capabilities of VAX allow it to handle the basic real-time
tasks of data acquisition and transmission , while also performing a wide range of other activities:
Looking from top to bottom , the diagram begins with a representation of an aircraft fuselage containing the cockpit
throttle levers and control panel which the student operates to simulate flight. The signals and control movements
of the student go through an analog to digital conversion,
and are then passed through a real-time interface device
into VAX memory. Before the system can respond to the
data generated by the student , complex flight-motion
equations must be called and combined with current aircraft data. This , in turn, produces a new set of circumstances with which the student must deal.
Additional inputs to the system are also provided by an instructor at a timesharing terminal (shown in the right central part of the diagram). By typing commands at the
terminal , the instructor can control the situations which the
student must face-for example, by injecting an engine
failure, weather change , or some other complication. The
instructor can also introduce additional variables into the
system by inputting predefined "scenarios" which have
been stored on libraries. The student's flight environment
can include "visual cues" (e .g., terrain , runway approaches , other aircraft, etc.) produced by soph isticated , realtime graphics modules.
In addition to performing basic flight simulation functions,
the system also performs a number of auxiliary functions
which are less time-critical in nature. These would include
such activities as monitoring of engine performance , testing of navigation and instrument landing systems, testing
of weapons systems , and monitoring cabin , hydraulic, and
electrical systems. The system also generates a file of stu3-13

dent performance statistics which can be analyzed at a later time .
The system also provides facilities for program development. As shown in the lower left-hand side of the diagram ,
programmers at terminals can write and test new
applications programs ( i n either batch or interactive
mode) using text ed itors, compilers, linkers, and debuggers. Because of the way the VAX architecture handles
real-time vs. normal processes (described below), program development can take place simultaneously with the
running of the fl ight simulator; no significant reduction in
real-time processing speed will result.
Design Considerations
Systems such as that shown in Figure 3-5 must be designed to operate at extremely high speeds, both in terms
of real-time 1/0 and computation . Many simulators require
turnaround times (data sampling , computation, and output) in the 25-50 millisecond range .

There are several elements of the VAX hardware and the
VAX/VMS operating system which aid in achieving such
speeds . Most important is VAX's context switching ability-a result of spec ial processor instructions wh ich relieve
the operating system software of having to individually
load or save the hardware registers which define the hardware context. Another element is the design and operation
of VAX/VMS device drivers. VAX/VMS drivers are forked
processes which are created dynamically in response to a
user 1/0 request or unsolicited device interrupt. They operate with minimal context, execute to completion when invoked, and remain memory resident throughout execution . (VAX/VMS device drivers can be written for userdeveloped devices wh ich interface to the VAX UNIBUS .)
In addition , system designers can utilize the VAX/VMS
scheduling and system service facilities to yield still higher
processing speeds. For example, the following design
schema could be employed.
Those processes which perform the basic flight simulation
functions (i.e., fl ight motion equations, visual graphics
modules, etc.) can be designed as a group of hibernating
processes capable of being immediately reawakened as
required by the student's control activities. This could be
accomplished via the use of system service calls such as
($HIBERNATE/$WAKE) or ($SUSPEND/$RESUME) or by
using interprocess control structures such as mailboxes
and common event blocks.
To further ensure that basic simulation functions will operate at the fastest possible speed, real-time processes can
be granted the highest scheduling priorities. Such processes can also be given special privileges which allow
them to eliminate pag ing and swapping and thus assure
memory residency. (Note that when a VAX process runs at
real-time priority its priority level will actually be higher
than system processes such as the Swapper, Linker, and
Symbionts).
Processes that are not time-critical can be assigned normal scheduling priorities . Those processes which perform
what is essentially a monitoring function (e .g., engine performance, electrical system , etc.) can also be hibernated,
then monitored periodically via the issuance of a programmed system timer service. Other activities such as

program development and statistical processing can take
full advantage of the VAX/VMS system with minimum impact upon the basic real-time core of the simulation system .

For more information on the concepts of jobs, processes
and program images, refer to the Operating System section .

3-14

4
The
VAX.
Processors
IIIIIIIIIIII111111111111111111111111111111 II 1111111111111111111111111111111111111111
l IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIII 111111111111111111111111111111111111111111

11111111111mu111111111111/IUllllJI-

,,,,,,,,,,111111

-- -- -.-.•••• 111111

lllflll!!!!llf

1111
11111111 '

·. _

A VAX processor executes variable-length instructions in native mode
and non-privileged PDP-11 instructions in compatibility mode. The
VAX processor includes integral memory management, sixteen 32-bit
registers , 32 interrupt priority levels, an intelligent console, a programmable real-time clock , and a time-of-day and date clock .
The VAX native instruction set provides 32-bit addressing enabling the
processor to address up to 4 billion ( 10 9 ) bytes of virtual address
space. The processor's memory management hardware includes mapping registers used by the operating system , page protection by access
mode, and an address translation buffer that eliminates extra memory
accesses during virtual to physical address translation .
VAX also provides sixteen 32-bit general registers that can be used for
temporary storage, as accumulators, index registers , and base registers. Four registers have special significance: the Program Counter and
three registers that are used to provide an extensive procedure CALL
facility . The processor offers a variety of addressing modes that use the
general registers to identify instruction operand locations, including an
indexed addressing mode that provides a true post-indexing capab ility .
The native instruction set is highly bit efficient. It includes integral decimal , character string, and floating point instructions, as well as integer ,
logical , and bit field instructions. Instructions and data are variable
length and can start at any arbitrary byte boundary or, in the case of bit
fields, at any arbitrary bit in memory. Floating point instruction execution can be enhanced by an optional floating point accelerator.
The 1/0 subsystem consists of the processor's internal bus and the UNIBUS and MASSBUS interfaces.

INTRODUCTION

registers. While native mode is the primary instruction execution state of the machine and compatibility mode the
secondary state , the ir instruction sets are closely related ,
and their programm ing characteristics are very similar. A
user process can execute both native mode images and
compatib ility mode images.

This section is divided into two parts . The first discusses
the architecture of a VAX processor, and the second discusses VAX processor implementation details.

VAX ARCHITECTURE
The following sections discuss the architecture, i.e., the
programming characteristics of the processing system as
seen from the general user's and the operating system 's
viewpoint.

A native instruction consists of an operation code (opcode) and zero or more operands , which are described by
data type and addressing mode. The native instruction set
is based on over 200 different kinds of operations, of each
operand of which can be addressed in any one of several
ways . Thus, the native instruction set offers a tremendous
number of instructions from which to choose.

PROCESSING CONCEPTS FOR USER
PROGRAMMING
A program is a stream of instructions and data that a user
can request the operating system to translate , link , and execute. An executable program is called an Image to distinguish it from source and object programs. When a user
runs an image, the context in which the image is executed
is called a process . A process is the complete unit of execution in this computer system and typically runs several
images, one after another. Process context includes the
state of the image it is currently executing and includes the
image's allowable limitations, which primarily depend on
the privileges of the user executing the image.

In spite of the large number of instructions, the native instruction set is a natural programming language that is
very easy to learn. Many of the instructions correspond
directly to high-level language statements, and the assembler mnemonics are readily associated with the instruction
function .
To choose the appropriate instruction , it is only necessary
to become famil iar with the operations, data types, and addressing modes. For example, the ADD operation can be
applied to any of several sizes of integer, floating point, or
packed decimal operands , and each operand can be addressed directly in a register , directly in memory, or indirectly through po inters stored in registers or memory
locations.

The next few pages introduce some of the concepts that
concern assembly language programmers in general, including addressing , data types, instruction sets , and other
programming aspects of the processor . Further details on
these programm ing characteristics follow this introduction.

Registers and Addressing Modes
A register is a location within the processor that can be
used for temporary data storage and addressing. The assembly language programmer has sixteen 32-bit general
reg isters available for use with the native instruction set.
Some of these reg isters have special significance. For example , one register is designated as the Program Counter,
and it contains the address of the next instruction to be executed . Three general registers are designated for use
with procedure linkages: the Stack Pointer, the Argument
Pointer, and the Frame Pointer.

Process Virtual Address Space
Most data are located in memory using the address of an
8-bit byte. The programmer uses a 32-b it virtual address
to identify a byte location. This address is called a virtual
address because it is not the real address of a physical
memory location . It is translated into a real address by the
processor under operating system control. A virtual address is not a unique address of a location in memory, as
are physical memory addresses. Two programs using the
same virtual address might refer to two different physical
memory locations, or refer to the same physical memory
location using different virtual addresses.

An instruction operand can be located in main memory, in
a general register, or in the instruction stream itself. The
way in which an operand location is specified is called the
operand addressing mode. The processor offers a variety
of addressing modes and addressing mode optim izations.
There is one addressing mode that locates an operand in a
register. There are six addressing modes that locate an
operand in memory using a register to :

The set of all possible 32-bit virtual addresses is called
virtual address space. Conceptually , virtual address space
can be viewed as an array of byte " locations" labelled from
Oto 2 32 - 1, an array that is approximately four billion bytes
in length . This address space is divided into sets of virtual
addresses designated for certain uses. The set of virtual
addresses designated for use by a process , including an
image it executes, is one half of the total virtual address
space. Addresses in the remaining half of virtual address
space are used to refer to locations maintained and protected by the operating system .

• point to the operand
• point to a table of operands
• point to a table of operand addresses
Additionally, there are six addressing modes that are indexed modifications of the addressing modes that locate
an operand in memory. Finally, there are two addressing
modes that identify the location of the operand in the instruction stream : one for constant data , and one for
branch instruction addresses.

Instruction Sets
At any one time, the processor's instruction interpretation
hardware can be set to either of two modes: native mode
or compatibility mode. In native mode the processor executes a large set of variable-length instructions, recognizes a variety of data types , and uses sixteen 32-bit general purpose reg isters . In compatibil ity mode the
processor executes a set of PDP-11 instructions, recognizes integer data, and uses eight 16-bit general pu rpose

Data Types
The data type of an instruction operand identifies how
many bits of storage are to be treated as a unit, and what
4-1

(F _floating and D_floating) are standard on all VAX family
processors . Two extended range formats (G_floating and
H_floating) are available as options on VAX family processors . F_floating and D_floating are 4 and 8 bytes lo ng respectively, with an 8- bit excess 128 exponent. The effective 24-bit fraction of F _floating yields approximately 7 decimal digits of precision. The 56-bit fraction of O_float ing
yie lds approximately 16 decimal digits of prec ision.
G_floating is also 8 bytes in length , but has an 11-bit excess 1024 exponent and effectively 53 bits of fraction . Its
precision is approximately 15 decimal digits. H_float ing is
16 bytes in length with a 15-bit excess 16384 exponent and
113-bit fraction . Its precision is approximately 33 deci mal
digits.

the interpretation of that unit is. The processor's native instruction set recognizes four primary data types : integer,
floating point, packed decimal, and character string. For
each of these data types, the selection of operation immediately tells the processor the size of the data and its
interpretation. The processor can also manipulate a fifth
data type, the bit field, in which the user defines the size of
the field and its relative position . In addition, the processor
supports two types of linked-list queue structures.
There are several variations on the four primary data
types. Table 4-1 provides a summary of the data types
available, each of which are illustrated in Figure 4-1. Integer data are stored as a binary value in either byte, word,
longword, or, in some cases, in quadword or octaword format. A byte is eight bits, a word is two bytes, a longword is
four bytes, a quadword is eight bytes, and an octaword is
sixteen bytes. The processor can interpret an integer as either a signed (2's complement) value or an unsigned value. The sign is determined by the high-order bit.

Packed decimal data are stored in a string of bytes . Each
byte is divided into two 4-bit nibbles. One decimal digit is
stored in each nibble. The first , or high-order, d igit is
stored in the high-order nibble of the first byte , the second
digit is stored in the low-order nibble of the first byte, the
third digit is stored in the high-order nibble of the second
byte , and so on . The sign of the number is stored in the
low-order nibble of the last byte of the string .

Floating point values are stored using a signed exponent
and a binary normalized fraction . The normalization bit is
not represented. Four types of floating point data formats
are provided . The two PDP-11 compatible formats

Character data are simply a string of bytes contain ing any
binary data, for example , ASCII codes . The first character

Table 4-1
Data Types

DATA TYPE

RANGE (decimal)

SIZE

Integer

Signed

Unsigned

Byte

8 bits

- 128 to +127

0 to 255

Word

16 bits

Longword

32 bits

- 32768 to +32767
- 2 31 to +2 31 -1

0 to 65535
Oto 232-1

Quadword

64 bits

Octaword

128

- 263 to + 263 - 1
-2127 to +2127

0 to 264 -1
Oto 2 128 -1

bits

-1

±2.9 x 10- 37 to 1.7 x 10 38

F_and D_floating point
F_floating point

32 bits

approximately seven decimal
digits precision

D_floating

64 bits

approximately sixteen decimal
digits precision
±0.56 x 10· 3os to 0.9 x 10308

64 bits

approximately fifteen decimal
digits prec ision
± 0.84 x 10- 4932 to ±0.59 x 10 4932

128 bits

approximately thi rty-three
decimal digits precision

Packed Decimal
String

Oto 16 bytes
(31 digits)

numeric , two dig its per byte
sign in low half of last byte

Character String

0 to 65535 bytes

one character per byte

Variable-length
Bit Field

Oto 32 bits

dependent on interpretation

Numeric String

0 to 31 bytes (DIGITS) - 103' -1 to +10 3' - 1

G_floating point
G_floating
H_floating point
H_floating

Queue

Queue entries at arbitrary
displacement in memory

~ 2 longwords/
queue entry

4-2

WORD
_1_5_

BYTE

-7-

lONGM:)R()
31

0
: A

QJJAOWORD

:.3_1

0
:A
: A+4

OCTAWORD
31

:A
: A+4
: A+8
:A+12

127

F_FLOATING

7 6

EXPONENT

D_ FLOATING
15

FRACTION

FRACTION
FRACTION

G FLOATING
15 14

EXPONENT

4 3

EXPONENT

0
FRACTION

FRACTION

:A+2

FRACTION

: A+4

FRACTION

: A+6

H FLOATING
15
14

0
EXPONENT

FRACTION

:A+2

FRA CTIO N

: A•4

FRACTION

: A•6

FRACTION

:A+ 8

FRACTION

:A+lO

FRACTION

:A+12

FRACTION

:A+14

PACKED DECIMAL STRING ( + 123)

. ' . •. '/ ::.,

CHARl>CTER STRING ( XYZJ

0
" X"

: A

"Y"

: A+l

" z"

: A+2

VARIABLE-LENGTH BIT FIELD
P+S

P- 1

P+ S-1

: A

S-1

Figure 4-1
Data Type Representations
4-3

call nesting by using a general register as a pointer to the
place on the stack where a procedure has its linkage data.
This record of each procedure's stack data, known as its
stack frame, enables proper returns from procedures
even when a procedure leaves data on the stack . In add ition , user and operating system software can use the stack
frame to trace back through nested calls to handle errors
or debug programs.

in the string is stored in the first byte, the second character
is stored in the second byte, and so on . A character string
that contains ASCII codes for decimal digits is called a numeric string .
The address of any data item is the address of the first byte
in which the item resides. All integer, floating point,
packed decimal, and character data can be stored starting
on an arbitrary byte boundary. A bit field, however, does
not necessarily start on a byte boundary. A field is simply a
set of contiguous bits (0-32) whose starting bit location is
identified relative to a giyen byte address or register. The
native instruction set can interpret a bit field as a signed or
unsigned integer.

Condition Codes
A user program can test the outcome of an arithmetic or
logical operation . The processor provides a set of cond ition codes and branch instructions for this purpose. The
condition codes indicate whether the previous arithmetic
or logical operation produced a negative or zero result, or
whether there was a carry or borrow, or an overflow . There
are a variety of branch on condition instructions: those for
overflow and carry or borrow, and those for signed and
unsigned relational tests .

The VAX queue data types consist of circular, doubly
linked lists. A queue entry is specified by its address. Each
queue entry is linked to the next entry via a pair of longwords. The first longword is the forward link; it specifies
the location of the succeeding entry. The second longword
is the backward link; it specifies the location of the preceding entry. VAX processors support two queue types according to the nature of the links: absolute and self-relative . An absolute link contains the absolute address of the
entry that it points to. A self-relative link contains a displacement from the address of the queue entry.

Exceptions
Certain situations may require that the results of an operation be tested either by the user or by the processor directly. The processor recognizes many events for which it
must test directly, and automatically changes the normal
flow of the user program when they occur. These events ,
called exceptions , are the direct result of executing a specific instruction . Exceptions also include errors automatically detected by the processor, such as improperly
formed instructions.

Stacks, Subroutines, and Procedures
A stack is an array of consecutively addressed data items
that are referenced on a last-in, first-out basis using a
general register. Data items are added to and removed
from the low address end of the stack. A stack grows toward lower addresses as items are added, and shrinks toward higher addresses as items are removed .

All exceptions trap to operating system software. There
are essentially no fatal exceptions. All exceptions either
wait for the instruction that caused them to complete before trapping or they restore the processor to the state it
was in just prior to executing the instruction that caused
the exception . In either case, the instruction can be retried
after the cause of the exception is cleared . Depending on
the exception, it may be desirable to correct the situation
and continue. If not, the operating system issues an appropriate message and aborts the instruction stream in progress. To continue, the user can request the operating system software to · start execution of a condition handler
automatically when an exception occurs .

A stack can be created anywhere in the user's program
address space. Any register can be used to point to the
current item on the stack. The operating system, however,
automatically reserves portions of each process address
space for stack data structures. User software references
its stack data structure, called the user stack, through a
general register designated as the Stack Pointer. When
the user runs a program image, the operating system automatically provides the address of the area designated
for the user stack.
A stack is an extremely powerful data structure because it
can be used to pass arguments to routines efficiently. In
particular, the stack structure supports re-entrant routines
because the processor can handle routine linkages automatically using the Stack Pointer. Routines can also be recursive because arguments can be saved on the stack for
each successive call of the same routine.

USER PROGRAMMING ENVIRONMENT
A process context includes the definition of the virtual address space in which it executes an image. An image executing within a process context controls its execut ion
through the use of one of the instruction sets , the general
purpose registers, and the Processor Status Word . These
hardware resources are discussed in detail in the following
sections.

The processor provides two kinds of routine call instructions: those for subroutines, and those for procedures. In
general, a subroutine is a routine entered using a Jump to
Subroutine or Branch to Subroutine instruction, while a
procedure is a routine entered using a Call instruction .

Process Virtual Address Space Structure
As mentioned earlier , certain sets of virtual addresses in
virtual address space are designated for particular uses.
The processor and operating system provide a multiprogramming environment by dividing virtual address space
into two halves: one half for addressing context-dependent code and data, the other half for addressing contextindependent code and data.

The processor provides more elaborate routine linkages
for procedures than for subroutines. The processor automatically saves and restores the contents of registers to be
preserved across procedure calls. The processor provides
two methods for passing argument lists to called procedures: by passing the arguments on the stack, or by
passing the address of the arguments elsewhere in memory. The processor also constructs a "journal" of procedure

The first half is called per-process space (or more simply,
" process space" ), which is capable of addressing approxi-

4-4

mately two billion bytes. An image executing in the context
of a process and the operating system on behalf of the
process use addresses in process space to refer to code
and data particular to that process context. A process cannot represent virtual addresses in any process space but
its own . Thus , code and data belonging to a process are
automatically protected from other processes in the system .

• The Frame Pointer (FP or R13), which contains the address of the base of a software data structure stored on
the stack called the stack frame, maintained for procedure calls.
• The Argument Pointer (AP or R12), which contains the
address of the base of a software data structure called
the argument list, maintained for procedure calls.
In addition, the first six registers, RO through RS, have special significance for character and packed decimal string
instructions, and the Cyclic Redundancy Check and Polynomial Evaluation instructions. These instructions use
these registers to store temporary results and, upon com_pletion , leave results in the registers that a program can
use as the operands of subsequent instructions.

The second half of virtual address space is called system
space . The operating system assigns specific meanings to
addresses in system space. The significance of any address in system space is the same for every process, independent of process context. Although most locations
referred to by system space addresses are protected from
access by user images, if two images executing in different
process contexts do use an address in system space, the
address always refers to the same physical location in
memory.
Process space is further subdivided into two equal regions. Addresses in the first region , called the program region , are used to identify the location of image code and
data. Addresses in the second region , called the control
region , are used to refer to stacks and other temporary
program image and permanent process control information maintained by the operating system on behalf of the
process. Program region addresses are allocated from
address 0 up, and control region addresses are allocated
from address 2 3 ' -1 down.

A register's special significance does not preclude its use
for other purposes, except for the Program Counter. The
Program Counter can not be used as an accumulator, as a
temporary reg ister , or as an index register. In general,
however, most users do not use the Stack Pointer, Argument Pointer, or Frame Pointer for purposes other than
those designated .

Addressing Modes
The processor's addressing modes allow almost any operand to be stored in a register or in memory, or as an immediate constant. Table 4-2 summarizes the addressing
modes.

System space is also subdivided into two equal regions .
The operating system assigns the system region addresses for linkages to its service procedures , for memory
management data, and for 1/0 processing routines. The
second region is presently unused.

There are seven basic addressing modes that use the general registers to identify the operand location . They include:
• Register mode, in which the register contains the operand.

General Registers _
Instruction operands are often either stored in the processor's general registers or accessed through them . The sixteen 32-bit programmable general registers are labelled
RO through R15 (in decimal). Registers can be used for
temporary storage , accumulators , base registers , and index registers. A base register contains the address of the
base of a software data structure such as a table, and an
index register contains a logical offset into a data structure .

• Register Deferred mode, in which the register contains
the address of the operand.
• Autodecrement mode, in which the contents of the
register are first decremented by the size of the operand , and then used as the address of the operand. The
size of the operand (in bytes) is given by the data type of
the instruction operand , and depends on the instruction.
For example, the Clear Word instruction uses a size of
two , since there are two bytes per word.

Whenever a register is used to contain data, the data are
stored in the register in the same format as would appear
in memory. If a quadword or double floating datum is
stored in a register, it is actually stored in two adjacent
registers . For example, storing a double floating number in
register R? loads both R? and RB.

• Autoincrement mode, in wh ich the contents of the register are used as the address of the operand , and then incremented by the size of the operand. If the Program
Counter is the specified register, the mode is called Immediate mode.
• Autoincrement Deferred mode, in which the contents of
the register are used as the address of a location in
memory containing the address of the operand, and
then are incremented by four (the size of an address). If
the Program Counter is the specified register, the mode
is called Absolute mode.

Some registers have special significance depending on
the instruction being executed . Registers R12 through R15
have special significance for many instructions, and therefore have special labels. These special registers are:
• The Program Counter (PC or R15), which contains the
address of the next byte to be processed in the instruction stream .

• Displacement mode, in which the value stored in the
register is used as a base address . A byte, word , or
longword signed constant is added to the base address,
and the resulting sum is the effective address of the operand .

• The Stack Pointer (SP or R14) , which contains the address of the top of a stack maintained for subroutine
and procedure calls.
4-5

Table 4-2
Addressing Modes : Assembler Syntax

Literal
(Immediate)

I ~1

#constant

Reg ister Deferred

(Rn)

Autodecrement

-(Rn)

Auto increment

(Rn) +
Indexed

Autoincrement Deferred

@(Rn) +

(Absolute)

@#address

Displacement

(~·}

displacement (Rn)
address

@ff}

displacement (Rn)
address

Displacement Deferred

[Rx]

n = Othrough 15
x = 0 through 14

used with integer operations, and can therefore express
the constants 0 through 63 (decimal). The 6-bit field is interpreted as a floating point constant when used in floating
point operations, where three bits express an exponent
and three a fraction .

• Displacement Deferred mode, in which the value stored
in the register is used as the base address of a table of
addresses. A byte, word , or longword signed constant is
added to the base address, and the resulting sum is the
address of the location that contains the actual address
of the operand .

The autoincrement and autodecrement modes enable automatic stepping through tables. Displacement mode enables generation of offsets into a table , with a choice of either short or long displacements. The deferred modes enable the user to maintain tables of operand addresses
instead of the operands themselves .

Of these seven basic modes, all except register mode can
be modified by an index register. When an index register is
used with a basic mode to identify an operand , the addressing mode is the name of the basic mode with the suffix " indexed ." For example, the indexed addressing mode
for register deferred is called "register deferred indexed "
mode. In addition to the seven basic addressing modes
that use registers, the processor recognizes six indexed
addressing modes.

The indexed addressing modes allow indexing into tables
with a step size automatically determ in ed by the operand.
As in autoincrement and autodecrement addressing , the
index is calculated in the context of the operand data type.
Th is means that the user can easily access several tables
of differing data types using the same index key.

In an indexed addressing mode, one register is used to
compute the base address of a data structure, and the other register is used to compute an index offset into the data
structure. To obtain the operand 's effective address in an
indexed addressing mode, the processor: 1) computes the
base operand address provided by one of the basic addressing modes (except register mode), 2) takes the value
stored in the index register and multiplies it by the given
operand size, and 3) adds the resultant value to the operand address. The index register contents are not affected
and can be used for subsequent processing operations.

Furthermore, because the indexed addressing modes enable specification of the base operand address using any
mode that generates an actual address (that is, any mode
except register or literal), the user has the ability to construct double indexing . The base address can be selected
from a table of base addresses using displacement deferred mode, and then use an index register to provide the
offset into the particular table selected .

The processor also provides literal mode addressing , in
which an unsigned 6-bit field in the instruction is interpreted as an integer or floating point constant.

Thus the processor's addressing modes allow cons iderable flexibility in the arrangement and processing of data
structures . A data structure's design does not have to be
tied to its processing method for efficiency.

The variety of addressing modes enables the assembly
language programmer to write , and high-level language
compilers to produce, very compact code. For example, literal mode is a very efficient way to specify small constants . The 6-bit field is interpreted as an integer when

Program Counter
A native mode instruction has a variable-length format,
and instructions are thought of as byte aligned . A variablelength format not only makes code more compact, it

4- 6

means that the instruction set can be extended easily. The
opcode for the operation is either one or two bytes , and is
followed by zero to six operand specifiers, depending on
the instruction. An operand specifier can be one or several
bytes long , depending on the addressing mode. Figure 4-2
illustrates the representation of an instruction as a string of
bytes. Just before the processor begins to execute an instruction, the Program Counter contains the address of
the first byte of the next instruction . The way in which the
Program Counter is updated is totally transparent to the
programmer .

ment on a stack without removing it, use register deferred
mode, and to reference other elements of the stack use
displacement mode addressing .
The processor designates Register 14 as the Stack Pointer
for use with both the subroutine Branch or Jump instructions , and the procedure Call instructions. On routine entry, the processor automatically saves the address of the
instruction that follows the routine call on the stack . It uses
the Program Counter and the Stack Pointer to perform the
operation . Before entering the subroutine, the Program
Counter contains the address of the instruction following
the Branch , Jump, or Call instruction; the Stack Pointer
contains the address of the last item on the stack . The
processor pushes the contents of the Program Counter on
the stack. Returning from a subroutine, the processor automatically restores the Program Counter by popping the
return address off the stack .

The Program Counter itself can be used to identify operands. The assembler translates many types of operand
references into addressing modes using the Program
Counter. Autoincrement mode using the Program Counter , or immediate mode, is used to specify in-line constants other than those available with literal mode addressing. Autoincrement deferred mode using the Program Counter, or absolute mode, is used to reference an
absolute address . Displacement and displacement
deferred modes using the Program Counter are used to
specify an operand using an offset from the current location.

Also for the procedure Call instructions, the processor designates Register 12 as an Argument Pointer, and Register
13 as a Frame Pointer. The Argument Pointer is used to
pass the address of the argument list to a called procedure, and the Frame Pointer is used to keep track of
nested Call instructions.

Program Counter addressing enables the user to write position-independent code. Position-independent code can
be executed anywhere in virtual address space after it has
been linked, since program linkages can be identified as
absolute locations in virtual address space and all other
addresses can be identified relative to the current instruction.

An argument list is a formal data structure containing the
arguments required by the procedure being called. Arguments may be actual values, addresses of data structures,
or addresses of other procedures. An argument list can be
passed in either of two ways: by passing only its address,
or by passing the entire list on the user stack . The first
method is used to pass long argument lists, or lists that are
to be preserved. The second method is generally used
when calling procedures that do not require arguments, or
when building an argument list dynamically.

Stack Pointer, Argument Pointer and Frame Pointer
The Stack Pointer is a register specifically designated for
use with stack structures. Autodecrement mode addressing using the Stack Pointer can be used to place items on
the stack . Auoincrement mode can be used to remove
items from the stack . To reference and modify the top ele-

When issuing a procedure Call instruction, the processor
uses the Argument Pointer to pass arguments to the procedure . If arguments were passed on the stack, the proc-

AUTODECREMENT MODE, MOVE LONG INSTRUCTION
MOVL-{R3) ,R4
BEFORE INSTRUCTION EXECUTION
ADDRESS SPACE
00001014

\0000101s \

\00000000

00001015
00001016
00001017
AFTER INSTRUCTION EXECUTION
\00001014

\cE543210

MACHINE CODE : {ASSUMED STARTING LOCATION 00003000)
00003000

OPCODE FOR MOVE LONG INSTRUCTION

00003001

AUTODECREMENT MODE , REGISTER R3

00003002

Figure 4-2
Instruction Representation
4-7

one of two ways. By clearing the exception enable bits in
the Processor Status Word, the processor can be directed
to ignore integer and decimal string overflow and floating
underflow. The user may check for these conditions either
by testing the condition codes (except for underflow) using
the Branch on Condition instructions or by enabling the
exception bits. By enabling the exception bits, the processor treats integer and decimal string overflow and floating
underflow as exceptions. In any case, floating overflow and
divide by zero exceptions are always enabled .

essor automatically pops the arguments off on return from
the procedure.
The importance of the way the Call instructions work is that
nested calls can be traced back to any previous level. The
Call instructions always keep track of nested calls by using
the Frame Pointer register. The Frame Pointer contains
the address on the stack of the items pushed on the stack
during the procedure call. The set of items pushed on the
stack during a procedure call is known as a call frame or
stack frame. Since the previous contents of the Current
Frame register are saved in each call frame, the nested
frames form a linked data structure which can be unwound
to any level when an error or exception condition occurs in
any procedure.

Handling Exceptions
When an exception occurs, the processor immediately
saves the current state of execution and traps to the operating system. The operating system automatically searches for a procedure that wants to handle the exception. Procedures that respond to exceptions are called condition
handlers. The user can declare a condition handler for an
entire image and for each individual procedure called . In
addition, because the processor keeps track of nested
calls using the Frame Pointer register, it is possible to declare condition handlers for procedures that call other procedures in which exceptions might occur. The operating
system automatically traces back through call frames to
find a condition handler that wants to handle an exception
that occurs.

Processor Status Word
The Processor Status Word (a portion of the Processor
Status Longword) is a special processor register that a
program uses to check its status and to control synchronous error conditions . The Processor Status Word, illustrated in Figure 4-3, contains two sets of bit fields:
• the condition codes
• the trap enable flags
The condition codes indicate the outcome of a particular
logical or arithmetic operation. For example, the Subtract
instruction sets the Negative bit if the result of the subtraction operation produced a negative number, and it sets the
Zero bit if the result produced zero. The Branch on Condition instructions can be used to transfer control to a code
sequence that handles the condition .

NATIVE INSTRUCTION SET
The instruction set that the processor executes is selected
under operating system control to either native mode or
compatibility mode. The native mode instruction set is
based on over 200 different opcodes. The opcodes can be
grouped into classes based on their function and use. Instructions used to manipulate the general data types include:

There are two kinds of exceptions that concern the user
process: trace faults and arithmetic exceptions. The trace
fault is used by debugging programs or performance evaluators. Arithmetic exceptions include:
• integer, floating point, or decimal string overflow, in
which the result was too large to be stored in the given
format

• integer and logical instructions
• floating point instructions
• packed decimal instructions

• integer, floating point, or decimal string divide by zero,
in which the divisor supplied was zero

• character string instructions
• bit field instructions

• floating point underflow, in which the result was too
small to be expressed in the given format

Instructions that are used to manipulate special kinds of
data include:

Of the arithmetic exceptions, integer overflow, floating underflow, and decimal string overflow may be handled in

• queue manipulation instructions

NOT USED

DECIMAL OVERFLOW TRAP ENABLE
FLOATING UNDERFLOW EXCEPTION ENABLE
INTEGER OVERFLOW TRAP ENABLE
TRACE FAULT ENABLE
NEGATIVE CONDITION CODE
ZERO CONDITION CODE
OVERFLOW CONDITION CODE
CARRY (BORROW) CONDIT ION CODE

Figure 4-3
Processor Status Word
4-8

• address manipulation instructions

Table 4-3 lists the basic instruction operations in order by
these classifications . Instructions that enable operating
system procedures to provide user mode processes with
services requiring privilege are listed in the table, but discussed in the system programming environment section .
Instructions that are singular in the functions they provide
are listed last. The following paragraphs describe the functions of most of the instructions within each class. ·For further information on the instruction set, refer to the VAX-11
Architecture Handbook .

• user-programmed general register control instructions
Instructions that provide basic program flow control , and
enable the user to call procedures are:
• branch , jump and case instructions
• subroutine call instructions
• procedure call instructions

Table 4-3
Instruction Set Summary

Integer and Floating Point Logical Instructions

Packed Decimal Instructions

MOV_'
MNEG_
MCOM _
MOVZ_
CLR_
CVT_

MOVP
CMPP3
CMPP4
ASHP
ADDP4
ADDP6
SUBP4
SUBP6
MULP
DIVP
CVTLP
CVTPL
CVTPT
CVTTP
CVTPS
CVTSP
EDIT PC

CVTR_L
CMP_
TST_
BIS _2
BIS _3
BIC_2
BIG_3
BIT_
XOR_2
XOR_3
ROTL

Move (B,W,L,F,D,G,H,Q,O)"
Move Negated (B,W,L,F,D,G,H)
Move Complemented (B ,W,L)
Move Zero-Extended (BW,BL ,WL)
Clear (B,W,L=F,Q=D=G ,O=H)
Convert (B ,W,L ,F,D,G,H)(B ,W,L,F,D,G,H)
except BB ,WW ,LL,FF,DD ,GG ,HH ,DG ,
and GD
Convert Rounded (F,D,G,H) to Longword
Compare (B,W ,L,F,D,G,H)
Test (B,W,L,F,D,G,H)
Bit Set (B ,W,L) 2-0perand
Bit Set (B,W,L) 3-0perand
Bit Clear (B,W,L) 2-0perand
Bit Clear (B,W,L) 3-0perand
Bit Test (B,W,L)
Exclusive OR (B,W,L) 2-0perand
Exclusive OR (B ,W,L) 3-0perand
Rotate Longword

Move Packed
Compare Packed 3-0perand
Compare Packed 4-0perand
Arithmetic Shift Packed and Round
Add Packed 4-0perand
Add Packed 6-0perand
Subtract Packed 4-0perand
Subtract Packed 6-0perand
Multiply Packed
Divide Packed
Convert Long to Packed
Convert Packed to Long
Convert Packed to Trailing
Convert Trailing to Packed
Convert Packed to Separate
Convert Separate to Packed
Edit Packed to Character String

Character String Instructions
MOVC3
MOVC5
MOVTC
MO VT UC
CMPC3
CMPC5
LOCC
SKPC
SCANC
SPANG
MATCHC

Integer and Floating Point Arithmetic Instructions
INC_
DEC_
ASH _
ADD_2
ADD_3
ADWC
ADAWI
SUB_2
SUB_3
SBWC
MUL--2
MUL_3
EMUL
DIV--2
DIV_3
EDIV
EMOD_
POLY_

Increment (B ,W,L)
Decrement (B,W,L)
Arithmetic Shift (L,Q)
Add (B,W ,L,F,D,G,H) 2-0perand
Add (B ,W,L,F,D,G,H) 3-0perand
Add with Carry
Add Aligned Word Interlocked
Subtract (B ,W,L,F,D,G,H) 2-0perand
Subtract (B,W,L,F,D,G,H) 3-0perand
Subtract with Carry
Multiply (B ,W,L,F,D,G,H) 2-0perand
Multiply (B ,W,L,F,D,G ,H) 3-0perand
Extended Multiply
Divide (B ,W,L,F,D,G ,H) 2-0perand
Divide (B ,W,L,F,D,G,H) 3-0perand
Extended Divide
Extended Modulus (F,D,G,H)
Polynomial Evaluation (F,D,G,H)

Move Character 3-0perand
Move Character 5-0perand
Move Translated Characters
Move Translated Until Character
Compare Characters 3-0perand
Compare Characters 5-0perand
Locate Character
Skip Character
Scan Characters
Span Characters
Match Characters

Variable-Length Bit Field Instructions
EXTV
EXTZV
INSV
CMPV
CMPZV
FFS
FFC

4-9

Extract Field
Extract Zero- Extended Field
Insert Field
Compare Field
Compare Zero-Extended Field
Find First Set
Find First Clear

Table 4-3 (Cont.)
Instruction Set Summary
Loop and Case Branch

Index Instruction
INDEX

Compute Index

ACB_
AOBLEQ
AOBLSS
SOBGEQ

Insert Entry in Queue
Insert Entry into Queue at Head , Interlocked
Insert Entry into Queue at Tail, Interlocked
Remove Entry from Queue
Remove Entry from Queue at Head ,
Interlocked
Remove Entry from Queue at Tail ,
Interlocked

SO BG TR
CASE_

Queue Instructions
INSQUE
INSQHI
INSQTI
REMQUE
REMQHI
REMQTI

---- --------------------

Add, Compare and Branch (B,W,L,F,D ,G,H)
Add One and Branch Less Than or Equal
Add One and Branch Less Than
Subtract One and Branch Greater
Than or Equal
Subtract One and Branch Greater Tha n
Case on (B ,W,L)

Subroutine Call and Return Instructions
BSB_
JSB
RSB

Branch to Subroutine with (B, W)
Displacement
Jump to Subroutine
Return from Subroutine

Address Manipulation Instructions
MOVA_
PUS HA_

Procedure Call and Return Instructions

Move Address {B,W,L=F,Q=D=G ,O=H)
Push Address (B,W,L=F,Q=D=G,O=H)
on Stack

CAL LG
CALLS
RET

-- -- -----------

Call Procedure with General Argument List
Call Procedure with Stack Argument List
Return from Procedure

General Register Manipulation Instructions
-- --------------

PUSHL
PUSHR
POPA
MOVPSL
BISPSW
BICPSW

Push Longword on Stack
Push Registers on Stack
Pop Registers from Stack
Move from Processor Status Longword
Bit Set Processor Status Word
Bit Clear Processor Status Word

---------

Protected Procedure Call and Return Instructions
CHM _
REI
PROBER
PRO BEW

Change Mode to (Kernel, Executive,
Supervisor, User)
Return from Exception or Interrupt
Probe Read
Probe Write

Unconditional Branch and Jump Instructions
Privileged Processor Register Control Instructions
BR_
JMP

Branch with {B , W) Displacement
Jump

SVPCTX
LDPCTX
MTPR
MFPR

Branch on Condition Code
BLSS
BLSSU
BLEQ
BLEQU
BEQL
(BEQLU)
BNEQ
(BNEQU)
BGTR
BGTRU
BGEQ
BGEQU
(BCC)
(BCS)
BVS
BVC

Less Than
Less than Unsigned
Less than or Equal
Less than or Equal Unsigned
Equal
(Equal Unsigned)
Not Equal
(Not Equal Unsigned)
Greater than
Greater than Unsigned
Greater than or Equal
Greater than or Equal Unsigned
(Carry Cleared)
(Carry Set)
Overflow Set
Overflow Clear

Special Function Instructions
CRC
BPT
XFC
NOP
HALT

Cyclic Redundancy Check
Breakpoint Fault
Extended Function Call
No Operation
Halt

The underscore following the instruction name implies that the
instruction will operate upon any data type contained in the
parentheses following that instruction .
•• B =byte
W =word
L =longword
Q =quadword
0 = octaword
F = F_floating
D = D_floating
G = G_floating
H = H_floating

Branch on Bit
-----------------

BLB_
BB_
BBS_
BBC_
BBSSI
BBCCI

Save Process Context
Load Process Context
Move to Process Register
Move from Processor Register

Branch on Low Bit (Set, Clear)
Branch on Bit (Set, Clear)
Branch on Bit Set and (Set, Clear) Bit
Branch on Bit Clear and (Set, Clear) Bit
Branch on Bit Set and Set Bit Interlocked
Branch on Bit Clear and Clear Bit
Interlocked

Instructions that operate on G, H, and 0 formats are only available on VAX family processors equipped with the extended
range floating point option .

4-10

The Extended Modulus (EMOD) instructions multiply a floating point number with an extended precision floating
point number (extended by eight bits for F_floating and
D_floating for an effective 9 or 19 digits of accuracy) and
returns the integer portion and the fractional portion separately. This instruction is particularly useful for the reduction of the argument of trigonometric and exponential
functions to a standard interval.

Integer and Floating Point Instructions
The logical and integer arithmetic instructions illustrate
how the opcodes , data types, and addressing modes can
be combined in an instruction . Most of the operations provided for integer data are also provided for floating point
and packed decimal data. Exceptions are the strictly logical operations for integer data (such as bit clear, bit set,
complement) , the multiword arithmetic instructions for integer data (such as Add/Subtract with Carry and Extended
Multiply and Extended Divide), and the Extended Modulus
and Polynomial instructions for floating point data.

The Polynomial Evaluation (POLY) instructions evaluate a
polynomial from a table of coefficients using Homer's
method . This instruction is used extensively in the highlevel languages' math library for operations such as sine
and cosine.

The arithmetic instructions include both 2-operand and 3operand forms that eliminate the need to move data to and
from temporary operands. The 2-operand instructions
store the result in one of the two operands , as in "Set A
equal to A plus B. " The 3-operand instructions effectively
implement the high-level language statements in which
two different variables are used to calculate a third , such
as " Set C equal to A plus B." The 3-operand instructions
are applicable to both integer and floating point data, and
equivalent instructions exist for packed decimal data.

Packed Decimal Instructions
Many of the operations for integer and floating point data
also apply to packed decimal strings. They include:
• Move Packed (MOVP) for copying a packed decimal
string from one location to another, and Arithmetic Shift
Packed (ASHP) for scaling a packed decimal up or
down by a given power of 10 while moving it, and optionally rounding the value .

To illustrate the instruction set and addressing modes,
consider the FORTRAN language statement:

• Compare Packed (CMPP) for comparing two packed
decimal strings. Compare Packed has two variations: a
3-operand (CMPP3) i nstruction for strings of equal
length , and a 4-operand instruction (CMPP4) for strings
of differing lengths.

A(I) = 8(1) * C(I)
where A, B, and C are statically allocated REAL •4 arrays
and I is INTEGER*4 . A code sequence that performs this
operation is:
MOVL

l,RO

MULF3

B[RO] ,C[RO],A[RO)

• Convert Instructions, including Convert Long to Packed
(CVTLP), Convert Packed to Long (CVTPL) , Convert
Packed to Numeric with Trailing sign (CVTPT), Convert
Numeric with Trailing sign to Packed (CVTTP), Convert
Packed to Numeric with Separate overpunched sign
(CVTPS) , and Convert Numeric with Separate overpunched sign to Packed (CVTSP) . These instructions
enable the conversion of our packed decimal format to
commonly used numeric formats . Numeric with trailing
sign allows various sign encodings including zoned and
overpunched .

;Move the longword I
;to a register
;3-operand floating
;multiply

The same code applies if A, B, and Care REAL *8, INTEGER*4 , INTEGER*2, or even INTEGER*1 data types: the
MULF3 instruction is simply changed to MULD3, MULL3,
MULW3, or MULB3, respectively.
If arrays A, B, and Care dynamically allocated arrays, the
code sequence could be:
MOVL
MULF3

• Add Packed (ADDP) and Subtract Packed (SUBP) for
adding or subtracting two packed decimal strings, with
the option of replacing the addend or subtrahend with
the result (ADDP4 and SUBP4), or storing the result in a
third string (ADDP6 or SUBP6).

l,RO
Bdisp(FP)[RO) ,Cdisp(FP)[RO),Adisp(FP)[RO]

If A, B, and C are arguments to a procedure, the code
could be:
MOVL
MULF3

l,RO
@Bargptr(AP)[RO],@Cargptr(AP)[RO),
;@Aarg ptr[ RO)

• Multiply Packed (MULP) and Divide Packed (DIVP) for
multiplying or dividing two packed decimal strings and
storing the result in a third string .

In fact, the locations of A, B, and C can be arbitrarily selected . For example, combining the above, if A is statically
allocated , B dynamically allocated, and C an argument,
then the code sequence could be:
MOVL
MULF3

In addition , the packed decimal instructions include a special packed decimal string to character string conversion
instruction that provides output formatting : the Edit instruction .

l,RO
Bdisp(FP)[RO) ,@Cargptr(AP)[RO] ,A[RO)

Edit Instruction
The Edit Packed to Character String (EDITPC) instruction
supplies formatted numeric output functions. The instruction converts a given packed decimal string to a character
string using selected pattern operators. The pattern operators enable the creation of numeric output fields with any
of the following characteristics:

Some of the arithmetic instructions are used for extending
the accuracy of repeated computations. The Extended
Multiply (EMUL) instruction takes longword integer arguments and produces a quadword result. The instruction effectively implements a high-level language statement such
as "Set D equal to (A times B) plus C. " The Extended Divide (EDIV) instruction divides a quadword integer by a
longword and produces a longword quotient and a longword remainder.

• leading zero fill
• leading zero protection
• leading asterisk fill protection
4-11

of a 256-byte table of character type definitions. Fo r each
character in the given string , the instruction looks up the
type code in the table for that character , and then AN D the
given mask with the character's type code . SPANC finds
the first character in the string which is of the type indicated by its mask. SCANC finds the first cha racter in the
string which is of any type other the one indicated by its
mask.

• a floating sign
• a floating currency symbol
• special sign representations
• insertion characters
• blank when zero
Character String Instructions
The character string instructions operate on strings of
bytes. They include:

The Index Instruction
The Index instruction (INDEX) calculates an index for an
array of fixed length data types (integer and floatin g) and
for arrays of bit fields, character str ings, and decimal
strings. It accepts as arguments: a subscript, lower and
upper subscript bounds, an array element size , a g iven index, and a destination for the calculated index. It incorporates range check ing within the calculation for hig h-level
languages using subscript bounds , and it allows index calculation optimization by removing invariant expressions.

• move string instructions, with translation options
• string compare instructions
• single character search instructions
• substring search instructions
There are two basic forms of Move instructions for character strings. The Move Character instructions (MOVC3 and
MOVC5) simply copy character strings from one location
to another. They are optimized for block transfer operations . The 5-operand variation provides for a fill character
(user-supplied) that the instruction uses to pad out the
destination location to a given size .

The COBOL statements:
01
01

The Move Translated Characters (MOVTC) and Move
Translated Until Character (MOVTUC) instructions actually
create new character strings. The user supplies a string
which the instruction uses as a list of offsets into a translation table. The instruction selects characters from the table
in the order that the offset list points to the table. The
MOVTC instruction allows the user to supply a fill character that the instruction uses to pad out the resultant string
to a given size with an arbitrary character. The MOVTUC
instruction allows the user to supply any number of escape
characters. When the next offset points to an escape character in the table, translation stops.

A-ARRAY
02
A PIC X(10) OCCURS 15 TIMES.
B PIC X(10) .
MOVE A(I) TO B.

are equ ivalent to :
INDEX I, #1 , #15 , #10 , #0, RO

; 1 less than or equa l I
; I less than or equal 15
; (0 + 1) *1 0 is
; stored in RO

MOVC3 #10, A-10[RO],B
The FORTRAN statements:
INTEGER*4A(L1 :U1 , L2 :U2), I, J
A(l ,J) = 1

The Compare Characters (CMPC) instructions provide
character-by-character byte string compares. CMPC has a
3-operand form and a 5-operand form . Both instructions
compare two strings from beg inning to end, informing the
user when they reach the first character that is different
between the strings, or when they get to the end of either
str ing . The 5-operand variation provides for a fill character
which it uses to effectively pad out a string when comparing it with a longer one.

are equivalent to :
INDEX J, #L2 , #U2 , #M1 , #0, RO
INDEX I, #L 1, #U1 , #1 , RO, RO
MOVL #1 , A-a[RO)

; M1=U1 - L1
; a= ((L2*M1) +L 1)*4

Variable-Length Bit Field Instructions
The bit field instructions enable the user to define , acce ss,
and modify those fields whose size and location were userspecified. Location is determined from a base address o r a
register and a signed bit offset. If the field is in memory,
the offset can reach bits located up to 23 ' bits (approximately 256 million bytes) away in either directio n. If the
field is in a reg ister, the offset can be large as 31 . Fields of
arbitrary lengths (0 to 32 bits) may be used for storing data
structure header information compactly , for status codes,
or for creating user data types . The field instructions enable manipulation of fields easily.

The Locate Character (LOCC) and Skip Character (SKPC)
instructions are search instructions for single characters
within a string . LOCC searches a given string for a character that matches the search character supplied by the
user. This is useful, for example, when search ing for the
delimiter at the end of a variable-length string . SKPC, on
the other hand , finds the first character in the string that is
different from the search character supplied . This is useful
for sk ipping through fill characters at the end of a field to
find the beginning of the next field.
The Match Characters (MATCHC) instruction is sim ilar to
the Locate Character instruction , but it locates multiplecharacter substrings. MATCHC searches a string for the
first occurrence of a substring supplied by the user.

The Insert Field and Extract Field instructions store data in
and retrieve data from fields. Insert Field (INSV) sto res data in a field by taking a specified number of bits of a longword (starting from the low-order bit) and writing them into
a field , which may start at any bit relative to a given base
address. The Extract Field instruction retrieves data from a
field by copying the bit field and storing it in the low-order

The Span Characters (SPANG) and Scan Characters
(SCANC) instructions are search instructions that look for
members of character classes. For these instructions the
user supplies a character string , a mask, and the address
4- 12

bits of a longword . The field can either be signed (EXTV) or
unsigned (EXTZV).
The Compare Field and Find First instructions enable the
user to test the contents of a field . Compare Field extracts
a field and then compares it with a given longword . The
field can be interpreted as signed (CMPV) , or as unsigned
(CMPZV). The Find First instructions locate the first bit in a
field that is clear (FFC) or set (FFS), scanning from low-order bit to high-order bit. These instructions are particularly
useful for scanning a status control longword . For example, the longword may represent a set of queues processed in order by priority 0 (high) to 31 (low) . Each set bit
represents an active queue. The Find First Set instruction
quickly returns the highest priority queue that is active.
Together with the SKPC instructions, the Find First instructions are also useful for scanning an allocation table (bit
map) of arbitrary length .

The Push Address instruction is useful for computing an
address to be passed to a called subroutine or procedure.
Move Address is useful for loading a base register and
performing run time position-independent address computation. It has some interesting uses because it is effectively an ADD instruction :
MOVAB disp(R1 )[R2] ,X

; sets X=R1 +R2+displacement
; (two adds in one instruction)

MOVA_disp(Rn)[Rn ], Rn

; multiplies Rn by 3
; (for MOVAW),
; 5 (for MOVAL),
; or 9 (for MOVAQ)
; and adds displacement to it.

General Register Manipulation Instructions
The general register manipulation instructions enable any
user program to save or load the general purpose registers in one operation , examine the Processor Status Longword , and set or clear status bits in the Processor Status
Word. (Processor register control instructions primarily
used by operating system software are covered later.)

Queue Instructions
The processor has six instructions that allow easy construct ion and maintenance of queue data structures .
Queues manipulated using the queue instructions are circular, doubly linked lists of data items.

The Push Longword (PUSHL) instruction pushes a longword on the stack . This instruction is the same as a Move
Longword using the Stack Pointer in register deferred
mode, but is a byte shorter. It is a consistent and convenient way to move data to the stack .

The first longword of a queue entry contains the forward
pointer to the next entry in the queue, and the next longword contains the backward pointer to the preceding entry
in the queue.
Two types of queues are provided : absolute and self-relative. Absolute queues use pointers that are virtual addresses, whereas self-relative queues use pointers that
are relative displacements.

The Push Registers (PUSHR) instruction pushes a set of
registers on the stack in one operation. The user supplies
a mask word in which each set bit (0-14) represents a
register (RO-R14) to be saved on the stack. (The only general register that cannot be saved using this instruction is
R15, the Program Counter .) Pop Registers (POPA)
reverses the operation , loading each register from successive longwords on the stack according to the given mask
word . The PUSHR and POPA instructions replace the
need to write a sequence of Move instructions to save and
restore registers upon entry and exit from a subroutine.

Two instructions are provided for manipulating absolute
queues: INSQUE, and REMQUE . INSQUE inserts an entry
specified by an entry operand into the queue following the
entry specified by the predecessor operand . REMQUE removes the entry specified by the entry operand. Queue entries can be on arbitrary byte boundaries. Both INSQUE
and REMQUE are implemented as non-interruptible instructions.

The Move from Processor Status Longword (MOVPSL) instruction allows examination of the contents of the processor's status register by loading its contents into a specified
location . The Bit Set (BISPSW) and Bit Clear (BICPSW)
Processor Status Word instructions enable the user to set
or clear the PSW condition codes and trap enable bits. The
mask bits represent the bits to be set or cleared .

Four operations can be performed on self-relative queues:
insert at head (INSQHI), remove from head (REMQHI) , insert at tail (INSQTI), and remove from tail (REMQTI). Furthermore, these operations are interlocked to allow cooperating processes in a multiprocessor system to access a
shared queue without additional synchronization . Queue
entries must be quadword aligned .

Branch, Jump and Case Instructions
The two basic types of control transfer instructions are
branch and jump instructions. Both branch and jump load
new addresses in the Program Counter. With branch instructions, the user supplied displacement (offset) is added to the current contents of the Program Counter to obtain the new address. The jump instructions allow the user
specified address to be loaded, using one of the normal
addressing modes.

Address Manipulation Instructions
Because the processor offers a variety of addressing
modes enabling access to data structures easily via base
addresses and indices in registers, addresses are often
manipulated . The processor provides two instructions enabling an address to be fetched without actually accessing
the data at that location :
• The Move Address (MOVA) instruction, which stores the
address of a byte, word, longword (and floating), or
quadword (and double floating) datum in a specified
register or location in memory.

Because most transfers are to locations relatively close to
the current instruction , and branch instructions are more
efficient than jump instructions, the processor offers a variety of branch instructions to choose from. There are two
unconditional branch instructions and many conditional
branch instructions.

• The Push Address (PUSHA) instruction, which stores
the address of a byte, word, longword (and floating) , or
quadword (and double floating) datum on the stack.
4-13

• Branch on Bit Set and Set (BBSS)

The unconditional branch instructions allow specification
of either a byte-size (BAB) or a word-size displacement
(BRW), thereby permitting displacements as far away from
the current location as 32,767 bytes in either direction. The
Jump instruction (JMP) should be used for transfer of control to locations greater in displacement than 32,767 bytes.

• Branch on Bit Clear and Clear (BBCC)
• Branch on Bit Set and Clear (BBSC)
• Branch on Bit Clear and Set (BBCS)
These instructions are particularly useful for keeping track
of procedure comp letion or initialization , and for signal ing
the completion or initialization of a procedure to a cooperating process. In addition, there are two Branch on Bit Interlocked instructions that provide control variable protection :

Most conditional branches allow only byte displacements,
although some of the more powerful, such as the Add
Compare and Branch instruction, allow word displacements. Conditional branch instructions include:
• branch on bit instructions

• Branch on Bit Set and Set Interlocked (BBSSI)

• set and clear bit instructions with a branch if it is already
set or cleared

• Branch on Bit Clear and Clear Interlocked (BBCCI)
The memory interconnect bus provides a memory interlock on these instructions . No other BBSSI or BBCCI
operation can interrupt these instructions to gain access to
the byte containing the control variable between the testing of the bit and the setting or clearing of the bit.

• loop instructions that increment or decrement a counter, compare it with a limit value, and branch on a relational condition
• computed branch instruction in which a branch may
take place to one of several locations depending on a
computed value

The processor offers three types of branch instructions
that can be used to write efficient loops. The first type provides the basic subtract-one-and-branch loop. A counter
variable (user-supplied) is decremented each time the
loop is executed . In the Subtract One and Branch Greater
Than (SOBGTR) instruction, the loop repeats until the
counter equals zero. In the Subtract One and Branch
Greater Than or Equal (SOBGEQ) instruction , the loop repeats until the counter becomes negative.

The Branch on Condition (B) instructions enable transfer
of control to another location depending on the status of
one or more of the condition codes in the Processor Status
Word (PSW). There are three groups of Branch on Condition instructions :
• The signed relational branches, which are used to test
the outcome of instructions operating on integer and
field data types being treated as signed integers, floating point data types, and decimal strings.

The counterpart to subtract-one-and-branch is add-oneand-branch . A counter and a limit must be supplied by the
user. The counter is incremented at the end of the loop. In
the Add One and Branch Less Than (AOBLSS) instruction,
the loop repeats until the counter equals the user-defined
limit. In the Add One and Branch Less Than or Equal (AOBLEQ) instruction, the loop repeats until the counter
exceeds the user defined limit.

• The unsigned relational branches, which are used to
test the outcome of instructions operating on integer
and field data types being treated as unsigned integers,
character strings, and addresses.
• The overflow and carry test branches, which are used
for checking overflow when traps are not enabled, for
multiprecision arithmetic, and for the results of special
instructions.

The third type of loop instruction efficiently implements the
FORTRAN language DO statement and the BASIC language FOR statement: Add Compare and Branch (ACB).
In this case, the user must supply a limit, a counter , and a
step value. For each execution of the loop, the instruction
adds the step value to the counter and compares the
counter to the limit. The sign of the step value determ ines
the logical relation of the comparison: the instruction loops
on a less than or equal comparison if the step value is positive , on a greater than or equal comparison if the step
value is negative.

The instruction mnemonics clearly indicate the choice
between a signed and unsigned integer data type interpretation for relational testing . The relational tests determine
if the result of the previous operation is less than, less than
or equal, equal, not equal, greater than or equal, or greater
than zero . For example, the Branch on Less than or Equal
Unsigned (BLEQU) instruction branches if either the Carry
or Zero bit is set. The Branch on Greater Than (BGTR) instruction branches if neither the Negative nor the Zero bit
is set.
General purpose Branch on Bit instructions similar to
Branch on Condition also exist. The Branch on Low Bit Set
(BLBS) and Branch on Low Bit Clear (BLBC) instructions
test bit 0 of an operand, which is useful for testing Boolean
values. The Branch on Bit Set (BBS) and Branch on Bit
Clear (BBC) instructions test any selected bit.

The processor provides a branch instruction that implements higher-level language computed GO TO statements: the CASE instruction. To execute the CASE instruction , the user must supply a list of displacements that
generate different branch addresses indexed by the value
obtained as a selector. The branch falls through if theselector does not fall within the limits of the list.

There are special kinds of Branch on Bit instructions that
are actually bit set/clear instructions. The Branch on Bit
Set and Set (BBSS) is an example. The instruction branches if the indicated bit is set, otherwise it falls through. In
either case, the instruction sets the given bit. The BBSS instruction can thus be thought of as a Bit Set instruction
with a branch side-effect if the bit was already set. There
are four permutations:

Subroutine Branch, Jump, and Return Instructions
Two special types of branch and jump instruction are
provided for calling subroutines: the Branch to Subroutine
(BSB) and Jump to Subroutine (JSB) instructions. Both
BSB and JSB instructions save the contents of the Program Counter on the stack before loading the Program
Counter with the new address. With Branch to Subroutine,

4-14

the user supplies either a byte (BSBB) or word (BSBW)
displacement. With Jump to Subroutine, regular addressing is used.

The Cyclic Redundancy Check (CRC) instruction calculates a cyclic redundancy check for a given string using
any CRC polynomial up to 32 bits long . The user supplies
the string for which the CRC is to be performed, and a table for the CRC function. The operating system library includes tables for standard CRC functions, such as CRC16.

The subroutine call instructions are complemented by the
Return from Subroutine (RSB) instruction. RSB pops the
first longword off the stack and loads it into the Program
Counter. Since the Branch to Subroutine instruction is either two or three bytes long , and the Return from Subroutine instruction is one byte long, it is possible to write extremely efficient programs using subroutines.

The Breakpoint Fault (BPT) instruction makes the processor execute the kernel mode condition handler associated
with the Breakpoint Fault exception vector. BPT is used by
the operating system debugging utilities, but can also be
used by any process that sets up a Breakpoint Fault condition handler.

Procedure Call and Return Instructions
Procedures are general purpose routines that use
argument lists passed automatically by the processor. The
procedure Call instructions enable language processors
and the operating system to provide a standard calling interface. They :
• save all the registers that the procedure uses, and only
those registers , before entering the procedure

The Extended Function Call (XFC) instruction allows
escapes to customer-defined instructions in writable control store. The NOP instruction is useful for debugging .
The HALT instruction is a privileged instruction issued only
by the operating system to halt the processor when bringing the system down by operator request.

• pass an argument list to a procedure

COMPATIBILITY MODE

• maintain the Stack, Frame, and Argument Pointer registers

Under control of the operating system, the processor can
execute PDP-11 instruction streams within the context of
any process. When executing in compatibility mode, the
processor interprets the instruction stream executing in
the context of the current process as a subset of the PDP11 instruction set.

• initialize the arithmetic trap enables to a given state
When issuing a Call procedure instruction , the address of
the procedure being called, must be included . The first
word of a procedure contains an entry mask that is used in
the same way as the entry mask defined for the Push
Registers instruction. Each set bit of the 12 low-order bits
in the word represents one of the general registers, RO
through R11, that the procedure uses. The Call instruction
examines this word and saves the indicated registers on
the stack. In addition, the Call instruction also automatically saves the contents of the Frame Pointer, Argument
Pointer, and Program Counter registers. This is an extremely efficient way to ensure that registers are saved
across procedure calls . No general register is saved that
does not have to be saved.

In general , compatibil ity mode enables the operating system to provide an environment for executing most user
mode programs written for a PDP-11 except stand-alone
software. The processor expects all compatibility mode
software to rely on the services of the native operating system for 1/0 processing, interrupt and exception handling ,
and memory management. There are some restrictions,
however, on the environment that the native operating system can provide a PDP-11 program. For example, the
PDP-11 memory management instructions Move To/From
Previous Instruction/Data Space can not be simulated by
the operating system since they do not trap to native mode
software.

The Call Procedure with General Argument List (CALLG)
instruction accepts the address of an argument list and
passes the address to the procedure in the Argument
Pointer register. The Call Procedure with Stack Argument
List (CALLS) passes the argument list, (placed on the
stack by the user) by loading the Argument Pointer register with its stack address.

PDP-11 Program Environment
PDP-11 addresses are 16-bit byte addresses. There is a
one-to-one correspondence between compatibility mode
virtual addresses and the first 64K bytes of virtual address
space available to native mode processes. As in the PDP11, a compatibility mode program is restricted to referencing only these addresses. It is possible for the operating
system to provide most of the PDP-11 memory management mechanisms. For example, compatibility mode automatically supports PDP-11 memory segment protection,
but in 512-byte rather than 64-byte segments.

When a procedure completes execution , it issues the Return from Procedure instruction (RET). Return uses the
Frame Pointer register to find the saved registers that it restores, and to clean up any data left on the stack, including
nested routine linkages. A procedure can return values using the argument list or other registers.

All of the PDP-11 general registers and addressing modes
are available in compatibility mode. Compatibility mode
registers RO through R6 are the low-order 16 bits of native
mode registers RO through R6 . Compatibility mode R? (the
Program Counter) is the low-order bits of native mode register 15 (the Program Counter). Native mode registers 8
through 14 are not affected by compatibility mode. Note
that the compatibility mode register R6 acts as the Stack
Pointer for program-local temporary data storage, but that
the program-local stack is allocated address space in the
program region , not the control region.

Miscellaneous Special Purpose Instructions
The processor has a number of special purpose instructions. They include:
• Cycl ic Redundancy Check (CRC)
• Breakpoint Fault (BPT)
• Extended Function Call (XFC)
• No Operation (NOP)
• Halt
4-15

To support multiprogramming for a high-performance
system, the processor enables the operating system to
switch rapidly between individual streams of code. The
stream of code the processor is executing at any one time
is determined by its hardware context. Hardware context
includes the information loaded in the processor's registers that identifies:

A subset of the PDP-11 Processor Status Word is defined
for compatibility mode. Only the condition codes and the
trace trap bit are relevant for the PDP-11 instruction
stream .
All interrupts and exceptions that occur when the processor is executing in compatibility mode cause the processor
to enter native mode. As in native mode, it is the operating
system's responsibility to handle interrupts and exceptions. There are a few types of exceptions that apply only
to compatibility mode. They include illegal instruction exceptions and odd address trap.

• where the stream's instructions and data are located
• which instruction to execute next
• what the processor status is during execution
A process is a stream of instructions and data defined by a
hardware context. Each process has a unique identification in the system. The operating system switches between
processes by requesting the processor to save one process hardware context and load another. Context switching
occurs rapidly because the processor instruction set includes save hardware context and load hardware context
instructions. The operating system 's context switching
software does not have to individually save or load the
processor registers which define the hardware context.

PDP-11 Instruction Set
The compatibility mode instruction set is that of the PDP11 with the following exceptions:
• The privileged instructions (HALT, WAIT , RESET, SPL,
and MARK) are illegal.
• The trap instructions (BPT, IOT EMT, and TRAP) cause
the processor to enter native mode, where either the
trap may be serviced, or the instruction simulated
• The Move From/To Previous Instruction/Data space instructions (MFPI, MTPI, MFPD, and MTPD) execute exactly as they would on a PDP-11 in user mode with instruction and data space overmapped. They ignore the
previous access level and act as PUSH and POP instructions referencing the current stack.

The actual scheduling mechanism for arbitrating among
processes competing for processor time is left to the operating system software itself to give the system flexibility.
Priority Dispatching
While running in the context of one process, the processor
executes instructions and controls data flow to and from
peripherals and main memory. To share processor, memory and peripheral resources among many processes , the
processor provides two arbitration mechanisms that support high-performance multiprogramming: exceptions
and interrupts. Exceptions are events that occur synchronously with respect to instruction execution, while interrupts are external events that occur asynchronously.

• PDP-11 floating point instructions are emulated through
software.
All other instructions execute as they would on a PDP11 /70 processor running in user mode.

PROCESSING CONCEPTS FOR
SYSTEM PROGRAMMING

The flow of execution can change at any time, and the
processor distinguishes between changes in flow that are
local to a process and those that are system-wide. Process-local changes occur as the result of a user software error or when user software calls operating system services.
Process-local changes in program flow are handled
through the processor's exception detection mechanism
and the operating system's exception dispatcher.

The processor is specifically designed to support a highperformance multiprogramming environment. The chief
advantage of a multiprogramming system is its ability to
get the most out of a computer that is being used for several different purposes concurrently. For example, multiprogramming enables the simultaneous execution of two
or more application systems, such as process control and
order entry. It is also possible to execute several application systems while simultaneously developing application
programs. The characteristics of the hardware system that
support multiprogramming are:

System-wide changes in flow generally occur as the result
of interrupts from devices or interrupts generated by the
operating system software. Interrupts are handled by the
processor's interrupt detection mechanism and the operating system's interrupt service routines. (System-wide
changes in flow may also occur as the result of severe
hardware errors, in which case they are handled either as
special exceptions or high-priority interrupts.)

• rapid context switching
• priority dispatching
• virtual addressing and memory management
As a multiprogramming system, VAX not only provides the
ability to share the processor among processes, but also
protects processes from one another while enabling them
to communicate with each other and share code and data.

System-wide changes in flow take priority over process- local changes in flow. Furthermore, the processor uses a
priority system for servicing interrupts . To arbitrate
between all possible interrupts, each kind of interrupt is
assigned a priority, and the processor responds to the
highest priority interrupt pending. For example, interrupts
from real-time 1/0 devices would take precedence over interrupts from mass-storage devices, terminals, line printers and other less time-critical devices.

Context Switching
In a multiprogramming environment, several individual
streams of code can be ready to execute at any one time.
Instead of allowing each stream to execute to completion
serially (as in a batch-only system), the operating system
can intervene and switch between the streams of code
which are ready to execute.

The processor services interrupts between instructions, or
at well-defined points during the execution of long, itera4-16

tive instructions. When the processor acknowledges an interrupt , it switches rapidly to a special system-wide
context to enable the operating system to service the interrupt. System-wide changes in the flow of execution are
handled in such a way as to be totally transparent to individual processes.

processor registers.
Processor Status Longword
A processor register called the Processor Status Longword (PSL) determines the execution state of the processor at any time. The low-order 16 bits of the Processor
Status Longword is the Processor Status Word available to
the user process. The high-order 16 bits provide privileged control of the system. Figure 4-4 illustrates the Processor Status Longword.

Virtual Addressing and Virtual Memory
The processor's memory management hardware enables
the operating system to provide an execution environment
that allows users to write programs without having to know
where the programs are loaded in physical memory, and
to write programs that are too large to fit in the physical
memory they are allocated.

The fields can be grouped together by functions that control :
• the instruction set the processor is executing
• the access mode of the current instruction

The processor provides the operating system with the ability to provide virtual addressing . A virtual address is a 32bit integer that a program uses to identify storage locations in virtual memory. Virtual memory is the set of all
physical memory locations in the system plus the set of
disk blocks that the operatlng system designates as extensions to physical memory.

• interrupt processing
The instruction set the processor executes is controlled by
the compatibility mode bit in the Processor Status Longword . This bit is normally set or cleared by the operating
system . For further information on compatibility mode,
refer to the Operating System section.
The following paragraphs discuss access modes, the native instructions primarily used by the operating system ,
memory management, and interrupt processing.

A physical address is an address that the processor uses
to identify physical memory storage locations and peripheral controller registers . It is the physical address that
the processor sends to the memory and peripheral adaptors.

Processor Access Modes
In a high-performance multiprogramming system , the
processor must provide the basis for protection and sharing among the processes competing for the system 's resources. The basis for protection in this system is the
processor's access mode. The access mode in which the
processor executes determines:

The processor must be capable of translating virtual addresses provided by the executing program into the
physical addresses recognized by the memory and peripherals . To provide virtual to physical address mapping,
the processor has address mapping registers controlled
by the operating system and an integrated address translation buffer.

• instruction execution privileges: what instructions the
processor will execute

The mapping registers enable the operating system to relocate programs in physical memory, to protect programs
from each other, and share instructions and data between
programs transparently or at their request. The address
translation buffer ensures that the virtual address to physical address translation takes place rapidly.

• memory access privileges: which locations in memory
the current instruction can access
At any one time , the processor is executing code in the
context of a particular process, or it is executing in the system-wide interrupt service context. In the context of a
process, the processor recognizes four access modes:
kernel, executive, supervisor, and user. Kernel is the most
privileged mode and user the least privileged .

SYSTEM PROGRAMMING ENVIRONMENT

The processor spends most of its time executing in user
mode in the context of one process or another. When user
software needs the services of the operating system,
whether for acquisition of a resource, for 1/0 processing,
or for information, it calls those services.

Within the context of one process , user-level software controls its execution using the instruction sets, the general
registers and the Processor Status Word. Within the multiprogramming environment, the operating system controls the system 's execution using a set of special instructions , the Processor Status Longword, and the internal
31

The processor executes those services in the same or one
0

11 ~ 11 .. ~ .... I
'-._-II '-._-I

PROCESSOR STATUS WORD

~~---INTERRUPT

PRIORITY LEVEL
.
' - - - - - - - - - - PREVIOUS ACCES S MODE
' - - - - - - - - - - - - C U R R E N T ACCESS MODE
' - - -- ----------EXECUTING ON THE INTERRUPT STACK
' - - - - - - - - - - - - - - - I N S T R U C T I O N FIRST PART DONE
' - - - - - - - -- - - - - - - - - - T R A C E PENDING
' - - - - - - - - - - - - - - - - - - - C O M P A T I B I L I T Y MODE

Figure 4-4
Processor Status Longword
4-17

instruction is useful for providing general purpose se rvices
for user mode software. The system manager ultim ately
grants the privilege to write any code that handles Change
Mode traps to more privileged access modes.

of the more privileged access modes within the context of
that process. That is, all four access modes exist with in the
same virtual address space. Each access mode has its
own stack in the control region of per-process space , and
therefore each process has four stacks : one for each access mode. Note that this makes it easy for the operating
system to context switch a process even when it is executing an operating system service procedure.

For service procedures written to execute in privileged access modes (kernel , executive, and supervisor), the processor provides address access privilege val idation instructions. The PROBE instructions enable a procedu re to
check the read (PROBER) and write (PROBEW) access
protection of pages in memory against the privileges of the
caller who requested to access a particular location . This
enables the operating system to provide services that execute in privileged modes to less privileged callers and still
prevent the caller from accessing protected areas of memory .

In any mode except kernel, the processor will not execute
the instructions that:
• halt the processor
• load and save process context
• access the internal processor registers that control
memory management, interrupt processing , the processor console, or the processor clock

The operating system 's privileged service procedures and
interrupt and exception service routines exit using the Return from Exception or Interrupt (REI) instruction . REI is
the only way in which the privilege of the processo r's access mode can be decreased . Like the procedu re and
subroutine return instructions, REI restores the Prog ram
Counter and the processor state to resume the process at
the point where it was interrupted.

These instructions are privileged instructions that are generally reserved to the operating system .
In any mode, the processor will not allow the current instruction to access memory unless the mode is privileged
to do so. The ability to execute code in one of the more privileged modes is granted by the system manager and
controlled by the operating system . The memory protection the privilege affords is enforced by the processor. In
general, code executing in one mode can protect itself and
any portion of its data structures from read and/or write
access by code executing in any less privileged mode. For
example, code executing in executive mode can protect its
data structures from code executing in supervisor or user
mode. Code executing in supervisor mode can protect its
data structures from access by code executing in user
mode. This memory protection mechanism provides the
basis for system data structure integrity.

REI performs special services, however, that normal return
instructions do not. For example, REI checks to see if any
asynchronous system traps have been queued for the currently executing process while the interrupt or exception
service routine was executing, and ensures that the process will receive them . Furthermore, REI checks to e nsure
that the mode to which it is returning control is the same as
or less privileged than the mode in which the processor
was executing when the exception or interrupt occ urred .
Thus REI is available to all software including user-written
trap handling routines , but a program cannot increase its
privi lege by altering the processor state to be restored .

Protected and Privileged Instructions
The processor provides three types of instructions that enable user mode software to obtain operating system services without jeopardizing the integrity of the system . They
include:

When the operating system schedules a context switching
operation , the context switch ing procedure uses the Save
Process Context (SVPCTX) and Load Process Co ntext
(LDPCTX) instructions to save the current process co ntext
and load another. The operating system 's context switc hing procedure ident ifies the location of the hardware context to be loaded by updating an internal processor reg ister .

• the Change Mode instructions
• the PROBE instructions
• the Return from Exception or Interrupt instruction
User mode software can obtain privileged services by calling operating system service procedures with a standard
CALL instruction. The operating system's service dispatcher issues an appropriate Change Mode instruction
before actually entering the procedure. Change Mode allows access mode transitions to take place from one mode
to the same or more privileged mode only. When the mode
transition takes place, the previous mode is saved in the
Previous Mode field of the Processor Status Longword , allowing the more privileged code to determine the privilege
of its caller.

Internal processor registers not only include those that
identify the process currently executing , but also the mem ory management and other reg isters , such as the co nso le
and clock control registers. The Move to Processor Regi ster (MTPR) and Move from Processor Register (MFPR ) instructions are the only instructions that can expl ic itly access the internal processor registers . MTPR and MFPR are
privileged instructions that can be issued only in kernel
mode.

A Change Mode instruction is simply a special trap instruction that can be thought of as an operating system
service call instruction. User mode software can explicitly
issue Change Mode instructions, but since the operating
system receives the trap, non-privileged users can not
write any code to execute in any of the privileged access
modes. User mode software can include a condit ion
handler for Change Mode to User traps , however, and this

Memory Management
The processor is responsible for enforcing memory protection between access modes. Memory protection , however, is only a part of the processor's memory management function . In particular, the memory management
hardware enables the operating system to provide an extremely flex ible and efficient virtual memory program ming
environment. Virtual and physical address space defi4-18

physical address. A physical address is the address exchanged between the processor and the memory and peripheral adaptors. Physical address space is the set of all
possible physical addresses the processor can use to express unique memory locations and peripheral control
registers. Figure 4-5 compares the structure of the common virtual address space with that of the VAX-11 /750 and
VAX-11 /780 physical address spaces.

nitions provide the basis for the virtual memory available
on a system.
Virtual address space consists of all possible 32-bit addresses that can be exchanged between a program and
the processor to identify a byte location in physical memory. The memory management hardware translates a virtual
address into a physical address. A physical address can
be up to 30 bits in length as in the case of the VAX-11 /780.
Other processor implementations may choose a smaller

VAX-111780
PHYSICAL
ADDRESS
SPACE

VAX-11
VIRTUAL
ADDRESS
SPACE

VAX-11/750
PHYSICAL
ADDRESS
SPACE
0

AVAIL .
PHYSICAL
MEMORY

AVAIL.
PHYSICAL
MEMORY
BM BYTE s

2M BYTES>--------<

PROG

PHYSICAL
) MEMORY
ADDRESSES

REGION

PER PROCESS
SPACE
512M BYTES=2

PHYSICAL
MEMORY
ADDRESSES

15M BYTESl------l

IIO

DEVICES
15 .25M BYTESl------l

CONTROL
REGION

110
) SPACE
ADDRESSES

UNIBUS
SPACE

110
) SPACE
ADDRESSES

15 . 75M BYTESl-----l

UNIBUS
SPACE

16M BYTES=2 24 ' - - - - - ' ./

I..)

SYSTEM
REGION

SYSTEM
SPACE

Figure 4-5
Virtual and Physical Address Space

4-19

On the VAX-11 /780, physical address space is an array of
addresses which can be used to represent 2 30 byte locations , or approximately one billion bytes. Half of the addresses in VAX-11 /780 physical address space can be
used to refer to real memory locations and the other half
can be used to refer to peripheral device control and data
registers . The lowest-addressed half of physical address
space is called memory space, and the highest-addressed half 1/0 space. On the VAX-11/750, physical address space is an array of addresses which can be used to
represent 2 24 byte locations, or approximately 16 million
bytes. The first 15M bytes are dedicated to physical memory addresses while the last 1M byte is dedicated to 1/0
space.
The following section describes the way in which the memory management hardware enables the operating system
to map virtual addresses into physical addresses to provide the virtual memory available to a process .

Virtual to Physical Page Mapping
Virtual address space is divided into pages, where a page
represents 512 bytes of contiguously addressed memory.
The first page begins at byte zero and continues to byte
511 . The next page begins at byte 512 and continues to
byte 1023, and so forth . For example, decimal and hexadecimal addresses of the first eight pages of virtual addresss space are:
PAGE

ADDRESS(10)
decimal

ADDRESS(16)
hexadecimal

0
1
2
3
4
5
6
7

0000-0511
0512-1023
1024-1535
1536-2047
2048-2559
2560-3071
3072-3583
3584-4095

0000-01 FF
0200-03FF
0400-05FF
0600-07FF
0800-09FF
OAOO-OBFF
OCOO-ODFF
OEOO-OFFF

A page table is a virtually contiguous array of page table
entries. Each page table entry is a longword representing
the physical mapping for one virtual page. To translate a
virtual address to a physical address, therefore, the processor simply uses the virtual page number as an index into
the page table from the given page table base address .
Each translation is good for 512 virtual addresses since
the byte within the virtual page corresponds to the byte
within the physical page .
Figure 4-7 shows the format of a page table entry. The
high-order bits are used to indicate the page's status and
protection. The page's protection can be set to prevent
read and/or write access by any mode (kernel, executive ,
supervisor, or user). The page's status indicates what the
remainder of the page table entry means. It may be , for example, a page address in physical address space , a disk
. sector, or a temporary pointer to a page shared by two or
more processes. The system's virtual memory is a dynamic memory that is defined by the physical memory and
disk pages that are virtually mapped by page table entries .
The operating system's memory management software
maintains the page table entry protection and status bits ,
with the exception of the modified page bit. The processor
sets the modified page bit to indicate that it has written into
a physical page in memory. This is used to keep disk 1/ 0 to
a minimum when paging a process .
The processor uses the page table base registers to locate
the page tables, and uses the length registers as a validity
check to ensure that any given virtual page is in the range
of defined page table entries. Figure 4-8 summarizes and
compares the page table structures.
All process page tables have virtual addresses in the system region of virtual address space, but the system region
page table is located by its address in physical memory.
That is, the system region page table base register contains the physical address of the page table base , while the
process page table base registers contain the virtual addresses of their page table bases. Because a per-process
page table entry is referred to by a virtual address in the
system region , the hardware translates its virtual address
using the system page table .

The size of a virtual page exactly corresponds to the size of
a physical page of memory, and the size of a block on disk.
To make memory mapping efficient, the processor must
be capable of translating virtual addresses to physical addresses rapidly. Two features providing rapid address
translation are the processor's internal address translation
buffer and the translation algorithm itself.

There are two advantages to using a virtual address as the
base address of a per-process page table. The first advantage is that all page tables do not have to reside in physical
memory. The system region page table is the only page
table that needs to be resident in physical memory. All
process page tables can reside on disk ; that is, process
page tables can themselves be paged and swapped as
necessary.

Figure 4-6 compares the virtual address format to the
physical address formats of the VAX-11 /780 and VAX11 /750 processors. The high-order two bits of a virtual address immediately identify the region to which the virtual
address refers. Whether the address is physical (processor specific) or virtual, the byte within the page is the
same. Thus, the processor has to know only which virtual
pages correspond to which physical pages.

The second advantage is that the operating system 's
memory management software can allocate per-process
page tables dynamically, because the per-process page
tables do not need to be mapped into contiguous physical
pages. And although the system region page table must be
mapped into contiguous physical pages , this requirement
does not restrict physical memory allocation. The region is
shared among processes, and therefore does not requ ire
redefinition from context to context.

The processor has three pairs of page mapping registers,
one pair for each of the three regions actively used . The
operating system's memory management software loads
each pair of registers with the base address and length of
data structures it sets up called page tables. The page tables provide the mapping information for each virtual page
in the system . There is one page table for each of the three
regions.

To illustrate the efficiency of this memory mapp ing
scheme, suppose that 16 processes, each of which is us4-20

VAX-1 l VIRTUAL ADDRESS

9 8

-----VIRTUAL PAGE NUMBER
0 0
0 l
l 0
l l

BYTE WITHIN PAGE-

PROGRAM REGION
CONTROL REGION
SYSTEM REGION
RESERVED

u
VAX-11/780 PHYSICAL ADDRESS

4-----PAGE FRAME N U M B E R - - - - - - - - BYTE WITHIN PAGE -

+
0 MEMORY ADDRESS
l

I/O SPACE ADDRESS

VAX-11/750 PHYSICAL ADDRESS

~23 22 212019

~ 1111

----PAGE FRAME NUMBER

•l 'l +l +
l 1/0 SPACE ADDRESS

---111•-•- B'Y'TE WITHIN PAGE -

ALL OTHER BIT)
COMBINATIONS MEMORY ADDRESS SPACE
Figure 4-6
Virtual and Physical Address Formats

4-21

'-----...,--)t._
-

ACCESS MODE READ/WRITE PROTECTION CODE

'--~~~~~~~~ST~A~TU~S=-=B~IT_'~~~~~~~~~~~~~~~~~~~-,
PHYSICAL ADDRESS VALID
BIT 26 : MODIFIED PAGE STATUS
BITS 21-25: RESERVED FOR DIGITAL

SOFTWARE

BITS 0-20: PAGE FRAME NUMBER

ACTIVATE PAGING SOFTWARE ON REFERENCE

(PAGE FAULT)

BITS 0-26: MAY INDICATE :
DISK SECTOR IDENTIFICATION
TEMPORARY POINTER TO SHARED PAGE

Figure 4-7
Page Table Entry

virtual address of the routine to execute. Note that vector
14 (hex) can be used as a trap to writable control store to
execute user-defined instructions, and the vector contains
information passed to microcode.

ing 4 million bytes of virtual address space, are known to
the system at the same time (for a total of 64 Mb of virtual
address space). One system page table entry maps one
page of per-process page table entries, and one page of
per-process page table entries maps 65,536 (64K) bytes of
virtual address space (since it is possible to store 128 page
table entries in a single page of memory) . Therefore one
page of system page table maps 128 pages of per-process
page tables, which in turn maps 8 Mb of process virtual
address space. Thus the system region page table needed
to map these 16 processes requires approximately 8
physical pages (4K bytes) of memory.

Interrupt Priority Levels
Exceptions do not require arbitration since they occur synchronously with respect to instruction execution . lnterru pts, on the other hand, can occur at any time. To
arbitrate between interrupt requests that may occur simultaneously, the processor recognizes 31 interrupt priority
levels.

Exception and Interrupt Vectors
The processor can automatically initiate changes in the
normal flow of program execution. The processor recognizes two kinds of events that cause it to invoke conditional
software: exceptions and interrupts. Some exceptions affect an individual process only, such as arithmetic traps,
while others affect the system as a whole, for example, machine check . Interrupts include both device interrupts,
such as those signaling 1/0 completion, and softwarerequested interrupts, such as those signaling the need for
a context switch operation.

The highest 16 interrupt priority levels are reserved for interrupts generated by hardware, and the lowest 16 interrupt priority levels are reserved for interrupts requested by
software . Table 4-4 lists the assignment of each level , from
highest to lowest priority. Normal user software runs at
process level, which is interrupt priority level zero .
To handle interrupt requests, the processor enters a special system-wide context. In the system-wide context , the
processor executes in kernel mode using a special stack
called the interrupt stack. The interrupt stack cannot be
referenced by any user mode software because the processor only selects the interrupt stack after an interrupt,
and all interrupts are trapped through system vectors .

The processor knows which software to invoke when an
exception or interrupt occurs because it references specific locations, called vectors, to obtain the starting address
of the exception or interrupt dispatcher. The processor
has one internal register, the System Control Block Base
Register, which the operating system loads with the physical address of the base of the System Control Block, which
contains the exception and interrupt vectors. The processor locates each vector by using a specific offset into the
System Control Block. Figure 4-9 illustrates the vectors in
the System Control Block. Each vector tells the processor
how to service the event, and contains the system region

The interrupt service routine executes at the interrupt priority level of the interrupt request. When the processor receives an interrupt request at a level higher than that of the
currently executing software, the processor honors the request and services the new interrupt at its priority level.
When the interrupt service routine issues the REI (Return
from Exception or Interrupt) instruction, the processor returns control to the previous level.
4-22

SYSTEM REGION PAGE TABLE
Page Table Entry for Virtual Page O (first entry)
PTE for VPN 1
PTE for VPN 2

-------

System Base Register
(contains the physical address of the first entry of the page
table)
System Length Register
(contains the number of page table entries , N)

Page Table Entry for Virtual Page N - 1 (last entry)

PER-PROCESS PAGE TABLES

-------------

PROGRAM REGION PAGE TABLE
Page Table Entry for Virtual Page O (first entry)
PTE for VPN 1
PTE for VPN 2
PTE for VPN 3

Program Region Base Register
(contains the virtual address of the first entry in the page
table)

Program Region Length Register
(contains the number of page table entries , N)

_____....

PTE for Virtual Page N-1 (last entry)

CONTROL REGION PAGE TABLE
Page Table Entry for Virtual Page 2 22 -N
PTE for VPN 2·22-(N-1)
PTE for VPN 2·22-(N--'2)_ _ _ ___.
PTE for VPN 2•22-(N-3)

-----------t

Control Region Base Register
(contains the virtual address of base of the page table)

------4

Control Region Length Register
(contains the virtual address of the first entry in the page
table for virtual page number 2 22 -N, where N is the number of page table entries.)

______...

PTE for VPN 2·22-1 (last entry)

Figure 4-8
Page Tables

1/0 Space and 1/0 Processing
An 1/0 device controller has a set of control/status and data registers . The registers are assigned addresses in
physical address space, and their physical addresses are
mapped , and thus protected, by the operating system's
memory management software. That portion of physical
address space in which device controller registers are located is called 1/0 space .

disabling on the set of controllers for which it is responsible. When interrupts are enabled, an interrupt occurs
when the controller requests it. The processor accepts the
interrupt request and executes the driver's interrupt service routine if it is not currently executing on a higher-priority interrupt level.
Process Context
For each process eligible to execute, the operating system
creates a data structure called the software process control block. Within the software process control block is a
pointer to a data structure called the hardware process
control block. The hardware process control block is illustrated in Figure 4-10. It contains the hardware process
context, that is, all the data needed to load the processor's

No special processor instructions are needed to reference
1/0 space. The registers are simply treated as locations
containing integer data. An 1/0 device driver issues commands to the peripheral controller by writing to the controller's registers as if they were physical memory locations. The software reads the registers to obtain the controller status. The driver controls interrupt enabling and
4-23

4
8

c
10
14
18
1C

20
24

28
2C
30

40
44

48
4C

Machine Check
Kernel Stack Not Valid
Power Fail
Reserved or Privileged Instruction
Customer Reserved Instruction
Reserved or Illegal Operand
Reserved or Illegal Addressing Mode
Access Violation
Translation Not Valid (p~e fault )
Trace Fault
Breakpoint Fa ult
Compatibility Mode Exce pt ion
Arithmetic Exception

Chan_g_e Mode to Kernel
Change Mode to Executive
Change Mode to Supervisor
Change Mode to User

Software Level 1
Software Level 2
Software Level F
Interval Timer

100
101

Device Level 14, device 0
Device Level 14, device 1

13F

Device Leve l 14, device 15
Device Level 15, device 0

88
BF

140

EXCEPTION VECTORS

INTERRUPT VECTORS

17F
180

1BF

Device Level 15, device 15
Dev1celeveTio,aev1celY

1CO

Device Level 16, device 15
Device Level 17, device 0

1FF

Device Level 17, device 15

Offset from System Control Block Base Register (HEX)

Figure 4-9
System Control Block

4-24

Table 4-4
Interrupt Priority Levels
PRIORITY
HARDWARE EVENT
Hex Dec imal
1F
31
Machine Check , Kernel Stack Not Vali d
1E
30 ~owerTaTT
10
29
} Prnce8'oc,
1C
28
1B
27
Memory, or
1A
26
19
25
Bus Error
18
24
Clock
17
23
UNIBUS BR?
"'I
16
22
UNIBUS BR6
15
21
UNIBUS BR6
14
20
UNIBUS BR4
>Device Interrupt
13
19
12
18
11
17
.,;
10
16
PRIORITY
OF
15
OE
14
OD
13
OC
12
OB
11
OA
10
09
09
08
08
07
07
06
06

]_[

--00

04
03
02
01
00

SOFTWARE EVENT
}

Reserved for
DIGITAL

}Device
Drivers
Timer Process
Queue Asynchronous System Trap (AST)
[B:eservedTor DTGTTAl
1/ 0 Post
Process Scheduler
AST Delivery
Reserved for DIGITAL
User Process Level

Furthermore , the process control block provides the mechanism for triggering asynchronous system traps to user
processes . The Asynchronous System Trap field enables
the processor to schedule a software interrupt to initiate an
AST routine and ensure that they are delivered to the
proper access mode for the process.

process-specific reg isters when a context switch occurs.
To give control of the processor to a process , the operating system loads the processor's Process Control Block
Base register with the physical address of a hardware
process control block and issues the Load Process Context instruction . The processor loads the process context
in one operation and is ready to execute code within that
context.

CONSOLE
The console is the operator's interface to the central processor . Using the console terminal , the operator can examine and deposit data in memory locations or the processor
registers , halt the processor, step through instruction
streams , and boot the operating system.

As can be seen from the illustration , a process control
block not only contains the state of the programmable
registers , it also contains the definition of the process
virtual address space . Thus , the mapping of the process is
automatically context-switched .

4-25

HARDWARE PROCESS CONTROL BLOCK
Kernel mode stack pointer
Executive mode stack pointer
Supervisor mode stack pointer
User mode stack pointer
Register O
Register 1
Register 2
Register 3
Register 4
Register 5
Register 6
Register 7
Register 8
Register 9
Register 10
Register 11
Registe~ 12

Program Region Base Register
Program Region Length Register
Control Region Base Register

·~

Control Region Length Register

*Enabl e performance monitor
**Asy nchronous System Trap pending

Figure 4-10
Hardware Process Control Block

4-26

111111111111111111111111111111111111111111 III llll Ill 11111111111111111111111111111111
111111111111111111111111111111 lllll IIll III 11111111111111111111111111111111111111 IIII

I l II Ill II 11111111111111111111 111 11I l l l I IIII

lllllllllllllllllllllllll!!ll!llllll'''''
1
11111111111111111111111 '

11111111111111111111 111 111

THEVAX~1R80PROCESSOR

INTRODUCTION

crocode diagnostics, which can be loaded from the console's floppy disk.

A VAX-11 processor is a specific set of hardware logic that
performs the operations of the computer system according to the VAX-11 architecture.

The processor will also support 12K bytes of user writable
control store (WCS). WCS is optionally available to the
customer for augmenting the speed and power of the basic machine with customized functions.

This section describes the implementation-specific details
of the VAX-11 /780 processor. Its integrated components
are:

The console enables the computer system operator to
control the processor operation directly. The console actua11y consists of an LSl-11 microcomputer with 24K bytes of
memory, a floppy disk system, and a terminal. A serial line
interface is optionally available for remote diagnosis.

• The Central Processing Unit (CPU) itself, including its
cache, writable diagnostic control store, optional floating point accelerator, clocks and console.

Two memory controllers can be be connected to the memory interconnect. Each controller handles up to 4096K
bytes of semiconductor memory, for a system total of
8192K bytes of memory. The memory controllers employ
an error detecting and correcting technique that ensures
correction of all single-bit errors and detection of all double-bit errors.

• Main memory and main memory controllers .
• Input/output bus adaptors.
• Optional multiport memory.
• Optional high performance 32-bit interface.
These components communicate over a high-speed internal bus called the memory interconnect. Figure 4-11 illustrates the major processor components.

In addition, VAX-11 /780 supports the multiported memory
option. Multiported memory supports very high throughput interprocessor communications. Multiported memory
is discussed more fully in the Peripherals section .

The Central Processing Unit performs the logical and
arithmetic operations requested of the computer system.
Its user programmable registers include sixteen 32-bit
general purpose registers for data manipulation, and the
Processor Status Longword for controlling the execution
states of the CPU. The processor's instruction set is interpreted by the microcode contained in its control store.

Three 1/0 bus adaptors can be interfaced to the memory
interconnect: an adaptor for the MASSBUS, which connects high-speed disk and magnetic tape devices to the
processor; an adaptor for the UNIBUS, which connects
lower-speed devices to the processor, including disks,
communications lines, and 1/0 peripherals such as terminals, line printers , and card readers; and an optional adaptor for the high performance 32-bit interface. The high per-

The processor includes 12K bytes of writable diagnostic
control store for updating the instruction set microcode.
The writable diagnostic control is also used for storing mi4-27

FLOPPY
DISK

SYSTEMS
CONSOLE

LA120

VAX- 111780
CPU

MEMORY
CONTROLLER
512 KB

CACHE

4MB MAX

110 ADAPTERS
MAX I/0 RATE : 13.JMB/SEC (WITH 2 MEMORY CONTROLLERS)

INDICATES
OPTIONAL
EQUIPMENT

1 STANDARD
3 OPTIONAL

4 OPTIONAL

1 OPTIONAL

Figure 4-11
VAX-11 /780 Processor

each cycle. Data transfers use two consecutive cycles to
transfer 64 bits at a time. The maximum memory interconnect transfer rate is 13.3 million bytes per second. The
memory interconnect provides an unusual degree of
throughput and reliability because it uses:

formance interface enables the user to interface custom
devices directly to the memory interconnect, or to connect
two VAX-11/780 systems together. The high performance
interface will be discussed more thoroughly in the Peripherals section.

• time-division multiplexing

VAX-11 /780 PROCESSOR COMPONENTS

• distributed priority arbitration

Described below are the major hardware components of
the VAX-11 /780 processor.

• parity and protocol checking on every transfer
• transaction history recording

VAX-11 /780 Console
The VAX-11 /780's integrated console consists of an LSl11 microcomputer with 16K bytes of read/write memory
and BK bytes of ROM (used to store the LSI diagnostic, the
LSI bootstrap, and fundemental console routines) , a floppy disk system (for the storage of basic diagnostic programs and software updates), a hard-copy terminal, and
an optional remote diagnosis port.

The protocol, or sequence in which operations occur on
the memory interconnect, is time-division multiplexed to
increase the effective bus bandwidth. Time-division multiplexing means that the transactions which constitute one
transfer operation are interleaved with the transactions
which constitute another transfer operation. Thus, several
operations can be in progress over the same period of
time. For example, the CPU can ask a memory controller
to read some data; the same memory controller might then
transfer previously requested data to an 1/0 device before
it transfers the requested data to the CPU.

The console is further used for updating the software with
maintenance releases and for loading optional software
products distributed on floppy disk.
The operator communicates with the VAX-11 /780 console
via a set of user-oriented, English-like commands known
as the console command language (CCL).
An EIA serial line interface and modem can be added to
the console to provide remote diagnosis.
VAX-11 /780 MEMORY INTERCONNECT
The memory interconnect is the system 's internal bus,
conveying addresses, data, and control information
between the processor and memory, and between memory and the 1/0 controllers. The memory interconnect has a
cycle time of 200 nanoseconds and can transfer 32 bits
4-28

In some systems, the processor bus can be tied up for
each transfer because a requester acquires the bus to
send an address and then keeps the bus while it waits for
the requested data. In VAX-11 /780, the bus is not held inactive during the data access time because bus ownersh ip
is relinquished after every cycle. A requester acquires the
bus to specify an operation and send an address , and then
relinquishes the bus . At some time later the responder acquires the bus to send back the requested data. In the interim , any number of other transactions can be initiated or
completed. This and the fact that transactions are buffered

make it possible for the bus to operate at its full
bandwidth .
Arbitration on the memory interconnect is distributed,
which ensures that no unit is critical to bus operation. Every unit on the memory interconnect has its own arbitration line. Arbitration lines are ordered by priority and every
unit monitors all the arbitration lines each cycle to determine if it will get the next cycle . Unlike some bus systems,
any unit on the memory interconnect (except the CPU
clock) can fail without causing a failure of the entire bus.
To ensure the integrity of the signals transmitted, the
memory interconnect includes several error checking and
diagnostic mechanisms, such as:
• parity checking on data, addresses , and commands

Write-through provides reliability because the contents of
main memory are updated immediately after the processor performs a write . Most write-through cache systems tie
up the processor while main memory is updated . However,
the VAX-11 /780 processor buffers data to be written to
memory to avoid waiting while main memory is updated
from the cache. Therefore, while providing the reliability of
a write-through cache, this system also provides much the
same performance as a write-back cache.
Memory cache significantly reduces processor wait time
since practically all of the time, (greater than 95%), the data are in the cache. The cache memory system carries byte
parity for both data and addresses for increased integrity.
Address Translation Buffer
The address translation buffer is a cache of virtual to
physical address translations. It significantly reduces the
amount of time spent by the CPU on the repetitive task of
dynamic address translation . The cache contains 128
virtual-to-physical page address translations which are divided into equal sections: 64 system space page translations and 64 process space page translations. Each of
these sections is two-way associative. There is byte parity
on each entry for increased integrity.

• protocol checking in each interface
• a history silo of the last 16 memory interconnect cycles
VAX-11 /780 MAIN MEMORY AND CACHE SYSTEMS
The processor includes both main memory systems and
cache memory systems. Transactions between main
memory and the processor take place over the memory interconnect. The cache memory systems are internal to the
processor.

Instruction Buffer
The 8-byte instruction buffer improves CPU performance
by prefetching data in the instruction stream. The control
logic continuously fetches information from memory or
cache, where possible , to keep the 8-byte buffer full. It effectively eliminates the time spent by the CPU waiting for
two memory cycles where bytes of the instruction stream
cross 32-bit longword boundaries. In addition , the instruction buffer processes operand specifiers in advance of execution and subsequently routes them to the CPU.

Main Memory
Main memory consists of arrays of MOS RAM integrated
circuits with a cycle time of 600 nanoseconds. A memory
controller can access a maximum of 4,194,304 bytes (4M
bytes). Two memory controllers can be connected to the
memory interconnect, yielding a maximum of BM bytes of
physical memory that can be available on the system . The
maximum total physical address space is 2 29 or approximately 512 million bytes. However, the minimum required
memory is 256K bytes, which is then expandable in increments of 256K bytes.

1/0 CONTROLLER INTERFACES

A memory controller will buffer one command while it
processes another to increase system throughput. Main
memory can also be interleaved (where two memory controllers are each addressing the same amount of memory)
to increase the available memory bandwidth . The memory
system employs error checking and correction (ECG) that
corrects all single bit errors and detects all double bit errors .

Peripherals can be connected to the processor's memory
interconnect bus in either of two ways: through the MASSBUS, for high-speed disk and/or magnetic tape devices,
or through the UNIBUS , for a variety of 1/0 devices,
including line printers , disks, card readers, terminals , and
interprocessor communication links.
VAX-11 MASSBUS Interface
The processor interface for a MASSBUS peripheral is the
MASSBUS adaptor. The MASSBUS adaptor performs
control, arbitration, and buffering functions. Up to four
MASSBUS adaptors can be connected to the memory interconnect. The MASS BUS is typically used to attach highspeed disk or magnetic tape devices.

When the system is powered down, an ac standby current
is normally used to retain the contents of memory. In case
of temporary AC power interruption , an optional backup
battery is also available to provide 1O minutes of power for
up to 4M bytes of memory so that the contents of main
memory are not destroyed. Two backup batteries provide
power for up to BM bytes of memory.

Each MASSBUS adaptor includes its own address translation map that permits scatter/gather disk transfers. In
scatter/gather transfers, physically contiguous disk blocks
can be read into or written from discontiguous blocks of
memory. The translation map contains the addresses of
the pages, which may be scattered throughout memory,
from or to which the contiguous disk transfer takes place.

Data are fetched from main memory 64 bits at a time (two
memory interconnect cycles) and cached in the processor's internal memory systems. The internal memory systems include a main memory cache, an address translation buffer. and an instruction lookahead buffer.
Memory Cache
The memory cache is the primary cache system for all data
coming from memory, including addresses, address
translations, and instructions. The memory cache is an 8K
byte, two-way set associative, write-through cache.

Each MASS BUS adaptor includes a 32-byte s·ilo data buffer. Data are assembled in 64-bit quadwords (plus parity) to
make efficient use of the memory interconnect bandwidth.
On transfers from memory to a MASSBUS peripheral, the·
4-29

MASSBUS adaptor anticipates upcoming MASSBUS data
transfers by fetching the next 64 bits from memory before
all of the previous data have been transferred to the peripheral.

The UNIBUS adaptor permits concurrent program interrupt, unbuffered , and buffered data transfers . The aggregate throughput rate of the Direct Data Path, plus the 15
buffered data paths , is 1.35 million bytes per second .

On-line diagnostics and built-in loop-back testing enable
fault isolation of the MASSBUS adaptor for any of its function circuits without a drive on the MASSBUS.

Data Throughput
VAX-11 /780 includes many features that support high data
throughput, including silo data buffers for MASS BUS peri pheral controllers, buffered direct memory access for the
UNIBUS peripherals, and 64-bit data transfers and prefetching .

VAX-11 /780 UNIBUS Interface
All devices other than the high-speed disk drives and
magnetic tape transports are connected to the UNIBUS, an
asynchronous bidirectional bus. These include all
DIGITAL- and user-developed real-time peripherals. The
UNIBUS is connected to the memory interconnect through
the UNIBUS adaptor. The UNIBUS adaptor does priority
arbitration among devices on the UNIBUS. Up to four UNIBUS adapters can be placed on the memory interconnect.

Memory bandwidth matches that of the processor's internal bus - 13.33 million bytes per second , including time
for refresh cycles . This is primarily because of the memory
controller request buffers, which substantially increase
memory throughput and overall system throughput, and
decrease the need for interleaving for most configurations.
Memory interleaving , which is enabled and disabled under
program control , can be used effectively when more than
two MASSBUS peripheral controllers are connected and
the MASSBUS and UNIBUS devices are transferring at
very high rates - greater than one million bytes per second .

The UNIBUS adaptor provides access from the VAX11 /780 processor to the UNIBUS peripheral device registers by translating UNIBUS addresses, data transfer requests, and interrupt requests to their memory interconnect equivalents, and vice versa. The UNIBUS adaptor address translation map translates an 18-bit UNIBUS
address to a 30-bit memory interconnect address. The
map provides direct access to system memory for nonprocessor request UNIBUS peripheral devices and permits scatter/gather disk transfers .

The operating system supports the hardware throughput
in its 1/0 request processing software. The software uses
the processor's multiple hardware priority levels to increase 1/0 response time, and keeps each disk controller
as busy as possible by overlapping seek requests with 1/0
transfers.

The UNIBUS adaptor enables the processor to read
and/or write the peripheral controller registers. In some
cases this constitutes the transfer .

VAX-11 /780 FLOATING POINT ACCELERATOR
The floating point accelerator (FPA) is an optional high speed processor enhancement. When included in the
processor, the floating point accelerator accelerates the
execution of the addition, subtraction , multiplication , and
division instructions that operate on single- and doubleprecision floating point operands. This includes the special EMOD and POLY instructions in both single- and double-precision formats. Additionally, the floating point accelerator enhances the performance of the 32-bit integer
multiply instruction _MUL.

To make the most efficient use of the memory interconnect
bandwidth , the UNIBUS adaptor provides buffered direct
memory access data paths for up to 15 nonprocessor request (NPR) devices. Each of these channels has a 64-bit
buffer (plus byte parity) for holding four 16-bit transfers to
and from UNIBUS devices. The result is that only one
memory interconnect transfer (64 bits) is required for every four UNIBUS transfers. The maximum aggregate
transfer rate through the buffered data paths is 1.35 million bytes per second. On memory interconnect-to-UNIBUS transfers, the UNIBUS adaptor anticipates upcoming
UNIBUS requests by pre-fetching the next 64-bit quadword from memory as the last 16-bit word is transferred
from the buffer to the UNIBUS. By the time the UNIBUS device requests the next word, the UNIBUS adaptor has it
ready to transfer.

The processor does not have to include the floating point
accelerator to execute floating point operand instructions;
the FPA increases the execution speed of floating point instructions. The floating point accelerator can be added or
removed without changing any existing software.
When the floating point accelerator is included in the
processor, a floating point register-to-register add instruction takes as little as 800 nanoseconds to execute. A register-to-register multiply instruction takes as little as one m icrosecond. The inner loop of the POLY instruction takes
approximately one microsecond per degree of polynom ial.

Any number of unbuffered direct memory access transfers
are handled by one Direct Data Path. Every 8- or 16-bit
transfer requires one 32-bit transfer on the memory interconnect. The maximum transfer rate through the Direct
Data Path is 500,000 bytes per second.

4-30

\\\l\\\\\\\\\\\\\UU\I•
7

THE VAX-11 /750 PROCESSOR

The console enables the computer system operator to
control the processor operation directly. The console subsystem consists of the console terminal (LA38 DECwriter),
the front panel , the user oriented console command language, and a TU58 Tape Cartridge Drive. Also optionally
available for the console is the remote diagnosis interface.

INTRODUCTION
A VAX-11 processor is a specific set of hardware logic that
performs the operations requested of the computer system according to the VAX-11 architecture.
This section describes the implementation specific details
of the VAX-111750 processor. Its integrated components
are:

The main memory subsystem consists of ECC MOS memory, which is interfaced to the system via the memory controller . MOS memory may be added to the system in
increments of 256K bytes to a maximum of 2M bytes.

• the Central Processing Unit (CPU) itself, including its
cache , optional user control store, clocks and console

The 1/0 subsystem consists of the UNIBUS and MASS BUS
devices connected to the system via special buffered interfaces called adaptors. Each VAX-11 /750 system contains
one UNIBUS adapter for standard peripherals and up to a
maximum of three MASSBUS adapters for high speed
peripherals .

• main memory and main memory controllers
• peripheral bus adaptors
Figure 4-12 illustrates the major VAX-11 /750 processor
components.
The central processing unit performs the logical and
arithmetic operations requested of the computer system .
Its user programmable registers include sixteen 32-bit
general purpose registers for data manipulation, and the
Processor Status Word for controlling the execution states
of the CPU. The processors instruction set is defined by
the microcode contained in its control store.

VAX-11 /750 PROCESSOR COMPONENTS
Described below are the major hardware components of
the VAX-111750 processor.
VAX-11 /750 Console
The console enables the computer system operator to
control the processor operation directly. The console subsystem consists of the console terminal (LA38), the front
panel, the user oriented console command language, and
a TU58 Tape Cartridge Drive. Simple console commands,
entered through the console terminal, replace the traditional toggle switches and provide operational control (i.e.,
bootstrapping, initialization, self testing, examining and
depositing data in memory.etc.). When not performing op-

The optional User Control Store includes 1 OK bytes
( 1Kbytes of 80 bit microwords) of writeable storage. This
allows customers to augment the speed and power of the
basic machine with customized microcode functions.
Digital offers a loadable microcode package for extended
precision floating point arithmetic operations (G- and
H-floating point data types) on the 11 /750 .
4-31

TU58
SYSTEMS
CONSOLE

VAX-111750
CPU

LA38

MEMORY
CONTROLLER
512KB

CACHE

I/0 ADAPTORS
MAXI/ORATE: 5MB/SEC

INDICATES
OPTIONAL
EQUIPMENT

1 STANDARD

3 OPTIONAL

Figure 4-12
VAX-11 /750 Processor
The VAX-11/750 memory cache is a 4K byte, direct
mapped, write-through cache. It is used for all data com ing from memory, including addresses and instructions.
The write-through feature protects the integrity of memory
because memory contents are updated immediately after
the processor performs a write. For increased integrity, the
cache memory system carries byte parity for both data
and addresses. Cache locations are allocated when data is
read from memory or when a longword is written to memory . Memory cache also watches 110 transfers and updates
itself appropriately. Therefore, no operating system overhead is needed to synchronize the cache with 1/0 operations , i.e., memory cache is transparent to all software.

erator functions or error logging, the same terminal can be
available to authorized users for normal system operations.
The VAX-11 /750 console subsystem and the console command language also facilitate the loading of diagnostics
and software updates from the TU58 Tape Cartridge. For
those customers subscribing to a DIGIT AL maintenance
contract, the console subsystem may also be equipped
with a remote diagnosis module (ADM) allowing the VAX11 /750 to interface to a host computer at a DIGITAL Diagnostic Center for remote fault detection or preventive
maintenance procedures.
VAX-11 /750 Main Memory
The VAX-11 /750 main memory is built using 16K MOS
RAM (random access memory) LSI chips. Physical memory is organized into an array of 32-bit longwords plus an
additional 7 bits per longword dedicated to ECG (error
correcting code). ECG allows the correction of all single-bit
errors and the detection of all double bit errors to insure
data integrity. Main memory is interfaced to the VAX system via the memory controller. The VAX-11 /750 can be
easily field upgraded to 2M bytes of main memory by simply adding 256K byte expansion modules.

• Instruction Buffer
The instruction buffer is an 8 byte buffer that enables the
CPU to fetch and decode the next instruction while the current instruction completes execution . The instruction
buffer in combination with the parallel data paths (which
can perform integer arithmetic and shifting operations simultaneously) significantly enhances the VAX-11 /750's
performance because the CPU is not held in a wait state.
• Address Translation Buffer
The address translation buffer is a cache of the most frequently used 512 physical address translations. It significantly reduces the amount of time the CPU spends on the
repetitive task of dynamic address translation . The cache
contains 512 virtual-to-physical page address translations
which are divided into equal sections: 256 system space
page translations and 256 process space page translations. Each of these sections is two-way associative and
has parity on each entry for increased integrity.

VAX-11 /750 Cache Systems
The VAX-11 /750 CPU provides three cache systems: the
main memory cache, the address translation buffer, and
the instruction buffer.
• Main Memory Cache
Memory cache (typically 90% hit rate) provides the central
processor with high-speed data access by storing frequently referenced addresses, data and instruction items.
The memory cache typically reduces memory access time
in half.

Peripheral Controller Interfaces
Peripherals can be connected to the processor in either of
two ways: through the MASSBUS, which conveys signals
4-32

to and from high-speed disks or magnetic tape devices, or
through the UNIBUS, which conveys signals to and from a
variety of 1/0 devices, including line printers, disks, card
readers , tapes , terminals , and interprocessor communication links.

addresses (16M byte physical address space) , 18-bit UNIBUS addresses must be translated to 24 bit memory addresses. This mapping function is performed by the UNIBUS adapter (a special hardware interface between memory and the UNIBUS) which translates UNIBUS addresses
to their memory equivalents, and vice versa.

VAX-11 MASSBUS Interface
The processor interface for a MASSBUS peripheral is the
MASSBUS adaptor. The MASSBUS adaptor performs
control , arbitration, and buffering functions. There may be
a total of three MASS BUS adapters on each VAX-11 /750
system .

The UNIBUS adapter performs priority arbitration among
devices on the UNIBUS, a function handled by the central
processor in PDP-11 systems. The address translation
map permits contiguous disk transfers to and from noncontiguous pages of memory (these are called scatter /gather operations). Interrupts on the VAX-11 /750 UNIBUS are directly vectored into the appropriate process
handler.

Each 11 /750 MASS BUS adaptor includes its own address
translation map that permits scatter/gather disk transfers.
In scatter/gather transfers, physically contiguous disk
blocks can be read into or written from discontiguous
blocks of memory. The translation map contains the addresses of the pages, which may be scattered throughout
memory, from or to which the contiguous disk transfer
takes place.

The UNIBUS adapter allows two kinds of data transfer's;
program interrupt and direct memory access (OMA). To
make the most efficient use of the memory bandwidth, the
UNIBUS adapter facilitates high-speed OMA transfers by
providing buffered OMA data paths for up to 3 high-speed
devices at one time. Each of these channels has a 32-bit
buffer (plus byte parity) for holding two 16-bit transfers to
or from UNIBUS devices. The result is that only one memory transfer (32 bits) is required for every two UNIBUS
transfers. The maximum aggregate transfer rate through
the buffered data paths is 1.5M bytes per second .

Each 11 /750 MASS BUS adaptor includes a 32-byte silo
data buffer. Data are assembled in 32-bit longwords (plus
parity) to make efficient use of the system bus. On transfers from memory to a MASSBUS peripheral, the MASSBUS adaptor anticipates upcoming MASS BUS data transfers by fetching the next 32 bits from memory before all of
the previous data are transferred to the peripheral.

Any number of unbuffered OMA transfers are handled by
one direct OMA data path . Every 8- or 16-bit transfer on
the UNIBUS requires a 32-bit memory transfer (although
only 16 bits are used). The maximum transfer rate through
the direct data path is 1M bytes per second .

On-line diagnostics and loop-back enable adaptor fault
isolation without requiring the use of a drive on the MASSBUS.
VAX-11 /750 UNIBUS Interface
General purpose peripherals and customer developed devices are connected to the VAX-111750 system via the UNIBUS . Since the 11 /750 memory deals in 24-bit physical

It should be noted that the UNIBUS adapter permits program interrupts, unbuffered and buffered data transfers to
occur concurrently.

4-33

5
The
Peripherals
t~
'

--,,

!
{

The VAX system supports high-performance mass storage devices for
on-line data retrieval , unit record equipment for data processing, terminals and line interfaces for the interactive user, direct memory access interfaces for real-time users and a line interface for interprocessor communications.
The mass storage systems provide large capacity and high throughput.
Each MASS BUS adapter can support up to eight disk drives or seven
disk drives and one magnetic tape controller. In addition , up to eight
medium-capacity disk drives can be connected to the system 's UNIBUS . VAX/VMS overlaps seeks on all multiple-drive disk
configurations, performs multiple-block 1/0 transfers, and allows the
user to control buffering , positioning, and blocking .
Card readers and line printers can be spooled input and output devices
managed by operator-controlled queues. The LP11 and LA 11 series
line printers provide a range of high-speed and low-cost printer models . Up to four LP11 printers and up to 16LA11 printers can be used
on the system .
The system supports full-duplex handling for both hard copy and video
terminals. The LA 120 is a hard-copy terminal which offers moderate
throughput and advanced print features ; the VT100 video terminal offers a variety of controllable character and screen attributes including
24 lines by 80 columns or 14 lines by 132 columns screen sizes, smooth
scrolling , and split screen . The system can support up to 96 terminals.
The DMC11 serial synchronous communications line provides highperformance point-to-point interprocessor connection using the
DIGITAL Data Communications Message Protocol (DDCMP) . The
DMC11 ensures reliable data transmission and relieves the host processor of the details of protocol operation. For very high-performance
interprocessor communications, the VAX-11/780 offers both multiport
memory (MA780) and a high-speed channel interface (DR780) . The
DR780 can also be used for interfacing customer devices which require
transfer rates of up to 6.67M bytes/second .
For real-time applications , VAX supports the LPA 11-K and DR11-B
direct memory access (OMA) interfaces. These devices reduce CPU involvement in 1/0 operations and speed the transfer of data between external devices and computer memory. The LPA 11-K is an intelligent
(dual-microprocessor) controller which provides high speed data sampling, operates in both dedicated and multirequest mode, and supports
a number of peripheral devices. The DR11-B is a general purpose interface which performs high speed block data transfers between the
VAX memory and user peripheral devices.
All equipment is integrated with the software system, and is supported
by both on-line error logging and diagnostics. Each component includes extensive error checking and correction features. The software
provides power failure and error recovery algorithms.

pheral devices. The AX02 floppy disk, the AL02, and AKO?
disk drives, and the TS 11 tape subsystem are UNIBUS
peripheral devices. Each MASSBUS can support up to
eight device controllers ; eight disk controllers with one
drive each or seven disk drives and one magnetic tape formatter with up to eight tape transports.

COMPONENTS
VAX supports four types of peripheral subsystems:

• Mass storage peripherals such as disk and magnetic
tape
• Unit record peripherals such as line printers and card
readers

To support the performance and reliability features of the
system 's disk and magnetic tape devices, the operating
system 's disk and magnetic tape device drivers provide:

• Terminals and terminal line interfaces
• Interprocessor communications links
All peripheral device control/status registers (CSAs) are
assigned addresses in physical 1/0 space. No special
processor instructions are needed for 1/0 control. In addition , all device interrupt lines are associated with locations
that identify each device's interrupt service routine. When
the processor is interrupted on function request completion , it immediately starts executing the appropriate interrupt service routine. There is no need to poll devices to determ ine which device needs service.

• overlapped seeks for increased throughput on controllers with multiple disk drives
• overlapped magtape operations (write on one transport
while another rewinds, for example)
• multiple block non-contiguous 1/0 transfers for filestructured devices
• read and write checks on a per-request, per-file , and/or
volume basis
• extensive error recovery algorithms (e.g., ECG and offset recovery for disk, NAZI error correction for magnetic
tape)

Devices use either one of two types of data transfer
techniques: direct memory access or programmed interrupt request. The mass storage disk and magnetic tape
devices and the interprocessor commun ications link are
capable of direct memory access (DMA) data transfers.
The DMA devices are also called non-processor request
(NPA) devices because they can transfer large blocks of
data to or from memory without processor intervention until the entire block is transferred .

• logging of all device errors
• dynamic bad block support for file-structured disk devices
• volume mount verification after a change in drive status
(off/on-line, powerfail)
• powerfail recovery for on-line drives, including repositioning of magnetic tape transports

The unit record peripherals and terminal interfaces are
called programmed interrupt request devices. These devices transfer one or two bytes at a time to or from assigned locations in physical address space . Software then
transfers the data to or from a buffer in physical memory.

----

MASS STORAGE PERIPHERALS
The mass storage peripherals include various capacity
moving head disk drives and various speed magnetic tape
transports :
• the high speed , large capacity AP06 and AMOS disk drives
• the high speed, medium capacity AM03 disk drive
• the medium speed, smaller capacity AL02, and AKO?
disk drives
• the AX02 floppy disk
• the TE16, TU4S, and TU77 magnetic tape transports
• the TS 11 magnetic tape transport
The AM03, AP06, and AMOS disks and the TE16, TU4S,
and TU77 magnetic tape controllers are MASSBUS peri-

Table 5-1
Disk Devices
DISK

RX02

RL02

RK07

RM03

RP06

RMOS

Pack capacity:

512 Kbytes

10.4 Mbytes

28 Mbytes

67 Mbytes

176 Mbytes

300 Mbytes

Peak transfer
rate (/sec):

55 Kbytes

512 Kbytes

538 Kbytes

1200 Kbytes

806 Kbytes

1200 Kbytes

Ave . seek time:

263ms

55ms

36.5ms

30ms

Ave . rotational
latency:

83ms

12.5ms

8.3ms

6ms

8.3ms

S-1

For applications requiring special data reliability checks,
the programmer can implement user written error recovery procedures without having to write unique device driver routines. The operating system driver's normal error
recovery retry and error logging operations can be inhibited. If any error occurs when the recovery functions are inhibited, the driver immediately terminates the 1/0 operation and returns a failure status. User software can then
perform its own recovery or logging procedures, since all
the hardware diagnostic operations are available to jobs
granted the diagnostic privilege by the system manager.

In addition to the driver's dynamic bad block handling , the
system includes an on-line static bad block utility and online diagnostics for verifying drive level functions .

Magnetic Tape
The TE16, TU45 , and TU77 are high performance MASSBUS tape storage subsystems, which share the following
characteristics:
• program-selectable 1600 or 800 bpi, 9-track data storage
• industry compatible data formats
• reading in reverse (as well as forward)
• parity, longitudinal, and cyclic redundancy checking

Disks
The disk subsystems provide high performance and reliability. They feature accurate servo positioning, error correction, and offset positioning recovery. Table 5-1
summarizes the capacities and speeds of the disk devices.

• NAZI error correction
The TU45 and TE16 are identical in capacity: both allow up
to 40 million bytes per tape reel and both allow up to 8 tape
drives per formatter. However, the TU45 offers substantially higher speed and throughput. Read/write speeds for
the TU45 and TE16 are 75 and 45 inches/second respectively; their data transfer speeds are 120K and 72K bytes
per second.

All disk drives use top-loading removable media. The
AM03, AP06, and AM05 disk drives can be mixed on the
same MASSBUS.
The UNIBUS accepts AK06, AKO?, and AL02 disk drives
and the AX02 floppy disk. The AX02 is the smallest capacity disk available, while the AKO? is the largest capacity
disk. Up to eight AX02, AL02, AK06, and AKO? disk drives
can be mixed in any combination on the same controller.
In small system configurations where the AK06 or RK07 is
used as the systems device, two drives are required in the
configuration.
To decrease the effective access time and increase
throughput, the operating system's disk device drivers
provide overlapped seeks for all disk units on a controller.
All 1/0 transfers, including write checks, are preceded by a
seek, except when the seek is explicitly inhibited by diagnostic software. On MASSBUS devices, seeks to any unit
can be initiated at any time and do not require controller
intervention. During seeks, the controller is free to perform
a transfer on any unit other than the one on which the seek
is active. If a data transfer was in progress at the time of
completion, the driver processes the attention interrupts
caused by seek completion when the controller is free.
The device unit notifies the driver when it detects a read
error that can be recovered using its error correction code
(ECG). It provides the position and pattern of any error
burst of up to 11 bits within the data field. The driver applies the error correction to the data in memory. The transfer continues as if the error had not occurred.
In addition to overlapping seeks with data transfers, the
driver also overlaps offset error recovery with normal controller operation. Offset recovery enables the driver to reposition the head on the track to pick up a stronger signal
on a sector during a read operation. Provided that retry is
not inhibited, the driver performs offset recovery automatically when a read error occurs that can not be corrected
using the hardware ECG.

The TU77 tape storage system can perform read/writes of
data at the rate of 125 inches/second, while allowing a
peak transfer rate of 200K bytes/second. These features
make the TU77 ideally suited for heavy duty cycle applications such as disk to tape backup and transaction processing.

The driver logs all errors, including those from which it
successfully recovers. The driver also supplies dynamic
bad block handling for virtual 1/0 (Files-11 file-structured)
operations. Wheo a bad block is detected, the information
is stored in the file header. The bad block is recorded in
the bad block file when the file is deleted.

The TS11 is a med ium performance UNIBUS magnetic
tape subsystem containing 9 tracks with a density of 1600
bpi. It performs read/writes at a speed of 45 inches/second . The TS11 subsystem can handle a maximum
data transfer rate of up to 72K bytes/second.
5-2

The operating system 's magnetic tape device driver supports the read reverse operation, which enables a program to request a sequential read of the block preceding
the block at which the tape is positioned . Writing occurs
only while the tape is moving forward .

7 matrix with horizontal spacing of 10 characters per inch
and vertical spacing of six lines per inch.
Adjustable pin-feed tractors allow for a variable-form
width of 3 to 14-7/8 inches (up to 132 columns) . A forms
length switch sets the top-of-form to any of 11 common
lengths, with fine adjustment for accurate forms placement. The printer can accommodate multipart forms (with
or without carbons) of up to six parts.

The operating system's file system can read and write filestructured magnetic tape volumes using the current ANS
magnetic tape standard . The system also supports multivolume files , program-controlled blocking factors, and
unlabeled magnetic tapes.

CR11 Card Reader
CR11 card readers can be connected to the system as programmed interrupt request devices. The CR11 reads up to
285 80-column punched cards per minute. The card reader has a high tolerance for cards that have been nicked ,
warped , bent, or subjected to high humidity. The card
reader uses a short card path, with only one card in the
track at a time . It uses a vacuum pick mechanism and
keeps cards from sticking together by blowing a stream of
air through the bottom half-inch of cards in the input hopper. The input hopper holds up to 400 cards, and cards
can be loaded and unloaded while the reader is operating .

UNIT RECORD PERIPHERALS
The operating system normally treats line printers and
card readers as spooled shareable devices managed by
multiple operator-controlled queues. The devices can also
be allocated to individual programs.
The operating system's line printer handling includes line
and page counting for job accounting . The user can specify carriage control as: one lin~ per record, FORTRAN conventions, contained within the record itself, or general preand post-spacing (within the limits of the hardware capabilities).

TERMINALS AND INTERFACES

The operating system 's card reader driver interprets the
encoded punched information using the American National Standard 8-bit card code. The driver uses a special
punch outside the data representation to indicate end-offile .

Interactive terminals can be connected to the VAX system.
The operating system's terminal driver provides full duplex handling for both hard copy and video terminals.
Programs can control terminal operations through the terminal driver. The terminal driver supports many special
operating modes for terminal lines. A program can enable
or disable the following modes by calling a system service:

LP11 Line Printers
LP11 series line printers can be connected to the VAX system. The LP11 series printers are impact-type, rotating
drum, serial interface line printers. They feature full line
buffering , a static eliminator, and a self-test capability .

• SLAVE All unsolicited data are discarded . This mode
is used to establish application-controlled terminals .
• NO ECHO Data entered on the terminal keyboard are
not printed or displayed on the terminal. This mode is
used, for example , to read passwords typed on the terminal.

All models are 132-column printers that can accept paper
4 to 16-3/4 inches wide with up to 6-part forms. They print
10 characters per inch horizontally, and 6 or 8 lines per
inch vertically (switch selectable). They include a vernier
adjustment for horizontal and vertical paper position . All
models are available with either upper (64) or upper/lower
(95) character sets (including numbers and symbols) .
Most models have optional scientific or EDP character
sets.

• PASS ALL All data entered on the terminal are transmitted to the program as 8-bit binary information without any interpretation, except where a line terminator or
terminators are specified. This mode enables programs
to perform their own interpretation of control characters
instead of using the VAX/VMS interpretation.

The low-cost models print one line every two revolutions
(300 lines per minute with the 64-character set, 230 lines
per minute with the 95-character set), or one line every
revolution (600 lines per minute with the 64-character set,
460 lines per minute with the 95-character set). A higherspeed version that includes a noise-reduction cabinet, ribbon guide, and a high-speed paper puller offers 900 lineper-minute printing with the 64-character set, or 660 lines
per minute with the 95-character set.

• ESCAPE Escape sequences entered on the terminal
are recognized as read terminators, validated, and
passed to a program for interpretation .
• TERMINAL/HOST SYNCHRONIZATION Data sent to
the terminal are controlled by terminal-generated XOFF
and XON. These functions are generated by typing
CTRL/S and CTRL/Q on command terminals and are
interpreted as requests to stop and resume output to
the terminal.

For systems requiring even greater printer throughput,
LP11 models are available that print up to 1200 lines per
minute with the 64-character set or 800 lines per minute
with the 95-character set.

• HOST/TERMINAL SYNCHRONIZATION All read operations are explicitly solicited with XON and terminated
with XOFF. XON and XOFF are also used to keep the
type-ahead buffer from filling .

LA 11 Line Printer
The LA 11 is an extremely low-cost, highly-reliable parallel
interface printer. The LA 11 prints at speeds up to 180
characters per second . The print set consists of the ASCII
characters, including 95 upper and lower case letters,
numbers, and symbols. Characters are printed using a 7 x

Input from a command terminal is always independent of
concurrent output. This capability is called type-ahead.
Data typed at the terminal are retained in a type-ahead
buffer until a program issues a read request. At that time
the data are transferred to the program buffer and echoed
5-3

Characters are printed using a 7 x 7 matrix with horizontal
spacing of ten characters per inch and vertical spacing of
six lines per inch. To ensure clear visibility of the printed
line, the print head automatically retracts out of the way
when not in operation . Adjustable pin-feed tractors allow
for a variable-form width from 3 to 14-7 /8 inches (up to
132 columns). The print mechanism will accommodate
multi part forms (with or without carbons) of up to six parts .

on the terminal (provided that echoing is not disabled). If a
read is already in progress, the echo and data transfer are
immediate. Deferring the echo until a read operation is active allows the program to specify the mode of the terminal, such as No Echo or Convert Lower Case to Upper, to
modify the read operation.
A line entered on a command terminal is terminated by
any of several special characters, for example, the RETURN key . A program reading from a terminal can optionally specify a particular line terminator or class of line terminators for read requests (including read PASS ALL requests).

The LA36 operates at speeds of 110, 150, or 300 baud (10,
15, or 30 characters per second). Printable characters are
stored in a buffer during the carriage return operation .
While more than one character is in the buffer, the printer
mechanism operates at an effective speed of 60 characters per second.

Terminal characteristics are initially established during
system generation. Users operating command terminals
can modify the characteristics of the particular terminal
being used. For example, the user can set the baud rate
(transmission speed) or change the terminal line width .

VT100 Video Terminal
The VT100 Video Terminal is an upper- and lower-case
ASCII terminal which offers a variety of controllable character and screen attributes. The VT100 features a typewriter-like detachable keyboard which includes a standard
numeric/function keypad for data entry applications. Also
featured are seven LEDs , four of which are program-controlled , used as operator information and diagnostic aids .

LA120 Hard Copy Terminal
The LA 120 is a hard copy terminal which offers exceptional throughput and a number of advanced keyboard-selectable formatting and communication features . It uses a
contoured typewriter-styled keyboard and includes an additional numeric keypad and a prompting LED display for
infrequently used features.

The VT100 offers a number of advanced features . The
most important of these are:

The LA 120 achieves high throughput owing to several features:

• ability to select either of two screen sizes: 24 lines by 80
columns or 14 lines by 132 columns

• 180 character per second print speed

• ability to select either double-width single-height characters or double-width double-height characters on a
line by line basis

• 14 data transmission speeds ranging up to 9600 baud
• 1K character buffer to equalize differences between
transmission speeds and print speeds

• smooth scrolling and split screen capability
• ability to set baud rates, tabs, and Answer Back messages from the keyboard and to store these in RAM
(Random Access Memory)

• smart and bidirectional printing so that printhead always takes shortest path to next print position
• high speed horizontal and vertical skipping over white
space

• special line drawing graphic characters providing the
ability to display simple graphics for business or laboratory applications

In addition to its throughput, the LA 120 is distinguished by
its printing features. The terminal offers eight font sizes,
ranging from expanded (5 characters per inch) to compressed font (16 .5 characters per inch). Hence a user
could, for example, select a font size of 16.5 cpi and print
132 columns onto an 8 1/2-inch-wide sheet. Other print features include six line spacings ranging from 2 to 12 lines
per inch, user-selectable form lengths up to 14 inches,
left/right and top/bottom margins and horizontal and
vertical tabs.

• ability to select black-on-white characters or white-onblack characters on a full screen basis

I""°"

The LA120 is designed for easy use. Terminal characteristics are selected via clearly labeled keys and simple mnemonic commands. Once the selections have been made,
the operator can check his settings by depressing the
STATUS key. The terminal will then print a listing of theselected settings.
LA36 Hard Copy Terminal
The LA36 is an exceptionally reliable hard copy terminal. It
is a lower-priced device than the LA 120 with lower
throughput (30 cps vs. 165 cps) and fewer print features.
The LA36 uses a typewriter-like keyboard which produces
128 ASCII characters, consisting of 95 upper- and lowercase printing characters and 33 control characters, and is
available with optional special character sets, including
various foreign language character sets.

5-4

In addition , several options further extend the capabilities
of the VT100. These include the advanced video option ,
which adds selectable blinking, underline, and dual intensity characters to the existing reverse video attribute;
the provision of space, power, and interconnects for the
future addition of a terminal processor ; and additional
RAM allowing 24 lines of 132 characters.

word can synchronize the sampling with a user event. In
multirequest mode, throughput is determined by the number and types of requests. The aggregate throughput rate
for all users is typically 15,000 samples per second .
In dedicated mode, one user can sample from analog-todigital converters , or output to a digital-to-analog converter. Two analog-to-digital converters can be controlled simultaneously. Sampl ing is initiated by an overflow of the
real-time clock , or by an external signal. Two sampling
algorithms are implemented . One, at each overflow, samples both analog-to-digital converters in parallel, allowing
two channels to be sampled simultaneously. The other algorithm samples the two converters on an interleaved basis, beginning with the first whose sampling begins on alternate clock overflows.

DZ11 Terminal Line Interface
The DZ11 is a serial line multiplexer whose character formats and operating speeds are programmable on a perline basis. A DZ11 connects the UNIBUS with up to a maximum of 8 or 16 asynchronous serial lines, depending on
the configuration. Each line can run at any one of 15
speeds .
Local operation with EIA terminals is possible at speeds up
to 9600 baud . Remote dial-up terminals can operate fullduplex at speeds up to 300 baud using Bell 103 or 113 modems, or up to 1200 baud using the Bell 212 modem.

The LPA 11-K supports the following 1/0 devices on VAX:
• AA 11 K (four-channel 12-bit D/ A converter)
• AD11K(eight-channel12-bit A/D converter)

The DZ11 optionally generates parity on output and
checks parity on input. Incoming characters are buffered
using a 64-character silo buffer. Outgoing characters are
processed on a programmed interrupt request basis.

• AM11 K (multiplexer board)
• DR11 K ( 16-bit parallel , general device interface)
• KW11 K (real-time clock)
DR11-B
The DR11-B is a general purpose, direct memory access
(OMA) digital interface to the UNIBUS . The DR11-B, rather
than using programmed controlled data transfers ,
operates directly to or from memory, moving data between
the UNIBUS and the user device. The DR11-B, like the
LPA 11-K, is a block transfer device. However, it is less expensive than the LPA 11-K, does not include a microprocessor, and can only handle a single task for a single programmer.

REAL-TIME 1/0 DEVICES
To enhance real-time performance, VAX supports the DR11 B, the LPA 11-K direct memory access (DMA) interfaces,
and the DR780 32-bit high performance parallel interface
(VAX/11-780 only) . These devices allow data to be transferred from a peripheral device to memory and vice-versa
without the intervention of computer programs except at
the initialization and completion of transfers. The result is
that CPU involvement in 1/0 operations is greatly reduced .
Further, since these devices are capable of "driving " large
blocks of information at high speeds, their usage can
greatly increase 1/0 bandwidth, i.e., the capacity of the
system to sustain a total data transfer load . 1/0 bandwidth
is an important performance measure in real-time applications, since such applications require data transfers
between external devices and computer memory.

The DR11-B interface consists of four registers: command
and status, word count, bus address , and data. Operation
is initialized under program control by loading word count
with the 2's complement of the number of transfers, specifying the initial memory or bus address where the block
transfer is to begin and by loading the command/status
register with function bits. The user device recognizes
these function bits and responds by setting up the control
inputs. If the user device requests data from memory of a
UNIBUS device, the DR11-B performs a UNIBUS Data In
transfer (DATI) and loads its data register with the information held at the referenced bus address. The outputs of
this register are available to the user device; this output
data is buffered . If the user device requests data to be written into memory, the DR11-B performs a UNIBUS Data
Out transfer (DATO) , moving data from the user device to
the referenced bus address; this input data is not buffered.
Transfers normally continue at a user-defined rate until the
specified number of words is transferred . The DA11-B has
the capability of transferring data at a rate of 500,000
bytes/second, but actual transfer rates depend upon the
particular configuration.

LPA11-K
The LPA 11-K is an intelligent (dual-microprocessor) direct
memory access controller that buffers real-time data and
transfers them to VAX memory in efficient blocks (rather
than a word at a time) . Since the LPA 11-K has automatic
buffer switching capability, transfers may occur continuously. Via a system call, the programmer can instruct the
LPA 11-K to take samples from a data source at specified
time intervals. Sampling is handled by the microprocessors , without the intervention of the CPU . Under VMS, the
LPA 11-K can be accessed via VAX-11 FORTRAN, VAX-11
BASIC, VAX-11 BLISS-32, and MACRO .
The LPA11-K operates in two distinct modes: dedicated
mode and multi request mode.

DR780
The DR780 is a high performance general purpose interface adaptor that enables users to directly interface custom devices to a VAX-11 /780 system or to connect two
VAX-11 /780 systems. This high performance, general purpose interface provides a 32-bit parallel data path capable
of transferring data up to 6.67 megabytes/second.

In multirequest mode, up to eight requests can be active
concurrently. Each user's sampling rate is a user-specified
multiple of the common real-time clock rate; thus independent rates can be maintained for each user. Each request
specifies the device so that AID, DI A or digital 1/0 can be
synchronously sampled; the transition of a bit in a digital
5-5

The architecture of the DR780 uses separate interconnect
structures for transfer of control information and data. The
control interconnect is an asychronous 8-bit bidirectional
path for transferring control information to and from the
user device. The 8-bit width of the control interconnect
makes it possible to have up to 256 individual registers in
the user device. The data interconnect is a synchronous
32-bit bidirectional path synchronized to a single clock (either the internal DR780 clock or a clock provided in the
user device). By using the DR780 internal clock, the transfer rate is selectable under program control from .156 to
6.67M bytes/sec.

bits per second over coaxial cable up to 6,000 feet long , or
speeds of up to 56 ,000 bits per second over coaxial cable
up to 18,000 feet long . The necessary modems for local interconnection are built in. Where the computers are located remotely and connected using common carrier facili ties, the DMC11 permits transmission of up to 19,200 bits
per second using an EIA interface. A DMC11 can interface
to synchronous modems such as the Bell models 208 and
209, or other synchronous modems conforming to the
RS232-C standard.
MA780
MA780 multiport memory is a bank of MOS semiconductor memory with error correcting code (ECG) that can be
shared by up to four VAX-11 /780 systems. Each system
can randomly access all of the shared memory in exactly
the same way it accesses its local memory.

The DR780 provides the high performance interface to utilize the system bandwidth of the VAX-11/780. However, to
achieve DR780 bandwidths over 2.0 Mb/second , it is required that the system include two interleaved memory
controllers .

Each MA780 can be expanded from a minimum of 256K
bytes to a maximum of 2M bytes. This storage is in addition to each system's local memory, which can be as large
as SM bytes. Since there can be up to two MA780s connected to a CPU, a VAX-11/780 system can now directly
address up to 12M bytes of physical memory.

Typical applications of the DR780 are high-speed data collection, CPU to CPU communications , signal processing,
and interfacing to graphics and array processors .

INTERPROCESSOR COMMUNICATIONS LINK

Extensions to VAX/VMS make access to the shared memory transparent to the programmer. That is, processes can
be moved from one CPU to another with transparency to
the programs involved .

VAX permits interprocessor communications via the
DMC11 communications link or via the MA780, multiport
(shared) memory. MA780 is supported by the VAX-111780
processor only.

The MA780 can be thought of as a very fast commun ication device between VAX-11 /780 systems. Specifically,
VAX/VMS provides support for interprocessor commun ications through the sharing of data regions, VMS mailboxes, and common event flags. VAX/VMS also allows code
to be shared among CPUs.

DMC11
The DMC11 communications link is designed for high-performance point-to-point serial interprocessor connection
based on the DIGIT AL Data Communications Message
Protocol (DDCMP). The DMC11 provides local or remote
interconnection of two computers over a serial synchronous link. Both computers can include the DMC11 and
DECnet software , or both computers can include the
DMC11 and implement their own communications software . For remote operations, a DMC11 can also communicate with a different type of synchronous interface provided that the remote system has implemented the DDCMP
protocol.

The MA780 can be used to configure multiple computer
systems for very high throughput. Depending on the appl ication , the CPUs can be arranged in either a parallel o r
pipeline manner, as described in Figure 5-1 below:

VA X

By implementing the DDCMP protocol in its high-speed
microprocessor , the DMC11 ensures reliable data
transmission and relieves the host processor of the details
of protocol operation (including character and message
synchronization, header and message formatting, error
checking , and retransmission control). The DDCMP protocol detects errors on the channel interconnecting the system using a 16-bit Cyclic Redundancy Check (CRC-16).
Errors are corrected by automatic retransmission. Sequence numbers in message headers ensure that messages are delivered in proper order with no omissions or
duplications.

VAX
PARALLEL

PIPELINE

Figure 5-1
Multiported Memory Configurations

The DMC11 supports full- or half-duplex operation . Fullduplex operation offers the highest throughput and is used
when the communications facilities permit two-way operation . The DDCMP protocol permits continuous simultaneous transmission of data messages in both directions
when buffers are available and there are no errors on the
channels.

In parallel systems, two or more appropriately programmed CPUs can divide a task. This allows the CPUs to
effectively pool their power to finish the job quickly. Pipeline systems can increase throughput by allowing instantaneous data exchange between CPUs that are handling
sequential parts of an application .

Where both computers are located in the same facility, the
DMC11 permits transmission at speeds of up to 1,000,000

CONSOLE STORAGE DEVICES
The VAX-11/780 console utilizes the RX01 floppy d isk
5-6

while the VAX-11 /750 console utilizes the TU58 magnetic
tape cartridge.

VAX-11/ 750 console subsystem . Because the TU58 is connected directly to the CPU, it maintains the capability to
administer diagnostics even with some system components inoperative. This feature significantly increases system reliability. The TU58 may also be used to boot the system , to load files into physical memory, and to store files
which describe and execute site-specific bootstrap procedures .

RX01 Floppy Disk Cartridge
The RX01 floppy disk is an integral part of the VAX-11 /780
console subsystem , storing microdiagnostics and system
software. This feature facilitates fast diagnosis (initiated
both locally and remotely) , simplifies bootstrapping and
initialization and improves software update distribution .

The tape cartridge is preformatted to store 2048 records
each containing 128 bytes. The controller provides'
random access to any record. The TU58 searches at 60
inches per second (ips) to find the file requested , then
reads at 30 ips. Data read from the tape are verified
through checksums at the end of each record or header.
All data transfers between the TU58 and the host are in
512 byte blocks , with the TU58 concatenating four 128
byte records to accomplish this. Data are transferred to
the CPU at approximately 2 KB per second.

The RX01 is a random access mass memory subsystem
that stores data in fixed length blocks on a flexible diskette
with preformatted, industry standard headers. The RX01 is
a single drive floppy capable of storing 256K bytes of data.
The RX02 floppy disk system can also read/write data formatted for the RX01 floppy disk.
TUSS tape cartridge
The TU58 Tape Cartridge Drive is an important part of the

5-7

6
The
Operating
System

VAX/VMS is the general purpose operating system for VAX systems . It
provides a reliable, high-performance environment for the concurrent
execution of multiuser timesharing, batch , and real-time applications.
VAX/VMS provides:
• virtual memory management for the execution of large programs
• event-driven priority scheduling
• shared memory, file , and interprocess communication data protection based on ownership and application groups
• programmed system services for process and subprocess control
and interprocess communication
VAX/VMS uses memory management features to provide swapping ,
paging , and protection and sharing of both code and data. Memory is
allocated dynamically. Applications can control the amount of physical
memory allocated to executing processes , the protection pages , and
swapping . These controls can be added after the application is implemented .
VAX/VMS schedules CPU time and memory residency on a pre-emptive priority basis. Thus , real-time processes do not have to compete
with lower priority processes for scheduling services. Scheduling rotates among processes of the same priority.
VAX/VMS allows real-time applications to control their virtual memory
paging and execution priority. Real-time applications can eliminate
services not needed to reduce system overhead . Processes granted
the privilege to ex·e cute at real-time scheduling levels , however, do not
necessarily have the privilege to access protected memory and/or data
structures.
VAX/VMS includes system services to control processes and process
execution , control real-time response , control scheduling , and obtain
information . Process control services allow the creation of subprocesses as well as independent detached processes. Processes can
communicate and synchronize using mailboxes , shared areas of memory, shared files or multiple common event flag clusters. A group of
processes can also communicate using multi ported memory.
Applications designers can use the VAX/VMS protection and privilege
mechanisms to implement system security and privacy. VAX/VMS provides memory access protection both between and within processes.
Each process has its own independent virtual address space which can
be mapped to private pages or shared pages . A process cannot access
any other process's private pages. VAX/VMS uses the four processor
access modes to read and/or write protect individual pages within a
process. Protection of shared pages of memory, files , and interprocess
communication facilities.such as mailboxes and event flags , is based
on User Identification Codes individually assigned to accessors and
data.

INTRODUCTION

device management for sharing, protection, and throughput. Devices can be shared among all jobs or reserved for
exclusive use by particular jobs. Input and output for lowspeed devices are spooled to high-speed devices to increase throughput. Files on mass storage devices can be
protected from unauthorized access on an individual,
group, or volume basis.

VAX/VMS is built for executing high-performance applications where :
• Event-driven interprocess communication and procedure and data sharing are important. Order entry and
teller transaction systems often consist of many cooperating processes that synchronize record creation and
modification .

Furthermore, the 1/0 request processing system is optimized for throughput and interrupt response. The operating
system provides the programmer with several data accessing methods, from logical record accessing for easy,
device-independent programming to direct 1/0 accessing
for extremely rapid data processing . Files can be stored in
any of several ways to optimize subsequent processing .

• Priorities of resource allocation can be set for currently
executing jobs. Both real-time processes and resourcesharing processes can execute in the same environment, as in a communications network. High-speed
links can be serviced on demand , while interactive terminal users and batch jobs share processor time and
peripherals .

VAX/VMS provides the programming tools , scheduling
services, and protection mechanisms for multiuser program development. Programmers can write, execute, and
debug programs on the system interactively, and also create batch command files that perform repetitive program
development operations without requiring their attention.

• Large programs can be developed to execute in a physical memory smaller than the program's total memory
requ irements . Engineering computation programs such
as simulators often build data arrays which require a
large address space to describe the arrays.

Although it provides a multiuser program development environment, VAX/VMS is unlike traditional program development timesharing systems . VAX/VMS is an
application-oriented operating system that optimizes total
system throughput and response to high-priority activities.
As in a timesharing system, interactive jobs can be given
equal opportunities for resource acquisition. In addition ,
the system can be executing real-time applications while
program development jobs run, since higher priority activities always have the ability to pre-empt lower priority activities.

The VAX/VMS operating system provides the run time
services for executing high-performance application systems. Operations managers and systems programmers
have considerable flexibility in designing and controlling
data and program flow.
Applications can be divided into several independent subsystems where data and code are protected from one
another, and yet which have general communication and
data sharing facilities . Jobs can communicate using general , group, or local communication facilities .
Applications which require an immediate response to
some external event can be scheduled as real-time processes. When a real-time process is ready to execute, it executes until it becomes blocked or another higher priority
real-time process needs the resources of the processor .
Normal jobs can be scheduled using a modified pre-emptive algorithm that ensures that they receive processor and
peripheral resources at regular intervals commensurate
with their processing needs.

COMPONENTS AND SERVICES
The operating system is the collection of software that organizes the processor and peripherals into a high-performance system. The operating system's basic
components include:
• processes that control initial resource allocation, communicate with the sy:.>tem operator, and log errors
• the command interpreters

If insufficient memory is available for keeping concurrently
executing jobs resident, the operating system will swap
jobs in and out of memory to allocate each its share of
processor time. Real-time processes can be locked in
memory to ensure that they can be started up rapidly when
they need to execute.

• user-callable process control services
• memory management routines
• shared run time library routines
• scheduling routines and swapper
• file and record management services

The operating system provides a dynamic virtual memory
programming environment. Large programs can be executed in a portion of physical memory that is considerably
smaller than the program's memory requirements, without
requiring the programmer to define overlays . The operating system optimizes its virtual memory system for program locality and provides tools that support optimization.
It makes program performance predictable and controllable by allowing the programmer to restrict physical
memory usage, and by bringing in large amounts of a program at one time. Processes executing under VAX/VMS
page against themselves and not against the entire system ; thus heavily paging processes executing large programs do not affect the paging of other processes.

• interrupt and 1/0 processing routines
• compatibility mode executive routines
• hardware and software exception dispatching
The operating system 's jobs run as independent activities
on the system . They include the Job Controller, which initiates and terminates user processes and manages spooling ; the Operator Communications Manager, which handles messages queued to the system operators ; the
Swapper, which controls the swapping of a processes
working set in and out of main memory; and the Error Logger, which collects all hardware and software errors detected by the processor and operating system.
A command interpreter executes as a service for interactive and batch jobs. It enables the general user to request

The operating system provides sophisticated peripheral
6-1

The operating system controls privilege and accounts for
resource allocation by job. A job can be performing processing operations under the direction of one user at a terminal , or it can be performing processing operations for
several users at multiple terminals. A job can consist of
one or several independently executing processes that
share the resource allocations for that job. Jobs can be
grouped into application subsystems that share files and
communication channels that are protected from other application subsystems.

the basic functions that the operating system provides,
such as program development, file management, and system information services.
Both hardware-detected and software-detected exception
conditions are tracked through the exception dispatcher.
The exception dispatcher passes control to user-programmed condition handlers or, in the case of systemwide exception conditions, to operating system condition
handlers.
The operating system's memory management routines include the image activator, which controls the mapping of
virtual memory to system and user jobs; the pager, which
moves portions of a process in and out of memory as required; and various system services, callable by users that
want to manage their virtual address space directly. They
respond to a program's dynamic memory requirements,
and enable programs to control their allocated memory,
share data and code, and protect themselves from one
another.

Programs and Processes
The four concepts important for understanding how the
operating system supports multiprogrammed application
systems are:
• image , or executable program
• process, or image context and address space
• job, or detached process and its subprocesses

• group, or set of jobs that can share resources

The scheduler controls the allocation of processor time to
system and user jobs. The scheduler always ensures that
the highest priority, ready-to-execute real-time process
receives control of the processor until it relinquishes it.
When no real-time processes are ready to execute, the
scheduler dynamically allocates processor time to all other
jobs according to their priorities and resource requirements. The swapper works in conjunction with the scheduler to move entire jobs in and out of memory when memory requirements exceed memory resources. The swapper
ensures that the jobs most likely to execute are kept in
memory.

These concepts are for the most part transparent to the
general user whose only contact with the system is the operating system's command language interpreter or an application's command interface. They are, however, significant concepts for the applications programmer. Figure 6-1
illustrates the concepts of groups, jobs, processes , and
images.
An image is an executable program. It is created by translating source language modules into object modules and
linking the object modules together. An image is stored in
a file on disk . When a user runs an image, the operating
system reads the image file into memory to execute the
image.

The operating system's 1/0 processing software includes
interrupt service routines, device-dependent 1/0 drivers,
device-independent control routines, and user-programmed record processing services. The 1/0 system ensures rapid interrupt response and processing throughput, and provides programming interfaces for both special
purpose and general purpose 1/0 processing .

The environment in which an image executes is its context.
The complete context of an image not only includes the
state of its execution at any one time (known as its hardware context) , it also includes the definition of its resource
allocation quotas, such as device ownership, file access,
and maximum physical memory allocation . These resource allocation quotas are determined by the quotas
given to the user who runs the image.

The next few sections discuss some of the concepts basic
to understanding the operating system's functions and
services. They are followed by descriptions of the services
available to individual and cooperating processes, and descriptions of memory management and scheduling for the
systems programmer.

Two or more users can execute the same image concurrently; that is, image code can be shared, in which case the
image is executing in two or more different contexts. An
image context, including the address space used by an image, is called a process. The operating system schedules
processes, and a process provides a context in which an
image executes.

PROCESSING CONCEPTS
To support high-performance multiprogramming application environments, the operating system provides the applications programmer with the tools to implement:

The distinction between an image and a process is a significant one. We can speak of two processes, each executing the FORTRAN compiler. There may be only one copy
in physical memory of the FORTRAN compiler's image,
but two different contexts in which the image executes. In
one context, the compilation may have just begun ; in the
other context, it may be almost complete. In one context,
the compiler may be reading and writing files listed in one
directory; in the other context, the compiler may be reading and writing files listed in another directory.

• shared programs
• shared files and data
• interprocess communication and control
To enable the programmer to write shared programs easily, the operating system treats a program independently of
the context in which a program executes. The context defines the privileges assigned by the system manager to a
particular user. Users with different privileges can share
programs, and the operating system will enforce protection independently of the program .

A process executes only one image at a time, but it provides the context for serially executing any number of d ifferent images. For example, when a user logs on the sys-

6-2

GROU P 1
JOBI

OWNER

Sl.SPROCES

JOB 2

(DETAC HED )
PROC ESS

SU BP ROCESS
OWN ER

SUBPROCESS

PROCES S

JO B 3

D
G ROUP 2

GROUP 3

Figure 6-1
Programs and Processes

processes created below it heirarchically share a common
pool of quotas. The owner process can control the scheduling of its subprocess , and it can delete the subprocess.
When an owner process terminates , all of the subprocesses it owns are terminated .

tern at an interactive terminal , the operating system
creates a process for that user. If the user edits a file, the
editor image executes in the context of that user's process .
If the user then compiles a program , the compiler image
executes in the context of that user's process . A process
thus acts as a continuous "envelope" for a user's activities.
An image executing in the context of a process can create
subprocesses. A subprocess can be thought of as an auxiliary process in which a given image executes. When an
image creates a subprocess , it identifies the image to be
executed in the context of the subprocess or the source of
commands to be interpreted in that process . An image executing in a subprocess context can in turn create other
subprocesses.

A detached process is the process created by the operating system on behalf of a user who logs onto the system
and requests services of the system using a command
interpreter. A detached process has no owner. Normally,
only the operating system can create detached processes,
but a suitably privileged application program could also
create a detached process and start up an application
command interface to execute images serially for the detached process or any subprocess it creates.

The process executing the image that creates a subprocess is an owner process . An owner process has com plete control over the execution of the subprocesses it
creates. It determ ines which of its privileges it will allow a
subp rocess to have. Each detached process and all sub-

A job consists of a detached process and all the subprocesses it creates, and all the subprocesses they create,
etc. Jobs are the accounting entities that the system uses
to control resource allocation . All processes constituting a
job are scheduled independently (they can compete for
6-3

processor time individually to overlap processing) , but
they share the total resources allocated to the detached
process . Each job has a set of resources that it can use,
i. e. , authorized quotas. Subprocesses share these quotas
with the detached process.
Jobs can be associated in groups. Groups are the basis
for the definition and development of application subsystems . Groups are mutually exclusive , that is, if a job belongs to one group, it does not belong to any other group.
A process with appropriate privilege can control the execution of other processes in the same group. Processes in
the same group can synchronize their activities using protected group communication facilities .
Resource Allocation
The resources of the system are the processor, memory,
and peripherals. The system handles many jobs
simultaneously, and each job can have different resource
requirements . The operating system enables jobs to share
the resources according to their ind ividual needs, and the
operating system protects each job and its data from other
jobs on the system .
The operating system controls resource allocation dynam ically through its scheduling, memory management,
device allocation, and 1/0 processing software , and statically through the authorization of users.

When the user executes a known image, the process has
the privileges and quotas granted to the user in the authorization file, plus those run time execution privileges
granted specifically to that image. While that image executes, the user may have the priv ilege to perform operations not granted when executing any other image. Fo r example , one known image is the operating system 's LO GIN
image, which enables a user to log on the system . The
LOGIN image has the privilege required to access the user
authorization file to obtain the user's privileges and q uotas .

The system manager is responsible for creating an authorization file entry for each user of the system . The authorization file provides the operating system with the resource quotas and limits for each job. For example , there
are quotas and limits that control :
• total processor time usage
• number of subprocesses a job can create
• number of simultaneously open files
• process virtual and physical memory usage
• number of simultaneous 1/0 transfers

Protection
The basis for data protection in the VAX system is the user
identification code (UIC) . A UIC consists of two numbers: a
group number and a member number. The system manager assigns each user a user identification code (UIC) in
the user authorization file . Images that the user executes
are g iven or denied data access privileges based upon the
user's UIC .

Separate authorization files located on each disk volume
control disk usage quotas.
Privileges
In addition to provid ing job quotas, the user authorization
file provides the base definition for each user's privileges .
There are potentially 64 distinct privileges that can be
individually granted or withheld . Among them are privileges that give the job the right to:

When a file or an interprocess communication fac ility is
created , it is assigned a UIC and a protection code. The
UIC determines which group of users or programs , an d
which members within the group, have controlled access
to that data. The protection code provides the access control.

• alter the priority of a process
• execute a user-written program at a more privileged access mode
• execute operator functions

The protection code applies to four types of access : read ,
write , execute , and delete . Each type of access can be
given or denied to :

• create detached processes
• set up the communication facilities used by cooperating
processes

• the owner: the user whose UIC is the same as the UIC
assigned to the data.

• control other processes in the same group
Whenever the user executes an image, the image can at
most acquire only those privileges and quotas granted
directly to that user's job by the authorization file, unless
the image is a known image. Known images are installed
by the system manager, and while they execute they provide a second , dynamic set of privileges granted a user.

• the group: every user whose UIC group number is the
same as that assigned the data.
• the world : every user whose UIC group number is diffe rent or the same from that assigned the data. (everyo ne
on the system)
6-4

• the system: every user or program with the privilege
SYSPRV, and those whose UIC group number is a system privileged group number (1-X, where Xis a number
specified by the sysgen parameter, MAXSYSGROUP).

When a user program is translated and linked, the image
is allocated addresses starting with address 512 and continuing up . The first page is not normally allocated (although it can be) because it helps catch programming
errors caused by improperly initialized pointers, by
branching or jumping to 0, or by passing O or other small
addresses as arguments. The linker allocates the remainder of address space to image sections according to
whether they are shared or private, position-independent
or position-dependent, and read-only or read/write , such
that memory protection can be used to full advantage in
preventing and isolating programming errors.

For example, in a common application of the protection
scheme, a user can create a program image file and assign it the same UIC as the user's own UIC (the default
case). The user can give it a protection code to:
• enable the user (and all other users with the same UIC)
to read, write , execute, and delete the file
• enable other users in the group to execute the program
image, but prevent them from reading , writing or deleting the file

The addresses in the control region are used to identify the
locations containing temporary image control information
and data such as the stacks, permanent process control
information such as 1/0 channel allocations, and code provided by the operating system. These addresses are allocated from address 2 31 -1 down . One reason this method
of allocating in reverse is convenient is because the control region contains the process stacks, and stacks grow to
lower addresses as data are added , and to higher addresses as data are removed .

• prevent all users outside the group (other than privileged system users) from reading , writing, executing, or
deleting the file
• enable the the privileged system users to read the file
(so that it can be backed up, for example)
Read and write access applies to both files and interprocess communication facilities. Delete access applies only to
files , and execute access applies only to program image
files. (The privileges and quotas granted in the user authorization file control creation and deletion for
interprocess communication facilities.)

There are four stack areas reserved in the control region ,
one for each access mode orotection level that the processor provides for software executing in the context of a
process. (Refer to the VAX Processors section for a description of access modes.) The stack seen by the image
executing in the program region is the user stack . All other
stack areas are protected from that image. These stacks
are used by operating system software executing in the
context of the process on behalf of the image the process
is executing. For example, command interpreters use the
supervisor stack , the record management services use the
executive stack , and the exception dispatcher and some
exception handlers use the kernel stack.

USER PROCESS ENVIRONMENT
The user program environment is the process, which is the
entity the operating system schedules for execution . Each
process has its own independent address space in which
an image executes. Each image executing in a process
can call system service procedures to acquire resources
and request special processing services from the operating system . The following paragraphs introduce program
virtual address allocation and the fundamental system service procedures available to user programs directly, as
well as indirectly through the more complex programmed
requests provided by the operating system .

The system region addresses, which start at address 2 31 ,
are used to identify the locations containing the entry vectors for system service procedures, followed by locations
containing privileged operating system code and data. The
system service entry vectors are permanently reserved
virtual addresses so that no relinking is required if system
services are modified . Other addresses in the system region are not generally used by the image allocated to the
program region , and access to areas mapped by these addresses is restricted .

Virtual Address Space Allocation
Process virtual address space is the set of 32-bit addresses that an image executing in the context of a process uses to identify byte locations in virtual memory. For
the purpose of allocating virtual memory to processes, the
operating system divides process virtual address space
into four sets of virtual addresses. The first three sets of
addresses are called the program region , the control region, and the system region . The fourth set of addresses is
unused .

System Services
An image requests services of the operating system
directly through calls to the system services. The system
services are to the operating system what the instruction
set is to the processor. They provide all the primary resource request activities, such as 1/0 processing and
interprocess communication . Other programmed requests available to the user are often derived from system
services. For example, the record management services
use the 1/0 processing system services as the basis for
logical record processing functions.

Figure 6-2 illustrates the general allocation of virtual address space for each process . Addresses in the first two
regions are used for process code and data, where the
first reg ion is generally used for image-specific code and
data, and the second for stacks, process permanent data,
buffers, and operating system code. Addresses in the system region are also used for code and data maintained by
the operating system, but in this case the addresses refer
to the same locations in every process context. The system
region addresses provide a set of locations whose addresses are independent of process context , and therefore
do not have to be context switched.

Images that use the system services can be written in assembly language or any native programming language
that has a Call statement. (Refer to the Languages section
and the section on the RSX-11 M Programming Environment for system services available for compatibility mode
6-5

OooOooOo -

NO ACCESS PAGE

'---------------<
2 NON PIC IMAGE

3 PIC SHARED AND
PIC IMAGf
!AS MUCH AS FITSI
l NON· PIC SHARED
CODE AND DATA
(PREVIOUSLY LINl(EDl

THIS PART OF PO SPACE

> - - - - - - - - - - - - - i lS DEFINED BY THE
4

PO
PROGRAM
SPACE

LINl(ER AND MAPPE D
8Y THE IMAGE ACTIVAT~

THE REST OF
PIC SHARED AND PIC
IMAGE\IF ANY)

RTL VECTORS

f-- -

RUN TIME LIBRARY
VMS RTL . EXE
IF NEEDED AND NOT EXCLUDED!
.5

.____________,_ -

80000000

- -

- - -

I ~'t'.W,Bm"
~·-.~ coo•"o"
lNDD• •
NONPAGED E-XECUTIVE DATA

DEBUGGER (LIB$ DEBUG I
(IF REQUESTED AT LINK.,
RUN . OR EXECUTION TIME!

NONPAGED EXECUT!VE CODE
PAGE.AIRE EXECUTIVE ROUTINES

TRACE BACI( (LIB & TRACE)
( IF NOT OVERR IDDEN AT LINI(
TIME ANO NEEDED)

>-------------!- - - -

* - - - -- -- --

NOT MAPPED

-N~~:£o-

- - ~--------------<-

SYSTEM
REGION

POOL OF UN MAPPED SVSTEM PAGES
RESTART PARAMETER BLOCK
PFN DATABASE

Cll IMAGE

DYNAMIC PORTION
OF Pl SPACE

Pl WINDOW TO PROCESS HEADER

'---------------'- CHANNEL CDNTROL BLOCK TABLE

STATIC POfHION
OF Pl SPACE

PAGED DVNAMIC MEMORl

NONPAGED DYNAMIC MEMORY

PROCESS I/0 SEGMENT
INTERRUPT STA(I(
PER PROCESS COMMON AREA

SYSTEM CONT ROL BLOCK

PER PROCESS COMMON AREA
BALANCE SLOTS
COMPATIBILITY MOOE DATA PAGE
PROCESS 1/0 SEGMENT
SYSTEM HEADER
PROCESS ALLOCATION REGION

SYSTEM PAGE TABLE

GENERIC CLI DATA PAGES

GLOBAL PAGE TABLE

IMAGE ACTIVATOft SCRATCH PAGES
DEBUGGER CONTEH
VECTORS FOR MESSAGES AND USER
WRITTEN SYSTEM SERVICES
IMAGE HEADER SUFFER
NO ACCESS

PAGE

l(ERNEl STACK
EXECUTIVE STACI(
SUPERVISOR STACI(
Pl POINTER PAGE

- -

7 FFFFFFF

-- -

I NITIALIZATIQN TIME]

SVSTEM MESSAGE FILE (S"YSMSG EXEi

OF Pl SPACE

I DYNAMlCALLl MAPPED AT

CLI SYMBOL TABLE

(CONTROL}
SPACE

5 T~T!f .!:QR..!!_O_~J S_..!...S _E.X!J _ _

SVS TEM VIRTUAL PAGES MAPPED
TO J/0 ADDRESSES

RMS IMAGE ( RMS . EXE)

IMAGE I/0 SEGMENT

>-------------!-- - - PU PROCESS MESSAGE SECTION (S)
PERMANENT POftTION

PER
PROCESS
SPACE

XDELTA(USUALLV UNMAPPED). !NIT

DEBUGGER SYMBOL TABLE
(NOT MAPPED IF DEBUGGER NOT
PRESENT]
-~· ~---------~

Figure 6-2
Virtual Address Space Allocation

6-6

:::

programming languages.) The Call interface is the same
independent of the programming language selected with
the exception of CORAL 66.

applications programmer, some of which are controlled by
privilege. The table also lists those system services used
primarily by the operating system but which can also be
used by suitably privileged application system code .

Table 6-1 summarizes the system services available to the

Table 6-1
System Services

1/0 SERVICES FOR DEVICE-DEPENDENT 1/0
Assign 1/0 Channel
($ASSIGN)

Send Message to
Symbiont Manager
($SNDSMB)

Establish a path for an 1/0 request.
Establish a path for network operations.

Deassign 1/0 Channel ($DASSGN)

Release linkage for an 1/0 path.

Get 1/0 Channel
Information
($GETCHN)

Provide information about a device
to which an 1/0 channel has been
assigned .

Get 1/0 Device
Information
($GETDEV)

Provide information about a physical
device.

Allocate Device
($ALLOC)

Reserve a device for exclusive use by
a process and its subprocesses.

Deallocate Device
($DALLOC)

Request symbiont manager to initialize, modify, or delete a printer, device,
or batch job queue.
Request symbiont manager to queue
a batch or print file, or delete or
change characteristics of a queued
file.

Release a path from the network .
Send Message to
Operator
($SN DO PR)

Write a message to designated operator(s) terminal(s).

Send Message to
Error Logger
($SN DERR)

Write arbitrary data to the system error log file.

Relinquish exclusive use of a device.

Get Message
($GETMSG)

Return text of system or application
error message from message file.

Queue 1/0 Request
($QIO)

Initiate an input or output operation
and continue processing while 1/0 is
in progress .

Put Message
($PUTMSG)

Write a system or application error
message to the current output and error devices.

Queue 1/0 Request
and Wait ($QIOW)

Initiate an input or output operation
and cause the process to wait until the
1/0 is complete before continuing execution.

Cancel 1/0 on
Channel
($CANCEL)

Cancel pending 1/0 requests on a
channel.

Formatted ASCII
Output ($FAO)

Perform ASCII string substitution, and
convert numeric data to ASCII representation and substitute in output.

Formatted ASCII
Output with List Parameter ($FAOL)

Create Mailbox and
Assign Channel
($CREMBX)

Create a temporary mailbox.

Delete Mailbox
($DELMBX)

Mark a permanent mailbox for deletion .

Broadcast
($BRDCST)

Send a high-priority message to a
specified terminal or terminals.

Send Message to
Accounting Manager ($SNDACC)

Control accounting log file activity.

LOGICAL NAME SERVICES
Create Logical
Name ($CRELOG)

Place logical name/equivalence
name pair in process logical name
table .
Place logical name/equivalence
name pair in group logical name
table.
Place logical name/equivalence
name pair in system logical name
table.

Delete Logical
Name ($DELLOG)

1/0 SERVICES FOR MAILBOXES AND MESSAGES

Enable or disable an operater's terminal, send a reply to a user request or
initialize the operator's log file .

Remove logical name/equivalence
name pair from process logical name
table.
Remove logical name/equivalence
name pair from group logical name
table.

Create a permanent mailbox.

Remove logical name/equivalence
name pair from system logical name
table.
Translate Logical
Name ($TRNLOG)

Write an arbitrary message to the accounting log file.

6-7

Search logical name table for a specified logical name and return its equivalence name when the first match is
found.

Table 6-1 (con 't)
System Services

AST (ASYNCHRONOUS SYSTEM TRAP) SERVICES

EVENT FLAG PROCESSING

Associate Common
Event Flag Cluster
($ASCEFC)

Create a temporary common event
flag cluster.
Create a permanent common event
flag cluster.
Create a common event flag cluster in
memory shared by multiple processors.
Establish association with an existing
common event flag cluster.

Disassociate Common Event Flag
Cluster ($DACEFC)

Cancel association with a common
event flag cluster .

Delete Common
Event Flag Cluster
($DLCEFC)

Mark a permanent common event flag
cluster for deletion .

Set Event Flag
($SETEF)

Turn on an event flag in a processlocal event flag cluster.

Turn off an event flag in a processlocal event flag cluster.
Turn off an event flag in a common
event flag cluster.

Read Event Flags
($READEF)

Establish AST routine to receive control following power recovery condition .

Set AST Enable
($SET AST)

Enable or disable the delivery of
AS Ts.

Declare AST
($DCLAST)

Queue an AST for delivery.

CONDITION HANDLING SERVICES

Turn on an event flag in a common
event flag cluster.
Clear Event Flag
($CLREF)

Set Power Recovery
AST ($SETPRA)

Set Exception Vector ($SETEXV)

Define condition handler to receive
control in case of hardware- or software-detected exception conditions .

Set System Service
Failure Exception
Mode ($SETSFM)

Request or disable generation of a
software exception condition when a
system service call returns an error or
severe error.

Unwind from Condition Handler Frame
($UNWIND)

Delete a specified number of call
frames from the call stack following a
nonrecoverable exception condition.

Declare Change
Mode or Compatibility Mode Handler
($DCLCMH)

Designate a routine to receive control
when change mode to user instructions are encountered .

Return the status of all event flags in a
process-local event flag cluster.

Designate a routine to receive control
when compatibility mode exceptions
occur.

Return the status of all event flags in a
common event flag cluster.
Wait for Single
Event Flag
($WAITFR)

PROCESS CONTROL SERVICES

Place the current process in a wait
state pending the setting of an event
flag in a process-local event flag
cluster.
Place the current process in a wait
state pending the setting of an event
flag in a common event flag cluster .

Wait for Logical OR
of Event Flags
($WFLOR)

Place the current process in wait state
pending the setting of any one of a
specified set of flags in a processlocal event flag cluster .

Create Process
($CREPRC)

Create a subprocess.

Set Process Name
($SETPRN)

Establish a text string to be used to
identify the current process.

Get Job/Process
Information
($GETJPI)

Return information about the current
process.

Create a detached process.

Return information about the current
process context of other processes in
the same group.
Return information about any other
process in the system .

Place the current process in a wait
state pending the setting of any one of
a specified set of flags in a common
event flag cluster .
Wait for Logical
AND of Event Flags
($WFLAND)

Designate a routine to receive control
when change mode to supervisor instructions are encountered .

Delete Process
($DELPRC)

Delete the current process, or a subprocess.
Delete another process in the same
group.

Place the current process in a wait
state pending the setting of all specified flags in a process-local event flag
cluster.

Hibernate ($HISER)

Place the current process in a wait
state pending the setting of all specified flags in a common event flag
cluster .

Make the current process dormant
but able to receive ASTs until a subsequent wakeup request.

Schedule Wakeup
($SCHDWK)

Wake a process after a specified time
interval or at a specific time.

Delete any process in the system .

6-8

Table 6-1 (con't)
System Services
Cancel Wakeup
($CA NW AK)

Cancel a scheduled wakeup request.

Wake ($WAKE)

Restore executability of the current
process or a hibernating subprocess.

TIMER AND TIME CONVERSION SERVICES

Restore executab ility of a hibernating
process in the same group .
Restore executability of any hibernating process in the system .
Suspend Process
($SUSPND)

Make the current process or a subprocess nonexecutable and unable to
receive ASTs until a subsequent resume or delete request.
Make another process in the same
group nonexecutable and unable to
receive ASTs until a subsequent resume or delete request.
Make any process in the system nonexecutable and non interruptible
until a subsequent resume or delete
request.

Resume Process
($RESUME)

GetTime
($GETTIM)

Return the date and time in system
format.

Convert Binary
Time to Numeric
Time ($NUMTIM)

Convert a date and time from system
format to numeric integer values .

Convert Binary
Time to ASCII String
($ASCTIM)

Convert a date and time from system
format to an ASCII string.

Convert ASCII
String to Binary
Time ($BINTIM)

Convert a date and time in an ASCII
string to the system date and time format.

Set Timer
($SETI MR)

Request setting of an event flag or
queuing of an AST , based on an absolute or delta time value .

Cancel Timer Request ($CANTIM)

Cancel previously issued timer requests.

Schedule Wakeup
($SCHDWK)

Schedule a wakeup for the current
process or a hibernating subprocess.

Restore executability of a suspended
subprocess.

Schedule a wakeup for a hibernating
process in the same group.

Restore executabil ity of a suspended
process in the same group.

Schedule a wakeup for any hibernating process in the system.

Restore executability of any suspended process in the system .
Exit ($EXIT)

Terminate execution of an image and
returns to command interpreter.

Force Exit
($FORCEX)

Cause image exit for the current process or a subprocess .

Cancel Wakeup
($CANWAK)

Cancel a scheduled wakeup request
for a hibernating process in the same
group.
Cancel a scheduled wakeup request
for any hibernating process in the
system .

Cause image exit for a process in the
same group.
Cause image exit for any process in
the system .
Declare Exit
Handler ($DCLEXH)

Designate a routine to receive control
when an image exits.

Cancel Exit Handler
($CANEXH)

Cancel a previously established exit
handling routine.

Set Priority
($SETPRI)

Change the execution priority for the
current process or a subprocess .
Change the execution priority for a
process in the same group.
Change the execution priority for any
process in the system.

Set Resource Wait
Mode ($SETRWM)

Request wait, or that control be returned immediately, when a system
service call cannot be executed because a system resource is not available.

Set Privileges
($SETPRV)

Allow a process to enable or disable
specified user privileges .

Cancel a scheduled wakeup request
for the current process or a hibernating subprocess .

Set System Time
($SETIME)

Set or recalibrate the current system
time.

MEMORY MANAGEMENT SERVICES
Adjust Working Set
Limit ($ADJWSL)

Change maximum number of pages
that the current process can have in
its working set.

Expand
Program/Control
Region
($EXPREG)

Add pages at the end of the program
or control region .

Contract
Program/Control
Region
($CNTREG)

Delete pages from the end of the program or control region .

Create Virtual
Address Space
($CRETVA)

Add pages to the virtual address
space available to an image.

Delete Virtual
Address Space
($DEL TVA)

Make a range of virtual addresses
unavailable to an image.

6-9

Table 6-1 (con't)
System Services
Create and Map
Section ($CRMPSC)

Delete Global Section ($DGBLSC)

Identify a disk file as a private section
and establish correspondence
between virtual blocks in the file and
the process' virtual address space.

Mark a permanent global section for
deletion .
Mark a system global section for
deletion.

Identify a disk file containing shareable code or data as a temporary global section and establish correspondence between virtual blocks in the
file and the process' virtual address
space.

Set Protection on
Pages ($SETPRT)

Control access to a range of virtual
addresses.

Lock Pages in
Working Set
($LKWSET)

Specify that particular page cannot be
paged out of the process' working set.

Identify a disk file containing shareable code or data as a permanent global section and establish correspondence between virtual blocks in the
file and the process' virtual address
space.

Unlock Pages from
Working Set
($ULWSET)

Allow previously locked pages to be
paged out of the working set.

Purge Working Set
($PURGWS)

Remove all pages within a specified
range from the current working set.

Identify a disk file containing shareable code or data as a system global
section and establish correspondence between virtual blocks in the
file and the process' virtual address
space.

Lock Page in Memory ($LCKPAG)

Specify that particular pages may not
be swapped out of memory.

Unlock Page in
Memory
($ULKPAG)

Allow previously locked pages to be
swapped out of memory.

Set Process Swap
Mode ($SETSWM)

Control whether or not the current
process can be swapped out of the
balance set.

Identify one or more page frames in
physical memory as a private or global section and establish correspondence between the page frames and
the process' virtual address space .
Update Section File
on Disk ($UPDSEC)

Write modified pages of a private or
global section into the section file.

Map Global Section
($MGBLSC)

Establish correspondence between a
global section and a process' virtual
address space.

CHANGE MOOE SERVICES

1/0 System Services
The operating system provides the programmer with two
request interfaces for performing input/output operations:
the 1/0 system services and the record management services. Record management services, discussed in the Data Management Facilities section, provides a general purpose file and record programming interface that satisfies
most 1/0 processing needs, and allows the programmer to
implement 1/0 processing quickly. The 1/0 system services provide the programmer with direct control over the
1/0 processing resources of the operating system. In
particular, the 1/0 system services enable the programmer
to:

Change Mode to Executive ($CM EXEC)

Execute a specified routine in executive mode.

Change Mode to
Kernel ($CMKRNL)

Execute a specified routine in kernel
mode.

Adjust Outer Mode
Stack Pointer
($ADJSTK)

Modify the current stack pointer for a
less privileged access mode.

grammed in the same manner. All devices can be sequentially accessed, including mass storage devices such as
disks and magnetic tapes, and record-oriented devices,
such as terminals, card readers and line printers. In addition, disk volumes can be accessed randomly.
Mass storage volumes can be either file-structured or nonfile-structured, according to the choice of the user. The 1/0
system services enable programmers to use either the
physical (device assigned) address or a logical (driver assigned) address for directly addressing blocks on foreign
mass storage vo lumes. A foreign volume can be either
non-file-structured, or structured with the user's own file
structure. If the volume is structured using the operating
system's Files-11 disk file or ANSI magnetic tape structure, the 1/0 system services enable the programmer to
address blocks directly using virtual (file system assigned)
addresses.

• perform both device-independent and device-dependent 1/0 processing
• read and write blocks on mass storage media using
physical (device-oriented), logical (volume-relative), or
virtual (file-relative) addressing

A special type of record-oriented device is the mailbox,
which is a virtual device that a process creates for the receipt of messages from other processes . Mailboxes are

The 1/0 system recognizes several types of devices, and
within the extents of their capabilities, all devices are pro-

6-10

treated like any other record-oriented device: they can be
read from and written to using either the 1/0 system services or record management services . Mailboxes are discussed further in the section on Interprocess Communication .

Each process has available for its own use two local event
flag clusters, each of which contains 32 event flags. Eight
flags in the first cluster are reserved by the operating system. A process can set, clear, and read individual event
flags, as well as wait for one or more event flags to be set.
The advantage to having two clusters of event flags is that
the flags in each cluster can be treated as a related group.
A process can wait until any of a specified set of flags in a
particular cluster is set, or wait until all of a specified set of
flags in a particular cluster are set.

Before a process requests 1/0 to a device, it obtains a
channel assignment from the operating system. A process
can use a device name or a logical name in a channel assignment request to identify the device for which the channel is desired .

Aside from their use with 1/0 processing and timer scheduling , a process can assign its own meanings to local
event flags . Event flags can be used to coordinate several
asynchronous events , such as multiple 1/0 request completions , or to simplify asynchronous processing. For example, a program may wish to know if a terminal user has
typed a CTRL/C (indicating the desire to interrupt execution) only at well-defined points during processing . An
asynchronous system trap routine can set an event flag to
indicate that a CTRL/C has been received .

A device name is a unique name assigned by the operating
system to a particular physical device. The name identifies
the type of device and its controller and line or unit number, as applicable. For example, DMA3: is the operating
system's device name for the RK06 disk drive unit 3 on
controller A, and TT A 12: is the operating system 's name
for the terminal on line 12 on multiplexer A.
A logical device name is any string of characters a user or
program assigns to a device name assigned by the operating system. The Create Logical Name system service not
only enables a process to define logical names for device
names, but it enables a process to assign logical names to
any portion of a file specification, or to other logical
names. Furthermore, logical names can be assigned on a
per-process , per-group , or system-wide basis. (For more
information on logical names, refer to the Data Management Facilities Section .)

Asynchronous System Traps
An asynchronous system trap (AST) is a software-simulated interrupt used for event notification within a process.
An asynchronous system trap routine is a procedure that
handles an AST. AST routines provide an efficient means
for processing events that can occur at any time during
processing (such as terminal input) because they eliminate the need for poll ing.

Once a channel is obtained, a process can issue 1/0 requests on that channel. The Queue 1/0 Request system
service is a general 1/0 request interface. All 1/0 using system services is asynchronous: both 1/0 and computation
can be taking place simultaneously. An 1/0 request is simply queued to the device driver and control is normally returned to the requesting process before the 1/0 operation
is complete.

For example , a program can specify AST routines for 1/0
request processing , timer scheduling , and power recovery . When the 1/0 operation completes , time interval expires , or power is restored, the operating system declares
an AST. When the AST is delivered, the operating system
interrupts the process and executes the AST routine. A
process can be hibernating and still receive ASTs declared for it.

The process is responsible for synchronizing with 1/0
completion. The process can simulate synchronous 1/0
processing by using the Queue 1/0 Request and Wait system service, or it can continue to execute during the 1/0
operation and request 1/0 completion notification using
the general purpose event flag or asynchronous system
trap notification mechanisms.

Code executing at one processor access mode can declare an AST for code executing at the same or a less privileged access mode. The operating system automatically
disables AST delivery while an AST routine is executing,
and code executing at a given access mode can explicitly
disable AST delivery. While ASTs are disabled, the operating system queues any ASTs waiting to be delivered to that
access mode in the order in which they were declared.
When AST delivery is again enabled for that acess mode,
the ASTs are delivered in the order in which they were
queued.

Real-time interface extensions (connect-to-interrupt and
map-to-1 / 0 page) provide the real-time programmer
(MACRO and BLISS-32) a technique of more simply interfacing to user-specialized devices. The connect-to-interrupt facility can be used to cause an interrupt to be delivered directly to the user's program. As a consequence of
this approach , response to the interrupt occurs in the
shortest time possible without writing an 1/0 driver.

Exception Conditions and Condition Handlers
A program may request the processor or a system service
to do operations they cannot perform correctly. For example , a program might inadvertently issue a divide instruction using a divisor of zero. Normally there is no way to recover and the program cannot continue. In this system ,
however, it is possible for a program to continue if it declares a condition handler that can correct the situation. If
a user program declares a condition handler, control
transfers to the condition handler when an exception condition occurs.

The map-to-1/0 page is a complement to the connect-tointerrupt facility by allowing the user program to access
the device registers. Before these extensions were available, the user had to write both a device driver and an application program to achieve the same results.
Local Event Flags
An event flag is a status bit used for posting an event, such
as 1/0 completion or elapsed time interval. Event flags are
an extremely efficient means of starting up or synchronizing procedures.

This system treats all errors or special events that occur
synchronously with respect to a program's execution as
6-11

and so on. The advantage of a complex job over a simple
job is that a complex job performs parallel processing operations because it has control over several images executing concurrently .

exception conditions, and provides a general purpose
mechanism for dispatching condition handlers. Exception
conditions include:
• errors from which the processor cannot normally recover, such as the divide by zero arithmetic trap

The following sections describe the services that enable a
process to control and communicate with other processes .

• special conditions for which a program does not wish to
test continually , for example, the floating point overflow
arithmetic trap or unsuccessful system service completion

Process Control Services
The system services provide process control by enabl ing a
process to :

Some of the exceptions detected by the processor are
handled automatically by the operating system . For example, the pager is a condition handler for translation-buffernot-valid faults.

• create and delete subprocesses
• hibernate, then reactivate, a process via the Hibernate/ Wake and Suspend/Resume system services
The ability to create subprocesses is granted to a user by
the system manager, where the number of subprocesses a
job can create is a resource limit. When a process creates
a subprocess, it can give the subprocess all or some of its
privileges, and its resource quotas and limits are shared
with the subprocess . Other resource quotas are shared
between the creator and the subprocess.

In addition to processor and system service detected exception conditions, any software procedure can define
cases for which it will fail or produce an exception by calling a system library procedure that signals an exception
cond ition . The search sequence for a condition handler is
independent of the nature of the exception condition : the
search sequence is the same whether an exception condition is detected by hardware or software.
A process can declare two kinds of condition handlers:
those that are process-wide and those that are applicable
to individual procedures . Process-wide condition handlers
are declared using the Set Exception Vector system service, which enables a process to declare a primary and a
secondary condition handler. Condition handlers applicable to individual procedures are declared by the procedure when it is called using one instruction.
When an exception condition occurs, the exception dispatcher does not differentiate between exception
conditions, it simply transfers control to the first condition
handler it can find that wants to handle the exception condition . This method for handling exception conditions is an
efficient means of transferring control to the appropriate
condition handler rapidly, since condition handling is defined by the module or modules in which an exception
condition may occur.
For programs written in high-level languages, each language may have different definitions of what is and what is
not an exception condition . As the user program calls language functions , the exception conditions for those functions can be handled locally with the procedure. And
where exception conditions should be handled on a process-wide basis , the primary and secondary exception vectors provide a top level exception condition trap . For example, when a user program is linked with the debugger,
the debugger uses the primary exception vector to declare
a process-wide condition handler.

The Hibernate/Wake and Suspend/Resume mechan isms
are methods of process control which are especially efficient in real-time applications. They allow the user to prepare an image for execution and then place it into a wait
state until some event occurs which requires its activation .
The Hibernate system service provides the greatest flexibility in sequencing processes for execution . When a Hibernate system service is invoked, normal execution can
be resumed only by issuance of a $WAKE system service
(or a variant , $SCHDWK, which allows wake-up at an
absolute time or at a fixed time interval). However, a hibernating process can be interrupted temporarily by the delivery of an AST (Asynchronous System Trap) . When the AST
service routine completes execution , the process continues hibernation. If, however, the process calls the $WAKE
system service during execution of the AST service routine, the process wakes itself after the service routine
completes. Figure 6-3 shows an example of a program
which uses the hibernate and wake system services.
Using the $SUSPEND system service, a process can place
itself or another process into a wait state similar to hibernation . However, a suspended process cannot be as easily
activated as a hibernating one. It cannot, for example, be
interrupted by delivery of an AST. Nor can it wake itself,
but can only resume normal execution following issuance
of a $RESUME system service by another process . Table
6-1 summarizes the differences between hibernation and
suspension.
The interprocess system services can be used by a process to control another process executing in the same
group. While only an owner process can create and delete
subprocesses , a process can be given the privilege to suspend , resume, and wake other processes in its group.

INTERPROCESS COMMUNICATION AND CONTROL
This system supports both simple and complex job definitions. A simple job is a detached process created by the
operating system on behalf of the user who logs in at a terminal, or for the purpose of executing a batch job . A simple job serially executes images, but it does not create
subprocesses.

Jobs with sufficient privilege can also create detached
processes, and delete, suspend, resume, or wake any
process in the system . These privileges are normally reserved for the operating system or the system manager.

A complex job is one in which a detached process creates
subprocesses in which designated images execute. These
subprocesses can also create their own subprocesses,

Interprocess Communication Facilities
In addition to providing process control services, the oper6-12

Process: GEMINI
ORION : .ASCID 'ORION'
FASTCOMP: .ASCID 'COMPUTE.EXE'

;SUBPROCESS NAME
;I MAGE

$CREPRG_SPRCNAM =ORION ,IMAGE = FASTCOMP

;CREATE ORION - HE'LL
;SLEEP
;BRANCH IF SERVICE ERROR
;CONTINUE

BLBC

RO ,ERROR

$WAKE_S
BLBC

PRCNAM =ORION
RO,ERROR

;WAKE ORION
;BRANCH IF SERV ICE ERROR

$WAKE_S
BLBC

PRCNAM =ORION
RO, ERROR

;WAKE ORION AGAIN
;BRANCH IF SERVICE ERROR

.WORD
$HIBER_S

BLBC

RO,ERROR

;ENTRY MASK
;SLEEP 'TIL GEMINI WAKES
;ME
;BRANCH IF SERVICE ERROR
;PERFORM ...

BAB

10$

Process: ORION
FASTCOMP :

2
10$

;BACK TO SLEEP

Notes :
1. Process GEMINI creates the process ORION , specifying the image name FASTCOMP .

2. The image FASTCOMP is initialized , and OR ION issues the $HISER system service.
3. At an appropriate time, GEMINI issues a $WAKE request for ORION. ORION continues execution following the $HISER service call.
When it finishes its job , it loops back to repeat the $HISER call and to wait for another wake.

Figure 6-3
Program Using Hibernate/Wake System Services

number of common event flag clusters. Each cluster contains 32 flags . The flags can be assigned any meaning for
the processes in the group.

ating system provides process communication facilities for
synchronizing execution, for sending messages, and for
sharing common data. The three techniques that cooperating processes can use to communicate are :

Each process in a group can associate with up to two of its
group's common event flag clusters at one time. A process
can read , set , clear, or wait for common event flags to be
set. The ability to read, set, or clear event flags is
controlled by the protection code and User Identification
Code assigned to the common event flag cluster.

• common event flags
• mailboxes
• shared areas of memory
Common event flags are available by group association to
processes with in jobs. Mailboxes and shared areas of
memory are more general purpose fac ilities which can be
limited or unlim ited in scope. They can be limited to a specific member family within a group or to a specific group of
jobs, or they can be extended to all jobs in the system.

Common event flag clusters can also be used by cooperating processes on different processors in a multi port memory configuration.

Mailboxes
A mailbox is a record-oriented virtual 1/0 device created
by a process . Mailboxes can be used to pass status information , return codes, messages , or any other data from
one process to another. A process can protect its mailboxes from read and/or write access by any process outside

Common Event Flags
In addition to the local event flags available to each process, cooperating processes can commun icate using common event flags . Every group in the system can define any
6-13

shared memory for access from multiple processors connected to the multiport memory. All three of the interprocess communication facilities, i.e., common event flags ,
mailboxes, and global sections, may reside in multiport
memory, thus providing interprocessor communication facilities. Through the use of logical names, common event
flags, mailboxes, and global sections may be placed either
in local or multiport memory, transparently to the program . Common event flags, mailboxes, and global sections are the communication facilities permitted in mult iport memory configurations.

its member family or outside its group . Mailboxes can also
be used by processes to communicate with other
processes on different processors in a multiport memory
configuration.
All of the 1/0 system services and record management
services can be applied to mailboxes. Other processes
write messages to a process' mailbox by queuing write requests for the device. A process reads messages in its
mailbox by queuing read requests for the device. A process can request AST notification when anything is written
to its mailboxes, and it can assign mailboxes logical
names.

MEMORY MANAGEMENT
In a multiprogramming system , many processes coexist
simultaneously in main memory. The system switches
between these processes, giving each some time to execute . In most multiprogramming environments , however,
the number, size , and kind of concurrently executing
processes change rapidly, while the amount of memory
available for processes remains constant. Users log on
and off the system, production activities vary periodically ,
and special production jobs occur. Since it is generally
inefficient to have available the maximum amount of memory that might ever be needed at one time, it becomes the
task of the operating system to provide a dynamic memory
that responds to the changing multiprogramming environment.

Shared Areas of Memory
The system supports a high degree of code and data sharing through the use of global sections. A global section is a
copy of all or a portion of an image or data file that can be
mapped in a process virtual address space at run time .
Global sections can be used for shared data structures, as
communication regions for cooperating processes , or they
can be used simply to eliminate multiple copies of image
code or data.
Global sections can be created dynamically by a process
or they can be permanently present in the system . Dynamically created global sections are mapped into processes that reference them, and deleted when no more
references are made to them . Permanent global sections
are created by a sufficiently privileged process, and remain until they are explicitly deleted . They are loaded into
and removed from memory dynamically as references are
made to them.

VAX/VMS uses two interdependent complementary techniques to allocate limited memory to competing processes: paging and swapping . These techniques relieve the
general programmer of concern for memory allocation
while still allowing system programmers to optimize program performance in limited configurations . This section
and the following section on scheduling discuss how this
system 's paging and swapping techniques extend limited
memory resources with minimum effect on the system or
programs when the system has sufficient memory to hold
all concurrently executing processes.

Each process that maps a global section into its virtual address space can have a different access privilege to a global section . When a global section is created, it is assigned
a User Identification Code (UIC) identifying the group and
member family to which the global section belongs, and a
protection code identifying the read and write access
privileges of processes in the system. Global sections can
be shared by all processes in the system, or shared only
by processes within a particular group, or shared only by
processes within a particular job. One or more controlling
processes can have write privileges while other processes
in the system , group, or job have only read privileges.

Mapping Processes into Memory
The operating system 's memory management software is
responsible for creating and maintaining the information
used to map the virtual addresses used in a program to
physical memory addresses. The unit mapped is the page ,
which is a block of 512 contiguous byte locations in physical memory.

A process can map to a global section explicitly by issuing
a Map Global Section system service, or it can be mapped
implicitly by referring to a shareable image. If an image
references a shareable image, the linker does not normally
include the shareable image in the image. The shareable
image is installed as a global section or set of global sections . When the image is executed , the image activator
calls the Map Global Section system service on behalf of
the image. For example, the Common Run Time Procedure Library is a shareable image consisting of library
procedures that is mapped as a system-wide permanent
global section. The use of permanent global sections significantly reduces the size of programs using common library procedures and the overall system memory requirements.

Virtual addresses are also grouped into 512-byte pages,
and each page of virtual addresses can be mapped to a
page of real memory locations. Any number of virtual
pages can be mapped to one physical page. Unlike systems that partition or statically allocate portions of physical
memory, this system dynamically allocates physical memory, with the result that pages of a process may be scattered anywhere throughout memory. It is never the concern of the programmer to determine how physica l
memory is allocated. To illustrate this, Figure 6-4 shows
how two processes might be mapped into physical memory .

Interprocessor Communication Facility
VAX/VMS support for the multiport memory subsystem
means that both user data and subroutines may reside in

When a process is created, the operating system sets up
its mapping information, called page tables. Each process
has its own page tables mapped by system region virtual
6-14

Process A

Phys ical Mem o ry

Worki ng Set

Pag es

software uses the linker's descriptions of code and data
areas to map image virtual pages to physical pages. The
operating system 's image activator builds the image's
mapping information in the process' page tables.

Pages
';y

'"l'-

3
4

I
I

A51

I
I
I

A3 7

9
32

830

A39

I
L
I

\Glob"

Section

A4 and 8 17

A1and814

A3 and816

At any one time, some of the pages of process virtual
memory may be mapped to disk and some to physical
memory. The physical memory requirement of a process
is the process working set. When a process is executing,
a process working set consists of all the pages of a process ' virtual memory residing in physical memory that the
process can directly access without incurring a page fault,
plus any actively used portions of the process page tables
and process header information .

A2 and 81 5

Process Virtual Memory and Working Set
The total virtual memory requirement of a process is called
process virtual memory. Process virtual memory consists
of all the pages of the process program region and control
region which are mapped by the process page tables .

1""' "

The working set is a dynamic characteristic of a process
that has both min imum and maximum size limits. The system designates a required minimum number of pages that
has to be in a process working set, and the system manager defines the maximum number of pages allowed in any
one job's working set in the user authorization file . The size
of a process working set affects its paging and swapping
performance , as well as affecting the number of process
working sets that can be resident when the process working set is resident.

A43
A44

Pr ocess 8
820

Work ing Set
Pages

I
I

819

I
I

A42

822

I
I
I

16
17

A process may increase or decrease its working set, within
the authorized limits, through the use of command
language commands or system service calls .

A32

I
I

81 2

Under version 2.0 of VAX/VMS , working set size adjustments are made automatically by the operating system.
This facility, when enabled by the system manager, monitors the page fault rate of a process and automatically increases or decreases the working set (again within authorized limits) to optimize performance and memory usage.
This automatic adjustment provides a more immediate response in system reaction/performance.

811

831

,..{/

Paging
Through its pag ing technique, the operating system can
execute programs that are too large to fit in the amount of
physical memory allocated to a process, without requiring
the programmer to define overlays. Inactive portions of a
program are automatically stored on disk while the active
portions are resident in memory. When the program references a disk-resident portion of the program, the operating system reads in , or pages in, the referenced portion,
moving out other portions of the program to disk if necessary . This system 's paging technique has several features
that distinguish it from other techniques:

Figure 6-4
Mapping Processes into Memory

addresses. (Refer to the Processor section on memory
management for a complete description .) Initially, a process page table simply maps those pages of the control region that define the permanent process context.

• clustering, or the ability to read in several pages at one
time

When a program is linked , the addresses the linker assigns in the image are always virtual addresses . The linker
has no knowledge of how physical memory is allocated . Its
primary function is to build descriptions of the size and
protection requirements of the program 's code and data
areas.

• paging processes against themselves, not against the
entire system
• maintaining an available page pool from which processes can recover recently discarded pages without incurring disk 1/0

When an image is executed , the memory management

6-15

• writing back to disk only the modified pages that are released from a process working set and only writing them
when several have accumulated
• activating a process waiting for page fault 1/0 to execute
AST routines when they are delivered
When the operating system activates an image for the first
time, a number of pages are read into memory from the
image file on disk. The number of pages read in the first
time can be controlled by a cluster factor the programmer
can assign optionally per image. The ability to read in several pages at once allows the image to execute for some
time without incurring page faults, and provides
significantly improved responsiveness in starting programs.

If all the pages are not read in initially, at some point the
image will reference the pages that have not been read in .
At that time, the process incurs a page fault, that is, a
reference to a page not mapped in the process working
set.
The operating system's pager is a condition handler that
executes when a process incurs a page fault. If the working
set size limit has not yet been reached, the pager reads in
the faulted page from disk, plus any additional pages,
again according to a cluster factor for that section of the
image.
If a page fault occurs when the working set size limit is
reached , the pager obtains a page from a pool of available
pages to read in the faulted page, and releases the least
recently faulted page from the process working set into the
pool and writes it to disk if it has been modified. Figure 6-5
illustrates two different size working sets for a process running the same program in each case. The illustration
shows the order in which the pages were faulted. (Refer to
Figure 6-2 to see how the pages appear in virtual address
space.)

A process is subsequently paged only when it executes an
image that needs more pages than the process is allowed
to have in its working set. If the number of pages in the image plus the number of pages for the remainder of the
process is less than the working set size limit, all the pages
are read in and the process is never paged.

7FFEFEOO
7FFEECOO
7FFEEAOO
7FFEE800
8008BOOO
8008B200
8008B400
7FFE8AOO
7FFE8600
8008F200
7FFE8400
7FFE8200
7FFEFCOO
7FFEF800
7FFE9200
8008B600
7FFE7800
7FFE9400
7FFEE400
7FFE4600
7FFE6800
7FFEFOOO
7FFE4800
7FFE4EOO
7FFE5000
7FFEFAOO
7FFE3EOO
7FFE1COO
00001800
00001000
00001600
00001400
OOOOOEOO
00001AOO
00001200

Pointers
Kernel
Stack
Process
Header
Command Interpreter Code
Control Re ion Pa e Table
Command Interpreter Code

7FFEFEOO
7FFEECOO
7FFEEAOO
7FFEE800

Pointers
Kernel
Stack

8008BOOO
8008B200
8008B400
8008B600
8008F200
7FFEFCOO
7FFEF800
7FFE1COO
00001800
00001000
00001600
00001400
OOOOOEOO
00001AOO
00001200
7FFE3EOO
7FFEE400
00000600
00000800
7FFE4600
7FFE5000

Process
Header

Supervisor Stack
Command Interpreter Code
Program Region Page Table
Command Interpreter Data
Command Interpreter Code
Image Header Buffer
Process 1/ 0 Segment
Process Logical Name Table
Executive Stack
Process 1/ 0
Segment
Supervisor Stack
1/ 0 Channels
User Stack

USER
PROGRAM

00000600
00000200
00000400

ooooocoo
Figure 6-5
Process Working Sets

6-16

Program Region Page Table
Control Region Page Table
Supervisor Stack
User Stack

USER
PROGRAM

1/ 0 Channels
Image Header Buffer
USER PROGRAM

Process 1/ 0 Segment

The pager pages a process only against itself. It does not
release pages of one process to satisfy another's needs.
This ensures that only those processes that need paging
are affected by paging . Other processes in the system
need not be affected by another process's memory requirements.
The list of available pages works as a cache of pages that
effectively extends a working set size above its limit when
few processes are competing for memory resources and
there are many pages in the list. If a process faults a page
that was released and is still in the list, the page does not
have to be read in from disk, it is simply taken from the list
and remapped into the working set.
When a page is released , it is placed on one of two lists:
the free page list or the modified page list. Modified pages
are pages the process has written into and , if they need to
be added to the free page list to be used by another process , must be saved on disk. Modified pages are only written to disk when the modified page list exceeds a threshold size or when an image's execution is terminated and
the files containing the modified pages must be closed .
When modified pages are written , they are writen in clusters to increase system performance.
Virtual Memory Programming
The processor provides the programmer with a large virtual address space and rapid address translation , and the
operating system provides the programmer with extremely
efficient mapping and paging algorithms. Furthermore,
these memory management mechanisms are totally transparent to the application programmer. It is not necessary
for a programmer to be concerned with address allocation
or page mapping: the high-level language compilers and
the linker take advantage of the memory management
mechanisms to set up the memory allocation optimal for
most programming requirements.

For those systems with limited memory or special processing requirements , however, this system enables users to
control and optimize memory management. The system
manager can control the memory allocation requirements
of the system as a whole by initialization parameters such
as desired and minimum acceptable number of available
pages, and of individual jobs by user authorization parameters such as paging file usage limit and maximum working set size. It is possible to have a process avoid paging
entirely by making its working set size equal to its virtual
memory requirements, or to reduce paging by choosing a
working set size that satisfies the average demand for
pages over time.

ing data structures as if they were disk-resident files. In
general , the programmer need not be concerned with data
structures such as tables and arrays whose elements are
virtually contiguous and sequentially processed.
Large data structures that are randomly accessed can,
however, be optimized . For example, processing down a
linked chain in which the chain elements are spaced far
apart with no useful data in between requires that an image reference a large number of pages in a short period of
time . If all of the pages cannot fit in the process working
set at the same time , the references to successive chain
elements will incur disk 1/0.
On the other hand , a large data structure can be efficiently
accessed using directory trees, where a page or set of
consecutive virtual pages contains all one kind of information . One page can contain all of the information that
points to randomly arranged, but virtually contiguous,
pages containing the data processed at that locality.
VAX/VMS Memory Management Services
For those who have special processing requirements,
there are system services that control memory management within the quotas and limits assigned by the system
manager. These memory management system services
enable a process to :

• modify the working set size limit
• add or delete pages from process virtual memory
• expand or contract the program region or control region
• lock pages in the working set
• lock pages in physical memory
A program can impose a limit on process working set size
anywhere between the minimum required by the system
and the maximum specified by the system manager. The
limit can be adjusted in accordance with program behavior and real-time requirements. By maintaining the smallest working set size consistent with an acceptable paging
rate, a program that temporarily requires a large working
set can reduce its impact on the system. For example , a
process control program or simulator might use a small
working set while processing interactive initialization commands. Once real-time processing is underway, the program can expand its process working set size to reduce
paging . When real-time processing is finished, the program can contract the working set.
A process can add selected pages to and delete selected
pages from its virtual memory dynamically. Deleting a
page is in effect saying that the image is no longer going to
use those virtual addresses, and the operating system
does not need to map them to pages in virtual memory.
Deleting read/write pages (such as those used for interprocess communication) as soon as they are no longer
used eliminates the need for the system to write them out
as modified pages to a paging file. When an image has
reached its paging file quota, it can delete pages in order
to map other pages in its virtual address space.

The programmer also has the ability to control memory allocation for images in two ways: through properly coded
programs and through the memory management system
services. This system 's memory management software optimizes for program locality. Programs that incur paging
infrequently are those in which the code and data used
during each stage of processing are contained in the fewest possible number of virtually contiguous pages.

A process can request an extension to the amount of virtual memory allocated to its program region. The operating
system will map zero-filled pages into the process virtual
address space following the highest addressed page allocated for the program region. This service is useful for dynamically creating data arrays whose size is not known be-

For the most part, the linker allocates virtual addresses so
that images require the minimum mapping information
possible. The programmer can also ensure that images
that process large data structures require the least possible mapping information and potential paging by organiz-

6-17

process is waiting . Whenever a system event occurs , the
scheduler adjusts the process state queues accordingly .
For example, the scheduler adds a process to the exec utable state queue when a resource for which it is waiting becomes available , or removes it when it requests an event
or resource for which it must wait.

forehand , and it eliminates the need for allocating a data
area in a program image. A process can also extend the
initial allocation of pages for the user stack by requesting
th8 operating system to map zero-filled pages into process
virtual address space preceding the lowest addressed
page allocated for the control region. In Version 2.0 of
VAX/VMS , the system will automatically extend the stack if
the process references unmapped addresses in the control region .

The executable state queue supplies the scheduler with a
list of processes that are eligible to execute. Priority determines which process among those eligible executes. Rescheduling occurs when a system event makes executable
a process with higher priority than the one current ly
executing .

In unusual situations , a process can lock pages in its working set. Locking a page in the working set is useful when a
process does not reference a particular page regularly ,
but the page needs to be in the working set to increase the
performance of the code in that page . For example , it
might be desirable to keep the page containing asynchronous system trap routines in a working set to ensure that
the routines are started up rapidly when an AST is delivered . Note, however, that locking a page in the working set
causes other pages to be paged more frequently, since the
page will not be paged out, no matter how long it has been
in the working set. A page can be unlocked when it is no
longer necessary to keep it in the working set.

Unlike timeshared scheduling, therefore , event-driven
scheduling is based on the activities of the processes
themselves, not on a time limit imposed by the schedule r.
Because scheduling intervals are determined by system
events , the interval between rescheduling is random.
Quantum keeping and requested timer events provide a
minimum level of event activity but, in practice, the average interval between events is determined by the duration
of the typical 1/0 operation .
Priority: Real-Time and Normal Processes
The scheduler recognizes 32 scheduling priorities , where
priority 31 is high and 0 is low. Priorities 31-16 are for real time processes, and priorities 15-0 are for normal processes. When a process is created, the system assigns it a
scheduling priority. A program image that the process executes can modify the process priority using a system service . The system manager grants jobs the privilege to
execute at real-time priorities.

It is also possible to lock pages in memory. A page locked
in memory is not only locked in the working set, it is not
swapped out with the process. This service is useful for
real-time processes that need to keep buffers in memory
for 1/ 0 transfers.
PROCESS SCHEDULING
VAX/VMS features event-driven scheduling based on
process priority. Unlike traditional timeshared scheduling
systems , this system 's ability to respond to events enables
it to dispatch real-time processes efficiently as well as to
share processing time among normal processes competing for resources . Furthermore, priority assignment enables the user to bias processor time allocation based on
process activity, to bias the allocation absolutely forcertain processes, or to mix both allocation methods.

The scheduler maintains a queue for each scheduling priority . Processes having the same priority are listed in the
same queue. The priority assigned to a process when it is
created is its base priority. The scheduler does not alter
the priority of a real-time process during execution . The
scheduler may temporarily increase the priority of a normal process during its execution, but its priority never
drops below its base priority.

The operating system's scheduler and swapper are responsible for ensuring that the processes executing in the
system receive processor time commensurate with their
priority, which is controlled by assignment, and with their
ability to execute, which is controlled by system events.

Scheduling by strict priority for real-time processes and by
potentially modifying priority for normal processes allows
the scheduler to achieve maximum overlap of compute
and 1/0 activities while still remaining responsive to highpriority real-time applications.

System Events and Process States
In VAX/VMS , dispatching a process for execution involves
little decision making. The selected process is always the
highest priority executable process. The real scheduling
decisions are made as the result of system events that
make processes executable.

Scheduling Real-Time Processes
When a system event occurs that makes a real-time process eligible to execute, it receives control of the processor
unless another higher priority process is currently executing . A real-time process retains control of the processor
until it finishes execution , enters a wait state, or is preempted by a higher priority process. (Note that under
VAX/VMS , real-time processes actually have a higher pri ority than system processes, thus ensuring that real-time
processing will never be encumbered by system overhead .)

A system event is an event that affects the ability of a process in the system to execute. System events include events
external to the process currently executing , such as 1/0
completion or timer interrupt. System events also include
events internal to the process currently executing . The
process may issue a wait request or a hibernate request,
or it may request or release a system resource, for example, a page of memory.

A higher priority real-time process can pre-empt any lower
priority process whenever a system event occurs that
makes it eligible to execute. For example, a device interrupt may occur that signals the completion of an 1/0 transfer requested by the higher priority real-time process.

Every active process in the system is listed in one of several state queues that identifies whether or not a process is
executable, and if not, the event or resource for which the

When a real-time process is pre-empted to dispatch a

6-18

process of higher priority, the pre-empted process is
placed at the end of its priority queue. This rotates processes within a priority , with the result that available processor time is distributed among processes of the same priority.

and is limited to moving few pages in and out of memory,
swapping balances the memory requirements of the system as a whole.
The swapper is activated whenever a system event occurs
that can make a nonresident process resident, a nonresident process executable, or a resident process nonexecutable. For example, a resident process might release
sufficient memory to enable the swapper to move in a nonresident process. An 1/0 completion event might make a
nonresident process executable. A resident process might
enter a wait state and become nonexecutable. In any case,
the swapper uses three conditions to determine which
processes should be swapped in and which should be
swapped out:

Scheduling Normal Processes
When no real-time processes are executing, the scheduler
distributes processor time among the processes on the
normal priority levels. As with real-time processes, the
scheduler selects the highest priority ready-to-execute
normal process . That process executes until it finishes execution, enters a wait state, or is pre-empted by a higher
priority process. Unlike real-time process scheduling ,
however, the scheduler modifies normal process priority
whenever a system event occurs for a normal process and
whenever a normal process is scheduled .

• which processes are executable and which are not (and
the reason for the wait state)
• what the process priorities are

When a system event occurs that affects a normal process,
the scheduler increases the priority of the normal process
(but not to more than the maximum priority of 15) and
places the process at the tail of the queue for its new priority. The amount of priority increment depends on the nature of the event. For example, the scheduler increases the
priority of a normal process on the following events:

• whether a process balance set quantum has expired
The balance set quantum effectively enforces a swapping
rotation for compute-bound normal processes. Every normal process is assigned a time quantum that provides a
guaranteed minimum amount of time in which the process
can perform useful work before it is eligible to be swapped
out of the balance set. A process can be pre-empted many
times before it has received its full quantum. It remains in
the balance set until it completes its first quantum unless a
real-time process that is swapped out becomes executable and no other processes can be swapped out to make
room for the real-time process.

• term inal input completed
• terminal output completed
• resource available
• wake, resume, delete request received
• nonterminal 1/0 completion, page fault completion, or
other event

VAXNMS Process Control Services
In addition to the VAX/VMS system services that enable
processes to create, delete, suspend, resume, and wake
other processes , or to hibernate and wake themselves, a
process can control the manner in which it is scheduled
by:

In this case, the terminal 1/0 events receive the highest priority increments to enable the system to be most responsive to the interactive terminal user. When the scheduler
increases a normal process's priority, that process gets
control of the processor if its new priority is higher than
that of the process currently executing.

• setting process swap mode

Each time a normal process is scheduled , the scheduler
decreases its priority by one (unless it is already in its base
priority queue) and places it at the end of that priority
queue. The effect of dynamically increasing and decreasing normal process priority ensures maximum overlap of
computation and 1/0.

• setting resource wait mode
A suitably privileged process can request that it not be
swapped out of the balance set, even when it becomes inactive. This is useful for high priority real-time processes
that need to be activated rapidly when they become executable.

Swapping and the Balance Set
It is the job of the swapper to keep the scheduler supplied
with the highest priority executable processes in configurations that do not have a sufficient amount of physical
memory to keep all process working sets memory-resident. The balance set is the set of all process working sets
that are currently in memory. The swapper ensures that
the balance set always contains the highest priority executable processes by moving low priority or nonexecutable
memory resident process working sets to a swap area on
disk , and moving high priority or executable process
working sets into memory.

Normally, when a process requires dynamic resources of
the system and they are not available, the process enters a
wait state until the resources become available. Dynamic
resources primarily include the buffer space needed for
mailboxes, 1/0 requests, etc. A process can request to be
notified when resources are not available and take alternative action instead of entering a wait state.
1/0 PROCESSING
The 1/0 processing system consists of several modular, interdependent components that enable programmers to
choose the programming interface and processing
method appropriate for their needs, without incurring run
time space or performance overhead for features not
used . In addition, the 1/0 request processing software
takes advantage of the hardware's ability to overlap 1/0
transfers with computation, switch contexts rapidly, and

Swapping is a very efficient way of extending limited memory resources when many processes are executing concurrently. Process working sets for small processes (less
than 64K bytes or 128 pages) can be swapped in and out
of memory in one disk 1/0 operation . Where paging extends limited memory resources on a per-process basis

6-19

generate interrupts on multiple priority levels to ensure the
maximum possible data throughput and interrupt re sponse . Figure 6-6 presents an overview of the major 1/0
processing system components and their relationsh ips.

The 1/ 0 system services provide both device-independ ent
and device-dependent programming . Users perform their
own record blocking on file-structured and non-file-structured devices. Virtual block addressing is used on Fil es-11
d isk or ANSI magnetic tape volumes . In add ition , users
with sufficient privilege can perform 1/ 0 operations usin g
either logical or physical block addressing for defining
their own file structures and accessing methods o n disk
and magnetic tape volumes .

Programming Interfaces
The 1/ 0 programming tools are the record management
system , VAX-11 RMS , for general purpose file and record
processing , and the Queue 1/0 system services , for direct
1/0 processing . Table 6-2 summarizes the programming
interfaces .

The 1/ 0 system services also provide device-dependent
programming of devices not supported by RMS , such as
real-time interfaces.

RMS (refer to the Data Management Facilities Section )
provides device-independent access to file-structured 1/0
devices . The most general purpose type of access enables
programs to process logical records; RMS automatically
provides record blocking and unblocking .

Ancillary Control Processes
Both RMS and the 1/0 system services use the same 1/ 0
control processes, called ancillary control processes
(ACPs) , for processing file- structured 1/ 0 requests . An
ACP provides file structuring and volume access control
for a particular type of device. Typical ACP fun cti ons
would include creating a directory entry or file, accessin g
or deaccessing a file , modifying file attributes, and deleting a directory entry or file header. There are three kin ds
of ACPs provided in the system : Files-11 disk, ANS I
magnetic tape, and network commun ications link .

RMS users can also choose to perform their own record
blocking on file-structured volumes such as disk and magnetic tape, either to control buffer allocation or optimize
special record processing . Users performing their own
record blocking address blocks using a virtual block number (which is the number of the block relative to the file being processed) for volume-independent processing .

USER
PROGRAM

OPERATING SYSTEM
Procedures

Processes

RECORD MANAGEMENT
SERVICES

-- - ~
file and
record

RMS OP EN, CLOSE
GET , PUT

OR
OR

ANCILLARY
CONTROL
PROCESS

RMS OPEN , CLOSE
and $0 10

file
access

- -- --1

$010
Physical Block
Device Relative

file-structured or
non-file-structured

DRIVER
FORK
PROCESS

- - - --j

1/ 0 SYSTEM
SERVICES

t!J7l7lA RMS provides
V//!i!JI record blocking / unblocking

user does own
record blocking and unblocking

Figure 6-6
User Interfaces to 1/0 Services

6-20

Disk .
Magtape, or
Networks

Table 6-2

1/0 Processing Interfaces

METHOD

PROGRAM
INTERFACE

1/0 COMPONENTS

PURPOSE

Record 1/0

RMS requests

RMS, ACP and
Driver

Use Files-11 disk or ANS
magtape file structure, use RMS
record access methods

File 1/ 0

RMS OPEN and
$QIO requests

RMS for OPEN,
ACP and Driver

Use Files-11 disk or ANS
magtape file structure,
implement own record access
methods

Device 1/0

$QIO requests

Driver

Fast dumps to disk or magnetic
tape , foreign file structure

The RMS and 1/0 system services programming interfaces
are the same regardless of the ACP involved, but since
ACPs are particular to a device type, they do not have to
be present in the system if the device is not present. There
is one network ACP process for all DECnet network communications links in the system, and none if the system is
not in a network.

These include both MASSBUS and UNIBUS devices. In
addition , the user can write additional drivers for non-standard UNIBUS devices.

1/0 Request Processing
All 1/0 requests are generated by a Queue 1/0 (QIO) Request system service. If a program requests RMS procedures, RMS issues the Queue 1/0 Request system service on the program 's behalf. Queue 1/0 Request processing is extremely rapid because the system can:

Device Drivers
Once the ACP sets up the information for file-structured
1/0 requests, a request can be passed on to a device driver. All non-file-structured 1/0 requests are passed directly
to a device driver.
A VAX/VMS driver:

• keep each device unit as busy as possible by minimizing
the code that must be executed to initiate requests and
post request completion

• defines the peripheral device for the rest of the
VAX/VMS operating system

• keep each disk controller as busy as possible by
overlapping seeks with 1/0 transfers

• defines the driver for the operating system procedure
that maps and loads the driver and its device data base
into system virtual memory

The processor's many interrupt priority levels improve interrupt response because they enable the software to have
the minimum amount of code executing at high priority
levels by using low priority levels for code handling request verification and completion notification. In addition,
device drivers take advantage of the processor's ability to
overlap execution with 1/0 by enabling processes to execute between the initiation of a request and its completion.
User processes can queue requests to a driver at any time,
and the driver immediately initiates the next request in its
queue upon receiving an 1/0 completion interrupt.

• initializes the device (and/or its controller) at system
startup time and after a power failure
• translates software requests for 1/0 operations into device-specific commands
• activates the device
• responds to hardware interrupts generated by the device
• reports device errors

All access validation and checking takes place before an
1/0 request is actually queued. For file-structured 1/0 requests, the Queue 1/0 Request system service obtains all
the virtual block mapping and volume access checking information from the ACP or directly from tables created by
the ACP . For example, on virtual block 1/0 requests for
multivolume files, the system service obtains from the
ACP's tables the mapping information that enables it to
queue requests to different drivers when the user's 1/0 request involves a transfer that spans volumes . The Queue
1/0 Request system service also checks the validity of the
function requested (read, write, rewind, etc.) for the particular device. Because all access validation and function
checking is performed before the request is queued, the
driver has little to do to initiate a request.

• returns data and status from the device to software
Device drivers work in conjunction with the VAX/VMS operating system. The operating system performs all 1/0
processing that is unaffected by the particular specifications of the target device (i.e., device-independent) processing. When details of an 1/0 operation need to be translated into terms recognizable by a specific type of device,
the operating system transfers control to a device driver
(i.e., device-dependent processing) . Since different peripheral devices expect different commands and setups,
each type of device on VAX/VMS requires its own supporting driver.
The VAX/VMS operating system contains device drivers
for a number of standard DIGITAL-supported devices.
6-21

IV IV AX to RSX, or PDP-11 MACRO program can be executed in compatibili ty mode, provided that it is first linked
by the RSX-11 M Ve rsion 3.2 task builder and that the resulting task image meets the following requirements:

Once the system service has verified the 1/0 request, it
raises the interrupt priority level to that of the driver. The
only activity it has to perform at this level is a test to see if
the driver is busy. If the driver is not busy, it calls the driver . Otherwise , it queues the request according to the priority of the requesting process and immediately returns to
the user process .

• it must not execute PDP-11 privileged instructions
• it must have been built for a mapped system
• it must not depend on 32-word memory granularity

When the driver is called, it initiates the request and returns to the user process. Because disk seeks do not require the controller once they are initiated , if a disk driver
receives a seek request and the controller is currently
busy with an 1/0 transfer request on some other disk unit,
the driver queues the request so that the controller will initiate the seek request before any pending 1/0 transfers
when it has finished the current transfer .

• it must not use the privileges that enable it to map into
the executive or 1/0 page
• it must not use the PLAS (program logical address
space) executive directives
• it must not rely on environmental features of RSX-11 M
that VAX/VMS does not support , e.g ., partitioning or
significant events

When the device subsequently generates its interrupt at
the hardware interrupt priority level , the interrupt dispatcher calls the appropriate interrupt service routine. An
interrupt service routine simply saves the device control / status registers , requests a software interrupt at the
driver's interrupt priority level , and returns to the interrupt
dispatcher which is then free to scan for unit attentions.
Because a disk controller cannot generate interrupts on
any unit performing a seek until the current transfer completes, the interrupt dispatcher will also dispatch seek
completion when dispatching a disk 1/0 transfer completion interrupt.

• it must not use DECnet
The task can be privileged to issue directives other than
memory management directives-direct volume access
using the QIO request executive directive, for example.
IAS or RSX- 11 D tasks that meet these requirements can
also be executed . They must first be built with the RSX11 M Version 3.2 task builder. For programs that do not
meet these requirements, VAX/VMS provides the program development utilities (for example , the MACRO assembler and the task builder) for modifying programs to
execute in compatibility mode.
For most RSX-11 M executive directives, the native mode
operating system executes a functionally equivalent system service. In most cases , the system service duplicates
the function . For example:

When the driver receives the completion interrupt, it prepares the 1/0 completion status for the requester , and requests a software interrupt. The driver is then free to process another request in its queue and , if the queue is not
empty , the driver begins again . All 1/0 completion notification takes place outside the driver , minimizing the
interrequest idle time. The 1/0 post routine notifies the
process of 1/0 completion and releases or unlocks buffers.

• A checkpoint enable/disable directive is interpreted as
the set swap mode system service .
• The send/receive directives are translated into mailbox
write/read system services. Native mode and compatibility mode images can communicate using mailboxes.
• The event flag directives are for the most part identical.
Native mode and compatibility mode images can
communicate using common event flags , provided they
are in the same group.

COMPATIBILITY MODE OPERATING ENVIRONMENT
The processor can execute user mode PDP-11 instruction
streams in the context of a process. The operating system
supplements this feature by substituting its functionally
equivalent system services for many of the RSX-11 Moperating system executive directives that user mode tasks
may call. This enables the system to execute such non-privileged RSX-11 M task images as:

• A Logical Unit Number (LUN) assignment directive is interpreted as a channel assignment for the appropriate
device.
In some cases the operating system cannot duplicate the
function , but it does what it can to let a program continue .
For example:

• the PDP-11 MACRO assembler
• the PDP-11 FORTRAN IV/VAX to RSX compiler

• A task image is allowed to declare a significant event ,
but the directive is ignored .

• the PDP-11 BASIC-PLUS-2/VAX compiler
• the RSX-11 M program development and file management utilities, including the task builder, text editor , etc.
In addition, the operating system supports the RMS-11
and RMS-11 K record management services procedures
for compatibility mode programs. Program and data files
can therefore be transported between VAX and RSX systems.
The operating system also supports the RSX-11 M Monitor
Console Routine (MCA) commands, either typed directly
on a terminal, or submitted as indirect command files.
User Programming Considerations
Any PDP-11 BASIC-PLUS-2/VAX, PDP-11 FORTRAN
6-22

• A set priority directive is ignored , since the scheduling
priority ranges are different. To run at a given priority ,
the image must be run in the context of a process given
that priority.
For the most part, however, many RSX-11 M and
VAX/VMS program environment characteristics correspond . For example, tasks can hibernate, receive asyn chronous system traps, and schedule wake requests . Synchronous system trap routines can be declared as condition handlers for trace traps, breakpoint traps, illega l
instruction traps, memory protection violations , and odd
address errors.

tion in ODS-2. Wh ile reading files stored on ODS-1 volumes, therefore , this protection field is ignored .

File System and Data Management
Both RSX- 11M and VAX/VMS recogn ize User Identification Codes as a protection mechanism. UICs provide the
defau lt user file directory in RSX-11 M systems , while , in
VAX/VMS , a UIC is not necessarily associated with an account name or default directory name. UIC- based file protection , however, is much the same in both systems. That
is, it is used in determining read , write , and delete privileges for system , owner , group, and world .

Command Languages
VAX/VMS users can select the MCA command interpreter,
which allows them to execute a language (VAX/VMS MCA)
that is similar to the RSX-11 M MCA command language.
Selecting the MCA command interpreter allows the
VAX/VMS user to perform the following :
• Run RSX-11 M images and VAX-11 images.

Tasks may use any of the RSX data management services
includ ing File Control Services (FCS) , RMS-11, and RMS11 K. Special versions of FCS and RMS- 11 /RMS-11 K are
supplied with VAX/VMS. A compatibility mode task built
on VAX/VMS is thus provided with the full file naming capabil ities of VAX/VMS , including logical names and multilevel d irectories. However, update of a file by multiple
tasks is not supported .

• Use VAX/VMS components for native program development, for example , VAX-11 MACRO or the linker.

Both magnetic tape and Files-11 disk volumes can be
transported between systems. VAX/VMS can read and
write both Files-11 Level 1 disk structures (ODS-1) and the
Level 2 disk structures (ODS-2). The Extend access protection field in ODS-1 is used for Execute access protec-

VAX/VMS will associate the MCA command interpreter
with a process, if " MCA " is the default command interpreter named in the user's authorization file entry or if the user
specifies /C LI= MCA following "username" in the LOGIN
statement (overrid ing the default).

• Use RSX-11 M components for RSX-11 M program development, for example, MACR0-11 or the task builder .

• Execute RSX - 11 M indirect command files. (The
VAX/VMS user can use this facility to execute those files
required for RSX-11 M or RSX-11 S system generation .)

6-23

7
The
Languages
'

,,,,,,,,,,,,,,j.·

111111

1'11111 '

···"

VAX/VMS includes a complete program development environment for
a wide range of languages. In addition to the native assembly language,
VAX/VMS offers many optional high-level programming languages
commonly used in developing both scientific and commercial applications: FORTRAN , COBOL, BASIC , PL/I , PASCAL, CORAL 66 , and
BLISS-32. It provides the tools necessary to write, assemble or com pile, and link programs, as well as build libraries of source, object, and
image modules .
Programmers can use the system for development while production is
in progress. They can interact with the system on-line , execute command procedures, or submit command procedures as batch jobs. Novice programmers can learn the system quickly because the command
language accepts standard defaults for invoking the editors, compilers,
and linker. Experienced programmers will appreciate the flexibility and
control each tool offers.

INTRODUCTION

FORTRAN or LINK commands (but it can be suppressed
by specifying NOTRACE with the LINK command) .

VAX/VMS provides a complete program development environment. In addition to the assembly language, MACRO,
it offers the optional higher level languages commonly
needed in engineering and scientific, commercial , instructional, and implementation applications-FORTRAN, COBOL, BASIC , PL/I, PASCAL, CORAL 66, and BLISS-32.
VAX/VMS provides the tools to write, assemble or compile, and link programs, as well as to build libraries of
source , object, and image modules. User applications may
employ more than one language, and the ability of languages to call one another allows concatenation of application segments written in a variety of languages, provided
they satisfy certain criteria.

When an error condition is detected, the error message is
displayed by the run time library indicating the nature of
the error and the address at which the error occurred
(user PC). This is followed by the traceback information,
which is presented in inverse order to the calls. For each
call frame , traceback lists module name, routine name,
source program line, and absolute and relative PC. Using
this information , the programmer can usually locate the
source of the error in a relatively short period of time.
Common Run Time Library
The VAX-11 Common Run Time Procedure Library contains sets of general purpose and language-specific procedures . User programs call these procedures to perform
specific tasks required for program execution. Both VAX11 MACRO and native mode high-level language programmers can use any of the Run Time Library procedures in any combination. Because all procedures
follow the same programming standards and make no
conflicting execution assumptions, a language-independent common run time environment is provided for user
programs. Such an environment encourages a user program to be composed of procedures written in different
languages, and thus increases programming flexibility.

These native mode language processors produce native
object code, and take advantage of the native instruction
set and 32-bit architecture of the VAX hardware.
In addition, there is the host development mode programming environment which provides support for PDP-11 BASJC-PLUS-2/V AX, PDP-11 FORTRAN IV IV AX to RSX, and
MACR0-11 . These produce compatibility mode object
code.

VAX COMMON LANGUAGE ENVIRONMENT
An important feature provided by VAX is a "common language" environment, i.e., the VAX languages adhere to a
specific set of standards, including :

VAX Calling Standard
The VAX-11 procedure calling standard defines and supports the mechanisms for passing arguments between
modules of major VAX-11 software subsystems such as
languages, VAX-11 RMS, and the VAX/VMS operating
system. The standard facilitates the calling of a procedure
written in one language from a program written in another
language.

• symbolic debugger interface
• use of the symbolic traceback facility
• use of the Common Run Time library
• conformance to the VAX calling standard which allows
calls among any set of VAX languages, to VAX/VMS
system services and to SORT and FMS subroutines

Exception Handling
The mechanisms defined by the VAX-11 calling standard
are also used by the condition handling facility to signal
the occurence of exceptions detected by hardware or software.

• common handling of exceptions
• use of VAX-11 RMS for record handling
Symbolic Debugger Interface
VAX/VMS provides facilities to aid the debugging of programs written in native mode. It accomplishes this via a
program known as the interactive symbolic debugger. The
debugger can be linked with a native program image to
control image execution during development. It can be
used interactively or can be controlled from a command
procedure file. The debugging language is similar to the
VAX/VMS command language. Expressions and data
references are similar to those of the source language
used to create the image being debugged. Debugging
statements can be conditionally compiled.

VAX-11 RMS
VAX-11 Record Management Services (RMS) is the technique programmers use to handle record 1/0 within programs. VAX-11 RMS routines are system routines that
provide an efficient and flexible means of handling files
and their data. Typically, VAX-11 RMS routines allow the
programmer to create a file and:
• accept new input
• read or modify data
• produce output in a meaningful form

Debugging commands include the ability to start and interrupt program execution, to step through instruction sequences, to call routines, to set break or trace points, to
set default modes, to define symbols, and to deposit, examine, or evaluate virtual memory locations.

High-level language programmers normally use the 1/0 facilities of their particular language to perform record and
file operations. These operations are implemented using
the VAX-11 RMS facilities. VAX-11 MACRO programmers
can use the VAX-11 RMS routines directly within their programs.

Symbolic Traceback Facility
VAX/VMS supports the Symbolic Traceback Facility. This
is a run time facility that aids programmers in finding errors by describing the call sequences that occurred prior
to the error. The traceback facility is automatic and does
not require that any special qualifiers be included with the

VAX-11 RMS routines are an integral part of the operating
system. The programmer need not perform any special
linking or declaring of global entry points for the routines.

7-1

Furthermore, calls to VAX-11 RMS routines are consistent
with the VAX calling standard.

the 1/0 list are converted to characters and written in a
fixed format according to the data type of the value.

The elements of the common language environment are
discussed more fully as they apply to each individual VAX
language. Introduced below are each of the VAX-supported languages, their attributes, characteristics, and sample
coding .

Character Data Type
A program can create fixed-length CHARACTER variables
and arrays to store ASCII character strings. The VAX-11
FORTRAN language provides a concatenation operator,
substring notation , CHARACTER relational expressions,
and CHARACTER-valued functions. CHARACTER constants, consisting of a string of printable ASCII characters
enclosed in string quotes , can be assigned symbol ic
names using the PARAMETER statement. Operat ions
which use CHARACTER strings are more efficient and
easier to use than their analogs using arithmetic data
types. VAX/VMS provides a set of character manipulation
instructions that are FORTRAN-callable (e.g., LIB$LOCC,
locate a character in a string).

VAX-11 FORTRAN
Introduction
VAX-11 FORTRAN is an optional language processing system whose language specifications are based on the
American National Standard FORTRAN X3.9-1978 (commonly called FORTRAN-77). The VAX-11 FORTRAN compiler supports this standard at the full-language level. At
the same time, it provides optional support for certain
FORTRAN features based on the previous ANSI standard ,
X3 .9-1966. The VAX-11 FORTRAN compiler performs the
following functions :

Figure 7-1 illustrates two VAX-11 FORTRAN subroutines.
This figure illustrates the use of the FORTRAN CHARACTER data type and some of the VAX-11 FORTRAN extensions to FORTRAN-77. The first subroutine, wh ich
reverses a character string, illustrates CHARACTER declarations (both fixed and passed length), the intrinsic
function LEN, substring manipulation, and the ENDDO
statement. The second subroutine locates a substring of a
character string and marks the starting position of the substring. This subroutine illustrates CHARACTER declarations (both fixed and passed length), assignments of character values to variables, the intrinsic function INDEX, substring manipulation, and the FORTRAN-77 block IF
statement.

• produces highly optimized VAX native object code
• makes use of the VAX floating point and character string
instructions
• produces shareable code
The VAX-11 FORTRAN language is upwardly compatible
with the PDP-11 FORTRAN language. Table 7-1 lists extensions to the ANSI FORTRAN-77 language. Table 7-2
lists the features of FORTRAN-77.
Some characteristics of VAX-11 FORTRAN are described
below.

SUBROUTINE REVERSE($)
CHARACTER T, S*(*)

File Manipulation
OPEN and CLOSE statements extend the file management
characteristics of the FORTRAN language. An OPEN statement can contain specifications for file attributes that
direct file creation or subsequent processing. Attributes
include: file organization (sequential, relative, indexed);
access method (sequential , direct , keyed); protection
(read-only, read/write); record type (formatted, unformatted) ; record size; and file allocation or extension . The program can also specify whether the file can be shared , and
whether the file is to be saved or deleted when closed . The
OPEN statement can contain an ERR keyword which specifies the statement to which control is transferred if an error is detected during OPEN .

J = LEN(S)
DO 1=1 , J/2
T = S(l:I)
S(l :I) = S(J:J)
S(J :J) = T
J = J-1
ENDDO
END
SUBROUTINE FIND_SUBSTRINGS(SUB, S)
CHARACTER*(*) SUB, S
CHARACTER*132 MARKS
I= 1
MARKS=' '

Of particular interest is the VAX-11 FORTRAN support for
the Indexed Sequential Access Method (ISAM), a powerful
keyed input/output file access capability. VAX-11
FORTRAN is able to create, read , and write indexed (and
relative) files. In addition, FORTRAN is able to reference a
relative or indexed file already created by another language (for instance, COBOL), provided the file and data
formats and the key information are compatible.

Simplified 1/0 Formats
List-directed input and output statements provide a
method for obtaining simple sequential formatted input or
output without the need for FORMAT statements. On input,
values are read, converted to internal format, and assigned to the elements of the 1/0 list. On output, values in

J = INDEX(S(I:), SUB)
IF (J .NE. 0) THEN
I= I+ (J-1)
MARKS(l:I) = '#'
I= 1+1
IF (I .LE. LEN(S)) GO TO 10
ENDIF
WRITE(6,91) S, MARKS
FORMAT( 2(/1X, A))
END
Figure 7-1
FORTRAN CHARACTER
Data Type Program

7-2

Table 7-1
Language Extensions to FORTRAN-77,
X3.9-1978
VAX-11 FORTRAN
31-character symbolic
names

CALL extensions

Hexadecimal and octal constants and field descriptors

Symbolic names used to
identify programs, subprograms, external functions
and subroutines, COMMON
blocks, variables, arrays,
symbolic constants, and
statement functions can be
longer than the standard six
characters. Symbolic names
can include letters, digits,
dollar sign , and underscore;
however, the first character
in name must be a letter.
Permit interfacing to
VAX/VMS system service
procedures using the VAX11 calling standards.
Both octal and hexadecimal
constants can be expressed
in DAT A statements. No conversion of the defined value
(such as sign-extension) is
performed . The Z field descriptor in FORMAT statements enables a program to
read and write hexadecimal
digits which are stored in an
internal format in an 1/0 list
element.

DO WHILE/END DO

Structured looping control
constructs.

Data initialization in type-declaration statements

Variables can be assigned
initial values in type declaration statements.

INTEGER data type defaults

A compiler command specification allows all
INTEGER and LOGICAL declarations without explicit
length specifications to be
considered as INTEGER*2
and LOGICAL *2 or INTEGER*4 or LOGICAL *4, respectively.

Keyed READ

Key types: INTEGER*2, INTEGER*4, CHARACTER with
generic, and approximate
key match.

Indexed File WRITE

New records can be written
to ISAM files with the write
statements.

REWRITE statement

Existing records in ISAM files
can be modified with the
REWRITE statement.

DELETE statement

Existing records can be
deleted from ISAM or relative files with the DELETE
statement.

UNLOCK statement

Single-record locking (in the
VAX environment) and bucket-level locking (in the PDP11 environment) for shared
file applications involving relative and indexed organization files.

Logical operations on
integers

The logical operators .AND. ,
.OR., .NOT., .XOR., and
.EQV . may be applied to integer data to perform bit
masking and manipulation.

INCLUDE statement

The INCLUDE statement incorporates FORTRAN
source text from a separate
file into a FORTRAN program.

VAX-11 FORTRAN, PDP-11 FORTRAN IV-PLUS, and PDP-11
FORTRAN IV
Array subscripts using general expressions of any numeric data type

Any arithmetic expression
can be used as an array subscript. If the value of the
expression is not an integer,
it is converted to integer format.

End-of-Line comments

Any FORTRAN statement
can be followed, in the same
line, by a comment that begins with an exclamation
point.

Conditional comp ilation of
debugging statements

Statements that are included
in a program for debugging
purposes can be so designated by the letter D in column 1. Those statements
are compiled only when the
associated compiler command option is set. They are
treated as comments otherwise.

Default FORMAT width

The programmer can specify
input or output formatting by
type and default width and
precision values will be supplied.

VAX-11 FORTRAN and PDP-11 FORTRAN IV-PLUS
Additional data types and
type declaration statements
(DOUBLE COMPLEX, COMPLEX*16, and CHARACTER*
n are VAX-11 FORTRAN only)
NOTE
Names appearing on the
same line above are synonyms. Those in boldface are
the ANSI standard ones.
Indexed File 1/0

BYTE, LOGICAL *1,
LOGICAL*2,
LOGICAL, LOGICAL *4,
INTEGER*2,
INTEGER, INTEGER*4,
REAL, REAL *4 ,
DOUBLE PRECISION, REAL *8,
COMPLEX, COMPLEX*8,
DOUBLE COMPLEX, COMPLEX*16,
CHARACTER*n
Extensions are provided to
allow FORTRAN language
access to RMS ISAM files.
7-3

Table 7-2
FORTRAN-77 Features
VAX-11 FORTRAN
Additional data types

The data type INTEGER*4
provides a sign plus 31 bits
of precision. INTEGER*4 allows a greater range of values to be represented than
INTEGER*2. Both data types
can be used in the same program.

Additional 1/0 statements

READ (u'r,fmt) and WRITE
(u'r,fmt) provide input and
output to direct access files.

DO control variable data
types

The control variable of a DO
statement can be a REAL or
DOUBLE PRECISION variable, as well as an INTEGER*
2 or INTEGER*4 variable.
The initial, terminal, and increment parameters can be
of any data type and are converted before use to the type
of the control variable if
necessary.

Additional data type

The data type CHARACTER
permits manipulation of
strings of ASCII characters
expressed as constants, variables, arrays, substrings,
symbolic names, or functions.

IF THEN ELSE statements

The FORTRAN-77 block-IF
statements are provided: IF,
ELSE IF, ELSE, and ENDIF.
These structured programming statements provide
more readable and reliable
methods for expressing con ditional statement execution.

Standard CALL facility

Provides standard argument
definitions for called procedures.

VAX-11 FORTRAN and PDP-11 FORTRAN IV-PLUS
ENTRY statement
ENTRY statements can be
used in SUBROUTINE and
FUNCTION subprograms to
define multiple entry points
in a single program unit.
PARAMETER statement

PARAMETER statements
can be used to give symbolic
names to constants .

Generic function selection

Function selection by argument data type is provided
for many FORTRAN library
functions.

7-4

Array dimension bounds

Lower bounds as well as upper bounds of the array dimension can be specified in
array declarators. The value
of the lower bound dimension declarator can be negative, zero or positive.

List-Directed 1/0 statements

The READ (u , *) , WRITE (u, *),
TYPE* , ACCEPT* , and
PRINT* statements provide
list-directed, or "free format," 1/0 without requiring a
FORMAT specification .

Additional 1/0 statements

OPEN and CLOSE statements provide file control
and attribute definition . ACCEPT, TYPE, and PRINT
statements provide deviceoriented 1/0. ENCODE and
DECODE statements provide
memory-to-memory formatting . DEFINE FILE, READ
(u'r) , WRITE (u'r) , and FIND
(u 'r) provide unformatted
direct access 1/0, which allows the FORTRAN programmer to read and write files
written in any format.

End-of-file or Error Condition transfer

The specifications END = n
and ERR=n (where n is a
statement label) can be included in any READ or
WRITE statement to transfer
control to the specified statement upon detection of an
end-of-file or error condition .
The ERR=n option is also
permitted in the ENCODE
and DECODE statements, allowing program control of
data format errors.

Additional data type

The byte data type (keyword
LOGICAL *1 or BYTE) is useful for storing small integer
values as well as for storing
and manipulating character
information.

IMPLICIT declaration

The IMPLICIT statement has
been added to redefine the
implied data type of symbolic
names.

DO loop iteration count

The terminal and increment
parameters can be modified
within a DO loop without affecting the iteration count.
The number of times a DO
loop is executed is determined at the initialization of
the DO statement and is not
re-evaluated during successive executions of the loop.
Consequently, the number of
times the loop is executed
will not be affected by changing the variables used in the
DO statement.

VAX-11 FORTRAN, POP-11 FORTRAN IV-PLUS, and
PDP-11 FORTRAN IV

Array dimensions

Arrays can have up to seven
dimensions.

Character literals

Character strings bounded
by apostrophes can be used
in place of Hollerith constants.

Mixed-mode expressions

Mixed-mode expressions
can contain any data type, including complex and byte.

General expression DO and
GO TO parameters

General expressions are permitted for the initial value, increment, and limit parameters in the DO statement, and
as the control parameter in
the computed GO TO statement.

DO increment parameter

The value of the DO statement increment parameter
can be negative.

Optional statement label list

The statement label list is an
assigned GO TO is optional.

General expressions in 1/0
lists

General expressions are permitted in 1/0 lists of WRITE,
TYPE, and PRINT statements.

class messages indicate errors that must be corrected
before compilation can be completed . Object code is not
produced. E-class messages indicate that an error was
detected that is likely to produce incorrect results; however, an object file is generated. W-class messages are
produced when the compiler detects acceptable but nonstandard syntax; or when it corrects a syntactically incorrect statement. The message indicates the existence of
possible trouble in executing the program.

Source Program Libraries
The INCLUDE statement provides a mechanism for writing
modular, reliable, and maintainable programs by eliminating duplication of source code. A section of program text
that is used by several program units, such as a COMMON
block specification , can be created and maintained as a
separate source file. All program units that reference the
COMMON block then merely INCLUDE this common file .
Any changes to the COMMON block will be reflected automatically in all program units after compilation .

Compiler Operations and Optimizstlons
The VAX-11 FORTRAN compiler accepts sources written
in the FORTRAN language and produces an object file
which must be linked prior to execution. The compiler
generates VAX-11 native machine language code. Figure
7-2 is an illustration of VAX-11 FORTRAN code and its
equivalent VAX-11 MACRO code .

Calling External Functions and Procedures
FORTRAN programs can call subroutines written in any
other VAX language, and also system services, using the
VAX-11 procedure calling standard. Special operators exist for passing arguments by immediate value, by reference, or by descriptor. A special operator also exists for
obtaining the location of argument values used by the system services procedures.

During compilation , the compiler performs many code optimizations. The optimizations are designed to produce an
object program that executes in less time than an equivalent nonoptimized program. Also, the optimizations are
designed to reduce the size of the object program.

Shareable Programs
The FORTRAN language can be used to create shareable
programs. FORTRAN subprograms can also be used to
create shareable image libraries, which can be available to
any program written in a native programming language.

The VAX-11 FORTRAN compiler performs the following
optimizations:
• Constant folding-constant expressions are evaluated
at compile-time.

Diagnostic Messages
Diagnostic messages are generated when an error or potential error is detected . Errors detected during compilation are reported by the compiler, and include source program errors, such as misspelled variable names, missing
punctuation marks, etc.

• Compile-time constant conversion.
• Compile-time evaluation of constant subscript expressions in array calculations.
• Constant pooling-only a single copy of a constant is allocated storage in the compiled program. Constants
that can be used as immediate mode operands are not

Source program diagnostic messages are classified according to severity: F (Fatal), E (Error), or W (Warning). F-

7-5

1-Apr-1980 11 :56:00 VAX-11 FORTRAN V2 .0-1
0001

SUBROUTINE RELAX2(EPS)

0002
0003
0004

PARAMETER M =40, N =60
DIMENSION X(O:M,O:N)
COMMON X

Page 1

0005

LOGICAL DONE

0006

DONE= .TRUE.

0007
0008
0009
0010
0011

DO 10J = 1,N-1
DO 10 I= 1,M-1
XNEW = ( X(l-1 ,J)+X(l+1,J)+X(l ,J-1)+X(l ,J+1) )/4
IF ( ABS(XNEW-X(l,J)) .GT . EPS) DONE= .FALSE.
X(l,J) = XNEW

0012

IF (.NOT . DONE) GO TO 1

0013
0014

RETURN
END

1-Apr-1980 11 :56:00 VAX-11 FORTRAN V2 .0-1

0000
0000
0000
0000
0000
0002

Page 2

.TITLE
.IDENT

RELAX2
01

.PSECT

$BLANK

.PSECT

$CODE

.WORD
MOVAL

tM<IV,R5,A6,R7 ,AS,R9 ,A10,R11 >
$LOCAL, R11

X:
RELAX2::

; 0006
0009
0009

.1:
MNEGL

#1, DONE(A11)
; 0007

oooc
OOOF
0016

MOVL
MOVAL

#1, R6
$BLANK, A5

MOVL
MULL3

#1, R9
#41, R6, R7

L$1 :

; ooos
0016
0019
0010

L$2:

0010
0021
0029
002F
0035

ADDL3
ADDF3
ADDF2
ADDF2
MULF3

003D
0042
0047
004B
004D
004F

SUBF3
BICW2
CMPF
BLEQ
CLRL

; 0009
R9, A7, R10
X+4(R5)(R10), X-4(R5)[R10], RO
X-164(R5)[A10], RO
X+164(A5)[R10], RO
#tX3FSO, RO, RS
; 0010
X(R5)[R10], R8, RO
#tXSOOO, RO
RO, tEPS(AP)
L$3
DONE(R11)

004F
0053
0057
005B
005F
0063

MOVL
AOBLEQ
AOBLEQ
MOVL
MOVL
MOVL

RS, X(A5)[R1 OJ
#39,R9,L$2
#59,R6,L$1
R6, J(R11)
RS, XNEW(R11)
R9,l(R11)

0067

BLBC

DONE(R11 ), .1

006A

RET
.END

L$3:
; 0011

Page 1 above illustrates, as a VAX-11 FORTRAN subroutine, a relaxation function often found in engineering
applications. This particular example is a planar (2 - dimensional) function that can be used to obtain the values of a variable at coordinates on a surface, for instance, temperatures distributed across a metal plate.
The algorithm illustrated here locates the array element
values relative to a given point in the plane.
Page 2 contains the equivalent VAX-11 MACRO
assembly code for this VAX-11 FORTRAN subroutine.
The line numbers in the comment just to the left of this
paragraph refer to the lines in the VAX-11 FORTRAN
subroutine listing above. Several VAX-11 FORTRAN
compiler optimizations are illustrated, including global
and local register assignment, removal of invariant
computations from the DO loop, recognition of common
subexpressions, branch instruction optimizations, in line ABS function , and peephole optimization.
The code for lines 7 and 8 contains the global register
assignments for the function. The multiply statement
just preceding the code for line 9 is an invariant computation (J ' 41) removed from the DO loop. DO loop control is provided by the Add One and Branch Less Than
or Equal (A OBLEQ) instructions in the code for line 11 .
The code for line 9 evaluates the common subexpression for the computation. The code contains a local register assignment (R10), and uses 2- and 3-operand
instructions and context switching ([R10]) to calculate
an array element value. The last instruction for line 9 is a
peephole optimization that increases execution speed
by using a "multiply by .25" in place of the FORTRAN
statement's "divide by 4."

; 0012
; 0013

Figure 7-2
VAX-11 FORTRAN Program
7-6

allocated storage. For example, logical, integer, and
small floating point constants are generated as immediate mode or short literal operands wherever possible.

variable can be printed to verify its contents. Therefore, by
including a number of such source lines at strategic points
throughout the program, debugging the program is greatly
simplified. FORTRAN provides a facility for conditionally
compiling such source lines so that they can be compiled
during the development stage but treated as comments
once the program has been debugged .

• Argument list merging-if two function or subroutine
references have the same arguments, a single copy of
the argument list is generated.
• Branch instruction optimizations for arithmetic or logical
IF statements.

Symbolic Traceback
Figure 7-3 illustrates a source VAX-11 FORTRAN program
and the symbolic traceback facility supported by
VAX/VMS. (Note that some of the entries in the list show
relative and absolute PC but no corresponding values for
module name and routine name; this indicates that the values refer to procedure calls internal to the run time library.)

• Elimination of unreachable code-an optional warning
message is issued to mark unreachable statements in
the source program listing.
• Recognition and replacement of common subexpressions.
• Removal of invariant computations from DO loops.
• Local register assignment-frequently referenced variables are retained (if possible) in registers to reduce the
number of load and store instructions.
• Assignment of frequently used variables and expressions to registers across DO loops.
• Reordering expression evaluation to minimize the number of temporary registers required .
• Delaying negation/not to eliminate unary complement
operations.
• Flow-Boolean optimizations.
• Jump/Branch instruction resolution-the Branch instruction is used wherever possible to eliminate unnecessary Jump instructions. This reduces code size.
• Peephole optimizations-the code is examined on an
operation-by-operation basis to replace sequences of
operations with shorter and faster equivalent operations.

0001
0002
0003
0004
0005
0006
0007
0008

1=1
CONTINUE
J= 2
CONTINUE
K=3
CALL SUB1
CONTINUE
END

0001
0002
0003
0004
0005

SUBROUTINE SUB1
1=1
J= 2
CALLSUB2
END

0001
0002
0003
0004
0005
0006

SUBROUTINE SUB2
COMPLEXW
COMPLEXZ
DAT AW / (0 .,0.)/
Z = LOG(W)
END

%MTH - F- INVARGMAT, invalid argument to math library
user PC 00000449
% TRACE-F-TRACEBACK, symbolic stack dump follows

Debugging Facilities
VAX-11 FORTRAN debugging facilities include diagnostic
messages, conditional compilation flags, and access to the
VAX/VMS DEBUG program. The DEBUG program lets the
programmer set breakpoints and trace points, and examine and modify the contents of locations dynamically when
executing the program .

module name

DEBUG understands FORTRAN data type representations
and syntax. It can examine and deposit locations using
floating point representation, and it can reference
FORTRAN symbols, statement labels, and line numbers
symbolically. It can also reference arrays symbolically, for
example:

routine name

line

relative PC

absolute PC

0000074C

0000081 c

SU82

00000011

00000449

SU81

00000017

00000437

T1$MAIN

00000018

00000418

Figure 7-3
FORTRAN Symbolic Traceback

EXAMINE A(l,J+3)
When debugging VAX-11 FORTRAN programs, the programmer can disable optimizations that would remove unreferenced statement labels, FORMAT statement labels,
and immediately referenced labels. This ensures that all
statement labels are available to the debugger .

VAX-11 COBOL
Introduction
VAX-11 COBOL is a new, high-performance implementation of COBOL. It is based on American National Standard
Programming Language COBOL, X3 .23-1974, the industry-wide accepted standard for COBOL. Some features
planned for the next COBOL (anticipated in 1981 ), are also
included. VAX-11 COBOL expands and enhances its predecessor, VAX-11 COBOL-74, and includes features that

Conditional Compilation of Statements
During the development stages of a program, it is often
useful to establish points in the program at which specified
values can be examined to insure that the program is functioning correctly. For example, if the value of a variable is
known after the execution of a specified statement, the
7-7

The object code produced by VAX-11 COBOL uses the
VAX/VMS traceback facility for determining the source of
run time errors. If a fatal error occurs at run time, an English error message is printed to identify the cause of the
error. Additionally, the traceback pinpoints the source of
the error to a specific line number in the COBOL source
module producing the error. The English error message
coupled with the traceback facility gives the user a powerful debugging tool for identifying fatal execution errors.

appeal to a wider range of COBOL users because it allows
more complex coding procedures to be accomplished
more simply.
It is anticipated that the new ANSI standard will call for
greater structured programming. This allows explicit delimiting of statements in the Procedure Division, a feature
which can simplify COBOL coding that previously required
additional GO TO statements and procedure names. In
meeting the requirement for structured programming, the
new VAX-11 COBOL includes-among other features-the in-line PERFORM statement, allowing a reduction of program complexity by putting all the logic of the
PERFORM in line.

Object modules produced by the compiler can be linked
with native mode object modules produced by other VAX11 language processors including BASIC, FORTRAN , and
MACRO .

Many features of VAX-11 COBOL make the programmer's
job easier, either by simplifying coding procedures or by
giving direct access to more VAX/VMS facilities. The
COBOL SORT and MERGE verbs are now available in
VAX-11 COBOL so that sorting and merging can be performed at the source language level rather than though
direct calls to the VAX/VMS utilities. VAX-11 COBOL supports symbolic characters so that the programmer can define non-printable characters simply and can generate
video display forms. Further, the REFORMAT utility allows
bidirectional conversion of COBOL source programs from
easy-to-enter DIGITAL terminal format to ANSI standard
format and vice versa.

Structured Programming
Structured programming adds some of the features of a
block-structured language (such as ALGOL) to the new
VAX-11 COBOL compiler. Thus, more complex programs
can be written in-line without recourse to subroutines. This
makes programs easier to write and to read.
The example below shows the READ and IF statements using structured programming. The statements after ENDREAD are executed regardless of whether the AT END
condition occurs. Similarly, the MOVE after END-IF is executed regardless of the value of FILE-END.
IF ITEMA = ITEMS
READ FILE-A AT END
MOVE 1 TO FILE-END

VAX-11 COBOL is properly defined as an implementation
of ANSI COBOL with full support of the following:
• full Level 2 Nucleus Module without the RERUN option
in the 1-0-CONTROL paragraph

CLOSE FILE-A
END-READ
MOVE ITEMS TO ITEMC
IF FILE-END = 1
DISPLAY ITEMC
END-IF
MOVE ITEMD TO ITEME.

• full Level 2 Table Handling Module
• full Level 2 Sequential 1/0 Module
• full Level 2 Relative 1/0 Module
• full Level 2 Indexed 1/0 Module
• full Level 2 Segmentation Module

Several COBOL verbs have structured programming delimiters. Among them are:

• full Level 2 SORT/MERGE Module
• full Level 2 Library Module

ADD
CALL
COMPUTE
DELETE
DIVIDE
IF
MULTIPLY
PERFORM
READ
RETURN
REWRITE
SEARCH
START
STRING
SUBTRACT
UNSTRING
WRITE

• full Level 2 lnterprogram Communication Module
Besides the VAX-11 object module, the compiler is capable of producing a machine language listing, a cross reference listing in either alphabetic sequence or order of declaration , and maps of file names, data names, procedure
names, and external program names.
General Characteristics
Most of the code in an object module is implemented with
in-line VAX-11 instructions. The object code produced by
the compiler takes advantage of such native mode features as:
• direct calls to the operating system
• transparent access to DECnet
• direct calls to VAX-11 SORT
• direct calls to the Common Run Time Library

Particularly, the PERFORM verb has been enhanced . The
resultant in-line PERFORM capability is similar to DO
WHILE and DO UNTIL in other high-level languages.

• direct calls to an external routine (written in a DIGITALsupported language) that conforms to the VAX-11
Procedure Calling Standard

In this example, if the first occurrence of ITEMS is not
equal to "X": (1) the in-line PERFORM statements are executed, moving an "X" to the first 10 occurrences of ITEMS;

• many of the VAX-11 string manipulation instructions

7-8

then , (2) the message is displayed .

• faster arithmetic operations than standard numeric display data type

IF ITEMS (1) NOT = "X"
PERFORM
VARYING ITEMA FROM 1 BY 1
UNTIL ITEMA > 10
MOVE "X" TO ITEMS (ITEMA)
END-PERFORM
DISPLAY " ARRAY INITIALIZED"

• compatibility with and migration from other COBOL
vendors
Figure 7-4 illustrates a record definition of a typical payroll
master file application in which the COMP-3 data type is
frequently used. In this record definition, all numeric fields
on which arithmetic operations are performed are defined
to be the COMP-3 data type.

Data Types
VAX-11 COBOL increases the number of data types available to the COBOL programmer , including floating point
and double floating point. The standard data types are:

Figure 7-5 illustrates a sample calculation of one such
COMP-3 data item in the record. Here, the year-to-date
net pay is calculated as a function of the gross pay, and all
voluntary and involuntary deductions to date.

• Numeric DISPLAY Data
Trail ing overpunch sign
Leading overpunch sign
Trailing separate sign
Lead ing separate sign
Unsigned
Numeric-edited

Infrequently, commercial applications arise in which the
utilization of floating point data (COMP-1 and COMP-2) is
useful. For example, a large corporation may want to survey its customers regarding its product quality. The corporation wishes to select a statistically valid sample of its customer base without going to the expense of contacting
each and every customer. Hence, it is necessary to randomly sample its customer base; a random number generator is used to select those customers to be sampled.

• Numeric COMPUTATIONAL Data
Word fixed binary
Longword fixed binary
Quadword fixed binary

Figure 7-6 illustrates a COBOL program fragment in which
a CALL to the VAX-11 run-time procedure library routine
MTH$RANDOM is made to generate random numbers.
This routine returns a random number in COMP-1
(F_floating) data type representation in the range from 0.0
to 1.0. Such numbers are then integerized and subsequently used to select those customers to be sampled in
the product quality survey.

• Packed-Decimal Data (COMP UT ATIONAL-3)
Unsigned packed decimal
Signed packed decimal
• Floating Po int Data
F floating (COMPUTATIONAL-1)
D~floating (COMPUTATIONAL-2)

The COMP-2 data type may be used in similar commercial
applications.

• Alphanumeric DISPLAY Data
Alphanumeric
Alphabetic
Alphanumeric-edited

Files and Records
VAX-11 COBOL's Sequential 1/0, Relative 1/0, and Indexed 1/0 modules meet the full ANSI Level 2 standard.
The language's Level 2 Indexed 1/0 module statements
enable VAX-11 COBOL programs to use the VAX-11 RMS
multi key indexed record management services to process
files . These files can be accessed sequentially, randomly,

As indicated previously, VAX-11 COBOL supports the
COMP-3 (packed decimal) data type (two decimal digits
per byte). This data type offers the following advantages:
• disk storage savings

FD
01

PAYROLL-MASTER
LABEL RECORDS ARE STANDARD.
PAYROLL-REC.
02 EMPLOYEE-NAME
02 EMPLOYEE-ID
02 YTD-GROSS-PAY
02 YTD-FED-WITHHOLD-T AX
02 YTD-FICA
02 YTD-STATE-WITHHOLD-TAX
02 YTD-LOCAL-WITHHOLD-TAX
02 YTD-VOLUNTARY-DEDUCTIONS
02 YTD-NET-PAY

PIC X(30).
PIC9(9)
PIC 9(5)V99
PIC 9(5) V99
PIC 9(4)V99
PIC 9(5)V99
PIC 9(5)V99
PIC 9(4)V99
PIC 9(5)V99

Figure 7-4
COMP-3 Record Definition
7-9

USAGE IS DISPLAY.
USAGE IS COMP-3.
USAGE IS COMP-3.
USAGE IS COMP-3.
USAGE IS COMP-3.
USAGE IS COMP-3.
USAGE IS COMP-3.
USAGE IS COMP-3.

•
•
•SUBTRACT YTD-FED-WITHHOLD-T AX,

SORT/MERGE Facility
The VAX-11 COBOL SORT/MERGE module meets the full
ANSI standard and permits performing sort and merge
operations at the COBOL source language level without
requiring the programmer to understand the VAX-11
SORT interface. The COBOL SORT/MERGE capability includes sorting and/or merging one or more files in the
same source module, specifying one or more sort/merge
key(s) (in ascending or descending order) for each file ,
and the option to use either standard or user-specified input/output procedures .

YTD-FICA
YTD-ST ATE-WITHHOLD-TAX,
YTD-LOCAL-WITHHOLD-T AX,
YTD-VOLUNTARY-DEDUCTIONS
FROM YTD-GROSS-PAY
GIVING YTD-NET-PAY.

•
•
•

Figure 7-7 illustrates how to sort a file with the USING and
GIVING phrases of the SORT statement. The fields to be
sorted are S-KEY-1 and S-KEY-2; they contain account
numbers and amounts. The sort sequence is amount within account number. Notice that OUTPUT-FILE is a relative
file.

F!gure 7-5
Arithmetic on COMP-3 Data Type

In Appendix B, the sample program is merging three identically sequenced regional sales files into one total sales
file. The program adds sales amounts and writes one record for each product-code.

or dynamically using one or more indexed keys to select
records. The RESERVE AREAS clause enables the user to
specify the number of 1/0 buffers for fast multi key processing. The APPLY clause allows the user to specify file
processing optimization attributes for fast record access.

Symbolic Characters Facility
It is often useful for the programmer to be able to construct
on a video terminal, the image of a form similar to a printed
form . This process involves imbedded or non-printing
characters (i.e., line feed, carriage return, escape key,
etc.). VAX-11 COBOL provides the user with the ability to
include, within the COBOL code, non-printing control
characters. Essentially, these characters control the position of the cursor during an interactive session utilizing a
video terminal (i.e., VT52, VT100, etc.).

VAX-11 COBOL has full variable-length record capability.
This is an improvement over VAX-11 COBOL-74, in which
variable-length records were only partially supported.
Reference modification-the ability to refer to parts of defined fields without redefining them-has also been included in VAX-11 COBOL.

Figure 7-8 illustrates a sample data entry form used as a
prompt for data input.

The language includes a facility to manipulate data strings.
The INSPECT verb allows the user to search for embedded character strings, tallying and/or replacing the occurrences of such strings. Additionally, the STRING and
UNSTRING verbs permit the user to join together and
break out separate strings with various delimiters.

01
01
01

•
•
•

The VAX-11 COBOL code used to generate this particular
form is listed in Appendix C. The sample code is used in
conjunction with the VT100 terminal. Code for the VT52 is
similar.

RAND-NUM
RAND-CUST-NUM
PIC 9(7).
CUST-NUM REDEFINES RAND-CUST-NUM.
02 DISTRICT
PIC 9(2).
02 WITHIN-DIST
PIC 9(5).
SEED
PIC 9(8)

CALL "MTH$RANDOM"

USAGE IS COMP-1 .

USAGE IS COMP .

USING SEED
GIVING RAND-NUM.
COMPUTE RAND-CUST-NUM ROUNDED= RAND-NUM * 10000000.

Figure 7-6
Example of COMP-1 Data Type
7-10

IDENTIFICATION DIVISION.
PROGRAM-ID.
SORT EXAMPLE.
ENVIRONMENT DIVISION.
CONFIGURATION SECTION.
SOURCE-COMPUTER .
VAX-11.
OBJECT-COMPUTER.
VAX-11 .
INPUT-OUTPUT SECTION.
FILE-CONTROL.
SELECT INPUT-FILE ASSIGN TO " INPFIL".
SELECT OUTPUT-FILE ASSIGN TO "OUTFIL"
ORGANIZATION IS RELATIVE.
SELECT SORT-FILE ASSIGN TO "SRIFIL".
DAT A DIVISION.
FILE SECTION .
SD
SORT-FILE.
01
SORT-REC.
03 S-KEY-1 .
05 S-ACCOUNT-NUM PIC X(8) .

module. The object module file can be linked with other
VAX-11 object modules, so as to produce an executable
image. Thus, COBOL programs can call external routines
written in other VAX-11 supported languages including
BASIC, FORTRAN, and MACRO.
The CALL statement facility has been extended by allowing the user to pass arguments BY REFERENCE (the default in COBOL), BY DESCRIPTOR, and BY VALUE. These
argument-passing mechanisms conform to the VAX-11
Procedure Calling Standard and allow COBOL programs
to call VAX/VMS operating system service routines. Also,
a COBOL program can receive a returned status value
from the routine it calls via the GIVING clause associated
with the extended CALL facility. Such an extended CALL
facility gives the user access to operating system specific
facilities and Common Run Time facilities. Figure 7-9 illustrates a sample program utilizing all three types of argument passing mechanisms.
In this program the system service routine $ASCTIM is
called, which converts binary time to an ASCII string representation . In this example, the buffer length as specified by "timbuf" plus the value of the item "dummy" determine the type of information which the service routine will
return to the COBOL program (e.g., specifying a length of
24 plus values of 0 in the following two arguments will
cause both current date and time to be returned; if a length
of 11 had been specified, then only the date would be returned) .

03
03

FILLER PIC X(32).
S-KEY-2.
05 S-AMOUNT
PIC S9(5)V99.
03 FILLER PIC X(53).
FD
INPUT-FILE
LABEL RECORDS ARE STANDARD.
01
IN-REC
PICX(100).
FD
OUTPUT-FILE
LABEL RECORDS ARE STANDARD. PIC X(100).
01
OUT-REC
PIC X(100).
PROCEDURE DIVISION.
000-00-THE-SORT.
SORT SORT-FILE ON ASCENDING KEY
S-KEY-1
S-KEY-2
WITH DUPLICATES IN ORDER
USING INPUT-FILE GIVING OUTPUT-FILE.

Source Library Facility
VAX-11 COBOL supports the full ANSI COBOL Library facility . All frequently used data descriptions and program
text sections can be stored in library files available to all
programs. These files can then be copied into source programs performing textual substitution (i.e., replacement)
in the process. This capability reduces program preparation time and eliminates a common source of error during
program development.

DISPLAY "END OF PROGRAM SORT EXAMPLE".
STOP RUN .

Shareable Programs
The COBOL language can be used to create shareable
programs. VAX-11 COBOL subprograms can be placed in
shareable image libraries created by the linker, which then
can be made available to any program written in a native
programming language.

Figure 7-7
Sample SORT Code

Debugging COBOL Programs
The VAX-11 COBOL compiler produces source language
listings with embedded diagnostics indicating line and position of error. Fully descriptive diagnostic messages are
listed at the point of error. Many error conditions are
checked at compile time, varying from simple informational indications to severe error detections. At the user's option, the compiler can also produce a machine language
listing, a file name map, a data name map, a procedure
name map, an external program name map, and a cross
reference listing.

CUSTOMER NUMBER:12345678
CUSTOMER NAME:ROLAND J. JONES----CUSTOMER ADDRESS :747 FIRST AVE.----CITY:ANYTOWN------ST ATE: NH ZIP:03061
Figure 7-8
Video Form

When a fatal error occurs at run time, an error message
identifying the cause of the error is displayed to the user.
Additionally, the traceback system facility prints the sequence of routine invocations active at the time of the fatal
error. For each routine invocation, traceback displays the

CALL Facility
The CALL statement enables a COBOL programmer to execute routines that are external to the source module in
which the CALL statement appears. The VAX-11 COBOL
compiler produces an object module from a single source
7-11

IDENTIFICATION DIVISION.
PROGRAM-ID. CALL TST2.
ENVIRONMENT DIVISION .
CONFIGURATION SECTION .
SOURCE-COMPUTER. VAX-11 .
OBJECT-COMPUTER. VAX-11.
DATA DIVISION.
WORKING-STORAGE SECTION.
01 TIMLEN
01 D-TIMLEN
01 TIMBUF
01 RETURN-VALUE

PIC 9(4) USAGE IS COMP VALUE IS 0.
PIC 9(4) VALUE IS 9999.
PIC X(24) VALUE IS SPACES.
PIC 9(9)USAGE IS COMP
VALUE IS 999999999.
PIC 9(9) VALUE IS 999999999.

01 D-RETURN-VALUE
PROCEDURE DIVISION .
PO.
DISPLAY "CALL SYS$ASCTIM" .
CALL " SYS$ASCTIM "
USING
BY REFERENCE TIM LEN
BY DESCRIPTOR TIMBUF
BY VALUE ZERO
BY VALUE ZERO
GIVING
RETURN-VALUE.
DISPLAY " DATE/TIME=" TIMBUF.
MOVE TIMLEN TO D-TIMLEN.
DISPLAY " LENGTH OF RETURNED= " D-TIMLEN .
MOVE RETURN-VALUE TOD-RETURN-VALUE.
DISPLAY "RETURN-VALUE=" D-RETURN-VALUE.
STOP RUN .
Figure 7-9
System Services Call

module name, routine name, and source line number in
which either an invocation to another user routine occurs
or the fatal error itself occurs.

Thus, the issuance of specific, English-like error messages
coupled with the traceback facility and interactive interrogation of the system to explain completely the run-time error offers the user a powerful debugging tool in identifying
programm ing errors.

As an example of the traceback facility, Figure 7-10 illustrates the printing of error messages and the subsequent
traceback for a COBOL module in which an 1/0 error occurs at run time. Specifically, a COBOL OPEN statement
failed because the file "DB2:[COBOL)MASTERFIL.DAT"
was not found on the OPEN operation. The " module
name" and " routine name" fields (of the traceback) identify the entry point, IOERRTEST, into the COBOL module.
The OPEN failure occurs on line number 22 of the source
module. The "relative PC" field specifies that the OPEN
failure corrrspondingly occurs at "67'' hexadecimal bytes
into the object code relative to the entry point IOERRTEST.
The " absolute PC" field also specifies that the OPEN failure occurs at absolute location "667" in the executable image containing IOERRTEST.

Also , the VAX-11 COBOL debugging facilities provide access to the VAX/VMS SYMBOLIC DEBUGGER. The SYM BOLIC DEBUGGER lets the programmer set breakpoints,
and examine and modify the contents of locations
dynamically while the COBOL program is executing.
Source Translator Utility
The source translator utility is helpful to those users migrating from PDP-11 COBOL and VAX-11 COBOL-74 to
the VAX-11 COBOL compiler. This utility produces a translated source program and a listing with flags indicating
those language elements which could not be mechanically
translated and which therefore require further investigation by the programmer.
Some of the differences between VAX-11 COBOL and
PDP-11 COBOL or VAX-11 COBOL-74 that require such a
translator are:

Additionally, the user can request a complete explanation
of the OPEN error by interrogating the system interactively
with the command "HELP ERRORS COB FILNOTFOU".
·This VAX/VMS command displays the information shown
in Figure7-11 .

• some changes in file status codes
• different specification for the storage of intermediate results
7-12

%COB-F-FILNOTFOU, file_ D32:[COBOL]MASTERFIL. DAT; not found on OPEN
-RMS-E-FNF, file not found
%TRACE-F-TRACEBACK, symbol stack dump follows
module
name

routine
name

line

relative PC

absolute PC

IOERRTEST

00000067

00000667

Figure 7-10
Example of Traceback Facility

• different methods of specifying file optimization attributes

Additional Features
Some additional features of the VAX-11 COBOL compiler
are:

Fortunately, most differences are transparent to the programmer, and moving programs from PDP-11 COBOL or
VAX-11 COBOL-74 requires little (in some cases, no) programmer work.

• Subscripts can be arithmetic expressions.
• Subscripting and indexing are interchangeable.
• The CONTINUE statement is included. It transfers control to the next executable statement and can replace
conditional or imperative statements.

Source Program Formats
The VAX-11 COBOL compiler accepts source programs
that are coded using either the ANSI standard (conventional) format or a shorter, easy-to-enter DIGIT AL terminal
format. Terminal format is designed for use with the interactive text editors. It eliminates the line number and identification fields and allows the user to enter horizontal tab
characters and short text lines.

• The AUTHOR , INSTALLATION, DATE-WRITTEN, DATECOMPILED, and SECURITY paragraphs are included .
• INITIAL clause on the Program-ID is included .
• User-defined alphabets are included.
• PADDING CHARACTER is supported in the FILE-CONTROL paragraph .

The REFORMAT utility reads COBOL source programs
that are coded using DIGITAL terminal format and converts the source statements to the ANSI standard format
accepted by other COBOL compilers throughout the industry. It also has the inverse option to accept programs
written in ANSI standard format and to convert the source
statements to DIGITAL terminal format. This offers the advantage of saving disk space and compile-time processing
when a user is initially migrating from a non-DIGITAL
COBOL system to VAX-11 COBOL.

• VALUE OF clause is included.
• Delimited scope statements are included (e.g. , ENDADD , END-IF) .
• All arithmetic statements with overlapping operands
function as if the operands did not overlap except for
operands specified in LINKAGE SECTION or as EXTERNAL.
• AL TEA statement is included.
• CALL data-name is included. Both ON OVERFLOW and
EXCEPTION are supported.
• CANCEL statement is fully implemented.

ERRORS

• INITIALIZE statement is fully implemented .

COB

• INSPECT statement is fully implemented including combined TALL YING and REPLACING format.

FILNOTFOU
file not found on OPEN

• SET statement supporting mnemonic-names and condition-names is included.

Explanation: The named file was not
found during the execution of the open
statement. The file status variable , if present, has been set to 97. No applicable
USE procedure has been found.

• Independent segments (segments 50 and above) of the
SEGMENTATION module are included .
• WRITE advancing mnemonic-name and associated
SPECIAL NAMES C01 is included.

User Action: The user should examine
the referenced directory to check for the
existence of the named file . Another
common source of this error is a mistake
in spelling the file specification for the
file.

• Use of source file libraries by the COPY statement is included.
This powerful, flexible, and easy-to-use compiler is layered with the VAX/VMS operating system and is available
to those customers who require COBOL with VAX/VMS,
V2 .0.
Sample VAX-11 COBOL Code
This sample VAX-11 COBOL code demonstrates some of

Figure 7-11
Interactive Explanation of Error

7-13

the powerful language elements of VAX-11 COBOL. It illustrates an interactive COBOL program which will generate
various types of reports depending upon user specified
options. The program operates on an indexed information
file via the dynamic access mode. Illustrated are three major COBOL verbs: ACCEPT, DISPLAY and INSPECT.

IF OPTION-STORAGE= ALL ZERO
DISPLAY "No options recognized"
STOP RUN.
DISPLAY "Selected options'.'
IF WANT-STATEMENTS
DISPLAY "Statements'.'
IF WANT-INVOICES
DISPLAY " Invoices"

In Figure 7-12, the program describes the file organization
and the access mode. Also described are the primary and
alternate keys used for accessing the file randomly.

Figure 7-13 (Can 't)
Procedure Division Using
Interactive COBOL Verbs

INPUT-OUTPUT SECTION .
FILE CONTROL.
SELECT CUSTOMER-FILE
ASSIGN TO "CUSTOM.DAT"
ORGANIZATION IS INDEXED
ACCESS MODE IS DYNAMIC
RECORD KEY IS CUST-CUST-NUMBER
ALTERNATE RECORD KEY IS
GUST-CUSTOMER-NAME
FILE STATUS IS CUSTOMER-FILE-ST A TUS .
SELECT STATEMENT-REPORT
ASSIGN TO " STATEM.REP"
FILE STATUS IS
STATEMENT-REPORT-STATUS.

Figure 7-14 illustrates the dynamic access method , i. e.,
shift from random to sequential access. The user moves
zero to the primary record key, searches the file randomly,
and commences sequential processing at the first non-zero number.

OPEN INPUT CUSTOMER-FILE.
MOVE " 000000" TO CUST-CUST-NUMBER.
START CUSTOMER-FILE
KEY IS> CUST-CUST-NUMBER.
OPEN OUTPUT STATEMENT-REPORT .

Figure 7-12
File Description

MAINLINE SECTION
SBEGIN .
READ CUSTOMER-FILE NEXT
ATEND
GO TO END-PROCESS.
ADD 1 TO RECORD-COUNT

In Figure 7-13, using the DISPLAY verb, the interactive
COBOL program requests the user to specify an options
selection. The user response is then transmitted to the
program via the ACCEPT verb. The program uses the INSPECT verb to check that a valid response has been received.

Print statement if required

DISPLAY " ENTER OPTIONS:".
DISPLAY "S = Print statements".
DISPLAY" I = Print invoices".
DISPLAY "CA= Mail all catalogs" .
DISPLAY "CO = Mail selective catalogs".
DISPLAY "CL= Credit limit letters".
ACCEPT OPTIONS-AREA.
MOVE ALL ZERO TO OPTION-STORAGE.
IF OPTIONS-AREA =SPACES
DISPLAY "Discrepancy Report Only"
GO TO CONFIRM-OPTIONS .
MOVE 0 TO A-COUNT.
INSPECT OPTIONS-AREA TALLYING
OPTION-ENTRY (1) FOR ALL "S"
OPTION-ENTRY (2) FOR ALL "I"
OPTION-ENTRY (3) FOR ALL "CA"
OPTION-ENTRY (4) FOR ALL "CO"
OPTION-ENTRY (5) FOR ALL "CL':

Figure 7-14
Random to Sequential Access

VAX-11 BASIC
Introduction
The new BASIC product gives the VAX user all the benefits
of a highly interactive programming environment and
high-performance development language. It combines the
best features of PDP-11 BASIC-PLUS-2 and RSTS/E BASIC-PLUS with the significant performance and addressing benefits provided by a native mode VAX language that
is fully integrated with the VMS environment.
VAX-11 BASIC is a highly extended implementation language. It provides powerful mathematic and string handling facilities, support for symbolic characters, and full
RMS indexed , sequential, and relative 1/0 operations.
There does not yet exist an ANSI standard comparable to
th is level of BASIC.

Figure 7-13
Procedure Division Using
Interactive COBOL Verbs

VAX-11 BASIC can be used as though it were either an interpreter or a compiler. A fast RUN command and support
7-14

for direct execution of unnumbered statements (immediate mode) gives the VAX-11 BASIC user the "feel " of an
interpreter. However, this product can also be used in a
compilation mode, where it generates native-mode object
modules like the other VAX compilers. In either case, VAX11 BASIC generates optimized VAX native-mode instructions which have extremely fast execution times. Typical
compilation speeds are up to 3,000 lines per minute and
computations will generally execute up to five times faster
than the same programs on a PDP-11 .

11 BASIC can have one or more statement modifiers. The
BASIC statement modifiers include FOR, IF, UNLESS, UNTIL, and WHILE constructs . Each of the statements in Figure 7-16 illustrates the use of a statement modifier.
Data Types
VAX-11 BASIC increases the number of data types available to the BASIC programmer by allowing the use of 32bit integer and 64-bit floating point data values. Tabel 7-3
below describes the data type supported by VAX-11 BASIC.

Following is a brief overview of the general characteristics
of VAX-11 BASIC .
General Characteristics
VAX-11 BASIC generates in-line native VAX-11 instructions in both its RUN and its compilation modes. The code
produced takes advantage of VAX/VMS native mode capabilities, including:

Table 7-3

Data Types

Data Type

Meaning

REAL

Specifies that the variable or constant
contains floating-point data. The precision depends on the COMPILE command qualifier you use. COMPILE/ SINGLE specifies 32-bit floating
point numbers; COMPILE/DOUBLE
specifies 64-bit floating point numbers.

WORD

Specifies that the variable or constant
contains word-length integer data, regardless of the COMPILE command
qualifier you use.

LONG

Specifies that the variable or constant
contains longword integer data, regardless of the COMPILE command qualifier
you use.

The code generated by VAX-11 BASIC uses the standard
VAX/VMS traceback facility for determ ining the source of
run-time errors. If a fatal program error should occur, an
English message is printed identifying the module and line
number where the error occurred . The English text, the
traceback , and the integrated BASIC HELP utility provide
a powerful program debugging environment.

INTEGER

Specifies that the variable or constant
contains integer data. This data type defaults to the qualifier used at compiletime. If you compile the program with the
/WORD qualifier, integers are 16 bits
long ; with the /LONG qualifier, 32 bits
long.

Object modules produced by VAX-11 BASIC can be linked
with native mode modules produced by other language
processors including BLISS, COBOL, FORTRAN , PASCAL, and MACRO.

STRING

Specifies that the variable or array contains string data.

• direct calls on operating system service routines, even
in immediate mode
• transparent access to DECnet
• direct calls to the Common Run Time Library and standard system utilities, including VAX-11 SORT /MERGE
• direct calls to separately compiled native mode procedures written in any language that utilizes the VAX
procedure call ing standard
• program sizes up to 2 billion bytes are allowed
• all modules are position-independent (PIG) and can be
run as fully re-entrant code
• the VAX-11 DEBUG facility has full support for VAX-11
BASIC

Declarations
VAX-11 BASIC allows inplicit declaration of variables. Unless specifically named in a declaration statement, a variable will be declared by its first reference. The PDP-11 BASIC-PLUS-2 convention is to implicitly type a variable or
value by the trailing character in its representation, e.g.
ABC$ is a STRING variable; XYZ% and 123% are INTEGER; T12, 314159, and 3.14 are implicitly REAL.

Structured Programming
Structured programming adds some of the features of a
block structured language (such as PASCAL) to BASIC to
allow complex programs to be written without recourse to
subroutines or obscure programming techniques. This
makes programs easier to write and maintain.

Variables can be declared in COMMON, MAP, or DECLARE statements. Both COMM.ON and MAP statements
are used to declare static storage areas-typically 1/0 records or shared data blocks. If a program has several
named common statements with the same name, the common program sections (PSECTs) are stored one after the
other. If several MAP statements have the same name,
they overlay the same PSECT.

Figure 7-15 below illustrates a record defined by a MAP
statement, successive retrievals by the use of a GET statement, and iteration controlled by a WHILE ... NEXT statement block.
The SUBPROGRAM and FUNCTION constructs in VAX-11
BASIC have structured END and EXIT statements. In addition, BASIC allows the use of statement modifiers which allow conditional or repetetive execution of the statement
without requiring the user to construct artificial loops or
block constructs. Any non-declarative statement in VAX-

The DECLARE statement is used to explicitly type variables , functions, and constants. Note that the appearance
of a variable name in a DECLARE statement implies that
7-15

&
&
&
&

EMPLOYEE RECORD DEFINITION (S)
LINE 100: THE " GENERAL DEFINITION"
LINE 200: THE " EXPANDED DEFINITION"
100

MAP (REC1)

STRING EMPLOYEE.RECORD = 36 ,
REAL RATE,
INTEGER ENDFLAG

&
&

110
200

MAP (REC1)

STRING LAST.NAME = 20,
STRING FIRST.NAME = 12,
STRING MID.INITS = 4,
REAL FILL ,
INTEGER FILL

&
&

21 0
298
299
300
310
320
330
400
410
411
490
498
499
500

&
&
&

----------------------------------------------!
FILE.NAME.1$ ="EMPLOYEE.DAT"

OPEN FILE.NAME.1$ AS FILE #1 ,SEQUENTIAL, ACCESS READ, MAP REC1
TOTAL.RATES= 000000.00

---------------

COMPUTE SUM OF RATES IN FILE-------------- -

WHILE NOT ENDFLAG
GET #1
TOTAL.RATE = TOTAL.RATE + RATE
NEXT

510
511
590
591
599
600
610
611
690
699
700
900
999

---------------

REPORT CUMULATIVE SUM BELOW

PRINT " TOTAL.RATE: $"; TOTAL.RATE

---------------

REPORT COMPLETED: CLOSE FILE(S)

CLOSE#1
END

Figure 7-15
Sample Structured Basic Program

100
110
200
210
300
310
400
410
500
510
600

A(I) = A(I) + 1

FOR

I= 1TO100

PRINT SUMMARY .DATA

OPTION .1 AND REPORT =" MONTHL Y"

PRINT FNHOUSE.PAYMENT

UNTIL

RATE < 123.45

GET#1

WHILE

EMPLOYEE.NUMBER < 76000

GOSUB 12300

UNLESS

ERROR.FLAG

PRINT " NORMAL EXIT"

TOTAL > 1000

Figure 7-16
Statement Modifiers

7-16

UNLESS ERRORS > 0

that variable will not be in static storage (see MAP, COMMON above).

The OPEN statement in VAX-11 BASIC allows specification of file crganization , access modes, file sharing, record
formats , record size, and file allocation. At the record level,
a BASIC program can FIND , GET, PUT, UPDATE, DELETE,
or RESTORE any record in a file either sequentially or randomly.

Finally, the EXTERNAL statement is provided to let the BASIC programmer explicitly declare data types for symbols
external to the current program unit, e.g. the name of a
VMS system service module, an external BASIC function,
or an external constant which is to be global in an application.

VAX-11 BASIC can access files created by other native
mode languages, assuming appropriate data representations are maintained with the records.

Figure 7-17 illustrates the use of COMMON , MAP, DECLARE, and EXTERNAL statements.

Symbolic Characters
BASIC now supports references to symbolic characters-those characters in the 96-character ASCII set which
do not print, e.g. NUL, SOH , FF, CR, etc. Figure 7-18 illustrates the use of symbolic characters in a BASIC program .

Files and Records
VAX-11 BASIC supports RMS sequential , indexed , and relative file organization . In addition, BASIC applications can
access virtual arrays (stored on files) , terminal-format
files , and block 1/0 files via RMS .

100
101
102

!
!

-------------------&
&

COMMON (DATASET1) LAST.NAME$=10, FIRST.NAME$=5
MAP STATEMENTS

203
204

205
300
301
302

305
306
307
308
309
310
311
312
401
402
404
405
406
407
408
410
500

COMMON STATEMENTS

COMMON (DATASET1) REAL A,B,C,D,E,F,G,H,O,P,Q,R,S,T,U,V,W,X,Y,Z,
INTEGER
1,J,K,L,M,N
STRING
S 1,S2,S3 ,S4

103
104
105
200
201
202

303
304

--------- -------------

-------------- -------

MAP (DATASET2)

REAL
INTEGER
STRING

PART .NUMBER, COST ,
VENDOR .CODE, QA.INDEX ,
VENDOR .ID=40

&
&

MAP (DATASET2)

REAL
INTEGER
STRING

FILL ,
FILL ,
FILL ,
FILL ,
VENDOR.NAME= 10, FILL,
VENDOR.TWX= 30

&
&
&

INTEGER
REAL
LONG
WORD
STRING

COUNTER .1, COUNTER .2,
STANDARD.DEVIATION ,
A.32 .BIT .VARIABLE ,
A.16.BIT.VARIABLE,
LAB .NAME= 20

&
&

INTEGER

CONSTANT

REAL
STRING

CONSTANT
FUNCTION

DECLARE STATEMENTS
DECLARE

!
DECLARE

DEBUG .MODE
MY.P = 3,
MY.Pl
CONCAT

&
&
&

= 0,

= 3.1416,

DEF CONCAT( STRINGY, STRING Z)
CONCAT = Y + Z
FNEND
PRINT CONCAT("THIS IS" ," THE RESULT" )
!

!-----------------------------

EXTERNAL ST A TEM ENTS

------- -------------------

!
CAN BE USED FOR VMS SERVICES
EXTERNAL

INTEGER FUNCTION SYS$ASSIGN

EXTERNAL

INTEGER FUNCTION SYS$TRNLOG

! LOGICAL TRANSLATIONS

EXTERNAL

INTEGER FUNCTION SYS$QIOW

! SYNCHRONOUS QIO CALL

!-------- --- - ----- ----- -- --------- -- -- ------ -- ------ ------- ----

Figure 7-17
Declaration Statements
7-17

• Direct execution of unnumbered BASIC statements , all owing quick verification of algorithms, inspection / change of data values , and invocation of subroutines or functions in a halted BASIC program .

PRINT " PROGRAM STARTS ... "; LF;LF;"AT "+ TIME$(0)

10
11
15
19
20

TITLE$= "SUMMARY REPORT "
PRINT TITLE$ ;CR ; FOR I = 1 TO 5

21
30
31
40
41
50
51
99

Bold copy
by overprinting

• An
integral
HELP
facility
helps
program
debug/development by providing online reference text
from the BASIC manual set.

PRINT
PRINT A(I) FOR I= 1TO10

Output report data

PRINT
END

Ready
RUN
TEST5

28-MAY-1980

17:20

PROGRAM STARTS ...
AT05:20 PM
SUMMARY REPORT
0
0
0
0
0
0
0
0
0

• The VAX-11 BASIC system can produce source language listings with embedded diagnostics indicating the
line and position of any errors. Fully descriptive diagnostic messages are provided at the point of an error.
Many error conditions are caught at compile time . At the
user's option, VAX-11 BASIC can also output a machine
language listing and/or a cross-reference listing .
• The VAX/VMS SYMBOLIC DEBUGGER lets the programmer set breakpoints , and inspect or change the
value of variables during execution of a program.
Figure 7-19 illustrates the use of several of these features .
The text appearing in blue type in Figure 7-19 corresponds
to user input, the remaining text is supplied by the BASIC
system.

The LOAD Command
A major goal of VAX-11 BASIC is to support a program development environment. The LOAD command allows a
user to stay in BASIC, even when a program under development involves several separately compiled BASIC subroutines. When a RUN command is issued, any BASIC
modules moved into memory by the previous LOAD command are automatically bound together with the module
under development and the resulting in-memory image
begins execution, i.e., the user is not required to leave BASIC, invoke the LINKER, and use the DCL $RUN command. This speeds program development considerably.

0
Ready

Figure 7-18
Symbolic Characters

CALL Facility
The CALL statement allows the BASIC programmer to invoke procedures and functions that are external to the current source module . By using the VMS native mode LINK
utility, procedures written in any of the VAX native mode
languages can be invoked, i.e., BASIC routines can call or
be called by procedures written in COBOL , CORAL ,
FORTRAN , PASCAL, etc.

Once an application has been checked out, a final call on
the LINKER can be used to create a shareable native mode
executable image for production use.
Error Handling
VAX-11 BASIC allows user-directed error and event handling . Occurence of an error can activate one or more routines which handle the error (or event) , and then return
control to the point where the error occurred (RESUME) ,
or to the calling program (ON ERROR GO BACK), or to the
BASIC system itself for standard cleanup and return of
control at the BASIC command level.

The CALL statement in VAX-11 BASIC has been extended
to allow a procedure to pass parameters BY REFerence,
BY VALUE, or BY DESCriptor. These mechanisms conform to the VAX-11 procedure calling standard and allow
BASIC programs to call VMS service routines and accept
returning status values .

In determining the cause of an error , the BASIC prog ram
can use the value of: ERR-the error message number assigned by BASIC, ERL-the line number where the error
occured, ERN$-the name of the BASIC module where the
error occurred , and ERT$(ERR)-the error message text
which the BASIC system would print if the error were not
trapped by the program .

Shareable Programs
Applications written in VAX-11 BASIC can be made shareable images by the VMS LINKER. BASIC now generates
fully re-entrant PIC code.
Developing BASIC Programs
VAX-11 BASIC delivers a high-productivity development
environment. The key features of this environment include:

Migration to VAX/VMS
During the VAX-11 BASIC Field Test , numerous sites
moved programs from BASIC-PLUS-2 and BASIC-PLUS
(on PDP-11 systems) to the VAX native BASIC . A typical
site converted literally hundreds of programs and generally had few difficulties. Minor changes were made to BASIC-PLUS-2 programs : the error checking in VAX-11 BA-

• Automatic line number generation via SEQUENCE command .
• Integral line editing with EDIT.
• A RUN command which allows a program to be placed
directly into execution without requiring a separate LINK
operation .
7-1 8

100
110
140
160
180
200
210
900

910
991
995
999

1------ --------------INPUT A FILE NAME, COUNT NUMBER OF LINES IN IT---- -- -- - ---- ----

LINPUT " What file to be opened ", FILE .NAME$
F.NAME$ = EDIT$(FILE .NAME$ ,32%)
OPEN F.NAME$ FOR INPUT AS FILE #1
ON ERROR GOTO 900
LINPUT#1 %, TEXT$
FOR I = 1to1000000
STOP
LINE . = ERL
NUMBER . = ERR
MESSAGE$= ERT$(NUMBER .)
RESUME LINE 910
PRINT ... END, FROM LINE "; LINE.;"WITH TEXT : "; MESSAGE$ ;
PRINT " - AFTER " ;! ;" RECORDS "
STOP
PRINT .. ... THE END •• •"
END

Ready
RUN NH
%BASIC-E-SYNERR, syntax error
at line 900 statement 4
RESUME LINE 910

t
Ready
HELP RESUME
RESUME
The RESUME statement marks the end of an error handling routine ,
and returns program control to a specified line number.
Format
RESUME [ < lin-num > ]
Examples
990 RESUME 300
or
990 RESUME
Ready
LIST 900
TEST6 28-MA Y- 1980 17:15
900

LINE. = ERL
NUMBER. = ERR
MESSGE$ = ERT$(NUMBER.)
RESUME LINE 910

Ready
EDIT 900 I LINE I I
900

LINE . = ERL
NUMBER . = ERR
MESSAGE$ = ERT$(NUMBER)
RESUME 910

Ready
RUN
TEST6 28-MAY-1980 17:16

? TEST6.BAS
What file to be opened
•END, FROM LINE 200 WITH TEXT: ?End of file on device - AFTER 17 RECORDS
%BAS-l-STO, Stop
-BAS-1-FROLINMOD, from line 991 in module TEST 6
Ready
PRINT MESSAGE$;" FROM FILE"; F.NAME$
?End of file on device FROM FILETEST6.BAS
Ready
PR INT F.NAME$ ;CR ; FOR I= 1 TO 5
TESTS.BAS
Ready

Figure 7-19
BASIC Program Deve lopment Features
7-19

Program A was taken from page 84 of the March , 1980 issue of BYTE magazine. Program B is very sim ilar and is
from page 130 of the June, 1980 issue of Interface Age .

SIC caught actual bugs in many "working" programs. BASIC-PLUS programs were converted to EXTEND mode (or
run through the BASIC-PLUS to VAX-11 BASIC translator)
and then modified as though they were in BASIC-PLUS-2.
Typically, these changes were made:

Finally, initial performance tests on VAX-11 BASIC were
samples of the "Towers of Hanoi " program and the Whetstone benchmark . These tests show VAX BASIC execution
speeds comparable to non-optimized VAX-11 FORTRAN .

• the MODE expression on an OPEN statement was
changed to the corresponding set of keywords, e.g .,
OPEN F$ AS FILE #1 MODE2%
becomes
OPEN F$ AS FILE #1, ACCESS APPEND

Additional Functions
The features listed below complete the promise of a BASIC that leads the competition in virtually every area. Th is
is not an exhaustive list, but does serve to indicate key capabilities of this new product.

• MAP and DIM statements were moved to occur before
OPEN statements
• RSTS/E SYS-CALLS were examined and removed if not
supported by VMS

• powerful string manipulation functions for creating , con verting , searching , editing , and extracting character
values

Files were then copied over on tape or by using DECnet,
and the programs were RUN under VAX-11 BASIC. In the
event errors were detected by BASIC, the on line HELP facility was used to determine any additional changes needed for correct compilation.

• variable names up to 30 characters long
• maxiumum length of a single string is 65 ,535 characters
• multiple statements on a line
• multiline IF ... THEN ... ELSE statements

Certain features were carried forward from PDP-11 BASIC-PLUS and PDP-11 BASIC-PLUS-2 to VAX-11 BASIC
in order to make the move to VAX easier. These include:

• optional use of line continuator " &" and statement separator " \ ", e.g .,

• BASIC-PLUS to VAX-11 BASIC Translator utility

100 PRINT
PRINT
PRINT

• Program RESEQUENCE utility from BASIC-PLUS-2
V1 .6

vs .

100
PRINT
\ PRINT
\ PRINT

&
&

• DCL pass-through in the BASIC command mode by
simply prefixing the DCL command line with a dollarsign , e.g. ,

• FIELD statement
• CVT, SWAP, and MAGTAPE functions
• Foreign buffer support

Ready
$DIR

• String arithmetic
• Numerous non-privileged RSTS/E SYS calls

*. BAS, *.OBJ

• Provision for up to ten individual BASIC object library
files for automatic use at RUN time when developing an
application using separately-compiled BASIC subrou tines .

• Virtual arrays
Performance
The programs in Figure 7-20 illustrate the level of compute-bound performance possible under VAX-11 BASIC .

PRIMES
29-MAY-1980 21 :08
90
OPEN " T.1" FOR OUTPUT AS FILE #1, RECORDSIZE
132
100
rem
Interface Age's benchmark program to
110
rem
discover the first 1000 prime numbers
120
rem
125
PRINT CHR$(7)
130
t1 = time(1)
FORn=1to1000
140
150
FOR k = 2 TO 500
160
m=n/k
170
l=int(m)
180
IF I = 0 THEN 230
190
IF I = I THEN 220
200
1Fm > ITHEN220
210
IF m =I THEN 240
220
next k
230
PRINT#1 , N;
240
NEXTn
250
t2 = time(1)
255
PRINT CHR$(7)
260
PRINT " Elapsed time: ";0.1*(t2-t1 );" seconds"
270
END

Figure 7-~.-A
Prog ram A
7-20

System

CPU

Run-time

TRS-80

Z80

1982 sec

Technico

Tl9900

585 sec

DEC-10

PDP-10

65 sec

BASIC-PLUS- 2

PDP-11 /70

11 sec

VAX-11 BASIC

11/780

2.7 sec

PRIME3
10

29-MAY-1980 21 :09
PRIME NUMBER PROGRAM #3 OF 3

&
&
&
&

FROM MARCH 1980 BYTE MAGAZINE
PAGE84
"TRS-80 PERFORMANCE...
EVALUATION BY PROGRAM TIMING
(INCLUDES 370 / 148 PL/I AND BAL TIMES
15
20
30
40
45
50
55
70
80
90
100
110
121
122
190
195
196
199
200
201

&
&

DECLARE INTEGER M,K
OPEN " PRIMES3.TMP " FOR OUTPUT AS FILE #1
PRINT " PRIME3 " + TIME$(0)
PRINT
T1=TIME(1)
PRINT #1 , 1;2;3;

C=O
M=3
M=M+2
FORK= 3 TO M / 2 STEP K-1
IF INT(M/K)'K-M = 0 THEN 190
NEXTK
PRINT#1 %, M ;
C=C+1
IF M < 10000 then 80
PRINT#1 %,"C= ";C
PRINT "C= ";C
T2 = TIME(1%)
P$ = " DONE: " +NUM1$(0.1'(T2-T1))+ " CPU SEC"
PRINT P$
PRINT #1, P$
END

TRS-80 BASIC

Z80

23470 sec

Assembler

Z80

1370 sec

Optimizing
PL/I

370/148

79 sec

BAL

370/148

56 sec

VAX-11 BASIC

11 /780

58.2 sec

Figure 7-208
Program B

declarations is provided to minimize the coding required
to use these services .

VAX-11 PL/I
Introduction
VAX-11 PL/ I is an extended implementation of the General
Purpose Subset (X3.74- "Subset G") of ANSI PL/I , X3 .531976. VAX-11 PL/I extensions to the subset language are
either full language PL/I features included because they
were highly desirable, or system-specific extensions intended to provide more complete access to VAX/VMS features . VAX-11 PL/I is a shareable compiler which runs under the VMS operating system and generates highly optimized positon-independent machine code .

Subset G is a rich language that combines the scientific
computing abilities of FORTRAN, the commercial data
handling of COBOL , the string manipulation of BASIC, and
the block structuring of ALGOL. Selected extensions further enhance these basic capabilities.
The remainder of this section :
• provides an overview of the G Subset
• lists the extension made to the language to provide enhancements for PL/I programs executing in the
VAX/VMS operating system environment

All compiler-generated code, with the exception of some
built-in functions calls and 1/0 operations , is in line. Out-ofline operations are performed by the VAX-11 Common
Run Time Library . Most high-level language operations
are supported directly by VAX hardware instructions.

• lists features of full PL/I that were excluded from the G
Subset but that have been incorporated in the implementation of VAX-11 PL/I
• lists the implementation-defined values that are used in
VAX-11 PL/I

VAX-11 PL/I supports the VAX Symbolic Debugger.
All VMS system services are available to programs written
in PL/I via the CALL statement. Furthermore , VAX-11 PL/I
fully supports RMS , the VAX/VMS record management
services. A set of ENVIRONMENT options provides access
to a large subset of RMS features . All RMS file organizations are supported : sequential , relative , and indexed .

THE G (GENERAL-PURPOSE) SUBSET
The G subset of PL/I was designed to be useful in scientific, commercial , and system programming, especially on
small and medium-size computer systems. Among the primary goals of the design of the subset were:

VAX-11 PL/I fully supports the VAX interlanguage calling
standard. Routines written in any other native mode language can call PL/I and vice versa . In addition, all
VAX/VMS system services and system utilities (Run Time
Library, SORT, etc.) are available via the PL/I CALL statement. For system services , a library of predefined ENTRY

• to include features that were easy to learn and to use
and to exclude features that were difficult to learn or
prone to error
• to provide a subset that would be easily portable from
one computer system to another
7-21

dynamically allocates storage for a based variable and
sets a pointer to the location of the allocated storage. For
example:

• to exclude features that were not often used and whose
implementation greatly increased the complexity of the
run time support required by the compiler

ALLOCATE LIST_ELEMENT SET(LIST _POINTER );

The essential elements of the subset are described below.

allocates storage for the based variable LIST_ ELEMENT
and sets LIST POINTER to the location of the allocated
storage . The allocated storage can be subsequently released by :

Program Structure
The G subset includes a complete character set , with comments, identifiers, decimal arithmetic constants, and sim ple string constants.

Begin blocks and DO-groups are included in the subset.
Each block or group in the program must be terminated
with an END statement. For example:
ON ENDFILE(INFILE) BEGIN ;
PUT SKIP LIST('End of input file ');
CLOSE FILE(INFILE) ;
END ;

FREE LIST POINTER-.LIST ELEMENT

Input/Output
The 1/ 0 statements are:

• OPEN and CLOSE.
• READ , WRITE, DELETE and REWRITE for record 1/0 .
Record 1/ 0 statements operate on an entire record in a
file.
• GET, and PUT , with FILE, STRING , EDIT, LIST , PAGE,
SKIP , and LINE options for stream 1/ 0 . Stream 1/0
statements operate on a stream of ASCII input or outp ut
data; the stream may be a file of such data or a character-string variable or expression .

Program Control
The fol lowing program control statements are included in
the subset: CALL, RETURN , IF, DO, GOTO, null , STOP ,
ON, REVERT , and SIGNAL.

The DO statement options supported are TO , BY , WHILE,
and REPEAT .

The file attributes , specified in DECLARE or OPEN, are
DIRECT , ENVIRONMENT , INPUT , KEYED , OUTP UT ,
PRINT , RECORD, SEQUENTIAL, STREAM, and UPDATE.

An IF statement may contain unlabeled THEN and ELSE
clauses. A null statement may be used to specify no action
for a given condition . For example :

The FORMAT statement is included . The format items are
E, F, P, A, X, R, PAGE, SKIP, COLUMN , TAB , and LI NE.
Format items, and the GET EDIT and PUT EDIT statements, provide a formatted 1/0 capability comparable to
that of FORTRAN . For example :

IF A < B THEN; ; • no action•;
ELSE PUT LIST( 'valid ');
An ON statement may specify a single condition . The condition names supported are ERROR, ENDFILE, EN DP AGE ,
FIXEDOVERFLOW, KEY, OVERFLOW, UNDEFINEDFILE,
UNDERFLOW, and ZERODIVIDE. For example , an attempt
to divide by zero can be detected and handled by an ONunit that specifies ZERODIVIDE:

PUT EDIT(A) (F(S ,2)) ;
writes out the value of A as a fixed-point decimal nu mber
of up to five digits, two of which are fractional.
Attributes and Pictures
The DECLARE statement is included in the subset. A ll
names must be declared , either by means of a DECLARE
statement or by means of a label prefix.

ON ZERODIVIDE BEGIN;
1• action to be taken

The attributes supported are : ALIGN ED, AUTOMATIC ,
BASED, BINARY, BIT , BUil TIN , CHARACTER, DECIMA L,
DEFINED, DIRECT, ENTRY , ENVIRONMENT, EXTERNAL,
FILE, FIXED, FLOAT, INITIAL, INPUT, INTERNAL, KEYED,
LABEL , OPTIONS, OUTPUT, PICTURE, POINTER, PRIN T,
RECORD , RETURNS , SEQUENTIAL, STATIC, STREAM ,
UPDATE, VARIABLE, and VARYING .

·1
END ;
Storage Control
The subset includes the assignment statement and the assignment of array and structure variables whose dimensions and data types match . The subset also permits aggregate promotion , that is, the assignment of a scalar
expression to every element or member of an aggregate
variable. For example:

The picture characters included are CR , DB, S, V, Z, 9, -,
+. $, and • . The picture insertion characters (. , I B) are
also included . Pictures allow special characters to be inserted in a fixed-point decimal number. The picture facility
(on output) is comparable to the PRINT USING statem ent
of BASIC. For example :

ARRAY= A,;
evaluates the expression A, and assigns the result to every
element of ARRAY .

PUT EDIT(A) (P '$99999V.99');

The subset also provides the INITIAL attribute , which
specifies initial values for variables when they are declared . For example:

Writes out the value of A in fixed-point format with five integral digits and two fractional digits, and precedes th e
number with a dollar sign .

DECLARE EOF BIT(1) STATIC INITIAL('O'B) ;
declares an "end-of-line" flag named EOF and sets its initial value to 'O'B (false) . In the subset, only static variables
may be initialized .

Built-in Functions and Pseudovariables
PL/I provides a set of built-in functions that perform
common calculations and data man ipulation . A built-i n
function may be used without declaration , wherever an
expression of the same data type is permitted . The bui lt-in

The ALLOCATE statement with the SET option and the
FREE statement are included in the subset. ALLOCATE
7-22

functions in the subset are: ABS, ACOS , ADDA, ASIN,
ATAN , ATAND , ATANH, BINARY , BIT , BOOL , CEIL,
CHARACTER , COLLATE , COPY, COS , COSD, COSH,
DATE , DECIMAL , DIMENSION , DIVIDE, EXP , FIXED ,
FLOAT, FLOOR, HBOUND, INDEX, LBOUND , LENGTH,
LINENO, LOG , LOG2, LOG10, MAX, MIN, MOD, NULL,
ONCODE , ONFILE , ONKEY , PAGENO, ROUND , SIGN,
SIN , SINO , SINH, SQRT, STRING, SUBSTR, TAN , TAND,
TANH , TIME , TRANSLATE , TRUNC, UNSPEC, VALID , and
VERIFY .
Subset G also provides pseudovariables that can be used
as the targets of assignment statements . The pseudovariables are PAGENO , STRING , SUBSTR , and UNSPEC . For
example, whereas the SUBSTR built-in function returns a
substring of a specified bit or character string, the
SUBSTR pseudovariable allows the assignment of an
expression to such a substring .
Expressions
The subset supports all infix and prefix operators , the locator qualifier, parenthesized expressions , subscripts, and
function references . Implicit conversion from one data
type to another is restricted to those contexts in which the
conversion is likely to produce the desired results. For example :

The READ ONLY attribute can be applied to a static
computational variable whose value does not change .
3. The VALUE attribute defines a variable that is, in effect, a constant whose value is supplied by the linker.
The following extensions to ON condition handling provide
support for condition handling in the VAX/VMS environment:
1.

3.
4.

Support of VAX-11 Record Management Services
The options of the ENVIRONMENT attributes provide support for many of the features and control values of the
VAX-11 Record Management Services (RMS). Additional
extensions have been made to the PL/I language to augment this support, as described below.
1.

DECLARE I DECIMAL ;
I= '1.2' ;
converts the character string ' 1.2' to the decimal equivalent and assigns it to the decimal variable I. However,
the following assignment:

The OPTIONS option is supported on the GET, PUT,
READ , WRITE, REWRITE, and DELETE statements.
For example :
GET LIST(A) OPTIONS(PROMPT('Enter A>'));

DECLARE B BIT(8) ;
B = 'A';
is not valid because, to be convertible to a bit string , a
character string must consist entirely of 0 and 1 characters .

displays the prompt " Enter A" on the user's terminal.
These built-in subroutines provide file handling and
control functions : DISPLAY, EXTEND, FLUSH,
NXTVOL, REWIND, and SPACEBLOCK.

Miscellaneous Extensions and Deviations
The following list summarizes miscellaneous extensions
and deviations.

VAX-11 EXTENSIONS TO THE
G SUBSET STANDARD

Procedure-Calling and Condition-Handling Extensions
The following extensions to PL/I were made to allow VAX11 PL/I procedures to call procedures written in any other
programming language that also supports the VAX-11
calling standard.

The attributes ANY and VALUE describe how data are
to be passed to a called procedure.
2. The VARIABLE option for the ENTRY attribute permits
a PL/I procedure to call a non-PL/I procedure with an
argument list of variable length . It also permits a procedure to omit arguments in an argument list.
3. The DESCRIPTOR built-in function may be used to
pass an argument by descriptor to a non-PL/I procedure .
The following new attributes provide storage classes for
PL/I variables. These attributes permit PL/I programs to
take advantage of features of the VAX-11 linker and to
combine PL/I procedures with other procedures that use
these storage classes .
1.

The ON statement supports the ANYCONDITION keyword . The ON-unit established by this keyword is executed when any condition occurs for which no explicit
ON-unit exists.
The ON statement supports programmer-named
conditions with the VAXCONDITION keyword .
The RESIGNAL built-in subroutine permits an ON-unit
to keep a signal active .
The ONARGSLIST built-in function provides an ONunit with access to the mechanism and signal arguments of an exception condition .

The RANK and BYTE built-in functions are supported ,
which return the ASCII code for a given character and
the ASCII character for a given code, respectively.
The expression in a WHILE clause or in an IF statement may be a bit string of any length . When evaluated , the expression results in a true value if any bit of
the string expression is a one and in a false value if all
bits in the string expression are zeroes.
The control variable and the expressions in the TO,
BY , and REPEAT options of the DO statement are not
restricted to integers and pointers.

FULL PL/I FEATURES SUPPORTED
The items listed below are features that are explicitly excluded from the subset standard but that have been implemented in VAX-11 PL/I. These features all exist in full PL/I.
1.

The GLOBALDEF and GLOBALREF attributes let you
define and access external global variables and optionally to place all external global definitions in the
same program section .

2.
3.
7-23

The ENTRY statement is supported. The ENTRY statement provides a means of defining alternate entry
points to a procedure . For example, such an alternate
entry point can be invoked by a function reference
even though the containing procedure is invoked as a
subroutine.
The ENVIRONMENT option is supported on the
CLOSE statement.
The picture characters Y, T, I, and R are supported,

and pictures may include iteration factors . These
characters provide an additional means of representing zeros in a number (Y) and allow a digit and a sign
to be represented by a single character (T, I, R) .
4. RETURNS CHARACTER(*) is valid. That is, a function
can return a character string whose length is determined by the program.
5. The FINISH condition is supported.
6. A REWRITE statement need not specify the FROM option if the most recent 1/0 operation on the file was a
READ statement with the SET option. (A READ SET
statement acquires a record from an input file and
sets a pointer to the location of the 1/0 buffer containing the contents of the record.)
7. The AREA and OFFSET attributes are supported . The
AREA attribute allows declaration and manipulation of
an entire region of memory (up to 500 million bytes) ,
and the OFFSET attribute declares variables that are
offsets into such an area. Offset variables may be
used in most contexts in which pointer variables are
permitted . Allocation within an area must be controlled by a user-written procedure .
8. The OFFSET and POINTER built-in functions are supported . These functions convert pointer values to offset values, and vice versa .
9. The POSITION attribute is supported. POSITION allows the declaration of a DEFINED variable to specify
the position within the base string at which the
definition begins.
10. Automatic variables may be initialized. The INITIAL attribute may contain scalar expressions and asterisks
with automatic variables. Asterisks indicate that the
corresponding variable or element in the declaration
is not assigned an initial value .
11 . The SET option is optional on the ALLOCATE statement if the allocated variable was declared with
BASED(pointer-reference) . The ALLOCATE statement then allocates storage for the based variable and
sets the referenced pointer variable to the storage location.
12. The character pair /* may be embedded in a comment. This feature permits a statement such as:
IF tVALID(P) THEN
PUT LIST('lnput error');/* check validity* I

The maximum record size for SEQUENTIAL files is
32,767 bytes minus the length of any fixed-length control area .
4. The maximum key size is 255 bytes for character keys.
5. The default value for the LINESIZE option is as fol lows. The line size is used by stream 1/0 statements to
determine when to go to a new line.
•
If the output is to a physical record-oriented device, such as a line printer or terminal , the default
line size is the width of the device.
•
If the output is to a print file, the default line size is
132.
•
If the output is to a nonrecord device (magnetic
tape), the default line size is 510.
6. The default value for the PAGESIZE option is as follows. The page size is used by stream output (PUT)
statements to determine when to signal the ENDPAGE
condition.
•
If the logical name SYS$LP _LINES is defined , the
default page size is the numeric value of
SYS$LP LINES - 6.
•
If SYS$LP - LINES is not defined, or if its value is
less than 30 or greater than 90 , or if its value is not
numeric, the default page is 60.
7. The values for TAB positions are columns beginning
with column 1 and every eight columns thereafter : 1,
9, 17, 25, .. 8*i + 1, where i is (line size)/8 . List-directed
output (by the PUT LIST statement) is positioned on
tab stops if the output is to a file with the PRINT attribute.
8. The maximum length allowed for a file title is 128 characters. File titles are VAX/VMS file specifications that
are associated with PL/I file constants and file variables .
9. The maximum number of digits in editing fixed-point
data is 34.
10. The maximum numbers of digits for each combination
of base and scale are:
FIXED BINARY -31
FIXED DECIMAL-31
FLOAT BINARY -113
FLOAT DECIMAL-34
11 . The default precision for integer values is 31 .
12 . The maximum number of arguments that can be
passed to an entry point is 253.

to be "commented out":
I* IF tVALID(P) THEN
PUT LIST( 'lnput error');/* check validity* I

VAX-11 PL/I Programming Example
Figure 7-21 illustrates a VAX-11 PL/I program . This program calls a VAX/VMS system service (SYS$TRNLOG) to
determine the equivalence string for a logical name.

IMPLEMENTATION-DEFINED VALUES
AND FEATURES
1.
2.

The program TRNLN calls a VAX/VMS system service ,
SYS$TRNLOG , to determine the equivalence string for a
logical name. SYS$TRNLOG is declared as an external entry constant, with three parameters:

VAX-11 PL/I supports the full 256-character ASCII
character set.
The default precisions for arithmetic data are :
FIXED BINARY (31)
FIXED DECIMAL (7)
FLOAT BINARY (24)
FLOAT DECIMAL (7)

1.
2.

where each precision is the number of binary or decimal digits, as appropriate.

7-24

A character string of any length, representing the logical name.
An integer representing the length of the translated
name.
A character string of any length representing the
translated name itself.

I* Translates logical names to equivalent strings• I

TRNLN : PROCEDURE OPTIONS (MAIN);
DECLARE SYS$TRNLOG ENTRY(

I* VAX/VMS system service •I
I Parameters: ·I
I* Logical name ·I
I * Length of translated name •I
I* Translated name• I

CHARACTER( *),
FIXED BINARY(15) ,
CHARACTER( *)
)

OPTIONS(VARIABLE)
RETURNS(FIXED BINARY(31)) ;

I* Some arguments optional •I
I * SYS$TRNLOG returns integer* I

DECLARE INPUT CHARACTER(63) VARYING ,
I* Input name* /
OUTPUT CHARACTERS(63),
/* Translated name • I
OUTPUT_LEN FIXED BINARY(15) ,
I* Length of translated name •I
RETURN _STAT FIXED BINARY(31),
I* Return status of SYS$TRNLOG *I
SUCCESS BIT(1)
I* Successful return •I
ALIGNED BASED (ADD(RETURN_STAT)) ;
DECLARE SS$ NO TRAN
GLOBALREF FIXED BINARY(31) VALUE ;
%REPLACE NOTEND BY '1'B;
ON ENDFILE(SYSIN) STOP;

I* Flag for undefined name*/

I* " Always true"--to control DO-group• I
I* Stop on CTRL/Z from terminal • I

DO WHILE (NOTEND);
PUT SKIP;
GET LIST( INPUT)

I* Terminated by ON-unit* /

OPTIONS(PROMPT('Enter logical name '));
/* Invoke sytem service as PL/I function reference :• I
RETURN _STAT = SYS$TRNLOG((INPUT),OUTPUT _LEN ,OUTPUT,,,);
IF RETURN _STAT = SS$_NOTRAN THEN
PUT SKIP LIST(INPUT: :'not defined') ;
ELSE IF SUCCESS THEN
PUT SKIP LIST(INPUT::'is ':: SUBSTR(OUTPUT, 1,0UTPUT _LEN)) ;
END;
END TRNLN ;
Figure 7-21
Sample VAX-11 PL/I Code

written argument for such a parameter is a varying-length
string, a dummy argument is created by the compiler . To
avoid the accompanying warning message, the argument
INPUT is enclosed in parentheses.

The declaration also states that SYS$TRNLOG returns an
integer and can have an argument list of variable length .
Note that the explicit declaration of SYS$TRNLOG is
shown here for clarity . VAX-11 PL/I is supplied with a library of predefined entries for VAX/VMS system services , so
the declaration of SYS$TRNLOG can be replaced by the
single statement:

When a dummy argument is created, the invoked procedure cannot modify the associated parameter . Therefore , the translated name (OUTPUT) is not declared as a
varying-length string . Instead , OUTPUT and OUTboth
are supplied
as arguments to
PUT _LEN
SYS$TRNLOG , wh ich then assigns the necessary values to
them . The translated name is written out with a reference .
to the SUBSTR built-in function:

%INCLUDE SYS$TRNLOG;
The program also uses a global reference to
SS$_N OTRAN ; the return value of the system service
equals SS$ _NOTRAN when there is no defined equivalence for the given logical name.

SUBSTR(OUTPUT, 1,0UTPUT_LEN);

SYS$TRNLOG actually requires that its first and third arguments be the addresses of character string descriptors.
VAX-11 PL/ I allows you to pass a character string by descriptor by declaring the correspond ing parameter as
CHARACTER(*) . Such a parameter can have an argument
that is a fixed-length character string of any length . If the
7-25

which writes out a substring beginning at character 1 of
OUTPUT and continuing for OUTPUT_ LEN characters.
A sample session with the program might be :
$DEF LNK$LIBRARY DB1 :[PROJECT]MYLIB,OLB<RET>
$ R TRNLN < RET >

Enter logical name > LNK$LIBRARY<RET>
LNK$LIBRARY is DB1 :[PROJECT]MYLIB.OLB<RET >
Enter logical name > GRACE < RET >
GRACE not defined
Enter logical name > tZ

• Standard functions and procedures
In addition , VAX-11 Pascal incorporates the following extensions to standard Pascal , some of which are common in
other Pascal implementations:
• Lexical
Upper- and lowercase letters treated identically except in character and string constants
New reserved words: MODULE , OTHERWISE, SEQUENTIAL , VALUE , %DESCR, % 1MMED , %INCLUDE, and %STDESCR
The exponentiation operator , ••
Hexadecimal and octal constants
DOUBLE constants
$and _ characters in identifiers
• Predefined data types

VAX-11 PASCAL
Introduction
VAX-11 Pascal , a re-entrant native mode compiler, is an
extended implementation of the Pascal language as defined by Jensen and Wirth in the Pascal User Manual and
Report(1974) .
Particularly suited to instructional use, Pascal is also an in creasingly popular general purpose language. It implements a well-chosen, compact set of general purpose language features. In addition, portability is easily achievable
in programs written in Pascal.

DOUBLE
SINGLE
• Predefined procedures

Block structuring and flexible data types make Pascal a
good language for commercial users. It is also suitable for
systems programming and research applications.

CLOSE (f)
FIND (f,n)
OPEN (f, ... )
DATE (a)
HALT
LINELIMIT (f,n)
TIME(a)
• Predefined functions

VAX-11 Pascal takes advantage of the VAX-11 hardware
floating point, character instruction sets, and virtual memory capabilities of the VAX/VMS operating system. Many
of the features com man to other languages of VAX/VMS
are available through VAX-11 Pascal, including :

LOWER(a,n)
SNGL (d)
UPPER (a,n)
EXPO (r)
CARD (s)
CLOCK
UNDEFINED (r)
• Extensions to procedures READ and WRITE

• separate compilation of modules
• standard call interface to routines written in other languages
• access to VAX/VMS system services
At compile time, options available to the process include:
• run-time checks for illegal assignment to set and subrange variables, and illegal array subscripts

READ (or READLN) of user-defined scalar type
READ (or READLN) of string
WRITE (or WRITELN) of user-defined scalar type
WRITE (or WRITELN) of any data us in g
hexadecimal or octal format
• %INCLUDE directive

• cross-reference listing of identifiers
• source program listing
• machine code listing
• generation of some DEBUG and TRACEBACK records
for the VAX-11 Symbolic Debugger
Though VAX-11 Pascal has access to the Common Run
Time Library routines of VAX/VMS, it also has Pascal-specific Run Time Library routines installed in ST ARLET.OLB .
Such routines primarily provide 1/0 interfaces to the Record Management Services (RMS) .

• VALUE initialization

Standard Pascal provides a modular, systematic approach
to computerized problem solving . Major features of the
language are:

• Extended parameter specifications

• OTHERWISE clause in CASE statement
• External procedure and function declarations
• Dynamic array parameters
%DESCR
%1MMED
%1MMED PROCEDURE and %1MMED FUNCTION
%STDESCR
• Separate compilation of procedures and functions . (A
separate compilation unit is termed a MODULE and several routines may be part of a MODULE. Each MODULE
is eventually embedded in a host or main program .)

• INTEGER , REAL , CHAR , BOOLEAN, user-defined, and
subrange scalar data types
• ARRAY, RECORD, SET, and FILE structured data types
• Constant identifier definition
• FOR, REPEAT, and WHILE loop control statements

The OPEN , CLOSE and FIND procedures extend the 1/ 0
capabilities of the VAX-11 Pascal language. The OPE N
procedure can contain file attributes that define the creation or subsequent processing of the file . A FIND procedure is another extension to the language for direct access to sequential files of fixed length records. The stan-

• CASE and IF-THEN-ELSE conditional statements
• BEGIN ...END compound statement
• GOTO statement
• GET, PUT, READ , WRITE, READLN , and WRITELN 1/0
procedures

7-26

dard 1/0 procedures of GET , PUT , READ, WRITE,
READLN and WRITELN are also available in VAX-11 Pascal.

• Source code listing-When the LIST qualifier is specified , the source code is listed by default.
• Machine code listing-To generate the machine code
listing , the MACHINE_CODE qualifier must be specified .

The extended parameter specifications %DESCR , %
IMMED, and %STDESCR are added to the Pascal language to denote the method of argument passing when
calling a system service, procedure, or function not written
in VAX-11 Pascal (for example, in VAX-11 FORTRAN or
MACRO.)

• Cross-reference listing-To generate cross references
for all identifiers used in the program , the
CROSS_REFERENCE qualifier must be specified.
The following sections describe the format of the user
requested listings. Note that the numbers in these sections
are keyed to the circled numbers in the listings of Figure 722 .

Sample VAX-11 Pascal Code
Figure 7-22 illustrates a VAX-11 Pascal program which a
user may write to calculate the hypotenuse of a right triangle. The inputs to the program are the length of the two
sides of the triangle and the output is the length of the hypotenuse. This version of the program has not yet been
debugged to illustrate how error messages are reported.

Title Line - Each page of the listing contains a title line.
The title line lists the module name (1 ). the date and time of
the compilation (2) , the Pascal compiler name and version
number (3), and the listing page number (4).
Source Code Listing
Each page of the source code listing contains a line under
the title line specifying the date and time of source file
creation (5) and the VAX/VMS file specification of the
source file (6) .

Compiler Listing Format
Figure 7-22 contains the compiler listing of the hypotenuse
program. The compiler listing contains the following three
sections:

(i>

0
EXAMPLE

23-MA Y-1980
11-0CT-1979

LEVEL
PROC STMT

LINE
NUMBERS

0
VAX-11 PASCAL VERSION V1 .1-29
DB1.[200,200]EXAMPLE PAS:4 (1)

23:00:05
11 :34:47

STATEMENT

program EXAMPLE(INPUT,OUTPUT) ;

100
1
200
2
(;\ 300 0 3
v 400 4
500
5
600
6
700
7
800
8
900
9
1000 10
1100 11
1200 12
%PAS-W-DIAGN

label 10;
var A,B,C:REAL ;
beg in
repeat
WRITELN('Enter triangle sides ');
if EOLN(INPUT) then goto 10;
@ 2
READLN(A ,B) ;
2
C := (SQR(A) + SQR(B)) •• 0.5;
2
t450
.. . WARNING 450 : nonstandard Pascal : Exponentiation
2
WRITELN('Hypotenuse is: ', C) ;
2
until FALSE;

0
2

1300 13
1400 14
1500 15
1600 16
1700 17
% PAS - F-DIAGN

")
10:
WRITELN( ' Done';
t20: ,4

17 == >

18
19
20

end .

Compilation time =

1.35 seconds (

889 lines / minute) .

/-:,;\

2 Errors

12 == >

;; .. • ERROR
4: ")" expected
._, • • • ERROR 20: " ," expected
1800
1900

PAGE 1

1 Nonstandard feature ~

Last error(warning) on line

17. @

Active options at end of compilation :
a
NODEBUG ,STANDARD ,LIST ,NOCHECK,WARNINGS ,CROSS REFERENCE, 0
MACHINE-CODE,OBJECT ,ERROR_LIMIT = 30
-

Figure 7-22
Sample VAX-11 Pascal Code
7-27

EXAMPLE 23-MAY-1980
MACHINE-CODE

GENERATED CODE
(PRIOR TO BRANCH OPTIMIZATION)
LINE
ADDRESS
OPCODE
OPERANDS
0002
0009
0010
0012
0014
0017
001E
0021
0025
002C
0030
0035
0038 @
003A
0041
0045
0049
0040
0054

MOVAB
MOVL
CLRL
CLRD
PUS HAL
CALLS
PUS HAL
PUS HAL
CALLS
MOVL
MOVAB
SUBL2
PUSHL
MOVAB
MOVL
MOVL
MOVL
CALLS
MOVAB

EXAMPLE 23-MAY-1980
CROSS REFERENCE

PAGE2

VAX-11 PASCAL VERSION V1 . 1-29

23:00:05

BYTESTREAM (HEXADECIMAL; READ FROM RIGHT TO LEFT)

·vAR(O , O),R11
R13,*VAR(O, 4)
-(R14)
-(R14)
(R11)Bf8
#1,PAS$1NPUT
(R11)Bf8
(R11 )Wf236
#2,PAS$0UTPUT @
R14 ,(R13)Bf-12
(R11)Wf236,R10
#16,R14
R10
·vAR(2, O) ,R9
R9,(R14)Bf4
#21 ,(R14)Bf12
#21,(R14)Bf8
#5,PAS$WRITESTR
(R11)Wf236,R10

23:00:05

58 00 00 00 00 00 9E
00 00 00 00 00 50 DO
7E 04
7E 7C
08 AB OF
00 00 00 00 00 01 FB
98 AB OF
00 EC CB OF
00 00 00 00 00 02 FB
F4 AD 5E DO
5A 00 EC CB 9E
@ 5E10C2
5A DD
59 00 00 00 00 00 9E
04 AE 59 DO
OCAE15DO
08 AE 15 DO
00 00 00 00 00 05 FB
5A 00 EC CB 9E

VAX-11 PASCAL VERSION V1. 1-29

PAGE4

CROSSREFERENCE LISTING
A
B@

INPUT
OUTPUT

4
4
@ 4
1
1

11
11
12
10

0
0
0
0
0
0

10
14
11
4
12
9

12
12 @
13

GLOBALLY DEFINED IDENTIFIERS:
EOLN
FALSE
READLN
REAL
SQR
WRITELN

0
12
13

Figure 7-22 (Con 't)
Sample VAX-11 Pascal Code

block. The procedure level number increases by one for
each nesting level of functions or procedures .

Source Code Listing - The lines of the source code are
printed in the source code listing. In addition, the listing
contains the following information pertaining to the source
code:

• Statement level (10)-The listing specifies a statement
level for each line of source code after the first BEGIN
delimiter. The statement level starts at 0 and increases
by 1 for each nesting level of Pascal structured statements. The statement level of a comment is the same
level as that of the statement that follows it.

• SOS line numbers (7)-lf the source lines were created
or edited in a Pascal module with the SOS editor, SOS
line numbers appear in the leftmost column of the
source code listing. SOS line numbers are irrelevant to
the Pascal compiler .

Errors and Warnings - The source code listing includes
information on any errors or warnings detected by the
compiler . A line beneath the source code line in which the
error is detected specifies whether the diagnostic is a
warning or an error. In addition, the error description can
contain the following information:

• Line numbers (8)-The compiler assigns unique line
numbers to the source lines in a Pascal module. The
symbolic traceback that is printed if the program encounters an exception at run time refers to these line
numbers.

• A circumflex (t) that points to the character position in
the line where the error was detected (11 ).

• Procedure level (9)-Each line that contains a declaration lists the procedure level of that declaration. Procedure level 1 indicates declarations in the outermost

• A numeric code, following the circumflex, that specifies
7-28

the particular error (12) . On the following lines of the
source listing , the compiler prints the text that corresponds to each numeric code (13). Note that one source
program error often causes the Pascal compiler to detect more than one error ( 14 ).

• Globally-defined identifiers (27)-This section lists the
Pascal predefined identifiers that the program uses.
Each line of the cross-reference listing contains an identifier (28) and a list of the source line numbers where the
identifier is used (29) . The first line number indicates
where the identifier is declared . Predefined identifiers are
listed as if they were declared on line 0. The cross-reference listing does not specify pointer type identifiers that
are used before they are declared.

• An asterisk (*) that shows where the compiler resumed
translation after the error (15).
• The line number in which the error was detected (16)
and the line number of the last line containing an error
diagnostic (17) . These error line numbers can be used
to trace the error diagnostics backwards through the
source listing .

VAX-11 BLISS-32
Introduction
BLISS-32 is a high level systems implementation language
for VAX-11. It is specifically designed for building software
such as compilers , real-time processors, and operating
systems modules. (For example, the VAX-11 FORTRAN
compiler is coded in BLISS, as is most of the VAX-11 RTL.)
It contains many of the features of high-level languages
(e .g., DO loops, IF-THEN-ELSE statements, automatic
stack and register allocations, and mechanisms for defining and calling routines) but also provides the flexibility, efficiency, and access to hardware which one would expect
from an assembly language. The BLISS compiler produces highly-optimized object code which is typically within 5% to 10% of the efficiency of code produced by an
experienced assembly language programmer. The VAX11 BLISS-32 compiler is written entirely in BLISS and runs
in native mode under the VAX/VMS operating system .

Summary - At the end of the source listing , the compiler
lists the amount of time required for the compilation (18) . If
program generated warning or error messages resulted ,
the compiler prints a summary of all the errors (20) and the
source line number of the last message (19) . Finally, the
comp iler lists the status of all the compilation options (21 ).
Machine Code Listing
The machine code listing (if requested with the MACHINE_CODE qualifier) follows the source listing. The machine code listing contains:
• Symbolic representation (22)-The symbolic representation , similar to a VAX-11 MACRO instruction, appears
for each object instruction generated . Because the Pascal compiler operates in one pass , it must generate
these instructions before it performs branch optimization. Branch optimization can cause certain BRW instructions to be deleted. Therefore, these instructions
will not be identical to those appearing in the executable
image.

Features of BLISS-32
VAX-11 BLISS-32 has several characteristics which set it
apart from other high-level languages:

• Source line number (23)-A source line number marks
the first object instruction that the compiler generated
for the first Pascal statement on that source line.

• Data-BLISS-32 is "type-free": all data are manipulated
as values from 1 to 32 bits in length . The interpretation
of any data item is provided by the operator applied to it.
A value can be fetched from or assigned to any contiguous field of from 1 to 32 bits located anywhere in the
VAX-11 virtual address space. The expression
Y <4,4 > =0 deposits zero into bits 4 through 7 of location Y. This BLISS expression generates the single VAX11 instruction : BICB2 #tXFO ,Y.

• Hexadecimal address (24)-The hexadecimal address
is an approximation of the address of the object instruction . Do not use these addresses for debugging purposes because they do not correctly correspond to the
locations in the executable image. The branch optimization mentioned above can change the addresses of the
object instructions.

• Value Assignments-all names in BLISS-32 represent
addresses. Contents of storage locations are accessed
by means of a fetch operator(.). Hence, the expression
X= .Y+3 is interpreted as adding 3 to the contents of location Y, then assigning the result to the storage location beginning at X.

• Hexadecimal instruction (25)-This is the hexadecimal
representation of the object instruction . The hexadecimal instruction should be read from right to left because
the rightmost byte has the smallest address. Again,
because of its one-pass operation , the compiler must
generate some object instructions before it can determine the address bytes of their operands. The addresses of these operands are printed as zeros . After
generating the hexadecimal representation of an instruction , but before writing the object code file , the
compiler places the correct values into the binary object
code.

• Operators-BLISS-32 supplies operators which interpret their operands as addresses, signed integers, unsigned integers or character-sequences.
• Expressions-BLISS-32 permits construction of
complex expressions in which several different kinds of
operations can be performed in a single program statement. For example, the expression 2*(8=.C+1) calculates 2*(.C+1) and simultaneously assigns the value of
.C+1toB.

Cross-Reference Listing
The cross-reference listing (if requested with the
CROSS_REFERENCE qualifier) appears after the machine
code listing . It contains two sections:

• Structures-BLISS-32 defines such data structures as
VECTOR, BLOCK, BITVECTOR, and BLOCKVECTOR.
In addition , the programmer can define arbitrary data
structures specifically designed for a given application .

• User-specified identifiers (26)-This section lists all the
user declared identifiers.
7-29

Other BLISS-32 features include:

• BUILTIN declarations allow use of VAX-11 machi nespecific functions for access to VAX-11 features not otherwise provided in the BLISS-32 language. The comp ilation of a machine specific function results in the generation of inline code , often a single instruction , rathe r
than a call to an external routine. Machine specific functions generally have the same name as t he ir
corresponding VAX-11 instructions (e .g ., ADAW I,
BISPSW , CRC , HALT , INDEX , MTPR , PROBER ,
REMQUE , MOVP, etc.). Over 50 such functions are provided . (The complete list is shown in Table 7-4) .

• CASE , SELECT, SELECTONE , and IF-THENELSE-providing for sequencing of instructions based
on evaluation of expressions at run-time.
• DO, WHILE, and UNTIL-providing for looping until a
particular condition is satisfied .
• INCR, DECR-providing for iterative looping , incrementing or decrementing a loop index.
• EXITLOOP and LEAVE-providing for early termination
of loops and for exiting a BEGIN-END block. (There is
no GO TO in BLISS .)
• Condition Handl ing-the BLISS-32 language fully supports a VAX/VMS condition-handling environment. The
ENABLE declaration establishes a condition-handler for
exceptions raised within the scope of a routine . The
SIGNAL , SIGNAL _STOP and UNWIND predeclared
functions allow a programmer to raise conditions and
process them .

Table 7-4
VAX-11 Machine Specific Functions

Processor Register Operations

• GLOBAL and EXTERNAL declarations-allowing code
and data to be shared among several modules.

MTPR
MFPR

• LOCAL , STACKLOCAL , and REGISTER declarations- allowing dynamic stack-like allocation using either the execution stack or the general registers.

Move to a Processor Reg ister
Move from a Processor Reg ister

Parameter Validation Operations
PROBER
PRO BEW

• REQUIRE and LIBRARY declarations-allowing source
material from subsidiary files to be included in the module at compile time .

Probe Read accessibil ity
Probe Write accessibility

Program Status Operations

• LEXICAL functions-allowing a variety of compile-time
operations such as concatenation of strings , construction of names, testing properties of macro parameters,
testing compiler switch options, generating compiler
diagnostic messages, and controlling macro expansion .

MOVPSL
BISPSW
BICPSW

Move from PSL
Bit set PSW
Bit clear PSW

Queue Operations

• Conditional Compilation-The programmer may include or exclude source text from compilation through
use of %IF-%THEN-%ELSE-%FI. In conjunction with the
lexical functions, this provides a powerful capability for
tailoring source code for distinct environments.

INSQUE
REMQUE

Insert entry in Queue
Remove entry from Queue

Bit Operations

BLISS-32 also provides a number of machine-dependent
features specifically designed to complement the VAX architecture and VAX/VMS. These include the following :
• LINKAGE declarations allow the programmer to make
full use of the VAX-11 call facilities; in particular, you
may specify either the CALLS/CALLG/RET or
JSB/BSB/RSB call and return sequences, pass parameters in general registers or on the stack , and control
the use of registers by a routine or across a set of routines .

TESTBITSS
TESTBITSC
TESTBITCS
TESTBITCC

Test for Bit Set, then Set bit
Test for Bit Set, then Clear bit
Test for Bit Clear, then Set bit
Test for Bit Clear, then Clear bit

TESTBITSSI
TESTBITCCI
FFS
FFC

Test for Bit Set, then Set bit Interlocked
Test for Bit Clear. then Clear b it Interlocked
Find First Set bit
Find First Clear bit

Extended Arithmetic Operations
ASHQ
EDIV
EMUL
INDEX
CRC

• PSECT declarations provide information to the linker regarding storage requirements for various sections of a
program . For example, a particular data segment may
be designated as READ or NOREAD, SHARE or NOSHARE, LOCAL or GLOBAL, and so on .

Arithmetic Shift Quad
Extended Divide
Extended Multiply
Index (Subscript) Calculation
Cycl ic Redundancy Calculation

Floating Point Conversion Operations

• System lnterfaces-STARLET.REQ (or STARLET.L32)
prov ides keyword macros for all VAX-11 RMS and
VAX/VMS system services , as well as symbolic definitions for system data structures and completion
codes. LIB.REQ (or LIB .L32) provides the defin itions in
STARLET , as well as definitions needed for writing
ACPs or privileged kernel-mode code.
7-30

CVTLF
CVTLD
CVTFL
CVTDL

Convert Long to Floating
Convert Long to Double
Convert Floating to Long
Convert Double to Long

CVTFD

Convert Floating to Double

Library and Require Files
BLISS-32 provides two methods for including commonly
used text into BLISS programs at compile time. These involve use of either library files or Require files:

Table 7-4 (con't)
VAX-11 Machine Specific Functions

CVTDF
CVTRDL
CVTRFL

Convert Double to Floating
Convert Rounded Double to Long
Convert Rounded Floating to Long

• Library Files-These are special files created by the
compiler in a previous library compilation and are invoked by the LIBRARY declaration in the BLISS source
program .

CMPF
CMPD

Compare Floating
Compare Double

• Require Files-These are source (text) files which are
invoked via the REQUIRE declaration in the BLISS
source program .

String Operations
MOVTUC
SCA NC
SPANG

Since Library files are "precompiled," lexical processing
and declaration parsing and checking need not be repeated each time these files are included in a compilation ; their
use results in a considerable reduction in total compilation
time.

Move Translated Until Character
Scan Characters
Span Characters

Decimal String Operations
MOVP
CMPP
CVTLP
CVTPL

Move Packed
Compare Packed
Convert Long to Packed
Convert Packed to Long

CVTPT
CVTTP
CVTPS
CVTSP

Convert Packed to Trailing Numeric
Convert Trailing Numeric to Packed
Convert Packed to Leading Separate Numeric
Convert Leading Separate Numeric to Packed

EDITPC

Edit Packed to Character

The contents of Require files must be fully processed each
time the file is used in a compilation . Hence, using Require
files will, in general , be less efficient than using Library
files . However, since these files operate under a less stringent set of syntactical rules, their use may be warranted in
situations where a higher level of flexibility is desired .
Macros
VAX-11 BLISS-32 provides an extensive macro-building
facility, allowing frequently used groups of declarations or
expressions to be expressed in an abbreviated way.
Macros are defined via MACRO declarations and are accessed by simple call statements. They are fully expanded
at compile time . BLISS-32 allows parameters to be specified in the macro definition, thus allowing each block of
text to be specialized by the actual parameters passed to
it. Macros may be positional or keyword , and may be simple, iterative, or conditional.

Miscellaneous Operations
HALT
ROT
ADAWI
BPT

Halt Processor
Rotate
Add Aligned Word Interlocked
Breakpoint

CHMx
CALLG
NOP

Change Mode
Call with General Argument List
No Operation

Debugging
The VAX-11 BLISS-32 compiler produces a list of error
messages showing the source program line on which the
error occurred followed by a description of the error. If the
error is recoverable , then the compiler will generate a
"warning " diagnostic and continue with the compilation
process. If the error is serious enough to invalidate the
compiler's internal representation of the module, then an
"error" diagnostic is generated, and processing ceases
following the syntax checking-no object module is produced .

The VAX-11 BLISS-32 Compiler
The VAX-11 BLISS-32 compiler performs a number of optimizations . These include common subexpression elimination , removal of loop invariants, constant folding, block
register allocation, peephole replacement, test instruction
elimination, jump vs . branch instruction resolution, branch
chaining, and cross-jumping.

If an error occurs at execution time, the process image can
access the VAX-11 Symbolic Debugger program. This
program may be accessed when the object module is
linked with the DEBUG option . The DEBUG program allows the programmer to examine and deposit values in
storage, set breakpoints, call routines, trace through a
program as it executes, and perform other operations useful in checking out a program. VAX-11 DEBUG understands BLISS syntax and permits the use of the user's
symbolic names. (See the section on the Symbolic Debugger for a further description of VAX-11 debugging facilities.)

The VAX-11 BLISS-32 compiler optionally produces
source text and generated code in a format closely resembling a VAX-11 assembly listing . Other options allow the
programmer to control the degree of optimization, suppress production of object code, determine types and formats of output listings, generate traceback information ,
and specify the types of information to be listed at the terminal.
BLISS-32 generates shareable, re-entrant and position-independent code. These features make it easy to use
BLISS-32 for writing software to be included in shareable
libraries for use by any native language. it is also useful for
writing connect-to-interrupt device handlers and userwritten system services.

Transportability Features
The BLISS-32 language is designed to facilitate transportability, that is, the writing of programs that can be executed on architecturally different machines with little or no
7-31

Time Procedure Library to print the current t i me on
SYS$0UTPUT.

modification . BLISS compilers also exist for the DECsystem-10 and DECSYSTEM-20 (BLISS-36), and for the PDP11 (BLISS-16). Several language features enhance transportability:

VAX-11CORAL66

• The high-level language constructs may be transferred
from one machine to another with little or no alteration .

The VAX-11 CORAL 66 compiler compiles in compatibility
mode and generates native mode object code unde r
VAX/VMS . The CORAL 66 language, derived from JOVIAL
and ALGOL-60 in 1966, is the standard language prescribed by the British government for military real-time applications and systems implementation . A governmen t
agency controls the language standard for CORAL 66 ,
which was first widely used in military projects beg inning
in 1970. Her Majesty's Stationery Office publishes the "Official Definition of CORAL 66."

• Machine-specific functions can be separated from the
common, mainline code via modularization , macros,
and Library and Require files.
• Parameterization allows machine-specific characteristics to be passed to BLISS data structures.
• The compiler can be instructed to perform transportability checking via the " LANGUAGE (COMMON) " module-head switch .
BLISS 's transportability makes it an ideal language for
system programming applications-and a desirable alternative to assembly language coding in those applications
in which extreme machine dependence is not involved.

The CORAL 66 language replaces assembly-level programm ing in a number of commercial , process control , research , and military applications. It is particularly adapted
to long-life products requiring flexibil ity and ease o f
maintenance.

BLISS-32 Sample Program
Figure 7-23 illustrates how VAX-11 BLISS-32 can call
VAX/VMS system services and the VAX-11 Common Run

0001
0002
0003
0004
0005
0006
M 0007
0008
0009
0010
0011
0012
0013
0014
0015
0016
0017
0018
0019
0020
0021
0022
0023
0024
0025
0026
0027
0028
0029
P0030
p 0031
P0032
0033
0034
0035
0036
0037
0038
0039

VAX-11 CORAL 66 is a block-structured language. A block
is a piece of a program that can be entered only at the be-

MODULE showtime( IDENT =' 1- 1' %TITLE 'SHOW TIME', MAIN = timeout)=
BEGIN
LIBRARY 'SYS$LIBRARY:STARLET';

! Defines System Services, etc .

MACRO
desc[] = % CHARCOUNT(%REMAINING ),
UPLIT BYTE( %REMAINING) %;

! A VAX-11 Style String descriptor

OWN
timebuf:
msgbuf:
msgdesc:

VECTOR[2].
VECTOR[80 ,BYTE].
BLOCK[8 ,BYTE]
PRESET(
[dsc$w length] = 80 ,
[dsc$a~po i nter ] = msgbuf );

! 64 bit system tim e
! Output msg . buffer
! String descriptor
! for output buffer

BIND
fmtdesc = UPLIT(DESC('At the tone, the time is ', %CHAR(?), '115% T' ));
EXTERNAL ROUTINE
lib$put_output : ADDRESSING_MODE(GENERAL);
ROUTINE timeout=
BEGIN
LOCAL
RSLT :

WORD;

! From VMS RTL

! Resultant string length

$GETTIM( TIMADR = timebuf );

! Get time as 64 b it integer

$FAOL(

! Format control -string address
! Resultant length (only a word! )
! Output buffer descriptor address
! Address of pointer to time block

CTRSTR=fmtdesc,
OUT LEN = nslt,
OUTBUF = msgdesc,
PRMLST= %REF(timebuf)) ;

MSGDESC[dsc$w_length] = .rslt;

! modify output descriptor

lib$put_output( msgdesc)

! print the formatted time

END ;

Figure 7-23
Sample VAX-11 BLISS-32 Code
7-32

ginning . Though internal structures cannot be "seen " from
the outside , statements inside a block can "see" out. Sorting is possible, so that programs may be written in which
information is accessible for only the time it is required ,
and no longer. In this way, unwanted interactions among
program parts are avoided , and out-of-date information is
very quickly forgotten .

assembly listing . VAX-11 MACRO is similar to PDP-11
MACRO, but its instruction mnemonics correspond to the
VAX native instructions. VAX- 11 MACRO is characterized
by:
• relocatable object modules
• global symbols for linking separately assembled object
programs

To satisfy real-time needs, CORAL 66 allows different
modules of the same suite of programs to be executed at
apparently the same time. A CORAL 66 compiler assumes
that any subroutine global to the whole program is likely to
be active at the same time as any other, so the compiler
assures that such subroutines do not share any local storage . Interactions, however, can be explicitly arranged by
the programmer. A program consists of communicators
and separately compiled segments. Each segment has the
form of an ALGOL-60 block, within which blocks and procedures may be nested to arbitrary depth . In the absence
of communicators, block structure would prevent different
segments from using common data, labels, switches, or
procedures . The purpose of a communicator is to specify
and name those objects which are to be commonly accessible to all segments. The presence of communicators imposes a modular and disciplined approach to programming larger systems where a team of programmers is employed .

• global arithmetic , global assignment operator, global label operator and default global declarations
• user-defined macros with keyword arguments
• multiple macro libraries with fast access structure
• program sectioning directives
• conditional assembly directives
• assembly and listing control functions
• alphabetized , formatted symbol table listing
• default error listing on command output device
• a Cross Reference Table (CREF) symbol listing
Symbols and Symbol Definitions
Three types of symbols can be defined for use within
MACRO source programs: permanent symbols, user-defined symbols and macro symbols. Permanent symbols
consist of the VAX instruction mnemonics and MACRO
directives and do not have to be defined by the user. Userdefined symbols are those used as labels or defined by
direct assignment. Macro symbols are those symbols used
as macro names.

In addition to the functionalities prescribed in the Official
Definition , the VAX-11 CORAL 66 compiler provides the
following features:
• BYTE, LONG (32-bit integer) and DOUBLE (64-bit floating point) numeric types

MACRO maintains a symbol table for each type of symbol.
The value of a symbol depends on its use in the program .
To determine the value of a symbol in the operator field ,
the assembler searches the macro symbol table , user
symbol table , and permanent symbol table in that order.
To determine the value of the symbol used in the operand
field , the assembler searches the user symbol table and
the permanent symbol table in that order. These search
orders allow redefinition of permanent symbol table entries as user-defined or macro symbols.

• generation of re-entrant code at the procedure level
• switch-selectable option to optimize generated code
• conditional compilation of defined parts of source code
• English text error messages at compile and (optionally)
run-time
• switch-selectable option to control listing output
• INCLUDE keyword to incorporate CORAL 66 source
code from user-defined files

User-defined symbols are either internal to a source program module or global (externally available). An internal
symbol definition is limited to the module in which it
appears. Internal symbols are local definitions which are
resolved by the assembler.

• switch-selectable option to read card format
VAX-11 CORAL 66 is essentially a high-level block-structured language possessing certain facilities associated
with low-level languages, and is designed for use on small
or medium-size dedicated computers . One of the main intentions is that programs written in CORAL 66 should be
fast to execute, taking up limited quantities of storage,
while being easy to write .

A global symbol can be defined in one source program
module and referenced within another. Global symbols
are preserved in the object module and are not resolved
until the object modules are linked into an executable program. With some exceptions, all user-defined symbols are
internal unless explicitly defined as being global.

The real-time applications of the language are implicit
rather than explicit, permitting the utilization of any hardware or special features . Procedures, optionally with parameters, permit communication with and reaction to external events. This is aided further by a direct code facility
which enables machine code to be included in the source
program for extra efficiency in any critical tasks.

Directives
A program statement can contain one of three different
operators: a macro call, a VAX instruction mnemonic, or
an assembler directive. MACRO includes directives for:
• listing control
• function specification

VAX-11 MACRO

• data storage

The VAX-11 MACRO assembler accepts one or more
source modules written in MACRO assembly language
and produces a relocatable object module and optional

• radix and numeric usage declarations
• location counter control
7-33

Macro Calls and Structured Macro Libraries
A program can call macros that are not defined in that program. A user can create libraries of macro definitions, and
MACRO will look up definitions in one or more given library files when the calls are encountered in the program .
Each library file contains an index of the macro definitions
it contains to enable MACRO to find definitions quickly.

• program termination
• program boundaries information
• program sectioning
• global symbol definition
• conditional assembly
• macro definition

Program Sectioning
The MACRO program sectioning directives are used to declare names for program sections and to establish certain
program section attributes. These program section attributes are used when the program is linked into an image.

• macro attributes
• macro message control
• repeat block definition
• macro libraries

The program sectioning directive allows the user to exercise complete control over the virtual memory allocation
of a program, since any program attributes established
through this directive are passed to the linker. For example, if a programmer is writing multiuser programs , the
program sections containing only instructions can be declared separately from the sections containing only data.
Furthermore, these program sections can be declared as
read-only code , qualifying them for use as protected , reentrant programs .

Listing Control Directives
Several listing control directives are provided in MACRO
to control the content, format, and pagination of all listing
output generated during assembly. Facilities also exist for
titling object modules and presenting other identification
information in the listing output.
The listing control options can also be specified at assembly time through qualifiers in the command string issued to
the MACRO assembler. The use of these qualifiers provides initial listing control options that may be overridden
by the corresponding listing control directives in the
source program.

PDP-11 BASIC-PLUS-2/VAX
PDP-11 BASIC-PLUS-2/VAX is an optional language
processing system that includes a compiler and an object
time system. PDP-11 BASIC-PLUS-2is also available as an
optional language processor for the RSTS/E, RSX-11 M,
RSX-11 M PLUS , and IAS operating systems. The PDP11 BASIC-PLUS-2/V AX compiler produces code that exec utes in PDP-11 compatibility mode.

Conditional Assembly Directives
Conditional assembly directives enable the programmer
to include or exclude blocks of source code during the assembly process, based on the evaluation of stated condition tests within the body of the program . This capability
allows several variations of a program to be generated
from the same source module.

BASIC-PLUS-2 is a PDP-11 applications implementation
language which features many programming fac iliti es
found in VAX-11 BASIC . These include:

The user can define a conditional assembly block of code,
and within that block, issue subconditional directives.
Subconditional directives can indicate the conditional or
unconditional assembly of an alternate or non-contiguous
body of code within the conditional assembly block . Conditional assembly directives can be nested.

• program formatting and commenting facilities
• long variable names
• indexed, sequential, and relative file 1/0

Macro Definitions and Repeat Blocks
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 list of real arguments (replacing the
corresponding dummy arguments in the macro definition)
generates the desired coding sequence or macro expansion . MACRO automatically creates unique symbols where
a label is required in an expanded macro to avoid duplicate label specifications. Macros can be nested; that is, the
definition of one macro can include a call to another.

• virtual arrays

An indefinite repeat block is a structure that is similar to a
macro definition, except it has only one dummy argument.
At each expansion of the indefinite repeat range, this dummy argument is replaced with successive elements of a
specified real argument list. This type of macro definition
does not require calling the macro by name, as required in
the expansion of conventional macros. An indefinite repeat block can appear within or outside of another macro
definition, indefinite repeat block, or repeat block.

The object time system (OTS) is a collection of library
modules used during program execution . The library routines include math and floating point functions , input/output operations, error handling , and dynamic string
storage functions . Since the OTS is an object module library , the task builder can select only those functions
needed at run time to be included in a program . Unnecessary routines are omitted from the program and memory
usage is reduced .

• variable-length strings
• PRINT USING statement
• COMMON statement
• a subprogram CALL statement
• extended debugging facilities
• integrated INQUIRE (HELP) facility
The compiler accepts source programs written in the BASIC-PLUS-2 language.
The programmer can edit the source if necessary, and
compile it to produce an object module which can be
linked with previously compiled object modules.

7-34

Program Format
The basic source program unit is a line. In its simplest
form , it consists of a line number, a keyword and statement, and a line terminator. In BASIC-PLUS-2, one or several spaces or tabs can be used to separate line numbers,
keywords, and variable names. Line numbering determines the order in which the program is processed ; the
programmer can enter BASIC-PLUS-2 program lines in
any order.
BASIC-PLUS-2 programs can be one or several lines long.
The programmer can place one statement on each line,
place several statements on any one line, or spread one
statement over several lines. Program comments can be
placed anywhere within a line using the REM (Remark)
statement or using comment field delimiters. These facilities enable the programmer to freely format a program to
make it more readable.
Long Variable and Function Names
Most BASIC languages limit the length of a variable or
user-defined function name to one character . BASICPLUS-2 recognizes variable names and function names as
long as 30 characters. The programmer can fully identify
variables and functions.

PRINT USING Output Formats
The PRINT USING statement allows the programmer to
control the appearance and location of data on an output
line to create complex lists, tables, reports, and forms. In
addition to the numeric field definitions provided by BASIC-PLUS , which allow the programmer to generate floating dollar sign , aligned decimal point, trailing minus, asterisk fill, and exponential format fields, BASIC-PLUS-2 provides string field definitions which allow the programmer
to generate left-justified, right-justified, centered , and
extended string fields .
Subprograms and the CALL Statement
The BASIC-PLUS-2 CALL statement enables a program to
access external BASIC-PLUS-2 or MACR0-11 subprograms. A programmer can therefore write a program in
several modular segments, each of which can be compiled
separately to speed program development. BASIC-PLUS2 provides a complete traceback on errors occurring in
subroutines.
COMMON Statement
The COMMON statement enables a program to pass data
to another program or subprogram . The BASIC-PLUS-2
COMMON statement format is compatible with PDP-11
FORTRAN COMMON. Strings passed in COMMON are
fixed length , which reduces string handling overhead .

Powerful File 1/0
The BASIC-PLUS-2 language supports use of RMS-11
block, indexed, sequential, and relative file operations. Although RMS-11 operates in RSX-11 compatibility mode, it
does not have support for file sharing under VMS. For applications requiring access to shared files , VAX-11 BASIC
should be used .

Debugging Statements
BASIC-PLUS-2 provides an interactive debugging mode
similar to the " immediate mode" facilities found in most
BASIC interpreters. During program development, the
programmer can use the compiler to create, save, edit,
and test the source program . The compiler checks syntax
immediately on input from a terminal so that many errors
can be found prior to compilation. The debugging statements can be used when executing and testing the program . The BREAK, LET, PRINT, UNBREAK, CONTINUE,
STEP, and STOP statements enable the programmer to
control and observe program execution interactively.

Powerful String Handling
The BASIC-PLUS-2 language enables the programmer to
manipulate strings of alphanumeric characters easily. As
in BASIC-PLUS, the BASIC-PLUS-2 relational operators
enable programmers to concatenate and compare strings,
string operators enable the programmer to convert strings
and numerics, and string functions add the ability to analyze the composition of strings. The BASIC-PLUS-2 language includes string functions that:

To set breakpoints , the programmer uses the BREAK
command just prior to running the program , or while it is
stopped . As many as ten breakpoints can be set during the
course of program execution. On reaching a breakpoint,
the program halts to allow the programmer to examine or
modify variables or set other breakpoints.

• create a string of a fixed length
• search for the position of a set of characters within a
string

To examine variables while a program is stopped , the programmer uses the PRINT statement. The LET statement
allows the programmer to modify the value stored in the
variable .

• insert spaces within a string
• trim trailing blanks from a string
• determine the length of a string
Unlike many BASIC languages, BASIC-PLUS-2 imposes
no limit on the size of string values or string elements of
arrays manipulated in memory, other than the amount of
available memory.
Virtual Arrays
Virtual arrays are random access disk-resident files. A
program can create and access virtual arrays in the same
way memory-resident arrays are accessed: using element
names. Explicit read/write programming is not required .
The last element in the array can be accessed as quickly
as the first. Because the arrays are stored on disk, however, the programmer can manipulate large amounts of
data without affecting program size .
7-35

Typing the CONTINUE command resumes execution until
the next breakpoint is reached. Before typing CONTINUE,
the programmer can issue an UNBREAK command to selectively disable one or all of the breakpoints set, and execution continues until a STOP statement is encountered in
the program or until the program completes .
When a program halts because a STOP statement is included in the program , or because a BREAK command
was issued interactively, the programmer can type the
STEP command on the terminal to let program execution
continue on a line-by-line basis. Typing a STOP command
in interactive debugging mode terminates program execution, just as if an END statement was encountered in the
program.

PDP-11 FORTRAN IVIV AX to RSX
FORTRAN IV is an extended FORTRAN implementation
based on American National Standard (ANSI) FORTRAN,
X3.9-1966. PDP-11 FORTRAN IV code is executed in compatibility mode under VAX/VMS. The FORTRAN IV language includes the following extensions to the ANSI standard;

• user-defined macros

• general expressions allowed in all meaningful contexts

• alphabetized , formatted symbol table listing

• mixed-mode arithmetic

• default error listing on command output device

• comprehensive system macro library
• program sectioning directives
• conditional assembly directives
• assembly and listing control functions at program and
command string levels

• BYTE data type for character manipulation

Symbols and Symbol Definitions
Three types of symbols can be defined for use within
MACR0-11 source programs: permanent symbols , userdefined symbols, and macro symbols. Accordingly ,
MACRO maintains three types of symbol tables: the Permanent Symbol Table (PST), the User Symbol Table
(UST), and the Macro Symbol Table (MST).

• ENCODE, DECODE statements
• PRINT, TYPE, and ACCEPT input/output statements
• direct-access unformatted input/output DEFINE FILE
statement
• comments allowed at the end of each source line
• PROGRAM statement

Permanent symbols consist of the PDP-11 instruct ion
mnemonics and MACRO directives. The PST contains all
the permanent symbols automatically recognized by
MACRO and is part of the assembler itself. Since these
symbols are permanent, they do not have to be defined by
the user in the sou rce program .

• OPEN and CLOSE file access control statements
• list-directed input/output
Additionally, virtual arrays are supported on target systems with memory management directives. Virtual arrays
are memory-resident and require enough main memory to
contain all elements of all arrays.

User-defined symbols are those used as labels or defined
by direct assignment. Macro symbols are those symbols
used as macro names. The UST and MST are constructed
during assembly by adding the symbols to the UST or MST
as they are encountered.

The PDP-11 FORTRAN IV compiler is a fast, one-pass
compiler. Compiler options allow program size versus execution speed (threaded code versus inline code) tradeoffs. FORTRAN IV compiler optimizations include:
• common subexpression elimination

Directives
A program statement can contain one of three d ifferent
operators: a macro call, a PDP-11 instruction mnemonic,
or an assembler directive. MACRO includes directives for :

• local code tailoring
• array vectoring
• optional inline code generation for integer and logical
operations

• listing control

MACR0-11 assembly language subroutines may be called
from FORTRAN IV programs. FORTRAN IV also includes a
set of object modules, called the Object Time System
(OTS), that are selectively linked with compiler-produced
object modules to produce an executable program.

• function specification
• data storage
• radix and numeric usage declarations
• location counter control
• program termination

MACR0-11
MACR0-11, the PDP-11 assembly language, is included in
the compatibility mode environment. Programs written in
MACR0-11 can be assembled to produce relocatable object modules and optional assembly listings. MACR0-11
provides the following features:

• program boundaries information
• program sectioning
• global symbol definition
• conditional assembly
• macro definition

• relocateable object modules

• macro attributes

• global symbols for linking separately assembled object
programs

• macro message control
• repeat block definition

• device and file name specifications for input and output
files

• macro libraries

7-36

Program
Development
and
Support
Facilities

VAX/VMS offers the user a powerful and sophisticated program development environment including several support facilities . Described in
th is section are the interactive and batch text editors, the linker, the
Common Run Time Procedure Library, the VAX- 11 interactive symbol ic debugger, the librarian utility, command language procedures, the
differences utility, and VAX-11 RUNOFF.
In addition to the interactive text editor SOS and the SLP batch editor,
VAX/VMS now supports the powerful interactive text editor EDT . EDT
supports many user oriented features including both line and character
editing facilities , an extensive HELP facility , a journaling facility , and the
window editing facility.
The VAX/VMS linker is the program development tool that takes the
output of a translator or compiler and produces a file that can be executed on the VAX-11 hardware.
The Common Run Time Procedure Library offers the user a set of common language routines that can be called from any of the native mode
languages via the VAX-11 calling standard .
The VAX-11 interactive symbolic debugger is extremely useful in isolating program errors by allowing the user to examine and modify the
contents of memory locations while the program is executing.
The librarian is a utility that allows the user easy access to the data
stored in any of the four libraries (object , macro , help , text). The librarian allows an executing program to initialize and open a library , and to
retreive , insert, and delete modules.
The command language procedure section describes the power and
flexibility of executing strings of frequently used sequences of DCL
commands, or creating new commands from the existing DCL command set. Command procedures can accept up to eight parameters
and can include extensive control flow.
By utilyzing the DIFFERENCES utility, the user can determine whether
or not two files are identical.
VAX-11 RUNOFF is a document formatter, offering the user such features as page and title formatting, subject-matter formatting , and the
ability to produce indexes and tables of contents easily.

INTRODUCTION

File Type

Default Use

VAX/VMS provides a complete program development environment for the user. Development tools supporting this
environment are :

.FOR

FORTRAN language source statements

.MAR

MACRO assembly source statements

• Interactive and batch text editors

.COB

COBOL language source statements

.BAS

BASIC language source statements

.PAS

PASCAL language source statements

.DAT

A data file

• Command Language Procedures

.LIS

An output listing from a compiler

• Differences Utility

.EXE

An executable image

• VAX-11 Runoff

.OBJ

An output file from a compiler

These tools, as well as program language compilers, are
available to the programmer via the command language
facility. In addition , VAX/VMS supports a complete set of
shareable routines collectively known as the common run
time procedure library. And finally, VAX/VMS supports the
user's ability to write programs utilizing the DCL command
set (similar to coding programs in other high level languages). These programs are known as command language procedures.

For example , a file containing FORTRAN source statements would possess the file type .FOR.

• Linker
• Common Run Time Procedure Library
• VAX-11 interactive symbolic debugger
• Librarian Utility

SOS EDITOR
SOS is a line-oriented, interactive text-editing program.
SOS has features that allow examination and modification
of text, character by character. SOS can be used to perform the following functions :

The text editors can be used to create memos, documentation, and text and data files, as well as source program
modules for any language processor. The linker, librarian,
debugger, and run time procedure library are used only in
conjunction with the language processors that produce
native code.

• examine , create, and modify ASCII files
• search for and/or change one or more arbitrary text
strings, with the option to verify each change before it is
made
• merge parts of one file into another
• create a file that is a subset of another file

TEXT EDITORS

SOS is line-oriented , so it usually operates with line-numbered text files. If a file is edited that does not contain line
numbers, the editor adds line numbers to the text lines.
For most SOS commands, a line number or range of line
numbers specifies the text to be operated on . When commanded to insert, delete, move, or copy text, SOS maintains line numbers in ascending order within each page of
text.

Text editing refers to the process of creating , modifying ,
and maintaining files . VAX/VMS supports three text editors : two interactive text editors (SOS and EDT) and a
batch-oriented text editor (SLP) .
The user invokes the SOS and EDT text editors interactively, i.e. , the user creates and processes files on-line. The
SLP text editor, on the other hand, allows direct modification to a file via a command file prepared by the user. SOS
is often used to create SLP command files. All editors are
invoked by the command EDIT. The default editor is SOS.
Therefore , to invoke SOS, enter the command EDIT or
EDIT /SOS; to invoke EDT , enter EDIT /EDT; for SLP, use
EDIT/ SLP .

Advanced features of SOS allow considerable flexibility in
searching for a string of text and allow specification of
blocks of text by content, or relative position from a known
location , instead of by line number. SOS has many operational features under user control.
Initiating and Terminating SOS
SOS is initiated by entering one of the following commands in response to the command language prompt($) :

Before describing the text editors, a short summary of file
naming conventions and default file types is presented.
File Names and File Types
By taking advantage of the default disk and directory, the
user can identify a file uniquely by specifying its file name
and file type , illustrated in the following format:

$ EDIT file-spec <RET>
If the user were to omit file-spec, SOS would immediately
prompt the user for the missing parameter.
To terminate SOS, enter the command E (EXIT) followed
by a carriage return after SOS's prompt(*).

filename.typ
The file name can be from one to nine alphanumeric characters , and can assume any name that is meaningful to the
user.

*E<RET >
[file-spec]

The file type is a 3-character identifier preceded by a period; it describes more specifically the kind of data in the
file . Although file type can consist of any three alphanumeric characters meaningful to the user, several file
types have standard meanings. Among these special file
types are:

Upon terminating , SOS writes an output file containing all
the modifications made in editing the file. The original file
is not changed . The specifier SOS uses for the output file
has a version number higher by 1 than the latest version of
the original file unless otherwise specified by the user.
8-1

SOS Examples
Copy command

------------------------FILE

C300,9000:9500
Make a copy of lines numbered 9000-9500 and
insert the lines after line 300.
Find command

CURSOR

---------------

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

-............._

Fmore<ESC>
Search for " more" from the current point in the
file .
2) Fmore<ESC>,1 :1000
Search for the first occurrence of "more " in the
range of lines from 1 through 1000.
Print command

22 LINE
WINDOW

--------------Figure 8-1
Window Editing

P500:800
Print lines 500 through 800.
2) P1800
Print line numbered 1800.
Substitute command
1)

}

HELP Facilities
The HELP facilities on EDT are extensive. Users can get
help on general EDT operations by typing HELP. Wh ile in
keypad mode, users can get help by pressing the help key,
which displays a picture of the keypad and provides additional information on each of the keypad keys .

Smore<ESC>less<ESC>,500:800
Change all occurrences of " more" into "less" on
lines numbered 500 through 800.

EDT EDITOR
EDT, an interactive text editor, is included with VAX/VMS
Version 2.0. This editor lets users enter and manipulate
text and programs. EDT, with its extensive HELP facility, is
designed to be learned easily by novices. In addition, EDT
provides many capabilities that will appeal to advanced
users.

The Keypad
The keypad is a special set of keys to the right of the main
keyboard. Figure 8-2 illustrates the functions of the VT100
keypad ; the VT52 keypad is similar.

What EDT Does
EDT is a powerful text editor that provides:
• both line and character editing facilities

FINO

UNO L

FNONXT
REPL

DEL EL
UNO W

APPEND
PASTE

DEL EW
UNO C

ADVANCE BACKUP
DEL EOL

CUT
INSCOD

DEL C

WORD
EL
OPEN LINE

CHAR
RESET

LINE

SELECT

SHIFT

• screen editing and keypad editing on the VT52 and
VT100 video terminals

HELP

~OM MAND

• ability to work on multiple files simultaneously

PAGE
BOTTOM

• a journaling facility, which protects against loss of edits
due to system crashes, or loss of carrier on a dial-up
line

SECT
TOP

SUBS

• an extensive HELP facility
• a start-up command file, which allows a choice of editing options to be set automatically
• a window into a file (on video terminals only) that lets
users view changes in file contents immediately

ENTER

Figure 8-2

EDT is also supported on hardcopy terminals and video
terminals other than the VT52 and VT100.

VT100 Keypad Functions

EDT SPECIAL FEATURES

Keypad functions allow the user to perform a variety of operations. Furthermore, the function of any keypad key can
be changed to meet the needs of the user via the DEFINE
command.

Editing with a Window
"Window editing" is a valuable feature that lets users edit
one 22-line window (screenful) at a time . This feature allows a user to see immediately how the edits made affect
his file. The user may position the window anywhere in the
file. Window editing is illustrated in Figure 8-1.

The commands in the keypad submode let users alter text
or change the cursor position in the file . Keypad functions
are available to advance or back up the cursor or move the
cursor to the top or bottom of the text. One can also move
the cursor any number of characters , words, lines, or
pages at a time .

Start-up File
When the editor is started, it executes commands from a
start-up file. In this file, one can insert editing options such
as SET NOKEYPAD and DEFINE KEY. These options take
effect automatically when an editing session begins.

Keypad keys let a user select a string of text and move it
elsewhere in any of his files. One can even find the next
8-2

occurrence of some text and delete or replace it. There is
also a key to press for help messages.

commands simply by pressing a key . Users can also redefine the function of these keys .

Redefining Keypad Keys
One can redefine any of the keypad keys, and most of the
control (CTRL) keys, on VT52 and VT100 terminals. This
feature lets the user assign a series of commands to a key;
EDT performs these commands when the keys are
pressed . Therefore , one can adapt the functions of keypad
and CTRL keys to meet special needs.

SLP EDITOR
SLP is the batch -oriented editing program used for source
file maintenance. SLP allows updating (deletion , replacement, addition) of lines in an existing file. The SLP command file provides a reliable method of duplicating the
changes made to a file, at a later time or on another system .

The SET and SHOW Commands
The SET command , with a variety of qualifiers , controls
EDT's ed iting capabil ities. SET controls such screen parameters as line width or lets a user determ in e the appearance of text, such as changing the window size to less than
22 lines. The SHOW command provides information on
the current state of the editor , such as terminal parameters , definitions of keypad keys, and the names of buffers
in use during the editing session .

Input to SLP consists of a correction input file that is to be
updated , and command input containing text lines andedit command lines that specify the update operations to be
performed . SLP locates lines to be changed by means of
locators (line numbers or character strings). Command input normally enters through an indirect file that contains
commands and text input lines to be inserted into the file.
Alternatively, commands can be entered from the terminal.

Journal Processing
Journal processing protects the user's work against unlikely system crashes . During an editing session , EDT
saves all the terminal input in a journal file. After a crash
and recovery, the user may choose to retrieve and execute
commands in this saved file with the / RECOVER EDIT
command qualifier. In this way the user can recover edited
files to the time of the crash.

SLP output is an optional listing file and an updated copy
of the corrected input file. SLP provides an optional audit
trail that helps keep track of the update status of each line
in the file . The aud it trail is provided in the listing and is
included permanently in the output file. When a given file is
updated with successive versions of an SLP command file,
different audit trails may be used to differentiate between
changes made at various times .

The EDT CAI Program
Also available with VAX/VMS V2.0 EDT is a ComputerAssisted Instruction (CAI) program on EDT. This interactive program presents the " Introduction to the EDT Editor"
min icourse , which demonstrates how to use EDT. The CAI
program runs on VT100 terminals and takes about three
hours.

SLP output qualifiers permit the user to create or suppress
an audit trail , eliminate an existing audit trail , specify the
length and beginning position of the aud it trail , or generate
a double-spaced listing.
Initiating and Terminating SLP
SLP is initiated via the command language EDIT
command . The normal way to use SLP is to specify an indirect command file that informs SLP what files to process,
and indicates what editing changes are to be made to the
correction input file. The indirect file can be specified on
the same line with the EDIT command , or on a separate
line. The indirect file must be created before running SLP.
An interactive text editor is normally used to create SLP indirect command files . If both new and old versions of the
file exist , the differences utility can be used to create a SLP
correction file that will change the old file into the new one.

EDT Modes of Operation
A " mode" in EDT is a state in which the editor lets a user
perform a specific set of functions . EDT has two basic
modes of operation : line mode and change mode.
Line mode allows users to establish editing parameters
and to d isplay and edit text by range specification . (One
can specify a range with such entities as line numbers and
character strings.)
One can modify the text with line ed iting commands such
as COPY, SUBSTITUTE, and REPLACE . Or one can move
about in the text by using the FIND and TYPE commands,
for example, or by pressing the RETURN key.

SLP Input and Output Files
SLP requires two types of input: a correction input file and
command input. The correction input file is the source file
to be updated using SLP. Command input consists of an
initialization line, followed by SLP edit commands that indicate how the file is to be changed.

Change mode lets users operate on such entities as characters , words, sentences , paragraphs, and lines. One can
also work with strings of text or delete and move whole
pages . EDT lets a user redefine these entities to tailor them
to specific applications , which can be as diverse as documentation or programming .

SLP output consists of a listing file and an output file. The
listing file is a copy of the output file with sequence numbers added ; it shows the changes SLP makes to the correction input file . The output file is the permanently
updated copy of the input file .

Change mode consists of a set of NOKEYPAD commands.
Typing any of these commands lets a user perform useful
functions . By typing FNDNXT, for example , one can find
the next occurrence of a string of characters .

Correction Input File
The correction input file is the file to be updated by SLP. It
can contain any number of lines of text. When SLP
processes the correction input file, it makes the changes
specified by SLP ed it commands in the output file.

With VT52 and VT100 terminals, one can also use KEYPAD commands . The set of keypad keys , as well as several CTRL keys, lets the user enter any of the NOKEYPAD

8-3

nal references have values that are available only to the
linker when all the input (e.g ., modules and library procedures) is gathered together . The VAX-11 object language provides the ability for a language translator to describe to the linker the external items required by a module. The linker maintains a description of the items of each
module that are available to other separately translated
modules. In the object language, all of these external items
are either references to global symbols or definitions of
global symbols.

SLP Output File
The SLP output file is the updated input file . All of the updates specified by the command input are inserted in this
file . An audit trail , unless suppressed, is applied to lines
changed by the update. The numbers generated by SLP
for the listing file do not appear in the output file.

LINKER
The VAX/VMS linker is a program development tool that
takes the output of language translators (object files or
modules) , such as the VAX-11 MACRO assembler or the
VAX-11 FORTRAN compiler, and produces a file that can
be executed on the VAX-11 hardware. This output file is
known as an image. To write an application in modules, it
is necessary to be able to link together the separately compiled modules. The linker is activated by the DCL LINK
command, which can be entered interactively or from within a command procedure. Linking consists of three basic
operations:

Image Initialization
After the linker allocates virtual memory and resolves external references, the linker fills in the actual contents of
the image. This image initialization consists mainly of copying the binary data and code that was written by the
compiler or assembler. However, the linker must perform
two additional functions to initialize the image contents :
• It must insert addresses into instructions that refer to
externally defined fields . For example , if a module contains an instruction moving FIELDA to FIELDS , and if
FIELDS is defined in another module, the linker must
determine the virtual address of FIELDS and insert it into the instruction .

• allocation of virtual memory addresses
• resolution of intermodule symbolic references
• initialization of the contents of a memory image
At the end of a linking operation, the program has virtual
memory addresses assigned , has intermodule references
resolved, and exists as an executable initialized entity in a
disk-resident image file .

• It must compute values that depend on externally defined fields . For example, if a module defines X as being
equal to Y plus Z, and if Y and Z are defined in an external module, the linker must compute the value of Y plus
Zand insert it in X.

The LINK Command
The DCL LINK command provides the interface between
the user and the linker. When the user requests the linking
of object modules, the command interpreter receives the
command and activates the linker.

Overview of Linker Interface to Memory Management
The linker describes the virtual address space required for
an image in such a way that the image activator function of
VAX/VMS can initialize the VAX memory management
hardware to place the image in a process virtual address
space. When a user requests execution of an image, the
image activator obtains a description of the image's virtual
address requirements from the image file produced by the
linker.

Virtual Memory Allocation
Language translators do not compute any addresses in
the program . At the time of translation, the allocation of
virtual address space is undecided. Each object module is
relocatable in virtual memory. The reason that language
translators cannot allocate virtual memory addresses is
that a translator can see only one module at a time: it cannot know how modules interrelate. As a result , it is the linker's function to perform the memory allocation, reference
resolution , and image initialization required to form one
executable program from a number of object modules.

The mechanism used to describe images to VAX/VMS is
an image section descriptor. The linker creates an image
section description (ISD) for each image section of a
shareable or executable image. The header of an image
contains the ISDs for the image. With the ISO, memory
management can determine the following information
about an image section:

VAX/VMS language translators use the object language to
describe a module to the linker. The output from a translator is an object module consisting of records describing
the module to the linker. The language translators define
each object module as a number of separate areas called
program sections. Some program sections contain data,
others contain instructions. Some can be modified during
execution, others cannot. Some are accessible to procedures in other modules, others are local to a module.
When determining the virtual memory allocation of a program, the linker must consider the attributes of each program section. The linker groups program sections with
similar attributes together in virtual memory.

• the starting block number of the image section in the im age file
• the starting virtual page number in the process's virtual
address space to which to map the image section
• characteristics of the image section , e.g., read -only,
read/write
• additional control information
Using the information in the ISD , memory management
sets the page table and other data structures used to bring
process pages into physical memory and to allow sharing
in physical memory.

Resolution of Symbolic References
VAX language translators provide the ability to call external procedures by name. They permit the use of other external items such as literals and variables by name. Exter-

Linker Input
The linker accepts the following types of files as input to a
binding operation :

8-4

1.
2.
3.
4.

COMMON RUN TIME PROCEDURE LIBRARY

Object module files
Libraries of object modules
Shareable images files
Symbol tables from shareable images

The VAX-11 Common Run Time Procedure Library (RTL)
is composed of a set of general purpose and languagespecific VAX procedures which establish a common run
time environment for all user programs written in any native mode language. Because all of the language support
procedures follow the same programming standards and
make nonconflicting assumptions about the execution environment, a user program can be composed of modules
written in different languages, including assembly language. Because of the VAX procedure calling standard ,
each native mode user module can call any other native
mode user module or any of the procedures in the Run
Time Library.

Object Module Files
The linker requires as a minimum one object file as input
to a binding operation . An object module contains four
types of information :
1.
2.

3.
4.

Compiled program code and data.
Descriptions of program code and data used by the
linker in performing relocation and link-time computations .
Identification of the object module and its history for
use by the librarian and patch utilities.
Description of the memory allocation requirements of
the module.

Most of the VAX-11 Run Time Library is constructed as a
separate shareable image which is accessed by users via
entry point vectors . This allows:

Object Module Libraries
The librarian creates and updates object module library
files. Each library file contains a catalog of the object modules and global symbols within it. The linker can access
modules in such libraries either explicitly or implicitly.

Installation of a new library without the need to relink a
user's program .
2. Implementation of new internal algorithms without relinking all user programs.
3. A single copy of the library to be shared by all
processes.
Each procedure entry point in the shareable image is at an
address defined relative to the base of the shareable section, and will never change, once it has been assigned .
New entry points are always added at the end of the list of
entry point vectors . The entry point vector contains the
procedure entry mask and a transfer of control to the procedure . Use of entry vectors permits a single position -independent copy of the library to be bound to different
virtual addresses in processes which are sharing it. Use of
entry vectors also permits a new release of the library to be
installed without requiring that user images be relinked .

Explicit extraction is performed on the basis of the name of
a particular module in the file or by naming the library file
and letting the linker extract any modules required to resolve undefined symbols .
Implicit access to object module libraries occurs after all
explicitly named input modules have been extracted, and
is done by loading modules which contain global symbols
that resolve undefined global symbols in the link .
Shareable Image Files
A shareable image is an image that comprises part of a
complete program . All references in the shareable image
are resolved when the shareable image is created. Shareable images are used as input to a later link to create an
executable image.

The VAX-11 Run Time Library is designed as a set of modular re-entrant procedures comprising several functional
groups. They are:

Shareable Image Symbol Tables
When the linker produces an image file, it appends the
symbol table to the file. The symbol table produced by the
linker has the same form as an object module. That is, it
defines those symbols available to object modules that are
outside the set of object modules that produced the shareable image. Such symbols are called universal.

• a resource allocation group (virtual memory, logical unit
number, and event flags)

Linker Output
The linker can produce three different types of images.

• a language-independent support group (error handling
and Record Management Services support functions)

1.
2.
3.

• language-specific support groups (file handling support
functions)

• a condition handling group (signaling exception conditions and declaring condition handlers)
• a general utility group (data type conversions)
• a mathematical group (single and double precision trigonometric, logarithmic, and exponential functions)

Executable images
Shareable images
System images

• a string handling group (static and dynamic string functions)

Executable images are the most common . As the name
suggests , an executable image is the type run in response
to a command given to the command interpreter. The second type, shareable images, is intended for use at link time
and , potentially, at run time. At link time, a shareable image can be linked with object modules to produce an executable image. The same shareable image can be shared
when executable images bound to it are run . A system image is a special type of image intended for stand-alone
operation on the hardware i.e., it does not run under the
control of the VAX/VMS operating system .

Resource Allocation group (LIB$)
The resource allocation group includes all procedures
which allow allocation of process-wide resources . Such resources include:
1.

8-5

Virtual Memory-one procedure to allocate and one
to deallocate arbitrary-sized blocks of process virtual
memory.

may call the operating system directly. However, since
some languages cannot easily pass arguments in the form
that system services require, and some languages use data types that system services cannot properly handle (i.e. ,
dynamic strings), LIB$ routines provide easy access to the
operating system directives.

Logical Unit Numbers-allow logical unit numbers to
be allocated in a modular manner.
3. Event Flags-allow event flags to be allocated in a
modular manner.
In most cases, the resource allocation procedures must be
used to allocate process-wide resources in order for all library , DIGITAL, and customer-written procedures to work
together properly within an image.

Compiled-Code Support Procedures
These routines complement the compiled code by performing operations too complicated or too cumbersome to
perform directly with inline code . Thus, the language support libraries support the code generated by the compiler.
For example, division of complex numbers is performed
by a library procedure .

Signaling and Condition Handling
The VAX-11 condition handling facility is a collection of library procedures and system services which provides a
unified and standardized mechanism for handling errors
internally in the operating system, the Run Time Library ,
and user programs. In some cases, the mechanism is also
used to communicate errors across these interfaces. In
particular, all error messages are printed using this mechanism . When an error condition is signaled, the process
stack is scanned in reverse order. Establishing a handler
provides the programmer with some control over fix-up,
reporting, and flow of control on errors. It can override the
standard error messages in order to give a more suitable
application-oriented user interface.

Error Processing Procedures
Errors detected by the Run Time Library are indicated by
returning an error completion status wherever possible.
This is especially true for the general utility library (LIB$).
However, the math library and the language support libraries indicate most errors by CALLing the VAX-11
LIB$SIGNAL or LIB$STOP procedures. The LIB$SIGNAL
procedures use a condition value as an argument which
has an associated error message stored in a system error
message file. The condition is signaled to successive procedure activations in the process stack . These procedures
may have established handlers to handle the conditions or
change the error message. Thus an application can tailor
its error messages to its own needs.

General Utility (LIB$)
General utility procedures are not mandatory in order to
use the rest of the library successfully. They are provided
for the convenience of the user only. General utility procedures include outputting a record to a logical device
(SYS$0UTPUT).

VAX-11 SYMBOLIC DEBUGGER

Mathematical Functions (MTH$)
The mathematical library consists of standard procedures
to perform common mathematical functions , such as taking the sine of an angle . The standard entry points have
one or two call-by-reference input parameters and a single
function value. Some frequently used procedures also
have call-by-value entry points that are called by the JSB
instruction.

The VAX-11 SYMBOLIC DEBUGGER is a language-independent, interactive program that can be linked with user
code written in all native mode languages supported by
VAX/VMS. Current languages with which the debugger
can be used are: VAX-11 FORTRAN, VAX-11 BASIC, VAX11 COBOL, VAX-11 BLISS-32 , VAX-11 CORAL66 , VAX-11
PASCAL, and the VAX-11 MACRO assembly language. After linking with the user program , the DEBUG facility is operative in the language of the first module of the image file .
If it is necessary to alter the language for a later module,
the SET LANGUAGE command may be used .

Language-Independent Support (OTS$)
The language support libraries support the code generated inline by compilers. As such, most of the procedures
are called implicitly as a consequence of a language construct specified by the user, rather than being called explicitly by the user with a CALL statement. Those language
support procedures which are independent of higher level
language use the facility prefix OTS$.

DEBUG enables dynamic examination and modification of
the contents of memory locations, which is useful in isolating program errors. Since user program execution is controlled by DEBUG once it is invoked , modifications may be
made to the program while it is executing.

Language-Specific Support (FOR$, BAS$)
Each of the language support libraries is composed of five
principal types of procedures:

The VAX-11 debugger includes many user oriented
functions that facilitate the use of the VAX-11 SYMBOLIC
DEBUGGER.

• 1/0 processing procedures

• The debugger is interactive-the user maintains control
of the program while conversing with the debugger via
the terminal.

• Language-independent initialization and termination
• System procedures

• The debugger is symbolic-program locations may be
referred to by the symbols the user has created in the
program . The debugger is also capable of displaying locations as symbolic expressions .

• Compiled-code support procedures
• error and exception-condition processing procedures
String Processing (STA$)
The string processing procedures allocate and deallocate
dynamic strings and perform a number of useful string
functions on any class of VAX strings.

• The debugger supports all native mode languages-the
debugger lets the user converse with the program via
the source program's language. Furthermore, the user
may change languages during the course of a debugging session by means of a simple command.

System Procedures
VAX-11 programs written in the higher-level languages

8-6

• The debugger permits a variety of data forms-the user
controls the way in which the debugger accepts and displays addresses and data. For example , an address can
be represented symbolically, or as a virtual address in
dec imal , octal , or hexadecimal. Also , data can be represented by symbols , expressions (X + 3) , VAX-11 MACRO
instructions, ASCII character strings , or numeric strings
in decimal , octal , or hexadecimal.

tains indexes that store information regarding the library's
content, including type, and location. The librarian is a
utility that allows the user easy access to the data stored in
libraries.
The librarian may be invoked in one of two ways ; via a set
of librarian routines that can be called from the user program directly, or interactively via the DCL command LIBRARY issued from the terminal or from within an indirect
command file . The DCL LIBRARY command enables the
user to replace and maintain modules in an existing library , or to create a new library. The librarian routines enable
an executing program to initialize and open a library, and
to retrieve, insert, and delete modules.

DEBUG Commands
DEBUG commands direct the execution of the program
and can be entered interactively from a terminal or from an
indirect command file. Typically, the DEBUG commands
can :

The four library types are defined as follows:

• Specify po ints at which execution will be suspended,
when and if they are encountered , by using the SET
BREAK command.

• Object libraries (file type OLB) contain frequently called
routines and are used as input to the linker. Tbe linker
searches the object module library whenever it encounters a reference it cannot resolve from the specified
input files .

• Trace the sequence of program execution by means of
the SET TRACE command. This command establishes
tracepoints in the program.

• Macro libraries (file type MLB) contain macro definitions
used as input to the MACRO assembler. The assembler
searches the macro library whenever it encounters a
macro that is not defined in the input source file.

• Display before-and-after values of a location whenever
that location is stored into, by means of the SET WATCH
command .
• Initiate or resume execution , by means of the GO command or the STEP command .

• Help libraries (file type HLB) contain help modules; that
is, modules that provide user information concerning a
program. The help message can be retreived by calling
the appropriate librarian routines .

• Determine the location of breakpoints, tracepoints, and
watchpoints by means of the commands SHOW BREAK,
SHOW TRACE, and SHOW WATCH , respectively.

• Text libraries (file type TLB) contain any sequential record files requ ired by the user program . A user program
can call library routines d irectly to retrieve text modu les.

• Erase breakpoints, tracepoints , and watchpoints in the
program , through use of the CANCEL command .
• Display the contents of memory locations, by using the
EXAMINE command .

Librarian Routines
The Librarian utility provides a set of 18 user-callable routines that:

• Change the value of the contents of memory locations,
by using the DEPOSIT command .

• initialize a library

• Obtain the value of an expression or the current address
of a symbol , or express a numeric value in a different
rad ix, by using the EVALUATE command.

• open a library
• look up a key in a library
• insert a new key in a library

• Call a subroutine at DEBUG time , by means of the CALL
command .

• return the names of the keys

• Change values of parameters for LANGUAGE, SCOPE,
MODE, and TYPE .

• delete a key and its associated text
• read text records

• Spec ify an arbitrary file name for the DEBUG log file by
means of the SET LOG command.

• write text records
The user program can call the librarian routines using the
VAX-11 standard calling sequence supported in all languages producing VAX-11 native mode code.

• Control DEBUG 1/ 0 at debug time, via the SET OUTPUT
command . This includes normal term inal output, log file
output , and command file verification .
• Find all current output attributes (VERIFY, TERMINAL
and LOG) by using the SHOW OUTPUT command . For
more limited needs, a SHOW LOG command is available that d isplays only the LOG data.

DCL LIBRARY Command
The LIBRARY command creates or modifies an object ,
help, text , or a macro library, or inserts, deletes, replaces ,
or lists modules, macros, or global symbol names in a library .

• Instruct DEBUG to take commands from a specified file
by m eans of @filespec.

To invoke the LIBRARY command , enter the following format:
LIBRARY

THE LIBRARIAN UTILITY

Ii brary[file-spec ,. ... ]

For example, to create an object library named TESTLIB,
and insert entries ERRMSG, and STARTUP, the user
would proceed as follows:

Libraries are indexed files that contain frequently used
modu les of code or text. There are four types of libraries;
object, macro, help, and text. The library type indicates the
type of module that the library contains . Each library con-

$LIBRARY / CREATE TESTLIB

8-7

ERRMSG,STARTUP

the logical commands , the user can alter the flow of execution of the command procedure. By using the IF, GOTO,
ON, EXIT, and STOP commands, the user can control the
execution sequence, conditionally execute lines, construct
loops, and handle errors.

COMMAND LANGUAGE PROCEDURES
A command procedure is a file containing DCL commands, command or program input data, or both. Command procedures may be used to catalog sequences of
commands frequently used during an interactive session
or to submit all jobs for batch processing.

Lexical Functions
The command interpreter recognizes a set of functions,
called lexical functions, that return information about character strings and attributes of the current process. Lexical
functions may be used in any context in which symbols
and expressions are used . Within command procedures,
lexical functions are used to translate logical names, perform character string manipulations, and determine the
current processing mode of the procedure.

In its simplest form, a command procedure consists of one
or more command lines that the command interpreter executes. In its most complex form, a command procedure
resembles a program written in a high level programming
language: it can establish loops and error checking procedures, call other procedures, pass values to other procedures and test values set in other procedures, perform
arithmetic calculations and input/output operations, and
manipulate character string data.

Command Procedure Example
The command procedure described in Figure 8-3, when
invoked, locks up a user terminal as being in use. If the
current user must leave the terminal for some time and
does not wish to have it disturbed, the user can invoke the
command procedure I NUSE.COM, rendering the terminal
inaccessible to any other user not knowing the access
password .

Passing Parameters to Command Procedures
The user can write generalized command procedures that
may perform differently each time they are executed . The
command interpreter defines eight special symbols for
use as parameters within command procedures. These
symbols are named P1, P2, P3 ... P8; they are all initially
equated to null strings. Either numeric or character string
values for these parameters may be passed when executing the procedure with the @ command or the SUBMIT
command when entering a batch job.

This command procedure illustrates several of the powerful features of DCL, including:
• Trapping of the Control-Y function.

For example, the procedure named EXECUTE contains
the following lines:

• Calling a command procedure from within a command
procedure (i.e., @COMMANDS:INUSE.TXT). ERASE,
ERASELINE, and TEXT are user-defined symbols that
also invoke command procedures.

$IF P2 .EQS. ""THEN $P2:="FORTRAN"
$ 'P2' 'P1'
$LINK 'P1'
$RUN 'P1'

• Referencing lexical functions (i.e., 'F$TIME, 'F$LOCATE,
and 'F$EXTRACT).

The command procedure EXECUTE accepts both the language compiler and the user program name as input. If the
user executes the procedure with the @ command, the
values for the command parameters P1 and P2 would be
entered as follows:

Upon invoking the command procedure I NUSE.COM:
• The current setting of VERIFY is retrieved via the lexical
function 'F$VERIFY, and stored in local variable VER
(line 100).

$@EXECUTE PAYROLL COBOL
In this sample run, the user chose the program name PAYROLL, and the COBOL compiler.

• The procedure will set NOVERIFY (line 200), i.e. , the
command lines are not echoed on the terminal during
execution (if verify was on when the command procedure was invoked, it will be turned on when the procedure exits successfully).

It is also possible to define a symbol as a local symbol , using a single equals sign(=) in an assignment statement.
For example, the user might have equated the symbol EXE
to the execution command @EXECUTE as follows:

• The address of the password routine (line 900) is stored
for later use if Control-Y is typed.

$ EXE*CUTE:=@EXECUTE
The asterisk (*) specifies that EXE, EXEC, EXECU, etc. are
abbreviations of EXECUTE. The minimum abbreviation is
three characters (in this case, EXE). A colon (:) in an assignment statement indicates a character string assignment. Now to execute the command procedure the use
can enter the following:

• The address of ERR_HNDLR (line 950) is stored for
processing any errors that might occur.
Execution then proceeds with the BEGIN code block (lines
1000-1300). The ERASE procedure (line 1100) is called ,
which clears the screen of all text. !NUSE.COM then calls
the command procedure INUSE.TXT (line 1300). This procedure prints in block letters, "IN USE," across the video
screen . Execution then proceeds to the LOOP section of
code (lines 1500-2300). This block of code retreives the
current date and time of day from VMS , using the lexical
function 'F$TIME. The date and time of day normally appear as follows :

$EXE STRESS
In this run, STRESS is the user program name and the
compiler is the default compiler, FORTRAN (i.e. , the second parameter in the EXE command was left blank).
Logical Commands
Normally, the command interpreter executes each command in a command procedure in sequential order, and
terminates processing when it reaches the end of the command procedure file. However, by using combinations of

dd-mmm-yyyy hh:mm:ss.cc
The 'F$LOCATE and ' F$EXTRACT lexical functions operate upon the date and time of day, reducing the time quantity to hours and seconds only. Therefore the final date and
8-8

100
200
300
400
500
600
700
800
900
950
1000
1100
1200
1300
1400
1500
1600
1700
1800
1900
2000
2100
2200
2300
2400
2500
2600
2700
2800
2900
3000
3100
3200
3300
3400
3500
3600
3700
3800
380 1
3802
3803
3900
4000
4100
4200
4300
4400
4500

$
VER =' F$VERIFY()
$
SET NOVERIFY
$
$!
THIS COMMAND PROCEDURE LOCKS A TERMINAL AS BEING IN USE . IT DISABLES
$!
CONTROL Y, SETS A CONTROL Y ENTRY POINT AND LOOPS ON A SHOW TIME
$!
COMMAND. A CONTROL Y TYPED AT THE TERMINAL WILL TRANSFER CONTROL
$!
TO A CHECK FOR A PASSWORD TO EXIT FROM THIS PROCEDURE.
$
$
ON CONTROL Y THEN $GOTO PASSWORD
$
ON ERROR THEN GOTO ERR HNDLR
$BEGIN :
$
ERASE
$
$
COMMANDS :INUSE.TXT
$
$LOOP:
$
TIMSTR :='F$TIME()
$
DOT = 'F$LOCATE(" .",TIMSTR)
$
DOT = DOT-3
!SHOWTIMEDOWNTOMINUTES
$
TIMSTR: =' F$EXTRACT(O,DOT,TIMSTR)
$
TEXT 5 32 ""TIMSTR '"
$
TEXT 1 1
$
WAI T 00:01
$
GOTO LOOP
$
$PASSWORD:
$
TEXT 1 1
$
INQUIRE MAGIC " ENTER THE PASSWORD TO CONTINUE"
$
IF "" MAGIC"' .EQS . ""P1 "' THEN $GOTO EXIT
$
IF "" MAGIC"' .EQS . "REFRESH " THEN $GOTO BEGIN
$
ERASELINE 1
$
ERASELINE 2
$
ERASELINE 3
$
ERASELINE 4
$
ERASELINE 5
$
ERASELINE 6
$
ERASELINE 7
$
GOTO LOOP
$
$ ERR HNDLR :
$
ON ERROR THEN GOTO ERR HNDLR
$
GOTO PASSWORD
$EXIT:
$
SET CONTROL Y
$
ERASE
$
IF VER THEN $SET VERIFY
$
EXIT
$
$!
END OF I NUSE.COM

Figure 8-3
Example Command Procedure

Another added feature is the REFRESH statement (line
2900). The REFRESH statement directly follows the PASSWORD check statement (line 2900). If the screen gets cluttered with garbage characters, any user may enter a CONTROL Y, and in response to the system prompt for a password (line 2700), type REFRESH . REFRESH is recognized
in line 2900, clearing the entire screen of unwanted
characters . This is followed by a new IN USE and date and
time of day message.

time of day appear as:
dd-mmm-yyyy hh :mm
This date and time of day function are printed on the
screen above the " IN USE" message. The date and time of
day are refreshed once every minute. The PASSWORD
code (lines 2500-3700) is entered only if a Control-Y is
typed at the terminal while the terminal is locked up. The
command procedure prompts for a password . If the entered password matches the initial password (declared by
the current user of the terminal), the flow of execution
drops to the EXIT code block (lines 3900-4300). If the
passwords do not match , execution drops through to line
3000 , which clears the top seven lines of the screen .

DIFFERENCES UTILITY
By invoking the DIFFERENCES command, the user can
determine if two files are identical and, if not, how they dif-

8-9

exceed the right margin. RUNOFF justifies the line by expanding the spaces between words to produce an even
right margin.

fer. The DIFFERENCES utility compares the contents of
two disk files on a record-by-record basis and creates a
listing of the records that do not match. By default, the
DIFFERENCES utility compares every character in each
file, and by default, the DIFFERENCES utility writes the
output in ASCII.

Page Formatting
The page formatting commands control the appearance of
each page of output. For example, there are page formatting commands to establish the style and location of chapter headings and subheads. Other page formatting commands engage or disengage page numbering , produce
and format titles and subtitles, or force the printer to advance to a new page.

VAX-11 RUNOFF
VAX-11 RUNOFF is a document formatter. A RUNOFFprocessed document can be updated without extensive
retyping because text changes, via the text editors, do not
affect the basic design. The input to RUNOFF is a file containing the text of the document and the RUNOFF instructions. Executing in default mode, RUNOFF provides:

Title Formatting
Title formatting commands provide page, title , and subtitle
information for all pages. Such actions as placing only the
chapter heading on the first page of a chapter; printing any
subtitles of designated words; and determining the num ber of header levels (up to six) that the document will have
are all provided by the title formatting commands.

• a standard typewriter page size of BW' x 11 "
• sequential page numbering for every page but the first
• page width of 60 characters
• single spacing
• automatic tab settings for every eighth print position ,
starting with the ninth column (9, 17,25, etc.)

Subject-Matter Formatting
Subject-matter formatting commands include manag ing
the design and appearance of text, as with ragged ri ghthand margin, indenting a paragraph, skipping a number of
lines, centering the text, underlining, hyphenation , and
overstriking. Of course, different parts of the text may be
formatted differently, and commands may be comb ined .
To illustrate, a user has the option to have lists justified or
to have them with ragged margins.

• automatic filling and justifying
The output file is the print-ready document. After RUNOFF
has processed the file, the original file remains available
for further editing .
VAX-11 RUNOFF contains commands to perform the following functions:
• filling and justifying text

Index and Table of Contents
RUNOFF has powerful facilities for creating indexes and
tables of contents easily. There is a command to generate
a one-column index. In addition, the TCX program generates two-column indexes, while the TOG program generates tables of contents . Both TCX and TOG create files that
can be edited or can be processed by RUNOFF; this adds
great flexibility to the preparation of indexes and tab les of
contents.

• page formatting
• title formatting
• subject-matter formatting
• graphic, list and note formatting
• index and table of contents
• miscellaneous formatting
Filling and Justifying
RUNOFF commands set left and right margins, so that the
user may enter text without concern for line width or variable spacing between words. The RUNOFF program will
fill and justify the text when it is run . Filling is the successive addition of words to a line until one more word would

Miscellaneous Formatting
A number of useful RUNOFF commands help the user to
re-establish all default values, to add nonprinted comments to the sou rce file, to gather externally located files
into the input, and to set time and date.

8-10

9
Data
Management
Facilities
ENVIRONMENT DIVISION.
[INPUT-OUTPUT SECTION.]
FILE-CONTROL.
SELECT file -name
ASSIGN TO device-name-1

[, device-name-2]

; ORGANIZATION IS INDEXED

[

; ACCESS MODE IS

{

SEQUENTIAL} ]
RANDOM
DYNAMIC

; RECORD KEY IS data-name-1
[ ;ALTERNATE RECORD KEY IS data-name-2 [WITH DUPLICATES]]
DATA DIVISION.
[FILE SECTION.
[FD fi le -name
[ ; BLOCK CONTAINS
[;RECORD CONTAINS

[integer-1 TO]
[integer-3 TO]

integer-2

} ]
RECORDS
{ CHARACTERS

integer-4 CHARACTERS]

} {STANDARD}
RECORD IS
; LABEL { RECORDS ARE
OMITTED

[ ; DATA

}
RECORD IS
{ RECORDS ARE data-name-3 [. data-name-4) ..

PROCEDURE DIVISION.
OPEN

INPUT file -name -1 [. file -name-2) .
OUTPUT fi le-name-3 [. file -name -4) . . .
{
1-0 fil e-nam e-5 [. file-name-6) ...

}

...

]

VAX/VMS data management includes a file system that provides volume structuring and protection, and record management services that
provide device-independent access to the VAX peripherals.
The VAX/VMS on-disk structure provides a multilevel hierarchy of
named directories and subdirectories . Files can extend across multiple
volumes and be as large as the volume set on which they reside . Volumes are mounted to identify them to the system . VAX/VMS also supports multivolume ANSI format magnetic tape files with transparent
volume switching.
The VAX/VMS record management input/output system (RMS) provides device-independent access to disks, tapes , unit record equipment, terminals , and mailboxes. RMS allows user and application programs to create, access, and maintain data files with efficiency and
economy. Under RMS, records are regarded by the user program as
logical data units that are structured and accessed in accordance with
application requirements .
RMS provides sequential record access to sequential file organizations , sequential ; random , or combined record access to relative file
organizations and sequential, random , or a combination using index
key access to multikey indexed files . Multikey indexed file processing
includes incremental reorganization .
VAX/VMS also supports several other data management facilities:
DATATRIEVE, VAX-11 SORT, and the Forms Management System
(FMS) utility package.

INTRODUCTION

appears to a program to be a virtually contiguous set of
blocks . The blocks of the file, however, may be scattered
anywhere on a volume. Mapping information is maintained
to identify all the blocks constituting a file . Figure 9-1 illustrates the file structure.

The operating system's data management services are
provided by the following facilities :
• utilities for data and file manipulation and inquiry
• file system

Disk files can be extended easily . The blocks of the file are
allocated in physically contiguous sets , called extents.
Users are not requ ired to preallocate space, although they
can do so . Users can specify placement on an allocation
request , and they can control automatic allocation . For example, when a file is automatically extended, it can be extended by any g iven number of contiguous blocks . If desired , a file can be created as a contiguous file , in wh ich
case it is both virtually and physically contiguous.

• record management services
• device drivers
• command interpreter
Utilit ies which VAX offers include the VAX-11
SORT/ MERGE for reordering data, DATATRIEVE for data
inquiry and report writing , and FMS for screen formatting
and forms generation.
The file system provides volume structuring and directory
access to disk and magnetic tape files . Programmers can
use the file system as a base to build their own record
processing system, or they can use the VAX/VMS record
management services.

The disk structure includes duplicates of its critical volume
information . The system detects bad disk blocks dynamically and prevents re-use once the files to which they are
allocated are deleted .
Magnetic tape volumes are single-user volumes . Magnetic
tape files consist of physically contiguous blocks . Record
blocking is under program control. Files have ANSI format
labels . VAX/VMS also supports unlabeled (non-f ilestructured ) magnetic tapes.

The record management services (RMS ) provide deviceindependent access to all types of 1/ 0 peripherals . The
RMS procedures enable a program to access records
within files, and provide the same programm ing interface
regardless of device characteristics . The system includes
utilities for RMS file creation and maintenance.

File Directories and Directory Structures
A directory is a file containing a list of files on a given volume. A directory entry contains the name, type , version ,
and unique file ID for a particular file. A directory can list
files having the same owner UIC or files having different
owner UICs. The entries are listed alphabetically.

The device drivers provide the basic 1/ 0 device handling
for all of the other data management services. Device dri vers and their features are described in the Peripherals
and Operating System sections .
As described in the Users section , the command interpreter enables a user to reserve devices for exclusive use, set
device and directory name defaults, and assign logical
names to file specifications. The command interpreter also
enables the user to execute file management utilities that
provide fi le copy, transfer , and conversion operations .

A disk volume contains at least one directory, called the
master file directory. The system manager is responsible
for creating a volume's master file directory. The master
file directory can (and normally does) contain a list of
directory files wh ich form a second level of directories. The
second level of directory files can list data files and / or other directory files , called subdirectories. Users can create
subdirectories within the directories they own . The subdirectories can also list other directory files and/or data files.
Figure 9-2 illustrates a multilevel directory structure.

The fo llowing paragraphs discuss some of the features
and functions of the file system , includ ing the file structures , file naming facilities, and the file management utility
programs . The remainder of this section describes the
record management services programm ing environment ,
and utilities for high-level data and file man ipulation .

Since directories of files on volumes are files themselves ,
they are assigned owner UICs and can be protected from
certain kinds of access depending on the relationship established by an accessor's UIC. In the special case of
directory files, the file protection fields control an accessor's ability to :

FILE MANAGEMENT
VAX/VMS provides two file structures: one for disk volumes and one for magnetic tape volumes. From the user's
point of view, the only differences between the two file
structu res are those imposed by the capab ilities of the
media. Volumes are mounted for identif ication , and files
can extend across multiple volumes . The practical limit to
file size is that they can be only as large as the volume set
on which they reside .

• look up f iles
• enter new files in the directory, including new versions
of existing files
• remove files from the directory
File Specifications
A file specification identifies which file is to be used in a
file processing operation . Programs use file specifications
to identify the file they want to create, access, delete, or extend , and users supply the command interpreter with a file
specification to identify the file they want to edit, comp ile,
copy , delete, etc. A complete file specification is a well-defined character string composed of the following fields :

Volume and file protection are based on User Identification Codes (UICs) assigned to accessors and the file or
volume. The UICs establish the accessor's relationship to
the data structure as owner , the owner's group , the system , or the world (all others). Depend ing on the relationsh ip , the accessor may or may not have read , write, execute , or delete access to any given file.
Disk volumes are multiuser volumes . They can contain a
multilevel directory hierarchy that is defined dynamically
by the users of the volume . The on-disk file structure

• Node Name - The node of the network in which the volume containing the file is stored . The node name is

9-1

LOGICAL
BLOCK
#221

DATA
AREA 1
VIRTUAL
BLOCK
#1

DATA AREA PTR 1 I - DATA AREA PTR 2 ,____ _

#222

DATA AREA PTR 3 1--

#223

#224

#225

FILE HEADER BLOCK
FILE I.D .

INDEX FILE
FILE HEADER# 1
FILE HEADER#2
FILE HEADER #3

•
•
•

D-FILE
FILE HEADER BLOCK

•

D-FILE

LOGICAL
BLOCK
#172

DATA
AREA 2
VIRTUAL
BLOCK

#6
#173

#174

FILE HEADER BLOCK

LOGICAL
BLOCK
#450

DATA
AREA 3
VIRTUAL
BLOCK

Figure 9-1
Disk File Structure

followed by two colons(::) to delimit it from the remainder of the file specification.

OBA 1 is the name of the device (DB for disk pack device, A
for disk controller, 1 for drive unit number), [JONES] is the
directory name, HANOI is the file name, FOR is the file type
(meaning that th·e file is a FORTRAN source file), and 2 is
the version number.

• Device Name - The device on which the volume containing the file is mounted. The device name is followed
by a single colon (:)to delimit it from the remainder of
the file specification .

Neither programs nor command language users need to
provide a complete file specification to identify files . The
system applies defaults to most fields of a file specification
when they are not present. For example, if the node name
is not present, the node is assumed to be the node on
which the program is executing. If the version number is
not present, the version is always assumed to be the latest
version. Device name and directory name defaults for
users and the programs they execute are supplied by the
system manager in the user authorization file, and users
can change the standard defaults at any time during their
session on the system.

• Directory Name - The directory in which the file is listed. A directory name begins with an opening bracket ( <
or[) and ends with a closing bracket(> or]). If the file is
listed in a subdirectory, the directories to be searched
are listed in the desired search order, with the names
separated by periods, e.g.:
[name1 .name2.name3)

• File Name - The user-assigned name of the file.
• File Type - The type identification for the file. The type
is preceded by a period (.)to delimit it from the remainder of the file specification.

Some commands (such as COPY, PRINT, and DELETE)
accept a wild card in one or more fields of a file specification . A wild card is an asterisk appearing in a file specification field and it means "all."

• File Version - the generation number of the file. The file
version is preceded by a semicolon (;) or period (.) to
delimit it from the remainder of the file specification.

File specifications also apply to non-file-structured devices such as line printers, card readers, and terminals. In
these cases, however, the user or program needs to supply only the node name and device name, as appropriate.

For example, a complete file specification might be:
NODE47::DBA 1:[JONES]HANOl.FOR;2
In this case, NODE47 is the name of the network node,

9-2

MASTER FILE
DIRE CTO RY
JONES . DIR

JONES . DIR

SMITH . DIR

HANOI FOR
HANOI. LIS

PAYROLL . DIR

TEST . COM

et c.
etc .

:
:

L___.

BACKU P. DIR
MASTER . DAT

PAYROLL. DIR
MASTER . DAT
WEEKLY. DAT
BACKUP. DIR

WE EK 0 9. DAT
W EEK 10 DAT
W EE K l l. DAT

et c.

etc .

lf'AYROLg MASTER . DAT

[PAYROLL. BACKU ~ MASTER . DAT

Figure 9-2
Multilevel Directory Structure

Logical File Naming
To provide both system and device independence, users
and programs are not limited to identifying files by their file
specifications. They can use logical names in place of a
complete file specification , or in place of a portion of a file
specification. For example, a user can assign a logical
name to the left-most three fields of a file specification:

file is stored may be mounted on disk pack drive unit number 1 one week, or on unit 2 another week .
The application programmer can write the program
without knowing which directory the data file is listed in, or
which device the volume is mounted on . A series of logical
name assignments provides the complete file specification. The assignments are the responsibility of the people
who know what directory the file is listed in, and what drive
the volume is mounted on .

$ASSIGN NODE47 ::DBA4:[JONES] to VOL
And then use the logical name VOL in a subsequent command :

Logical name assignments can be made on a process,
group, or system-wide basis. Logical names can also be
recursive, that is, a logical name can be assigned to another logical name, or to a logical name and a portion of a file
specification.

In the example shown in Figure 9-3, the application program contains an OPEN statement for the payroll data file
logical
name
WEEKLY _PAYROLL_
using
the
CHANGES (note that underscore is a legal character) . The
application systems designer has created a command
procedure file called PAYRUN that controls the production
run . The command procedure file includes a logical name
assignment that obtains the file name as a parameter supplied by the operator or production clerk who starts the
production run . The logical name used by the application
program is given a value that consists of another logical
name (WEEKLY _PAYROLL) and the file name and type
specifications.

For example, suppose a company's weekly payroll production run includes an application program that uses the
current week's payroll changes data file . That data file may
be located in the directory named [PAYROLL) one week ,
or in the payroll backup subdirectory, [PAYROLL.BACKUP) , another week. The volume on which the

To complete the series of logical name assignments , the
payroll group operations manager makes a group-wide
logical name assignment: the payroll data files this week
are stored in the PAYROLL.BACKUP subdirectory. The
logical name assignment provides the directory name, using another logical name (PAY _PACK) known to the oper-

$TYPE VOL: HANOI.FOR
Defaults also apply when translating logical names, so that
the user could have made the assignment:
$ASSIGN NODE47: :(JONES] to VOL
In this case, the user's default device name would be used
to derive the complete file specification.

9-3

Application Programmmer:
OPEN (" WEEKLY _PAYROLL_CHANGES" )

Application System Programmer:
Command Procedure: PAY _RUN .COM
accepts one parameter (P1 ): Week Number
$ASSIGN WEEKLY _PAYROLL:' P1 '.WPY

WEEKLY_PAYROLL_CHANGES

$RUN APPLICATION

Production Clerk:
$ @PAY _RUN WEEK09

Payroll Group Operations Manager:
$ASSIGN/GROUP PAY _PACK:[PAYROLL.BACKUP]

WEEKLY PAYROLL

Local Operator:
$ASSIGN / SYSTEM DBA2:

PAY PACK

Figure 9-3
Logical Naming

on a Files-11 disk volume . It can also display the number
of available blocks in a volume, locate files that could not
otherwise be accessed , and list the names of files on the
volume.

ator who mounts the payroll data files volume. The operator makes the system-wide logical name assignment when
mounting the pack before the production run . Given the
assignments shown in the example, the logical name used
to open the file is translated to:

Bad Block Locator - The bad block locator utility determines the number and location of bad blocks on Files-1 1
disk volumes and stores this information in the bad block
file on the volume so that the blocks can not be allocated .
Running this utility before initializing a Files-11 volume is
useful in ensuring a disk's integrity.

DBA2:[PA YROLL.BACKUP]WEEK09.WPY
(The local system node name and the latest version number are used as defaults to complete the file specification .)
Should the directory name change, or the pack be mounted on another device that day, the only changes made are
the logical name assignments. There is no need to modify
either the application program or the command procedure
controlling the production run.

RMS Utilities - The record management services procedures are complemented by a number of utilities designed especially for RMS file creation and maintenance.
They allow the user to :

File Management
The VAX/VMS system includes many services that aid in
data management and maintenance. Some of these are
described in the following paragraphs .

• create an RMS file and define the attributes of the file
• list the attributes of a single file or a group of files , or list
the contents of a backup magnetic tape

Sorting Files - The SORT/MERGE program allows the
user to rearrange, delete, and reformat records in a file .
The user can arrange the records in the ascending or descending sequence of one or more fields within the records
for subsequent sequential processing . SORT can also create several different index files for accessing a file according to these indexes without reordering the data itself.

• convert a file wi th any file organization or record format
to a file with any other file organization or record format

Comparing Files - A file differences command contrasts
two files by automatically aligning matching text, and optionally ignoring comments, empty records , trailing
blanks, or multiple blanks. The output can be a file-by-file
list of differences, an interleaved list of differences, a list
with change bars, or a batch editor command input file .

RECORD MANAGEMENT SERVICES
The record management services (RMS) are a set of system procedures that provide efficient and flexible facilit ies
for data storage, retrieval , and modification. When writing
programs, the user can select processing methods from
among the RMS file organizations and accessing tech niques. The following sections discuss RMS :

• back up a single file or group of files in a compact fo rmat (optionally by creation or revision date)
• restore files previously backed up (optionally by
creation or revision date)

Backing Up Files and Volumes - The Disk Save and Compress (DSC) utility enables a user to back up entire disk
volumes to magnetic tape or to other disks. When backing
up disk volumes to other disk volumes, or restoring disk
volumes from magnetic tape, DSC combines unused
blocks on disks into contiguous areas.

• file organizations
• file attributes
• program operations
• run-time environment

Verifying File Structures - The file verification utility
checks the consistency and accuracy of the file structure

The manner in which RMS builds a file is called its organization . RMS provides three file organizations:
9-4

• sequential

exist within the information in the files.

• relative

RMS ensures that every record written into a file can subsequently be retrieved and passed to a requesting program as a single logical unit of data. The structure, or organization , of a file establishes the manner in which RMS
stores and retrieves records. The way a program requests
the storage or retrieval of records is known as the record
access mode. The organization of a file determines which
record access modes can be used .

• indexed
All three file organizations are available in both compatibility mode (using RMS-11) and in native mode (using VAX11 RMS ).
The organization of a file establishes the techniques one
can use to retrieve and store data in the file. These techniques are known as record access modes. The record
access modes that RMS supports are:

Sequential File Organization
In sequential file organization , records appear in
consecutive sequence. The order in which records appear
is the order in wh ich the records were originally written to
the file by an appl ication program or RMS utility. Sequential organization is the only file organization permitted for
magnetic tape and unit record devices. Most VAX/VMS
system utilities that deal with files , deal with sequentially
organized files. All system editors and language processors, for instance, operate on sequentially organized files.
Figure 9-4 illustrates sequential file organization .

• sequential
• random
• Record's File Address (RFA)
An application program or a RMS utility can be used when
creating a RMS file to specify the organization and characteristics of the file . Among the attributes specified are:
• storage medium
• file name and protection specifications
• record format and size

Relative File Organization
When relative organ ization is selected, RMS structures a
file as a series of fixed-size record cells . Cell size is based
on the maximum length permitted for a record in the file.
These cells are numbered from 1 (the first) ton (the last). A
cell's number represents its location relative to the beginning of the file.

• file allocation information
After RMS creates a file according to the specified attributes, application programs can store , retrieve and modify
data. These program operations take place on the logical
records in a file or the blocks comprising the file.
RMS FILE ORGANIZATIONS
A file is a collection of related information . For example , a
file might contain a company's personnel information (employee names, addresses , job titles) . Within this file, the information is divided into records. All the information on a
single employee might constitute a single record . Each
record in the personnel file would be subdivided into discrete pieces of information known as fields. The user defines the number, locations within the record , and logical
interpretations of these fields.

Each cell in a relative file can contain a single record.
There is no requ irement, however, that every cell contain a
record . Empty cells can be interspersed among cells containing records. Figure 9-5 illustrates a relative file organization.
Since cell numbers in a relative file are unique, they can be
used to identify both a cell and the record (if any) occupying that cell. Thus , record number 1 occupies the first cell
in the file , record number 17 occupies the seventeenth
cell, and so forth . When a cell number is used to identify a
record , it is also known as a relative record number.

The user can completely control the grouping of fields into
records and records into files. The relationship among
fields and records is embedded in the logic of the programs. RMS does not know the logical relationships that

Indexed File Organization
The location of records in indexed file organization is

END OF F I L E !

RE CO RD

RECORD

I RECORD I

Figure 9-4
Sequential Fiie Organization

CELL N O.

RECORD

Figure 9-5
Relative File Organization
9-5

999
RECORD
999

1000

Sequential Record Access Mode
Sequential record access mode can be used to access all
RMS files and all record-oriented devices, including mailboxes. Sequential record access means that records are
retrieved or written in the sequence established by the organization of the file.

transparent to the program. RMS completely controls the
placement of records in an indexed file. The presence of
keys in the records of the file governs this placement.
A key is a field present in every record of an indexed file.
The location and length of this field are identical in all records. When creating an indexed file, the user decides
which field or fields in the file's records are to be a key. Selecting such fields indicates to RMS that the contents (i.e.,
key value) of those fields in any particular record written to
the file can be used by a program to identify that record for
subsequent retrieval.

Sequential Access to Sequential Files - When using sequential record access mode in a sequentially organ ized
file , physical arrangement establishes the order in wh ich
records are retrieved . To read a particular record in a file,
say the fifteenth record , a program must open the file and
access the first fourteen records before accessing the
desired record . Thus each record in a sequential file can
be retrieved only by first accessing all records that physically precede it. Similarly, once a program has retrieved
the fifteenth record , it can read all the remaining records
(from the sixteenth on) in physical sequence. It cannot ,
however, read any preceding record without closing and
reopening the file and beginning again with the first re cord .

At least one key must be defined for an indexed file : the
primary key . Optionally, additional keys or alternate keys
can be defined . An alternate key value can also be used as
a means of identifying a record for retrieval.
As programs write records into an indexed file , RMS
builds a tree-structured table known as an index. An index
consists of a series of entries containing a key value copied from a record that a program wrote into the file. Stored
with each key value is a pointer to the location in the file of
the record from which the value was copied. RMS builds
and maintains a separate index for each key defined for
the file. Each index is stored in the file. Thus, every indexed file contains at least one index, the primary key index. Figure 9-6 illustrates an indexed file organization with
a primary key. When alternate keys are defined, RMS
builds and stores an additional index for each alternate
key.

Sequential Record Access to Relative Files - During the
sequential access of records in the relative file organization , the contents of the record cells in the file establish the
order in which a program processes records . RMS recognizes whether successively numbered record cells are
empty or contain records.
When a program issues read requests in sequential record
access mode for a relative file , RMS ignores empty record
cells and searches successive cells for the first one containing a record . When a program adds new records in sequential record access mode to a relative file, RMS places
a record in the cell whose relative number is one higher
than the relative number of the previous request, as long
as that cell does not already contain a record . RMS allows
a program to write new records only into empty cells in the
file .

RMS RECORD ACCESS MODES
The methods of retrieving and storing records in a file are
called record access modes. A different record access
mode can be used to process records within the file each
time it is opened. A program can also change record access mode during the processing of a file. RMS permits
only certain combinations of file organization and recor'.d
access mode. Table 9-1 lists these combinations.

Sequential Record Access to Indexed Files - A program
can use the sequential record access mode to retrieve rec-

KEY
DEFINITION

1 ~----PRIMARY

ABLE

ELM AV

24379

INDEX(EMPLOYEE

NAME---~

SMITH

JONES

JO NE S

MAIN ST

19 72 4

' - - - - - - - -- -- - - - -- -- - DATA RE CORDS--

SMITH
-

Figure 9-6
Indexed File Organization

9-6

HO LT RD

- - -- -- - - --

35888

- - -- /

Table 9-1
Record Access Modes and File Organizations
File Organization
Sequential
Sequential
Relative '
Indexed '

Record Access Mode
Random
Record#
Yes 2
Yes
No

Yes
Yes
Yes

No
No
Yes

RFA
Key Value
Yes '
Yes
Yes

1 Disk files only.
2 Fixed length record format d isk files only.

ords from an indexed file in the order represented by any
index. The entries in an index are arranged in ascending
order by key values . If more than one key is defined for the
file, each separate index associated with a key represents
a different logical ordering of the records in the file.

index, etc.) that RMS must search. When RMS finds the
key value in the specified index, it reads the record that the
index entry points to and passes the record to the user
program.
Program requests to write records randomly in an indexed
file do not require the separate specification of a key value.
All key values (primary and , if any, alternate key values)
are in the record itself. When an indexed file is opened,
RMS retrieves all definitions stored in the file. RMS knows
the location and length of each key field in a record . Before
writing a record into the file, RMS examines the values
contained in the key fields and creates new entries in the
indexes. In this way RMS ensures that the record can be
retrieved by any of its key values .

When reading records in sequential record access mode
from an indexed file, a program initially specifies a key
(primary key , first alternate key , second alternate key, etc.)
to RMS. Thereafter, RMS uses the index associated with
that specified key to retrieve records in the sequence represented by the entries in the index. Each successive record RMS returns in response to a read request contains a
value in the specified key field that is equal to or greater
than that of the previous record returned .

Record's File Address (RFA) Record Access Mode
Record 's File Address (RFA) record access mode can be
used to retrieve records in any file organization as long as
the file res ides on a disk volume. Like random record access mode, RFA record access allows a specific record to
be identified for retrieval , using the record's unique address. The actual format of this address depends on the
organization of the file.

When writing records to an indexed file , RMS uses the definition of the primary key field to place the record in the
file .
Random Record Access Mode
In random record access mode, the program establishes
the order in which records are processed . Each program
request for access to a record operates independently of
the previous record accessed. Each request in random
record access mode identifies the particular record of interest. Successive requests in random mode can identify
and access records anywhere in the file.

After every successful read or write operation, RMS returns the RFA of the subject record to the program . The
program can then save this RFA to use again to retrieve
the same record . It is not required that this RFA be used
only during the current execution of the program. RFAs
can be saved and used at any subsequent time.

Random Record Access to Sequential Files - Native programs can access sequential files on d isk using relative record number to randomly locate a record , provided that
the records are in fixed-length record format.

Dynamic Access
Dynamic access is not strictly an access mode. It is the
ability to switch from one record access mode to another
while processing a file. For example, a program can access a record randomly, then switch to sequential record
access mode for processing subsequent records. There is
no limitation on the number of times such switching can
occur. The only limitation is that the file organization must
support the record access mode selected.

Random Record Access to Relative Files - Programs can
read or write records in a relative file by specifying the relative record number. RMS interprets each number as the
corresponding cell in the file . A program can read records
at random by successively requesting , for example, record
number 47 , record number 11, record number 31, and so
forth . If no record exists in a specified cell , RMS notifies
the requesting program. Similarly, a program can store
records in a relative file by identifying the cell in the file that
a record is to occupy. If a program attempts to write a new
record in a cell already containing a record , RMS notifies
the program .

FILE AND RECORD ATTRIBUTES
When creating an RMS file, a program or user defines its
logical and physical characteristics, or attributes. These
characteristics are defined by source language statements
in an application program or by an RMS utility. The program or user assigns the file a name, the owner's User
Identification Code, and a protection code, and selects the
file organization . The program or user also defines or selects other attributes, including:

Random Record Access to Indexed Files - For indexed
files, a key value rather than a relative record number
identifies the record . Each program read request in random record access mode specifies a key value and the index (primary index, first alternate index, second alternate
9-7

• device

data record . Although stored together , the two parts are
returned to the program separately when the record is
read. The size of the fixed control area is identical for all
records of the file . The contents of the fixed control area
are completely under the control of the program and can
be used for any purpose. For example, fixed control areas
can be used to store the identifier (relative record number
or RFA) of related records. Indexed file organizations do
not support VFC record format.

• file size
• file location
• record format and size
• keys (for indexed files only)
Selection of device is related to the organization of the file.
Sequential files can be created on Files-11 disk volumes or
ANSI magnetic tape volumes. Sequential files can also be
read from mailboxes, terminals, and card readers, and
written to mailboxes, terminals, and line printers. Relative
and indexed files can be created on Files-11 disk volumes.

Key Definitions for Indexed Files
To define a key fo r an indexed file , the user specifies the
position and length of particular data fields within the records. At least one key, the primary key, must be defined
for an indexed file. Additionally, up to 254 alternate keys
can be defined. In general, most files have two or three
keys. Because indexes require storage space and RMS
updates indexes as records are added or modified , no
more than six to eight keys should be defined where storage space or access time is important.

The logical limit on file size is 2 31 -1 blocks, with a more
realistic limit being the volume set on which a file can reside. When creating an RMS file on a disk volume, the user
can specify an initial allocation size. If no file size is given,
RMS allocates the minimum amount of storage needed to
contain the defined attributes of the file. The initial size can
be extended dynamically. The user can let RMS locate the
file , or the user can allocate the file to specific locations on
the disk to optimize disk access time. The file's starting location can be specified optionally using a volume-relative
block number, or a physical cylinder address.

Each primary and alternate key represents from 1 to 255
bytes in each reco rd of the file. RMS permits six key field
data types.
• string

When creating a file on a magnetic tape volume, a user or
program does not specify an initial allocation size. The
blocks are simply written one after another down the tape ,
beginning after the last file, if any, written on the tape.
Once a tape file has been created, another file can replace
it or be appended to it, but all subsequent files on the tape,
if any, are lost.

• signed 15-bit integer
• unsigned 16-bit binary
• signed 31-bit integer
• unsigned 32-bit binary
• packed decimal
The string key field can be composed of simple or segmented keys . A simple key is a single, contiguous string of
characters in the record ; in other words, a single field . A
segmented key, however, can consist of from two to eight
fields within records. These fields need not be contiguous .
When processing records that contain segmented keys ,
RMS treats the separate fields (segments) as a logically
contiguous character string. The integer , binary , and
packed decimal data types can only be simple keys.

Record Formats
The user provides the format and maximum size specifications for the records the file will contain. The specified format establishes how each record appears in the file. The
size specification allows RMS to verify that records written
into the file do not exceed the length specified when the
file was created.
Fixed length record format refers to records of a file that
are all equal in size. Each record occupies an identical amount of space in the file. All file organizations support
fixed length record format.

When defining keys at file creation time, two characteristics for each key can be specified :
• duplicate key values are or are not allowed

Variable-length record format records can be either equal
or unequal in length . All file organizations support variable-length record format. RMS prefixes a count field to
each variable-length record it writes. The count field describes the length (in bytes) of the record . RMS removes
this count field before it passes a record to the program .
RMS produces two types of count fields, depending on the
storage medium on which the file resides:

• key value can or cannot change
When duplicate key values are allowed, more than one
record can have the same value in a given key. For example, the creator of a personnel file could define the department name field as an alternate key. As programs wrote
records into the file, the alternate index for the department
name key field would contain multiple entries for each key
value (e .g., PAYROLL, SALES, ADMINISTRATION) , since
departments are composed of more than one employee.
When such duplication occurs, RMS stores the records so
that they can be retrieved in first-in/first-out (FIFO) order.

• Variable-length records in files on Files-11 disk volumes
have a 2-byte binary count field preceding the data field
portion of each record. The specified size excludes the
count field .

If key values can change, records can be read and then
written back into the file with a modified key value. For example , this specification would allow a program to access
a record in the personnel file and change the contents of a
department name field to reflect the transfer of an em ployee from one department to another. This characteristic can be specified only for alternate keys. If key values
can change, the user must also specify that the duplicate

• Variable-length records on ANSI magnetic tapes have
4-character decimal count fields preceding the data
portion of each· record. The specified size includes the
count field. In the context of ANSI tapes, this record format is known as. D format.
Variable-with-fixed-control (VFC) records consist of two
distinct parts, the fixed control area and a variable-length

9-8

key values are allowed . If the primary key value can
change, the user may not change the record length.

File Processing
At the file level , that is, independent of record processing ,
a program can:

Figures 9-7 and 9-8 show excerpts from a COBOL program which operates upon an indexed customer information file via the dynamic access method . The program
searches through the file and generates various reports
based upon the customer's financial status and additional
input typed in by the user at the terminal. In Figure 9-7 , the
program describes the organization of the file and specifies the access method to be used. In Figure 9-8, the program searches for the first non-zero customer number.
Using the " approximate key" match facil ity (greater than) ,
the program searches for the first non-zero customer.
When RMS has located the first non-zero customer number , the program changes access method and the file is
read sequentially.

• create a file
• open an existing file
• modify file attributes
• extend a file
• close the file
• delete a file
Once a program has opened a file for the first time, it has
access to the un ique internal ID for the file. If the program
intends to open the file subsequently, it can use that internal ID to open the file and avoid any directory search .
Record 1/0 Processing
The organization of a file, defined when the file is created ,
determines the types of operations that the program can
perform on records . Depending on file organization , RMS
permits a program to perform the following record operations:

INPUT-OUTPUT SECTION .
FILE-CONTROL.
SELECT CUSTOMER-FILE
ASSIGN TO "CUSTOM.DAT"
ORGANIZATION IS INDEXED
ACCESS MODE IS DYNAMIC
RECORD KEY IS GUST-CUSTOMER- NUMBER
ALTERNATE RECORD IS KEY IS GUST-CUSTOMER-NAME
FILE STATUS IS CUSTOMER-FILE-STATUS .

• Read a record . RMS returns an existing record within
the file to the program .
• Write a record . RMS adds a new record that the program constructs to the file . The new record cannot replace an already ex isting record .

Figure 9-7

• Find a record . RMS locates an existing record in the file .
It does not return the record to the program , but establishes a new current position in the file.

ISAM File Description

• Delete a record . RMS removes an existing record from
the file. The delete record operation is not valid for the
sequential file organ ization .

OPEN INPUT CUSTOMER-FILE.
MOVE "000000" TO CUST-CUST-NUMBER.
START CUSTOMER-FILE.
KEY IS> CUST-CUST-NUMBER.
OPEN OUTPUT STATEMENT-REPORT.

• Update a record . The program modifies the contents of
a record in the file. RMS writes the modified record into
the file, replacing the old record. The update record operation is not val id for sequential file organizations ,
except for sequentially organized disk files .
Sequential File Record 110 - In a sequential file organization , a program can read existing records from the file using sequential , RFA, or, if the file contains fixed-length records, random record access mode. New records can be
added only to the end of the file and only through the use
of sequential or random record access mode.

....................................................•
MAINLINE SECTION.
SBEGIN .
READ CUSTOMER-FILE NEXT
AT END
GO TO END-PROCESS .
ADD 1 TO RECORD-COUNT .
Figure 9-8

The find operation is supported in both sequential and
RFA record access modes. In sequential record access
mode the program can use a find operation to skip records . In RFA record access mode, the program can use
the find operation to establish a random starting point in
the file for sequent ial read operations.

Dynamic Access Processing

PROGRAM OPERATIONS ON RMS FILES
After RMS has created a file according to the user's description of file characteristics , a program can access the
file and store and retrieve data.

The sequential file organization does not support the delete operation, since the structure of the file requires that
records be adjacent in and across virtual blocks. A program can, however, update existing records in sequential
disk files as long as the modification of a record does not
alter its size.
Relative File Record 110 - Relative file organization perm its programs greater flexibility in performing record
operations than does sequential organization. A program
can read existing records from the file using sequential ,
random , or RFA record access mode.

When a program accesses the file as a log ical structure
(i. e., a sequential , relative , or indexed file) , it uses record
1/0 operations such as add , update, and delete record.
The organization of the file determines the types of record
operations perm itted .
If the record accessing capabilities of RMS are not used ,
programs can access the file as an array of virtual blocks.
To process a file at this level, programs use a type of access known as block 1/0.
9-9

New records can be sequentially or randomly written as
long as the intended record cell does not already contain a
record . Similarly, any record access mode can be used to
perform a find operation . After a record has been found or
read, RMS permits the delete operation. Once a record
has been deleted , the record cell is available for a new record . A program can also update records in the file. If the
format of the records is variable, update operations can
modify record length up to the maximum size specified
when the file was created.

Indexed File Record 110 - Indexed file organization provides the greatest flexibility in performing record operations. A program can read existing records from the file in
sequential , RFA, or random record access mode. When
reading records in random record access mode, the program can choose one of four types of matches that RMS
performs using the program-provided key value. The four
types of matches are:
• exact key match
• approximate key match
• generic key match
• approximate and generic key match
Exact key match requires that the contents of the key in the
record retrieved precisely match the key value specified in
the program read operation.
The approximate match facility allows the program to select either of the following relationships between the key of
the record retrieved and the key value specified by the
program:
• equal to or greater than
• greater than
The advantage of this kind of match is that if the requested
key value does not exist in any record of the file, RMS returns the record that contains the next higher key value.
This allows the program to retrieve records without knowing an exact key value.
Generic key match means that the program need specify
only an initial portion of the key value. RMS returns to the
program the first occurrence of a record whose key contains a value beginning with those characters. This allows
the program to retrieve a class of records, for example, all
employee records in the personnel file with a name field
beginning with M.

mode. When finding records in random record access
mode, the program can specify any one of the four types of
key matches provided for read operations.
In addition to read, write, and find operations, the program
can delete any record in an indexed file and update any
record . The only restriction RMS applies during an update
operation is that the contents of the modified record must
not violate any user-defined key characteristic (e .g., key
values cannot change and duplicate key values are not
permitted).

The final type of key match combines both generic and approximate facilities. The program specifies only an initial
portion of the key value, as with generic match.
Additionally, a program specifies that the key data field of
the record retrieved must be either:

Block 1/0 Processing
Block 1/0 allows a program to bypass the record processing capabilities of RMS entirely. Rather than perform ing
record operations through the use of supported record access modes, a program can process a file as a structure
consisting solely of blocks.

• equal to or greater than the program-supplied
value
• greater than the program-supplied value
RMS also allows any number of new records to be written
into an indexed file. It rejects a write operation only if the
value contained in a key of the record violates a user-defined key characteristic (e.g., duplicate key values not permitted).

Using block 1/0, a program reads or writes blocks by identifying a starting virtual block number in the file and a
transfer length. Regardless of the organization of the file,
RMS accesses the identified block or blocks on behalf of
the program.

The find operation, similar to the read operation , can be
performed in sequential, RFA, or random record access
9-10

Since RMS files, particularly relative and indexed files,
contain internal information meaningful only to RMS itself,
DIGIT AL does not recommend that a file be modified by
using block 1/0. The presence of the block 1/0 facility,
however, does permit user-created record formats on a
Files-11 disk volume or ANSI magnetic tape volume.

RMS RUN TIME ENVIRONMENT
The environment within which a program processes RMS
files at run time has two levels, the file processing level and
the record processing level.
At the file processing level , RMS and the operating system
provide an environment permitting concurrently executing
programs to share access to the same file. RMS ascertains
the amount of sharing permissible from information provided by the programs themselves. Additionally , at the file
processing level, RMS provides facilities allowing programs to exercise as little or as much control over buffer
space requirements for file processing as desired.

with the corresponding information another program provides , both programs can access the file concurrently.
Buffer Handling - To a program, record processing under
RMS appears as the direct movement of records between
a file and the program itself. Transparently to the program,
however, RMS reads or writes the blocks of a file into or
from internal memory areas known as 1/0 buffers. Records within these buffers are then made available to the
program . Users can control the number and size of buffers. For sequential record access, users can choose an
optional 1/0 read-ahead and write-behind buffer management. For magnetic tape file access, they can control the
number of buffers for multiple buffering. For sequential
disk files, users can specify the number of blocks that are
to be transferred whenever RMS performs an 1/0 operation .

Run Time Record Processing
After opening a file , a program can access records in the
file through the RMS record processing environment. This
environment provides three facilities:

At the record processing level, RMS allows programs to
access records in a file through one or more record access
streams. Each record access stream represents an independent and simultaneously active series of record operations directed toward the file. Within each stream, programs can perform record operations synchronously or
asynchronously. That is, RMS allows programs to choose
between receiving control only after a record operation request has been satisfied (synchronous operation) or receiving control before the request has been satisfied
(asynchronous operation).

• record access streams
• synchronous or asynchronous record operations
• record transfer modes
Record Access Streams - In the record processing environment, a program accesses records in a file through a
record access stream . A record access stream is a serial
sequence of record operation requests. For example, a
program can issue a read request for a particular record ,
receive the record from RMS, modify the contents of the
record, and then issue an update request that causes RMS
to write the record back into the file. The sequence of read
and update record operation requests can then be performed for a different record, or other record operations
can be performed , again in a serial fashion. Thus , within a
record access stream, there is at most one record being
processed at any time .

For both synchronous and asynchronous record operations, RMS provides two record transfer modes, move
mode and locate mode. Move mode causes RMS to copy a
record to/from an 1/0 buffer from/to a program-provided
location. Locate mode allows programs to process retrieved records directly in an 1/0 buffer.

For relative and in dexed files, RMS permits a program to
establish multiple record access streams for record operations to the same file. The presence of such multiple record access streams allows programs to process in parallel
more than one record of a file . Each stream represents an
independent and concurrently active sequence of record
operations.

Run Time File Processing
RMS allows executing programs to share files rather than
requiring them to process files serially. The manner in
which a file can be shared depends on the organization of
the file. Program-provided information further establishes
the degree of sharing of a particular file.
File Organization and Sharing - With the exception of
magnetic tape files, which cannot be shared , an RMS file
can be shared by any number of programs that are reading, but not writing, the file. Sequential disk files can be
shared by multiple readers and multiple writers, but they
are responsible for any record locking required to handle
multiple readers and writers properly.

As an example of multiple record access streams, a program could open an indexed file and establish two record
access streams to the file. The program could use one
record access stream to access records in the file in random access mode through the primary index. At the same
time, the program could use the second record access
stream to access records sequentially in the order specified by an alternate index.

Program Sharing Information - A program specifies what
kind of sharing actually occurs at run time . The user controls the sharing of a file through information the program
provides RMS when it opens the file. First, a program must
declare what operations (e.g., read, write, delete, update)
it intends to perform on the file. Second , a program must
specify whether other programs can read the file or both
read and write the file concurrently with this program.
These two types of information allow RMS to determine if
multiple user programs can access a file at the same time.
Whenever a program's sharing information is compatible

Synchronous and Asynchronous Record Operations Within each record access stream, a program can perform
any record operation either synchronously or asynchronously. When a record operation is performed synchronously, RMS returns control to a program only after the record operation request has been satisfied (e.g., a record
has been read and passed on to the program).

If the programming language allows asynchronous processing, RMS can return control to a program before the
9-11

instruction. The NLK bit specifies that the record accessed with either a $GET or $FIND macro instruction is
not to be locked , while the RLK bit specifies that a record may be accessible for read purposes but may not
otherwise be a<;:cessed .

record operation request has been satisfied . A program
can use the time required for the physical transfer to perform other computations. A program cannot, however, issue a second record operation through the same stream
until the first record operation has completed. To ascertain
when a record operation has actually been performed , a
program can specify completion routines or issue a wait
request and regain control when the record operation is
complete.

UTILITY LANGUAGES
VAX/VMS supports a number of data management fac il,ties : DATATRIEVE, VAX-11 SORT, and the Forms Management System (FMS) utility package.

Record Transfer Modes - A program can use either of
two record transfer modes to gain access to each record in
memory:

DATATRIEVE

• move mode

DATATRIEVE is user application software that provides
direct, easy, and fast access to data contained in VAX-11
RMS (Record Management System) files. The system is
designed for relatively unsophisticated computer users;
everyday use of DATATRIEVE requires no programm ing
skills. While providing the user with an inquiry language
and a report writing facility, DAT A TRI EVE also suppo rts a
user-specifiable Data Dictionary which describes VAX-11
RMS record formats. DATATRIEVE data management facil ities include interactive retrieval , sort, update, and display of data records, in addition to maintenance com mands for the Data Dictionary.

• locate mode
Move mode means that the individual records are copied
between the 1/0 buffer and a program. For read operations , RMS reads a block into an 1/0 buffer, finds the desired record within the buffer, and moves the record to a
program-specified location in its work space . For write operations, the program builds or modifies a record in its
own work space and RMS moves the record to an 1/0 buffer. RMS supports move mode record operations for all file
organizations.
Locate mode enables programs to read records directly in
an 1/0 buffer. Locate mode reduces the amount of data
movement, thereby saving processing time. RMS provides
the program with the address and size of the record in the
1/0 buffer. RMS supports locate mode record transfers on
all file organizations for read operations only.

OATATRIEVE Inquiry Facility
DATATRIEVE accepts English-like commands from the
user, and reacts by modifying, updating , or extracting data
from the specified VAX-11 RMS file . In those cases where
certa in sequences of commands need to be issued on a
recurring basis, DAT A TRI EVE provides a feature that per mits the definition and use of procedures. A procedure is a
group of DAT A TRI EVE statements and commands identifiable (callable) by a unique procedure name. At any time
during the interactive session, this group of DAT A TRI EVE
statements and commands can be invoked simply by calling the procedure name.

RMS Record Locking
VAX-11 RMS provides a record locking capability for files
that use the relative and indexed organization . In addition ,
RMS record locking is supported for sequential files with
512-byte fixed length records. This provides control over
operation when the file is being accessed simultaneously
by more than one program and/or more than one stream
in a program . Record locking makes certain that when a
program is adding, deleting, or modifying a record on a
given stream , another program or stream is not allowed
access to the same record or record cell. RMS-11 executing in compatibility mode does not support record locking
and file sharing . There are two varieties of record locking
and unlocking:

DATATRIEVE Report Writer Facility
In addition to query commands , DATATRIEVE provides a
report facility to generate reports from VAX-11 RMS files.
The report facility allows the user to specify the following
parameters:
• Spacing
• Titles

• Automatic Record Locking - The lock occurs on every
execution of a $FIND or $GET macro instruction , and
the lock is released when the next record is accessed ,
the current record is updated or deleted, the record
stream is disconnected , or the file is closed . The $FREE
macro instruction explicitly unlocks all records previously locked for a particular record stream. The $RELEASE macro instruction explicitly unlocks a specified
record in a record stream .

• Column headings
• Page headings and footnotes
• Report headings
Commands to the report facility are simply an extension of
query facility commands. Although the report facility provides extensive formatting capabilities, its default settings
are suitable for many applications, further simplifying its
use. Furthermore, errors in commands are discovered immediately (as in the query facility), so the user can correct
the commands before printing wrong or incomplete reports.

• Manual Record Locking - In manual record locking,
varying degrees of locking may be specified by setting
bits in the record processing options field (ROP) of the
RAB . The ULK bit specifies manual (as opposed to automatic) locking and unlocking. This bit specifies that
locking will occur on the execution of a $GET, $FIND, or
$PUT macro instruction and that unlocking may take
place explicitly only via a $FREE or $RELEASE macro

Basic Commands
DATATRIEVE uses a simple English-like command lan guage for data retrieval , modification , and display .
Prompting is automatic for both command and data entry.
9-12

The major commands are:

The term fil es refers to the log ically related groups of data
that are kept by RMS . For example, we might put all of the
yacht records for a current inventory of yachts into one file.

• HELP - provides a summary of each DATATRIEVE
command .
• READY - identifies a domain for processing and controls the access mode to the appropriate file.

Domains are named groups of data containing records of
a single type. A DATATRIEVE domain consists of all the
records in a particular RMS file, in addition to a record definition of this file contained in the Data Dictionary. In this
case, we could say that all the yacht records for the current
inventory are kept in the YACHTS domain . The number of
records in any domain may change as new records are
stored or old records are erased.

• FIND - establishes a collection (subset) of records contained in either a domain or a previously established
collection based on a Boolean expression .
• SORT - reorders a collection of records in either the
ascending or descending sequence of the contents of
one or more fields in the records .

A record collection is a subset of a domain. It may consist
of no records , one record, or up to all the records in the
domain . Using our previous example, we could say that all
the yachts manufactured by Grampian could be made to
form the Grampian collection , while those yachts manufactured by Islander could be used to form the Islander
collection . To carry this example one step further, if the inventory is currently out of stock of yachts manufactured by
Seaworthy, the Seaworthy collection will be empty, or null.

• PRINT - pr ints one or more fields of one or more records. Output can optionally be directed to a line printer
or disk file. Format control can be specified . A column
header is generated automatically.
• SELECT - identifies a single record in a collection for
subsequent individual processing .
• MODIFY - alters the values of one or more fields for either the selected record or all records in a collection .
Replacement values are prompted for by name.

The Data Dictionary is a location where the definitions for
procedures , records , and domains are kept in a standard
fashion by DATATRIEVE. The data administrator will be
concerned with the creation and maintenance of Data Dictionary information . Certain users will be able to display
certain information from this dictionary, but only management will be concerned with defining it.

• STORE - creates a new record . The value for each field
contained in the record is prompted for by name.
• ERASE - removes one or more records from the RMS
file corresponding to the appropriate domain.
• FOR - executes a subsequent command once for each
record in the record collection, providing a simple looping facility.

Keywords
DAT A TRI EVE util izes language elements called keywords
which have a spec ific denotation and associated function .
If they are used in any other context, they may serve to
confuse the system about user intentions. Thus, it is good
policy to avoid the use of these words as names of domains, procedures , records, fields, and collections.

• EXTRACT - copies domains, records, procedures , and
tables from the Data Dictionary to an external file.
• SHOW FIELDS - prints field names and data types for
all fields in ready domains.
• DEFINE DICTIONARY tionaries.

allows creation of private dic-

In add ition to the simple data manipulation commands, a
number of more complex commands are available for the
advanced user. These commands, such as REPEAT , BEGIN-END, and IF-THEN-ELSE, may be used to combine
two or more DATATRIEVE commands into a single compound command . These , in turn , may be stored in the Data
Dictionary as procedures for invocation by less experienced users.

Additional DATATRIEVE Features
Among the many DATATRIEVE features supported by
VAX/VMS are:
• Application Des ign Tool (ADT)
ADT allows less experienced users to set up simple DAT A TRI EVE applications. Through a simple dialogue,
ADT generates a command file containing the record ,
domain , and file definitions.

DATATRIEVE provides a full set of arithmetic operators
(addition , subtraction , multiplication , division , and negation) , a set of statistical operators (total , average, maximum , minimum , and count) , and provides automatic conversion between data types used i n the FORTRAN ,
COBOL, DIBOL, and BASIC languages.

• Nested Procedures
Procedures may contain references to other procedures
(nested procedures) provided that no procedure invokes itself either directly or indirectly. The maximum
depth of nesting varies from about 10 to about 30 depending on the amount of memory available, number
and size of established collection, etc.

Terminology
Files, domains, collections , records, and fields are terms
of fundamental importance to the file structure of DATATRIEVE .

• Data Hierarchies
Use of hierarch ies allows manipulation of complex data
containing lists and sublists. A hierarchy may be defined
as a single file with a repeating group or multiple domains automatically cross-linked. Extensions related to
hierarchies include the inner print list (to override defau It formatting of a sublist) and the ANY Boolean
expression, which allows DATATRIEVE to search a sublist for the existence of a particular record.

Records are groups of related items of data that are treated as a unit. For example, all the pieces of data describing
a model of a yacht in a marina could be grouped to constitute the record for that yacht.
Each of the individual pieces of data in a record is referred
to as a field . The yacht's model number, length , and price
are all potential fields in its record.
9-13

• Views
Views can be used to restrict the set of fields accessible,
to apply an automatic selection criterion to a file, or to
cross-link a number of elementary domains to/from an
apparent hierarchy. Once defined, a view is indistinguishable from an RMS domain to the user.

• Procedure Editor
A DATATRIEVE procedure editor has been added . The
editor is invoked by the command:

• DATE Data Type
This allows easy inclusion of dates in DAT A TRI EVE records, direct comparison of dates, computation of
elapsed dates. Dates may be formatted for printing in
virtually any form. Similarly, DATATRIEVE accepts the
entry of dates in virtually any form . The DATATRIEVE
date format is compatible with the VAX/VMS date standard.

The command EDIT procedure-name invokes an editor
which can insert, replace, or delete text from procedures defined in the data dictionary.

EDIT procedure-name
where "procedure-name" is the name of an existing
procedure.

VAX-11 SORT/MERGE
VAX-11 SORT /MERGE is a native mode utility that may be
run interactively, as a batch job, or it can be callable from a
user-written VAX-11 native mode program.

• Tables
Tables are generally used to translate encoded values
into something that can be edited by the DAT A TRI EVE
editor. Table lookups are performed by the VIA value
expression; table searches (for table membership) are
specified with the Boolean IN expression.

The SORT utility allows the user to reorder data from any
input file into a new file in a sequence based upon key
fields within the input data records. A user can specify up
to ten input files and SORT will produce one sorted output
file . The sorting sequence is determined by user-specified
control fields, also known as key fields, within the data
themselves. If the user does not wish to reorder the data
base , SORT can still be used to extract key information ,
sort that information, and store the sorted information as a
permanent file. Later that file can be used to access the
data base in the order of the key information in the sorted
file. The contents of the sorted file may be entire records,
key fields, or record file addresses which point to the position of each record within the file.

• TOT AL Statement
The TOTAL statement allows very simple computation
of totals and subtotals.
• CONTAINING Relational Operator
CONTAINING is used in a record selection expression
to retrieve records with a field containing a particular
substring. The substring may be anywhere in the field,
and need not match the case (uppercase/lowercase) of
the search string. For example, the command:

SORT provides four sorting techniques:

FIND BOOKS WITH TITLE CONTAINING "LASER"

• Record Sort produces a reordered data file sorted by
specified keys, moving the entire contents of each record during the sort. A record sort can be used on any
acceptable VMS input device and can process any valid
VAX-11 RMS format.

will find all records in BOOKS with the word "LASER"
somewhere in the field TITLE.
• OCCURS Clause
Use of the OCCURS clause permits definition of records
containing a repeating group (sublist). The sublist may
be of fixed or variable length.

• Tag Sort produces a reordered data file by sorting
specified keys, but moving only the record keys during
the sort. SORT then randomly reaccesses the input file
to create a resequenced output file according to those
record keys. The tag sort method conserves temporary
storage, but can accept only input files residing on disk.

• Value Validation
A Boolean validation expression may be included as
part of a field description in a record definition. If specified on a field, the validation expression is automatically
executed every time the field is modified, to insure that
only legal values are stored in a data base. If a validation
error is detected, the user is re-prompted for a new value if possible, or the DATATRIEVE statement is aborted.

• Address Sort produces an address file without reordering the input file. The address file contains RFAs, a
pointer to each record's location in the file which can later be used as an index to read the data base in the desired sequence. Any number of address files may be
created for the same data base. A customer master file,
for instance, may be referenced by customer-number
index or sales territory index for different reports . Address sort is the fastest of the four sorting methods.

• COMPUTED BY Fields
A field in a record definition may be defined as a COMPUTED BY field by specifying a value expression to be
computed for its value. A COMPUTED BY field takes no
space in the actual RMS record, and is computed on
reference. A COMPUTED BY field may be used in conjunction with a table to provide completely automatic table lookup.

• Index Sort Index sort produces an address file containing the key field of each data record and a pointer to its
location in the input file. The index file can be used to
randomly access data from the original file in the desired sequence.

• Tutorial Software (Guide Mode)
A CRT-based tutorial is included in DATATRIEVE. The
tutorial feature can be used only by VT52, VT52-compatible, and VT100 terminals. A tutorial session is entered by the DATA TRI EVE command:

The MERGE utility permits the user to merge data from as
few as two, to as many as ten similarly sorted input files.
The MERGE utility merges the data according to key
field(s) defined by the user and generates a single output
file. The input files to be merged must be in sorted order,
i.e., the SORT and MERGE key fields must be the same.

SET GUIDE
The software is self-documenting.

9-14

The following example illustrates the sorting of a sales record file by customer last name. The name of the initial file
is SALES .DAT. Each record contains six fields: date of
sale, department code, salesperson , account number,
customer name , and amount of sale. The numerical
ranges listed below the set of records indicate the position
and size of each information field within the record .
DATE

DPT SALESP

ACCT

GUST- NAME

AMT

091580
091580
091580
091580
091580
091580
091580

25
25
25
19
28
25
19

980342
643881
753735
166392
272731
829582
980342

Cool idge Carol
McKee Michael
Rice Anne
Wilson Brent
Karsten Jane
Olsen Allen
Cool idge Carol

24999
2499
10875
1298
4000
3350
7200

Fielding
Sanchez
Bradley
Arndt
Meredith
Bradley
Erkkila

L._.}'-.--) 1...-..,-J
1-7

8-10

11-21

29-58

In this command sequence , the user is defining the SORT
key to be the customer's last name and the output file to be
BILLING.LIS
The user may now obtain a listing of the sorted data file by
using either the TYPE or PRINT commands.
$TYPE BILLING .LIS

GUST-NAME

AMT

091580
091580
091580
091580
091580
091580
091580

19
25
28
25
25
25
19

Erkkila
Fielding
Meredith
Sanchez
Bradley
Bradley
Arndt

980342
980342
272731
643881
829582
753735
166392

Cool idge Carol
Coolidge Carol
Karsten Jane
McKee Michael
Olsen Allen
Rice Anne
Wilson Brent

7200
24999
4000
2499
3350
10875
1298

358419
980342
669011
848105
561903
643881
454389

Beaulieu Ronald
Coolidge Carol
Fernandez Felicia
Kingsfield Stanley
Landsman Melissa
McKee Michael
VanDerling Julie

20
04
25
35
04
20
19

Oconnor
Doc us
Fielding
Leith
Kramer
Oconnor
Erkkila

1598
575500
12000
5550
230000
995
5480

$TYPE GENTA.FIL < er>

$SORT /KEY=(POSITION=29 ,SIZE=30) SALES .DAT
BILLING.LIS <er>

ACCT

091580
091580
091580
091580
091580
091580
091580

AMT

The user has indicated in the above command sequence
that the files are to merged via the alphabetical order of
the customer's last name. The user can examine the output file via the PRINT or TYPE commands.

The user may now rearrange the sales records in file
SALES.DAT according to any of the file 's information
fields. For instance, to sort the file in alphabetical order of
customer's last name, the user would type the following
command sequence:

DPT SALESP

GUST-NAME

$ MERGE/KEY=(POSITION=29,SIZE=30)
STORE1 .FIL,STORE2.FIL GENTA.FIL <er>

59-65

DATE

ACCT

To merge the two data files, the user must type the following command sequence:

1...-..,-J '----v----J L._.)
22-28

STORE2.FIL
DATE
DPT SAL ESP

DATE

DPT SAL ESP

ACCT

GUST-NAME

091580
091580
091580
091580
091580
091580
091580
091580
091580
091580
091580
091580
091580
091580

20
19
25
04
25
28
35
04
25
20
25
25
19
19

Oconnor
Erkkila
Fielding
Doc us
Fielding
Meredith
Leith
Kramer
Sanchez
Oconnor
Bradley
Bradley
Erkkila
Arndt

358419
980342
980342
980342
669011
272731
848105
561903
643881
643881
829582
753735
454389
166392

Beaulieu Ronald
Coolidge Carol
Coolidge Carol
Coolidge Carol
Fernandez Felicia
Karsten Jane
Kingsfield Stanley
Landsman Melissa
McKee Michael
McKee Michael
Olsen Allen
Rice Anne
VanDerling Julie
Wilson Brent

AMT
1598
7200
24999
575500
12000
.4000
5550
230000
2499
995
3350
10875
5480
1298

VAX-11 SORT/MERGE FEATURES
VAX-11 SORT can perform the following functions:
• reorder data files (records are sorted in ascending or
descending order by up to ten keys which can be in any
order)

To perform the MERGE function, the MERGE utility expects presorted data files upon which to operate. In the
following example , MERGE is operating upon two presorted (by alphabetical order) sales data files , STORE1 .FIL
and STORE2.FIL.

• merge up to ten sorted input files into one sorted output
file
• create reordered address files of RFAs and keys for
software use
• SORT/MERGE fixed , variable, and VFC records

STORE1.FIL
DATE
DPT SAL ESP
091580
091580
091580
091580
091580
091580
091580

19
25
28
25
25
25
19

ACCT

Erkkila
980342
Fielding
980342
Meredith 272731
Sanchez 643881
829582
Bradley
753735
Bradley
Arndt
166392

GUST-NAME

AMT

Coolidge Carol
Coolidge Carol
Karsten Jane
McKee Michael
Olsen Allen
Rice Anne
Wilson Brent

7200
24999
4000
2499
3350
10875
1298

• SORT /MERGE ASCII character keys in ASCII or
EBCDIC sequence
• SORT/MERGE sequential, relative, indexed-sequential
files
• SORT /MERGE character, decimal, binary, unsigned binary, F_, D_, G_, and H_floating data types
• SORT can determine its own work file requirements
based on input file RMS information received
9-15

• SORT can be controlled by a command string or specification file
• SORT can be tuned for maximum efficiency
• SORT provides four processing techinques: record, tag ,
address, index
• SORT /MERGE input files from any VAX/VMS input device

Programming Considerations
Any program can use either SORT /MERGE subroutine interface with any of the VAX-11 native mode languages.

SORVMERGEPERFORMANCEFEATURES

• Output sorted data files to any VAX/VMS output device

• SORT /MERGE compiles a key comparison routine specific to each sort or merge. This results in a substantial
reduction of CPU usage.

• SORT automatically prints out statistics upon completion

• Work files are not created until they are needed . Th is reduces overhead when sorting small files .

• Be invoked by a single command string, or can prompt
the operator for input and then output file specification

• The internal record size has been reduced , therefore ,
less 1/0 is required to do intermediate merge passes for
SORT.

• Respond with unique SORT/MERGE error messages in
VAX/VMS format
• Optional sequence checking of input files on merge
VAX-11 SORT /MERGE supports the following key formats:

• For some sorts, the number of intermediate merge
passes has been reduced , thereby providing a substan tial increase in speed of the SORT.

• character data are ASCII

VAX-11 FORMS MANAGEMENT SYSTEM (FMS)

• ASCII and EBCDIC collating sequence

VAX-11 Forms Management System (FMS) is a utility
package used to provide video form support for applications on the VT100 video terminal. VAX-11 FMS provides a
flexible , easy-to-use interface between the form appl ication program and the terminal user. FMS allows a term inal
user to develop form applications using any native mode
language processor .

• binary data are VAX representation
• packed decimal data are VAX representation
• zoned decimal data are VAX representation
• unsigned binary and F_, D_, G_,and H_floating
• string decimal data format can be:
leading separate sign
leading overpunched sign
trailing separate sign
trailing overpunched sign

Using Forms in an Application
VAX-11 FMS forms include a video screen image comprising data fields and constant background text, along with
protection and validation attributes for individual data
fields . The data fields and background text can be highlighted using any combination of the VT100 video attributes: reverse video, underline, blink, and bold characters .
Split screen and scrolling capabilities allow the user to
view more data than can be displayed on the screen at one
time.

SORT/MERGE as a Set of Callable Subroutines
SORT /MERGE can be used as a set of callable
subroutines from any native VAX language. This subroutine package provides two functional interfaces to choose
from: a file 1/0 interface and a record 1/0 interface. Both
interfaces share the same set of routines, and the same
calls are used for all languages.
SORT and MERGE subroutines are callable from VAX-11
COBOL using the standard COBOL SORT and MERGE
verbs.

Individual data fields can be display-only, enter-only (no
echo), or can be restricted to modification by privileged
users. Data fields can be formatted with fill characters, default values, and other formatting characters-such as the
dash in a phone number-which assist the terminal operator, but which are not visible to the application program.
Fields may be right- or left-justified or may use a special
fixed decimal data field type to normalize floating point decimal numbers into fixed point for easier use in computation .

For either interface, the user can supply a key comparison
routine. This feature allows the user the flexibility of departing from the key types supported by SORT /MERGE.
With this release of VAX/VMS, full MERGE capability has
entered the list of SORT features callable as a subroutine.
This allows a much increased file flexibility.
File 1/0 Interface
The file 1/0 interface allows the user to specify the input
files and an output file to SORT or MERGE. SORT then
reads the data from the input file(s) and sorts the data into
the output file. MERGE also reads the data form the input
file(s) and merges it into one output file .

Field validation includes checking each keystroke in a field
for the proper data type (e.g., alphabetic, numeric, etc.).
Fields may also be defined as "must enter" or " must
complete. "

Record 1/0 Interface
The record 1/0 interface allows the user to pass each individual data record to SORT/MERGE, let SORT/MERGE
order them and then receive each record back in the correct order, individually, from SORT /MERGE.

A line of HELP information may be associated with each
field , and a chain of one or more HELP forms may be associated with each form. If people need additional instructions while using a form, they press the HELP key to display the HELP line for the current field. A second HELP

9-16

keystroke displays the first HELP screen for the current
form , so that from any point in the application form the
user can get to an entire series of HELP forms. In this way
the entire user manual for an application can be put online, automatically keyed to the appropriate user form.
Almost any class of application can benefit from using
VAX-11
FMS . Source data entry and inquiry/
response/update are the most obvious types of forms-oriented uses, but other types of programs can benefit
equally well. For example, a simulation or numerical analysis program could use FMS forms to exp lain and accept
input parameters , and then to format and scroll through
the output of the run. Forms might constitute the front end
of a student registration or a computer-aided instruction
system . Alternatively, they could be used to review data
acquired from laboratory or factory instrumentation , or to
format the operator input of control parameters to such
processes. Almost any application which uses alphanumeric video terminals can be enhanced by using FMS
forms to talk to the terminal user.

Developing Applications with VAX-11 FMS
VAX-11 FMS forms are created and modified interactively
on the screen using a special FMS utility called the Form
Editor (FED). Because the video image is typed and manipulated directly on the screen , there is no need to lay out a
form on a paper chart or to code complex specifications
into a form definition program written in a difficult language. Rather , the form creator always sees the form on
the screen exactly as it will appear to the application user.
A set of 24 ed iting and data manipulation functions invoked through the function keypad of the VT100 terminal
allow easy alteration of the form description .

also generate COBOL Data Division code to correspond to
the form description .
Once the form has been stored in a form library, one or
more application programs to use it must be coded . These
application programs control the interaction between
themselves , the form , and the operator by making calls to
a library of VAX-11 FMS subroutines called the Form Driver (FD'/). Under the direction of the calling application
program, the Form Driver displays forms, performs all
screen management an application requires, handles all
terminal input and output, and validates each operator entry by checking it against the field description for the field .
A broad selection of subroutine calls allows the program to
communicate with the screen on either a full-screen or
field-by-field basis. While the terminal operator is typing
data, all data validation and formatting, error messages,
and HELP requests occur completely transparently to the
application program.
Maintaining VAX-11 FMS Applications
Several features of VAX-11 FMS make maintenance of applications using forms both rapid and reliable . The most
obvious such feature is the interactive editing capability of
the Form Editor. Capabilities such as Open Line for insert,
Cut and Paste, etc., make modifications to existing forms
quick and easy . Furthermore , when fields are moved
around on the screen , their attributes are preserved, so
that re-entry of the form is not required .
Perhaps the most important contribution to application
maintainability comes from the fact that the application
makes all references to screen data by field name. These
field name references are not resolved until the program
executes, so that the form description is actually independent of the appl ication program . This means that it is easy
to write programs that do not know or depend upon the
specific order of the fields , or even know the names or
each field . This capability, combined with the storage of
forms in libraries, means that in many cases the form can
be rearranged , or new fields added, without requiring that
the application program be recompiled, or even relinked!

Fields are defined interactively via the function keypad and
by typing the COBOL-like picture character for each position of the field directly on the screen. The remaining field
attributes, such as field name, default value, and the contents of the HELP line, are described by interactively filling
in a questionnaire form on the screen . Another form is
used to define certain characteristics which apply to the
form as a whole, such as the name of the form and of the
first HELP form associated with it, and whether the screen
is to be placed into reverse video or 132-column mode. A
third form is used to specify application constants , called
" named data," which can be stored with the form instead
of hard-coded into the appl ication program . This last feature allows appl ication parameters such as small data tables , file names, names of subsequent forms , etc. , to be
stored with the form and edited almost interactively.

The last feature of VAX-11 FMS that promotes application
maintainability is the ability to store application parameters with forms as named data. 1-'arameters such as the
names of related forms or programs, file specifications,
small look-up tables, range check boundaries, and error
messages specific to the particular form may be stored
with the form and ed ited with the same rapidity and ease.
For example , imagine an application that will be used by
operators who speak a variety of different languages. The
first thing the operator would do when logging into the application would be to select a language. The program
would simply open the form library with all the forms translated into that language. Named data would be used to
store all other language-dependent application parameters, such as program-generated error messages, abbreviations (Y=YES or O=OUI or S=SI), etc. The same program code would then be executed regardless of the language the application would "speak." A new language
could be added by simply modifying the initial language
selection form and creating a form library with the forms
and named data translated into the new language.

When the application developer is sat isfied with the
appearance and content of the form , the Form Editor
writes the form out into a work file . The Form Utility (FUT)
is then used to insert the form into a new or existing form
library, from which it will be retrieved when the application
program is executed. The Form Utility may be used to perform other maintenance functions on form libraries, as
well as to generate hard-copy descriptions of forms suitable for inclusion in application documentation. FUT can

9-17

10
Data
Communications
Facilities

DECnet is a family of _network products developed by Digital Equipment
Corporation that adds networking capability to DIGIT Al's computer
families and operating systems. Using DECnet, various DIGIT AL computer systems can be linked together to facilitate remote communications, resource sharing, and distributed computation. DECnet is highly
modular and flexible. It can be viewed as a set of tools or services from
which a user selects those appropriate to build a network to satisfy the
requirements of a particular application.
DIGITAL Network Architecture (DNA) provides the common network
architecture upon which all DECnet products are built. The architecture
is designed to handle a broad range of application requirements because all the functions of the network from the user interface to physical link control are completely modular. DNA allows nodes to operate
as switches, front-ends, terminal concentrators, or hosts.
DECnet-VAX:
• provides an interprocess communication facility that is highly transparent and easy to use
• provides a higher-level language programming interface
• allows programs to access files at other systems
• allows users and programs to transfer files between systems
• allows users to transmit command files to be executed on other systems
•allows an operator to down-line load RSX-11S system images into
other systems
VAX/VMS also supports protocol emulators (lnternets), which enable
DIGITAL systems to communicate with other vendors' systems.

INTRODUCTION

• Resource-Sharing Networks. These networks exist to
permit sharing expensive computer resources among
several computer systems. Shared resources not only
include peripherals such as mass storage devices, but
they can also include logical entities, such as a centralized data base wh ich is made available to other systems
in the network. Such networks are often characterized
by the concentration of high-performance peripherals,
extensive data bases, and large programs on one or two
host systems in the network , while the satellite systems
have less expensive peripherals and smaller programs.
Figure 10-2 illustrates a resource-sharing network .

DIGITAL computers can commun icate with other DIGITAL
computers either remotely or locally via a network . By utilizing protocol emulators (lnternets), they can communicate with computers from other suppl iers.
DECnet is the family of products that allow DIGIT AL systems to participate in a cooperative multiprocessing environment known as a network. A network is a configuration
of two or more independent computer systems, called
nodes, linked together to facilitate remote communications, share resources, and perform distributed processing. Network nodes are not all required to run on the same
type of operating system . Within the scope of a single network , several nodes with different operating systems and
different features can interact to provide increased processing flexibility.

PDP-1 i

Adjacent network nodes are linked together via carriers
known as physical links. Physical links can be relatively
permanent bonds , such as telephone lines or cable wires
laid from one node to another, or they can be temporary
connections that change with each use, such as dialed-up
telephone calls.

PDP - 11

t---

In a network of DECnet nodes, several tasks can use the
same physical link to exchange data. That is, more than
one data path can be handled simultaneously by a physical link . This data path is known as a log ical link . A task is
an image running in the context of a process.

--t

VA X - 111780 1 - - - - - - - - l

PD P-11

PDP -1 l

With DECnet, a variety of computer networks can be implemented. They typically fall into one of three classes:

Figure 10-2

• Commun ications Networks. These networks exist to
move data from one, often distant, physical location to
another. The data may be file-oriented (as is often the
case for remote job entry systems) or record-oriented
(as occurs with the concentration of interactive terminal
data). Interfaces to common carriers, using both
switched and leased-line facil ities , are normally a part of
such networks. Such networks are often characterized
by the concentration of all user applications programs
and data bases on one or two large host systems in the
network. Figure 10-1 illustrates such a network .

Resource-Sharing Network

• Distributed Computing Networks. These networks coordinate the activities of several independent computing
systems and exchange data among them. Networks of
this nature may have specific geometries (star, ring,
hierarchy), but often have no regular arrangement of
links and nodes. Such networks are usually configured
so that the resources of a system are close to the users
of those resources . Distributed computing networks are
usually characterized by multiple computers with applications programs and data bases distributed throughout the network. Figures 10-3 and 10-4 illustrate two
such networks.

PLANT INTERFACE

Figure 10-1

Figure 10-3

Communications Network

Typical Manufacturing Network

10-1

,----------------

COMPUTAT IONAL SERV ICE BUREAU
O R IN HOUSE DATA CENTER
VAX-111780

PDP- 11

' ,)
'

User Layer
This layer contains all user and DIGITAL supplied functions . Modules in this layer include network remote file
access modules, a remote file transfer utility, and a remote
system loader module. The protocol used for remote file
access and file transfer is the Data Access Protocol (OAP).

~PLon
==
ER~!

:
:
I

GRAPH ICS TERMINALS
IG T - 4 0)

Network Service Layer
This layer provides a location-independent commun ication mechanism for the user layer. The means by which
they communicate is called a loglcal link. Two network
application modules may communicate with each other by
means of the network service layer regardless of their
network locations. The protocol used between network
service modules is the Network Services Protocol (NSP).

L _____________ - - - - - - - - - ___ _J
ENGI N EER ING FIRM

Figure 10-4
Computational Network

Data Link Layer
This layer provides the transport layer with an error-free
communication mechanism between adjacent nodes. The
data link module specified for this layer implements the
DIGITAL Data Communications Message Protoco l
(DDCMP). The functions provided by this layer are independent of communication facility characteristics . For
DECnet-VAX, DDCMP is incorporated into the microprocessor of the communications interface.

When DECnet is used to connect heterogeneous systems,
each node of the network has both common DECnet attributes and system-specific attributes. Programs executing in native mode can access the following network facilities:
• Interprocess
(Task-to-Task)
Communication : Programs executing on one system can exchange data with
programs executing on other systems.
• lntersystem File Transfer: A program or user can
transfer an entire data file from one system to another.
• lntersystem Resource Sharing : Programs executing on
one system can access files and devices physically located at other systems in the network. Access to devices
in other systems is provided through the file system of
the target node and is subject to that node's file system
restrictions.

Physical Link Layer
This layer, the lowest layer in the DNA structure, provides
the data link layer with a communication mechanism
between adjacent nodes. Several modules are specified
for this layer, one for each type of communication device
that can be used in a DECnet network.

• Network Command Terminal: A terminal on one VAX
system can appear to be connected to another VAX system in the network.

DNA is system-independent. It enables a variety of
DIGITAL computers running a variety of DIGITAL operating systems to be tied together in a DECnet network .

• Down-Line System Loading: Initial load images for
RSX-11 S systems in the network can be stored on the
host VAX system , and be loaded into adjacent PDP-11
systems configured for the RSX-11 S operating system.

A DECnet network can grow both in size and in the number
of functions it provides. It can , therefore, be adapted to
new technological developments in both hardware and
software. Existing DECnet implementations can take advantage of these new technologies . DECnet components
can be replaced if better communications hardware becomes available or if technical innovations in networking
occur. A DECnet network can accommodate the change of
a function from software into hardware.

• Down-Line Command File Loading: Command language users can send command files to a remote node
to be executed there. However, no status information or
error messages are returned.

DIGITAL NETWORK ARCHITECTURE
DECNE~VAXFEATURES

The DIGIT AL Network Architecture (DNA) is a set of protocols (rules) governing the format, control, and sequencing
of message exchange for all DECnet implementations.
DNA controls all data that travel throughout a DECnet network and provides a modular design for DECnet.

The capabilities offered the DECnet-VAX programmer and
terminal user extend through a wide range of network
functions.
File Handling Using a Terminal
By using DECnet-VAX DIGITAL Command Language
(DCL) commands, the user can copy files from one node to
another, delete files stored on a remote node, take directories of files on remote nodes, and transfer a command file
to another node and then execute the command file on the
remote node.

Its functional components are defined within four distinct
layers: User Layer, Network Service Layer, Data Link Layer, and Physical Link Layer. Each layer performs a welldefined set of network functions (via network protocols)
and presents a level of abstraction and capability to the
layer above it.

10-2

File Handling Using Record Management Services
A wide range of VAX/VMS Record Management Services
(RMS) can be used to handle files and records stored on
remote nodes. At the file level, these operations include
opening, closing , creating, deleting , and updating files
stored on a remote node. Also, at the record level , RMS
can be used to read, write, update, and delete records
stored on a remote node.

The process can send optional data along with the connect
request; for example, the size or number of messages that
it wants to send . The receiving process or task can accept
or reject the connect initiation. A process can accept multiple connect requests.
A process can send or receive mailbox messages to or
from another process or task . Mailbox message traffic is
essentially no different from data message traffic except
that it uses a mailbox associated with the 1/0 channel over
which the logical link was created. (This is the same mechanism used, for example, for telling programs that unsolicited terminal data are available.) A logical link, therefore, has two subchannels over which messages can be
transmitted : one for normal messages and another for
high-priority messages. In DECnet-VAX, an interrupt message is written to a mailbox that a process supplies for that
purpose.

Network Command Terminal
The network command terminal facility allows a local terminal to operate as if it were physically connected to a remote computer .

Intertask Communications
Any native-mode language programmer can write programs that perform intertask (interprocess) communication. Intertask communication is a method of creating a
logical link between two tasks, exchanging data between
the tasks, and disconnecting the link when the communication is complete.

In DECnet-VAX, a program using nontransparent access
normally opens a control path directly to a Network Ancillary Control Process (NET ACP), and designates one or
more mailboxes for receiving information from the NET ACP about the logical or physical links over which the
process is communicating . The NETACP can notify a process when:

Intertask communication routines can be coded using one
of two methods, transparent or nontransparent.

• a partner requests a synchronous disconnect
• a partner requests a disconnect abort

Transparent Intertask Communication - The program
opens the network interchange as if it were preparing for
device access, and then performs a series of reads and
writes just as it would to a pair of serial devices, one for
input and the other for output.

• a partner exits

By its very nature, transparent access has no calls specifically associated with DECnet. The calls used for interprocess communication are the same as the calls used for accessing a sequential file in a higher-level language: OPEN,
CLOSE, READ , WRITE, etc. The programmer can choose
to include the target node name in the OPEN statement, or
can defer assignment using logical names.

DIGITAL COMMAND LANGUAGE (DCL)
FILE HANDLING

• a physical link goes down
• an NSP protocol error is detected

A VAX/VMS DCL user can transfer files from one node to
another and delete files at other nodes. However, to perform operations on files stored on a remote node, the user
must prefix the file specification with the remote node's
name, and an optional access control string as follows :
nodename"access control string"::filename.filetype;version
where:
A 1- to 6-character name (numerics
nodename =
or uppercase alphabetics) identifying
the remote network node.

Nontransparent Intertask Communication - In nontransparent access, a program can obtain information
about the network status to control the nature of its com munication with other processes or tasks . This method of
coding intertask communications is available to the
MACRO programmer. And if you don't do AST processing
or attempt to accept multiple connects, you may program
in any language. Nontransparent access is available only
through calls to operating system service procedures. A
program can issue the following requests :

access control
string=

Typically, a "username password." If
omitted, default login information
comes from an entry located in the local configuration data base.
The double colon(::) following the nodename separates the nodename
from the file specifier.

• CONNECT-establish a logical link (the analog of
OPEN)
• CONNECT REJECT -reject a connect initiation

filename=
filetype
version

• RECEIVE-receive a message (the analog of GET or
READ)
• SEND-transmit a message (the analog of PUT or
WRITE)
• SEND INTERRUPT MESSAGE-transmit a high-priority
message
• DISCONNECT -terminate a conversation (the analog of
CLOSE)

Use the following format for a DECnet-VAX node:
device:[directory]filename.filetype;
version

DECnet-VAX supports a subset of VAX/VMS (DCL) commands. They are:
APPEND
ASSIGN
COPY
10-3

agree to accept the request. The task requesting the logical link is called the source task; the one agreeing to accept the logical link request is called the target task .

DEASSIGN
DEFINE
DELETE
DIRECTORY
SUBMIT
TYPE

Sending and Receiving Messages
After the logical link has been created, the tasks must
"cooperate" with each other. That is, for each message
sent by a task (WRITE statement), the receiving task must
issue a corresponding receive (READ statement).

The following examples illustrate the COPY and SUBMIT
commands:
$COPY BOSTON: :DBA1:TEST.DAT DENVER::DMA2:

In addition, the tasks must ensure that enough buffer
space is allocated for messages, must ensure that the end
of dialogue can be determined, and must determine which
of the tasks will disconnect the logical link (CLOSE statement).

transfers a file named TEST.DAT from the disk (DBA 1:) at
the node named BOSTON to the disk (DMA2:) at the node
named DENVER.
Using the VAX/VMS command SUBMIT, a terminal user
can have a command file executed at another node in the
network. For example, the command:

Disconnecting the Logical Link
Either task can disconnect the logical link by calling
CLOSE. CLOSE aborts all pending sends and receives ,
disconnects the link immediately, and frees the channel
number associated with the logical link.

$SUBMIT/REMOTE WASHDC: :INITIAL.COM
preceded by a DCL COPY command will transfer the command file named INITIAL.COM from the host system to the
node named WASH DC, where the command file is executed. The SUBMIT command assumes that the file already
exists at the remote node. Command files must be written
in the command language of the system. No status information or messages are returned to the sender.

VAX-11 FORTRAN Intertask Communication Example
Figure 10-5 illustrates intertask communications using
normal VAX-11 FORTRAN 1/0 instructions.

RECORD MANAGEMENT SERVICES
FILE HANDLING

MACRO TRANSPARENT INTERTASK
COMMUNICATION

By using a subset of the VAX-11 Record Management Services (RMS), the user can manipulate records and files
stored on remote DECnet nodes. However, before using
VAX-11 RMS to perform operations on files and records
stored on a remote node, the user must prefix the file
specification with the node name of the remote node and
an optional access control string, just as with any other remote file application.

This section describes the fundamentals of coding a
MACRO program for transparent intertask commun ications utilizing a subset of the existing macro calls available
under VAX/VMS system services. These macro calls allow
the user to perform intertask communications in much the
same way as normal 1/0 operations are performed .

Much of the VAX-11 RMS functionality is supported by
DECnet-VAX, including managing sequential, relative, and
indexed file organizations. A large number of the VAX-11
RMS macros are available to network users.

Thus, communication with another task is performed in
much the same way as an 1/0 channel is assigned to a device ($ASSIGN). Reads and writes are then performed as if
to a pair of sequential devices (that is, $010 with the
10$_WRITEVBLK function or $OUTPUT, and $010 with the
10$_READVBLK
function or $INPUT). Finally ,
$DASSGN the device when communication is complete.

The term "transparent" simply implies that the calls are
identical in format to all other 1/0 calls.

SAMPLE VAX-11 FORTRAN INTERTASK
COMMUNICATION

Three major functions in transparent intertask commun ication are:

This section describes the basic communication protocol
involving VAX-11 FORTRAN intertask communications .
The user communicates with another task in much the
same way as an access to a sequential file, i.e., via OPEN ,
READ, WRITE and CLOSE statements. Similar capabilities
exist in any of the native mode languages.

1.
2.
3.

Three major steps in VAX-11 FORTRAN intertask communication are:
1.
2.
3.

Create a logical link between tasks.
Send and receive messages (each message can be O
to 65,535 bytes long).
Delete the link at the end of the message dialogue.

Creating a Logical Link Between Tasks
A logical link between tasks can be created only if the
tasks agree to cooperate with each other. That is, one task
must request that a logical link be created, and the other
task must agree to accept the request.

Creating a logical link between tasks.
Sending and receiving messages (each message can
be 1to512 bytes in length).
Destroying the link at the end of the message dialogue.

A logical link is requested by including a task specifier in
the source task's $ASSIGN call. A task specifier identifies
the remote node and the target task to which it will be connected .

Creating a Logical Link Between Tasks
A logical link between tasks can be created only if they
agree to cooperate, with each other. That is, one task must
request that a logical link be created, and the other must

The target task identifies the source task requesting the
logical link connect request by specifying SYS$NET as the
10-4

Source Task Code
PROGRAM DEM02.FOR

C
C
C
C
C
C

This program prompts the user for a request, communicates
with a remote task to obtain the requested data, and displays
the answer for the user. The remote task is referenced by
the logical name TASK . If the remote task is named DEM03.EXE
at node TULSA , the following procedure is used to run the
two programs:

C
C

$DEFINE TASK TULSA::""TASK=DEM03""
$RUN DEM02

c
c
c

100
200
300
400

c
c
c
c

LOGICAL *1 CODE(4),BUFFER(20)
FORMAT
FORMAT
FORMAT
FORMAT

('$Enter request code: ',4A 1)
(4A1)
(Q,20A1)
('OStock number for code ',4A 1,'is: ', 20A 1)

Request the remote task to be run and establish a logical
link into it.
OPEN

(UNIT= 1,NAM E='TASK,'ACCESS= 'SEQUENTIAL',FORM = 'FORMATTED')

c
C
C
C
C

Prompt the user for a request code, send the code to the
remote task, read the reply from the remote task, and display it to
the user.
Repeat the cycle until the user enters 'Exit' as his request code.

ACCEPT
100,CODE
IF(CODE(1) .EQ.' E') GOTO 20
WRITE
(1 ,200END=20)CODE
READ
(1,300) NCHAR,(BUFFER(K),K= 1,NCHAR)
TYPE
400,CODE,(BUFFER(K),K= 1,NCHAR
GOTO
10

c
c
c
20

Finished .
(UNIT=1)

CLOSE
END

Target Task Code
PROGRAM DEM03.FOR

c
C
C
C
C
C

This is the companion task for DEM02. For each request it
receives from the remote task it replies with a 1- to 20character response. This program does not know the name of the
requesting task . To complete the logical link with its initiator, it
opens the 'file ' specified by the logical name SYS$NET.

c
LOGICAL *1 CODE(4),BUFFER(20)

c
100
200

FORMAT
FORMAT

(4A1)
(20A1)

c
Figure 10-5
VAX-11 FORTRAN Intertask Communications
10-5

Establish a communication path with the remote task .

c
OPEN

(UNIT=1,NAME='SYS$NET,:'ACCESS='SEOUENTIAL' ,FORM='FORMATTED')

c
c
c

Process requests until end-of-file encountered .

READ

c
c
c
c

(100,END=20)CODE

Perform appropriate processing to obtain result to
transmit back to the requesting task .
WRITE
GOTO

(1,200) (BUFFER(K),K= 1,NCHAR)
10

c
C

Finished.

c
20

CLOSE
END

(UNIT=1)

Figure 10-5 (Con 't)
VAX-11 FORTRAN Intertask
Communications

two or more tasks interacting to establish a logical link . After establishing the logical link, the tasks exchange messages over the link , then disconnect the link when communication is complete .

devnam argument in the $ASSIGN statement. This action
completes the creation of the logical link.

Sending and Receiving Messages
After the logical link is created, the tasks must "cooperate"
with each other. That is, for each message sent by a task
($010 with the 10$_WRITEVBLK function or $OUTPUT),
the receiving task must issue a corresponding receive
($010 with the 10$_READVBLK function or $INPUT).

The MACRO system service calls discussed in this section
provide the user with greater flexibility and control over
network operations. The following features can be used
when performing nontransparent intertask communication:

In addition, the tasks must ensure that enough buffer
space is allocated for messages, must ensure that the end
of dialogue can be determined, and must decide which of
the tasks will disconnect the logical link ($DASSGN).

• Associate a mailbox with the 1/0 channel (over which
the logical link will be created) . The mailbox can then receive mailbox messages sent by a remote task , or notifications affecting the status of the logical link . For example, status returned through a mailbox includes whether
the remote task accepted or rejected a connect , or the
cooperating task disconnected or destroyed the link.

Disconnecting the Logical Link
Either task can disconnect the logical link by calling
$DASSGN. $DASSGN aborts all pending sends and receives, disconnects the link immediately, and frees the
channel number associated with the logical link.

• A task can declare itself as a network task to accept
multiple logical link connect requests.
• A source task can send a logical link connect request to
the target task. The source task can optionally send up
to 16 bytes of data to the target task at the same time it
issues the connect request.

MACRO CALLS
Listed below are the VAX/VMS system service macro calls
that can be used for transparent intertask communications.

• The target task can accept or reject the connect request. It can send up to 16 bytes of optional data back to
the source task at the same time it accepts or rejects the
connect request.

• $ASSIGN-Assign 1/0 Channel
• $010-Send a Message to a Remote Task $010
(10$ _WRITEVBLK)
• $010-Receive a Message from a Remote Task $010
(10$ _READVBLK)

• A task using the nontransparent interface can also accept or reject connect requests received from tasks
written using transparent intertask communication system service calls.

• $INPUT-Read a Message
• $OUTPUT -Write a Message
• $DASSGN-Disconnect the Logical Link

• Either task can send or receive a 1- to 16-byte interrupt
message after the logical link is created.

MACRO NONTRANSPARENT INTERTASK
COMMUNICATION

• Either task can abort the link immediately, or issue a
synchronous disconnect. The task disconnecting or
aborting the logical link can send up to 16 bytes of op-

Nontransparent intertask communication may consist of
10-6

tional data to the remote task at the same time it disconnects or aborts a logical link.

tending existing mainframe facilities to include powerful
minicomputer data processing .

Task Messages
There are two types of messages in nontransparent intertask communications : data messages and mailbox messages .

VAX-11 2780/3780 Protocol Emulator
This product provides the VAX/VMS user with a mechanism for transferring files between the VAX system and
another system equipped to handle IBM 2780 or 3780
communications protocols. It does this by emulating the
synchronous line protocol used by a 2780 or 3780 Remote
Batch Terminal.

Data Messages - A data message is a message sent by
one task, and expected by the cooperating task. That is,
for each message sent by a task $010 with the
10$_WRITEBLK function or $OUTPUT, the receiving task
must issue a receive $010 with the 10$ READVBLK function or $INPUT.
-

The emulator may be invoked either interactively or by a
command procedure. The emulator's command set is designed to facil itate sharing a communication line among
several users. With the appropriate modem options , the
emulator is capable of automatically answering incom ing
calls.

Thus, a data message in nontransparent intertask communications is the same as a data message sent in transparent communication .

Sophisticated operations can be performed by a combination of command procedures, allowing, for example, unattended operation . Th is would include the capability to detect an incoming call , establish the connection, and then
transmit and receive files and recover from transmission
failures, all without the intervention of the operator.

Mailbox Messages - All other messages received by a
task employing a nontransparent interface are classified
as mailbox messages. These include any one of the following message types :
1.

A logical link connect request-this message is received by the target task. It requests a logical link connection to the source task .
2. A connect accept-this message is received by the
source task. The message confirms that the target
task accepted the logical link connect request.
3. A connect reject-this message is also received by the
source task . The message informs the source task
that the target rejected the logical link connect request.
4. An interrupt message-either task can receive a 1- to
16-byte interrupt message sent by a cooperating task .
The 1 to 16 bytes of data are placed in the task 's mailbox.
5. A synchronous disconnect-this message informs the
task that the cooperating task synchronously disconnected the logical link.
6. A disconnect abort-this message informs the task
that the cooperating task aborted the link. The link is
destroyed immediately.
7. A network status message-this message informs the
task of some unusual network occurrence, for example, the data link has been restarted .
After a logical link is created between cooperating tasks,
DECnet places a received mailbox message into the mailbox associated with the channel representing the logical
link to which the mailbox message applies.

Several data formats are supported with the use of a particular format selected by user command . Users may
select various forms to control translation schemes (records can be padded with spaces to card images before
transmission), translation to and from EBCDIC, and BSC
transparency . All file 1/0 is performed through the
VAX/VMS record management facility . Print and punch
stream recogn ition is implemented in such a way that the
data manipulation scheme can differ with each stream .
The following remote batch terminal features are supported :
• 2780 Extended and Multiple Record Option
• Variable Horizontal Forms Control
• BSC Transparency
• 3780 Space Compression
All of the above features are supported on a simultaneous,
multiline basis. The product can concurrently run up to
four physical lines, each with a different set of attributes
(e.g ., some may be 2780, others, 3780) at speeds up to
9600 baud per line.
MUX200/V AX Multiterminal Emulator
MUX200/VAX is a VAX-based software package wh ich
provides communication with a CDC6000, CYBER series,
or other host computer systems capable of using 200 UT
mode 4A commun ications protocols.

In the case of a task that can accept multiple inbound connect requests, inbound connect requests are placed into
the mailbox associated with the 1/0 channel over which the
network name was declared .

Any VAX interactive terminal may be used to control remote job entry or to communicate at command level with
the host system . Input files may be sent from , and output
files received onto , any VAX-supported mass storage, unit
record , or term inal device.

Note that the mailbox was previously created using the
$CR EM BX system service call. The task must then explicitly retrieve the unsolicited message from the mailbox using
the $010(10$ _READVBLK) system service call.

MUX200/VAX communicates with the host using the Mode
4A communications protocol as defined in CDC publication 82128000. The software package can be configured to
support either the ASCII or the external BCD versions of
the protocol.

PROTOCOL EMULATORS (INTERNETS)
VAX/VMS supports a number of software emulators that
enable and promote coexistence between the VAX family
and products supplied by other vendors . In this way, VAX
computers become even more flexible, particularly in ex-

MUX200/VAX provides for one synchronous communication circuit to a host computer system . The product sup10-7

ports a single switched or dedicated leased line two- or
four-wire common carrier facility at speeds up to 9600
baud .
MUX200/VAX enables several users to communicate simultaneously with a host system over a single line. The
VAX/VMS system, though using a single physical drop,
appears to the host as a number of multidrops and terminals on the circuit.
M UX200/V AX features include:
• Output received from the host system may be spooled
to the line printer upon detection of a text string predefined by the user.

• Up to eight VAX/VMS files may be specif i ed fo r
transmission to the host in a single command .
• VAX/VMS terminals may be detached for other use
while the software package is operating. Data received
from the host directed to a terminal are saved for pri nting until the terminal is reattached .
• In many applications the host system can be offloaded
by taking advantage of the local processing power of the
VAX/VMS system. This reduces host processing and
line costs; for example, file editing can be performed locally rather than on the host.
Figure 10-6 illustrates a schematic of the MUX200.

CDC 6000
OR
CY BER
HOST SYSTEM

VAX-111780 - - - - - - - -

CARD
READER

PRINTER

UP TO 16 TERMINALS

Figure 10-6
MUX200 Schematic

10-8

APPENDIX A

COMMONLY USED MNEMONICS

ACP
ANS
ASCII
AST
ASTLVL
CCB
CM
CRB
CRC
OAP
DOB
DDCMP
DDT
DV
ECB
ECC
ESP
ESR
F11ACP
FAB
FCA
FCB
FCS
FDT
FP
FPO
FU
GSD
GST
IDB
IPL
IRP
ISECT
ISO
ISP
IS

ISR
IV
KSP
MBA
MBZ
MCR
MFD
MFPR
MME

MTPR
MUTEX
NSP
OPCOM
POBR
POLR
POPT
P1BR
P1LR
P1PT
PC
PCB
PCBB
PFN
PIO
PME
PSECT
PSL
PSW
PTE
QIO
RAB
RFA
RMS
RWED
SBI
SBR

Ancillary Control Process
American National Standard
American Standard Code for Information Interchange
Asynchronous System Trap
Asynchronous System Trap level
Channel Control Block
Compatibility Mode bit in the hardware PSL
Channel Request Block
Cyclic Redundancy Check
Data Access Protocol
Device Data Block
DIGITAL Data Communications Message Protocol
Driver Data Table
Decimal Overflow trap enable bit in the PSW
Exit Control Block
Error Correction Code
Executive Mode Stack Pointer
Exception Service Routine
Files-11 Ancillary Control Process
File Access Block
Fixed Control Area
File Control Block
File Control Services
Function Decision Table
Frame pointer
First Part (of an instruction) Done
Floating Underflow trap enable bit in the PSW

see
SCBB
SLR
SP
SPT
SSP
SVA
TP
UBA
UCB
UETP
UFO
UIC
USP
VCB
VPN

Global Section Descriptor
Global Symbol Table
Interrupt Dispatch Block
Interrupt Priority Level
1/0 Request Packet
Image Section
Image Section Descriptor
Interrupt Stack Pointer
Interrupt Stack bit in PSL
Interrupt Service Routine
Integer Overflow trap enable bit in the PSW
Kernel Mode Stack Pointer
MASSBUS Adapter
Must Be Zero
Monitor Console Routine
Master File Directory
Move From Process Register instruction

wee
wcs
woes

Memory Mapping Enable
A-1

Move To Process Register instruction
Mutual Exclusion semaphore
Network Services Protocol
Operator Communication Manager
Program region base register
Program region length register
Program region page table
Control region base register
Control region limit register
Control region page table
Program Counter
Process Control Block
Process Control Block Base register
Page Frame Number
Process Identification Number
Performance Monitor Enable bit in PCB
Program Section
Processor Status Longword
Processor Status Word
Page Table Entry
Queue Input/Output Request system service
Record Access Block
Record 's File Address
Record Management Services
Read , Write, Execute, Delete
Synchronous Backplane Interconnect
System Base Register
System Control Block
System Control Block Base register
System Length Register
Stack Pointer
System Page Table
Supervisor Mode Stack Pointer
System virtual address
Trace trap Pending bit in PSL
UNIBUS Adapter
Unit Control Block
User Environment Test Package
User File Directory
User Identification Code
User Mode Stack Pointer
Volume Control Block
Virtual Page Number
Window Control Block
Writable Control Store
Writable Diagnostic Control Store

APPENDIX B

IDENTIFICATION DIVISION .
MERGE EXAMPLE.
PROGRAM-ID.

*******************************•·························································
This program MERGEs three identically sequenced
regional sales files into one total sales file.
The program adds sales amounts and writes one
record for each product-code .

....••••.................••••••••..•.........••••••••••••.•••.........••••••.........•.••
ENVIRONMENT DIVISION .
CONFIGURATION SECTION .
VAX-11 .
SOURCE-COMPUTER .
OBJECT-COMPUTER.
VAX-11 .
INPUT-OUTPUT SECTION .
FILE-CONTROL.
SELECT REGION1-SALES
ASSIGN TO "REG1SLS".
ASSIGN TO "REG2SLS ".
SELECT REGION2-SALES
ASSIGN TO "REG3SLS ".
SELECT REGION3-SALES
ASSIGN TO "MRGFILE".
SELECT MERGE-FILE
ASSIGN TO "TOTLSLS".
SELECT TOTAL-SALES
DATA DIVISION.
FILE SECTION.
FD
REGION1-SALES
LABEL RECORDS ARE STANDARD.
PIC X(100) .
01
REGION1-RECORD
FD
REGION2-SALES
LABEL RECORDS ARE STANDARD.
PIC X(100) .
01
REGION2-RECORD
FD
REGION3-SALES
LABEL RECORDS ARE STANDARD.
PIC X(100) .
01
REGION3-RECORD
SD
MERGE-FILE.
01
MERGE-REC.
PICXX.
03 M-REGION-CODE
PIC X(10) .
03 M-PRODUCT-CODE
PIC S9(7)V99.
03 M-SALES-AMT
PIC X(79).
03 FILLER
FD
TOT AL-SALES
LABEL RECORDS ARE STANDARD.
PIC X(100) .
01
TOTAL-RECORD
WORKING-STORAGE SECTION.
VALUE " Y".
01
INITIAL-READ
PIC X
01
THE-COUNTERS.
PIC S9(7)V99.
03 PRODUCT-AMT
PIC S9(9)V99.
03 REGION1-AMT
PIC S9(9)V99.
03 REGION2-AMT
PIC S9(9)V99.
03 REGION3-AMT
PIC S9(11 )V99.
03 TOT AL-AMT
01
SAVE-MERGE-REC.
PICXX.
03 S-REGION-CODE
PIC X(10) .
03 S-PRODUCT-CODE
PIC S9(7)V99.
03 S-SALES-AMT
PIC X(79) .
03 FILLER
PROCEDURE DIVISION .
000-START SECTION .

B-1

010-MERGE-FILES .
OPEN OUTPUT TOT AL-SALES.
MERGE MERGE-FILE ON ASCENDING KEY M-PRODUCT-CODE
USING REGION1 -SALES REGION2-SALES REGION3-SALES
020-BUILD-TOTAL-SALES
OUTPUT PROCEDURE IS
THAU
100-DONE-TOTAL-SALES.
DISPLAY "TOTAL SALES FOR REGION 1 "
REGION1-AMT .
DISPLAY "TOTAL SALES FOR REGION 2 "
REGION2-AMT.
DISPLAY "TOT AL SALES FOR REGION 3 "
REGION3-AMT.
DISPLAY "TOTAL ALL SALES "
TOTAL-AMT.
CLOSE TOT AL-SALES .
DISPLAY " END OF PROGRAM MERGE01 ".
STOP RUN .
020-BUILD-TOTAL-SALES SECTION.
030-GET-MERGE-RECORDS.
RETURN MERGE-FILE AT END
MOVE PRODUCT-AMT TO S-SALES-AMT
WRITE TOTAL-RECORD FROM SAVE-MERGE-REC
GO TO 100-DONE-TOTAL SALES.
IF INITIAL-READ = "Y"
MOVE " N" TO INITIAL-READ
MOVE MERGE-REC TO SAVE-MERGE-REC
PERFORM 050-T ALLY-AMOUNTS
GO TO 030-GET-MERGE-RECORDS .
040-COMPARE-PRODUCT-CODE.
IF M-PRODUCT-CODE = S-PRODUCT-CODE
PERFORM 050-T ALLY-AMOUNTS
GO TO 030-GET-MERGE-RECORDS.
MOVE PRODUCT-AMT TO S-SALES-AMT.
MOVE ZEROES TO PRODUCT-AMT.
WRITE TOTAL-RECORD FROM SAVE-MERGE-REC.
MOVE MERGE-REC TO SAVE-MERGE-REC.
GO TO 040-COMPARE-PRODUCT-CODE.
050-T ALLY-AMOUNTS .
ADD M-SALES-AMT TO PRODUCT-AMT TOT AL-AMT.
IF M-REGION-CODE = " 01 "
ADD M-SALES-AMT TO REGION1-AMT.
IF M-REGION-CODE = " 02"
ADD M-SALES-AMT TO REGION2-AMT.
IF M-REGION-CODE = " 03"
ADD M-SALES-AMT TO REGION3-AMT .
100-DONE-TOTAL-SALES SECTION.
120-DONE.
EXIT.

8-2

APPENDIX C
VT100 examples:
IDENTIFICATION DIVISION .
PROGRAM-ID.
VIDE03.
ENVIRONMENT DIVISION.
CONFIGURATION SECTION.
SOURCE-COMPUTER.
VAX-11 .
OBJECT-COMPUTER.
VAX-11 .
SPECIAL-NAMES.
SYMBOLIC CHARACTERS
ARE

ESCAPER

PARM-1

PARM-2

PARM-3

PARM-4

103

75.

...........••.••••••••.••.•••••..•...•..•...................•....•......•.•••.••••••••••..•..•..
ESCAPER = ESC

PARM-1 = [ PARM-2 = ;

PARM-3=f

PARM-4 = J

................................................................................................
INPUT-OUTPUT SECTION .
FILE-CONTROL.
SELECT NAME-FILE ASSIGN TO " NAM FIL".
DAT A DIVISION.
FILE SECTION .
FO
NAME-FILE
LABEL RECORDS ARE STANDARD.
NAME-REC.
01
PICX(8) .
03
N-CUST-NUM
PIG X(25)
03
N-CUST-NAME
PIG X(25)
03
N-ADDRESS
PIG X(20)
03
N-CITY
PICXX.
03
N-STATE
PIG X(5) .
03
N-ZIP
WORKING-STORAGE SECTION .
01
HOME-UP.
PICX
03
FILLER
PICX
03
FILLER
PIC99
03
H-LINE
PICX
03
FILLER
PIC99
03
H-COL
PICX
03
FILLER
01
CLEAR-SCREEN .
PIG X
03
FILLER
PICX
03
FILLER
PICX
03
FILLER
THE-FORM .
01
FORM-LINE-1 .
03
PICX
FILLER
04
PICX
FILLER
04
PIC99
LINE-4
04
PICX
FILLER
04
PIC99
COL-3
04
PICX
FILLER
04
PIG X(24)
04
FILLER
"CUSTOMER NUMBER: _ __
FORM-LINE-2.
03
PICX
04
FILLER
PICX
04
FILLER
PIC99
04
LINE-6
PICX
04
FILLER
PIC99
COL-3
04
PICX
04
FILLER
PIG X(39)
04
FILLER
" CUSTOMER NAME:". _ _ _ __
C-1

VALUE ESCAPER.
VALUE PARM-1 .
VALUE 00.
VALUE PARM-2.
VALUEOO.
VALUE PARM-3.
VALUE ESCAPER.
VALUE PARM-1.
VALUE PARM-4.

VALUE ESCAPER.
VALUE PARM-1 .
VALUE04 .
VALUE PARM-2.
VALUE03.
VALUE PARM-3 .
VALUE

VALUE ESCAPER.
VALUE PARM-1 .
VALUE 06.
VALUE PARM-2.
VALUE 03.
VALUE PARM-3.
VALUE

FORM-LINE-3.
PICX
FILLER
04
PIC X
FILLER
04
PIC99
LINE-8
04
PICX
FILLER
04
PIC 99
COL-3
04
PICX
FILLER
04
PIC X(42)
FILLER
04
"CUSTOMER ADDRESS :
FORM-LIN E-4.
03
PICX
04
FILLER
PICX
04
FILLER
PIC99
04
LINE-10
PICX
04
FILLER
PIC99
COL-3
04
PICX
04
FILLER
PIC X(48)
04
FILLER
STATE:
ZIP :
"CITY:_ _ _ _
CURSOR-POSITIONER.
01
PICX
03
FILLER
PICX
03
FILLER
PIC 99.
03
LINE-POS
PICX
03
FILLER
PIC 99.
03
COL-POS
PICX
03
FILLER
01 INPUT-AREA
PROCEDURE DIVISION.
000-0PEN .
OPEN OUTPUT NAME-FILE.
005-BEGIN.
DISPLAY HOME-UP WITH NO ADVANCING .
010-CLEAR-SCREEN.
DISPLAY CLEAR-SCREEN WITH NO ADVANCING .
020-PAINT-FORM .
DISPLAY THE-FORM.
GO TO 050-GET-INPUT-DATA.
050-G ET-IN PUT-DAT A.
MOVE SPACES TO INPUT-AREA.
03

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

• Position cursor to accept customer number.•

.........•....••.•••••••.•..........•..•••.••••

MOVE 4 TO LINE-POS .
MOVE 19 TO COL-POS.
DISPLAY CURSOR-POSITIONER WITH NO ADVANCING.
ACCEPT INPUT-AREA.
IF INPUT-AREA= "DONE"
PERFORM 005-BEGIN THRU 010-CLEAR-SCREEN
CLOSE NAME-FILE STOP RUN.
MOVE INPUT-AREA TO N-CUST-NUM .
MOVE SPACES TO INPUT-AREA .

•..•••.•.......•..............•..••••••.••.••

• Position cursor to accept customer name.•
MOVE 6 TO LINE-POS.
MOVE 17 TO COL-POS.
DISPLAY CURSOR-POSITION ER WITH NO ADVANCING .
ACCEPT INPUT-AREA.
MOVE INPUT-AREA TO N-CUST-NAME.
MOVE SPACES TO INPUT-AREA .

C-2

VALUE ESCAPER.
VALUE PARM-1.
VALUE OS .
VALUE PARM-2.
VALUE03.
VALUE PARM-3.
VALUE

VALUE ESCAPER.
VALUE PARM-1 .
VALUE 10.
VALUE PARM-2 .
VALUE03.
VALUE PARM-3 .
VALUE

VALUE ESCAPER.
VALUE PARM-1.
VALUE PARM-2.
VALUE PARM-3.
PIC X(25) .

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

* Position cursor to accept customer address.*

.................••.....................••......
MOVE 8 TO LINE- POS.
MOVE 20 TO COL-POS.
DISPLAY CURSOR-POSITIONER WITH NO ADVANCING .
ACCEPT INPUT-AREA.
MOVE INPUT-AREA TON-ADDRESS.
MOVE SPACES TO INPUT-AREA.

.....•...•.•••..•..........•.•••••••
* Position cursor to accept city .
********** ** ** ** *** * ***** * **********

MOVE 10 TO LINE-POS.
MOVE 8 TO COL-POS.
DISPLAY CURSOR-POSITIONER WITH NO ADVANCING .
ACCEPT INPUT-AREA.
MOVE INPUT-AREA TON-CITY .
MOVE SPACES TO INPUT-AREA.
*** ** ************* *** ****************

* Position cursor to accept state.
MOVE 38 TO COL-POS.
DISPLAY CURSOR-POSITIONER WITH NO ADVANCING.
ACCEPT INPUT-AREA.
MOVE INPUT-AREA TON-STATE.
MOVE SPACES TON-STATE .

* Position cursor to accept zip.
MOVE 46 TO COL-POS.
DISPLAY CURSOR-POSITIONER WITH NO ADVANCING .
ACCEPT INPUT-AREA.
MOVE INPUT-AREA TON-ZIP.
WRITE NAME-REC.
GO TO 005-BEGIN.

C-3

The
Glossary

glos·sa·rY

abort An exception that occurs in the middle of an
instruction and potentially leaves the registers and
memory in an indeterminate state, such that the instruction cannot necessarily be restarted .

address access type The specified operand of an
instruction is not directly accessed by the instruction . The address of the specified operand is the
actual instruction operand. The context of the
address calculation is given by the data type of the
operand.

absolute indexed mode An indexed addressing
mode in which the base operand specifier is addressed in absolute mode.

addressing mode The way in which an operand is
specified ; for example, the way in which the effective
address of an instruction operand is calculated using the general registers. The basic general register
addressing modes are called: register, register deferred, autoincrement, autoincrement deferred, autodecrement, displacement, and displacement deferred . In addition , there are six indexed addressing
modes using two general registers, and literal mode
addressing. The PC addressing modes are called:
immediate (for register deferred mode using the
PC), absolute (for autoincrement deferred mode using the PC) , and branch .

absolute mode In absolute mode addressing, the
PC is used as the register in autoincrement deferred
mode. The PC contains the address of the location
containing the actual operand.
absolute time Time values expressing a specific
date (month, day, and year) and time of day. Absolute time values are always expressed in the system
as positive numbers.
access mode 1. Any of the four processor access
modes in which software executes. Processor access modes are, in order from most to least
privileged and protected : kernel (mode 0) , executive
(mode 1), supervisor (mode 2), and user (mode 3).
When the processor is in kernel mode, the executing
software has complete control of, and responsibility
for, the system . When the processor is in any other
mode, the processor is inhibited from executing privileged instructions. The Processor Status Longword contains the current access mode field . The
operating system uses access modes to define protection levels for software executing in the context of
a process. For example, the Executive runs in kernel
and executive mode and is most protected . The
command interpreter is less protected and runs in
supervisor mode. The debugger runs in user mode
and is no more protected than normal user programs . 2. See record access mode.

address space The set of all possible addresses
available to a process. Virtual address space refers
to the set of all possible virtual addresses. Physical
address space refers to the set of all possible physical addresses sent out on the SBI.
allocate a device To reserve a particular device
unit for exclusive use. A user process can allocate a
device only when that device is not allocated by any
other process .
alphanumeric character An upper or lower case
letter (A-Z, a-z) , a dollar sign ($), an underscore(_),
or a decimal digit (0-9).
alternate key An optional key within the data records in an indexed file ; used by VAX-11 RMS to build
an alternate index. See key (indexed files) and primary key.
American
Standard
Code for Information
Interchange (ASCII) A set of 8-bit binary numbers
representing the alphabet, punctuation, numerals,
and other special symbols used in text representation and communications protocol.

access type 1. The way in which the processor
accesses instruction operands. Access types are:
read , write, modify, address, and branch . 2. The way
in which a procedure accesses its arguments. 3. See
also record access type.

Ancillary Control Process (ACP) A process that
acts as an interface between user software and an
1/0 driver. An ACP provides functions supplemental
to those performed in the driver, such as file and
directory management. Three examples of ACPs
are: the Files-11 ACP (F11ACP), the magnetic tape
ACP (MTACP) , and the networks ACP (NET ACP).

access violation An attempt to reference an address that is not mapped into virtual memory or an
attempt to reference an address that is not accessible by the current access mode.
account name A string that identifies a particular
account used to accumulate data on a job's resource
use. This name is the user's accounting charge number, not the user's UIC .

area Areas are VAX-11 RMS maintained regions
of an indexed file . They allow a user to specify placement and/or specific bucket sizes for particular portions of a file . An area consists of any number of
buckets, and there may be from 1 to 255 areas in a
file .

address A number used by the operating system
and user software to identify a storage location. See
also virtual address and physical address.
G-1

Argument Pointer General register 12 (R12). By
convention, AP contains the address of the base of
the argument list for procedures initiated using the
CALL instructions.

the address of the actual operand. The contents of
the register are incremented by 4 (the number of
bytes in a longword). If the PC is used as the register ,
this mode is called absolute mode.

assign a channel To establish the necessary software linkage between a user process and a device
unit before a user process can transfer any data to or
from that device. A user process requests the system to assign a channel and the system returns a
channel number.

autoincrement indexed mode An indexed addressing mode in which the base operand specifier
uses autoincrement mode addressing.
autoincrement mode In autoincrement mode addressing, the contents of the specified register are
used as the address of the operand, then the contents of the register are incremented by the size of
the operand.

asynchronous record operation A mode of
record processing in which a user program can continue to execute after issuing a record retrieval or
storage request without having to wait for the request to be fulfilled.

balance set The set of all process working sets
currently resident in physical memory. The
processes whose working sets are in the balance set
have memory requirements that balance with available memory. The balance set is maintained by the
system swapper process.

Asynchronous System Trap A software-simulated interrupt to a user-defined service routine. ASTs
enable a user process to be notified asynchronously
with respect to its execution of the occurrence of a
specific event. If a user process has defined an AST
routine for an event, the system interrupts the process and executes the AST routine when that event
occurs. When the AST routine exits, the system
resumes the process at the point where it was interrupted.

base operand address The address of the base of
a table or array referenced by index mode addressing .
base operand specifier The register used to calculate the base operand address of a table or array
referenced by index mode addressing.
base priority The process priority that the system
assigns a process when it is created. The scheduler
never schedules a process below its base priority.
The base priority can be modified only by the system
manager or the process itself.

Asynchronous System Trap level (ASTLVL) A
value kept in an internal processor register that is
the highest access mode for which an AST is pending. The AST does not occur until the current access
mode drops in priority (raises in numeric value) to a
value greater than or equal to ASTLVL. Thus, an AST
for an access mode will not be serviced while the
processor is executing in a higher priority access
mode.
authorization file

base register A general register used to contain
the address of the first entry in a list, table, array, or
other data structure.

See user authorization file.

autodecrement indexed mode An indexed addressing mode in which the base operand specifier
uses autodecrement mode addressing.

binding

See linking.

bit string

See variable-length bit field.

block 1. The smallest addressable unit of data that
the specified device can transfer in an 1/0 operation
(512 contiguous bytes for most disk devices). 2. An
arbitrary number of contiguous bytes used to store
logically related status, control, or other processing
information.

autodecrement mode In autodecrement mode
addressing, the contents of the selected register are
decremented, and the result is used as the address
of the actual operand for the instruction. The contents of the register are decremented according to
the data type context of the register: 1 for byte, 2 for
word, 4 for longword and floating, 8 for quadword
and double floating.

block 1/0 The set of VAX-11 RMS procedures that
allow you direct access to the blocks of a file regardless of file organization.
bootstrap block A block in the index file on a
system disk that contains a program that can load
the operating system into memory and start its execution.

autoincrement deferred indexed mode An indexed addressing mode in which the base operand
specifier uses autoincrement deferred mode addressing.

branch access type An instruction attribute which
indicates that the processor does not reference an
operand address, but that the operand is a branch
displacement. The size of the branch displacement

autoincrement deferred mode In autoincrement
deferred mode addressing, the specified register
contains the address of a longword which contains
G-2

is given by the data type of the operand.

character A symbol represented by an ASCII
code. See also alphanumeric character.

branch mode In branch addressing mode, the instruction operand specifier is a signed byte or word
displacement. The displacement is added to the
contents of the updated PC (which is the address of
the first byte beyond the displacement), and the result is the branch address.

character string A contiguous set of bytes. A character string is identified by two attributes: an address
and a length. Its address is the address of the byte
containing the first character of the string . Subsequent characters are stored in bytes of increasing
addresses. The length is the number of characters in
the string.

bucket A storage structure, consisting of from 1 to
32 blocks, used for building and processing relative
and indexed files. A bucket contains one or more
records or record cells. Buckets are the unit of contiguous transfer between VAX-11 RMS buffers and
the disk.
buffered 1/0

character string descriptor A quadword data
structure used for passing character data (strings).
The first word of the quadword contains the length of
the character string . The second word can contain
type information. The remaining longword contains
the address of the string .

See system buffered 1/0.

bug check The operating system 's internal diagnostic check . The system logs the failure and
crashes the system.

cluster 1. A set of contiguous blocks that is the
basic unit of space allocation on a Files-11 disk volume. 2. A set of pages brought into memory in one
paging operation . 3. An event flag cluster.

byte A byte is eight contiguous bits starting on an
addressable byte boundary. Bits are numbered from
the right , 0 through 7, with bit 0 the low-order bit.
When interpreted arithmetically, a byte is a 2's complement integer with significance increasing from
bits 0 through 6. Bit 7 is the sign bit. The value of the
signed integer is in the range -128 to 127 decimal.
When interpreted as an unsigned integer, significance increases from bits 0 through 7 and the value
of the unsigned integer is in the range 0 to 255 decimal. A byte can be used to store one ASCII character.

command An instruction , generally an English
word , typed by the user at a terminal or included in a
command file which requests the software monitoring a terminal or reading a command file to perform
some well-defined activity. For example, typing the
COPY command requests the system to copy the
contents of one file into another file.
command file A file containing command strings.
See also command procedure.
command interpreter Procedure-based system
code that executes in supervisor mode in the context
of a process to receive, syntax check, and parse
commands typed by the user at a terminal or submitted in a command file.

cache memory A small, high-speed memory
placed between slower main memory and the processor. A cache increases effective memory transfer
rates and processor speed . It contains copies of data
recently used by the processor, and fetches several
bytes of data from memory in anticipation that the
processor will access the next sequential series of
bytes.
call frame

command parameter The positional operand of a
command delim ited by spaces, such as a file specification, option , or constant.
command procedure A file containing commands
and data that the command interpreter can accept in
lieu of the user typ ing the commands individually on
a terminal.

See stack frame.

call instructions The processor instructions
CALLG (Call Procedure with General Argument List)
and CALLS (Call Procedure with Stack Argument
List) .

command string A line (or set of continued lines),
normally terminated by typing the carriage return
key, containing a command and, optionally, information modifying the command. A complete command
string consists of a command, its qualifiers, if any,
and its parameters (file specifications, for example) ,
if any, and their qualifiers, if any.

call stack The stack , and conventional stack structure, used during a procedure call . Each access
mode of each process context has one call stack,
and interrupt service context has one call stack.

common A FORTRAN term for a program section
that contains only data.

channel A logical path connecting a user process
to a physical device unit. A user process requests
the operating system to assign a channel to a device
so the process can transfer data to or from that device.

common event flag cluster A set of 32 event flags
that enables cooperating processes to post event
notification to each other. Common event flag clusG-3

console The manual control unit integrated into
the central processor. The console includes an LSl11 microprocessor and a serial line interface connected to a hard copy terminal. It enables the operator to start and stop the system, monitor system operation, and run diagnostics.

ters are created as they are needed. A process can
associate with up to two common event flag clusters.

compatibility mode A mode of execution that enables the central processor to execute non-privileged PDP-11 instructions. The operating system
supports compatibility mode execution by providing
an RSX-11 M programming environment for an RSX11 M task image. The operating system compatibility
mode procedures reside in the control region of the
process executing a compatibility mode image. The
procedures intercept calls to the RSX-11 M Executive
and convert them to the appropriate operating system functions.

console terminal The hard copy terminal connected to the central processor console.
control region The highest-addressed half of perprocess space (the P1 region). Control region virtual
addresses refer to the process-related information
used by the system to control the process, such as:
the kernel, executive, and supervisor stacks , the
permanent 1/0 channels, exception vectors, and dynamically used system procedures (such as the
command interpreter and RSX-11 M programming
environment compatibility mode procedures). The
user stack is also normally found in the control region, although it can be relocated elsewhere.

condition An exception condition detected and
declared by software. For example, see failure exception mode.
condition codes Four bits in the Processor Status
Word that indicate the results of previously executed
instructions.

Control Region Base Register (P1 BR) The processor register, or its equivalent in a hardware process control block, that contains the base virtual address of a process control region page table.

condition handler A procedure that a process
wants the system to execute when an exception condition occurs. When an exception condition occurs,
the operating system searches for a condition
handler and , if found, initiates the handler immediately. The condition handler may perform some action to change the situation that caused the exception condition and continue execution for the process that incurred the exception condition. Condition
handlers execute in the context of the process at the
access mode of the code that incurred the exception
condition.

Control Region Length Register (P1 LR) The
processor register, or its equivalent in a hardware
process control block, that contains the number of
non-existent page table entries for virtual pages in a
process control region.
copy-on-reference A method used in memory
management for sharing data until a process
accesses it, in which case it is copied before the
access. Copy-on-reference allows sharing of the initial values of a global section whose pages have
read/write access but contain pre-initialized data
available to many processes.

condition value A 32-bit quantity that uniquely
identifies an exception condition.
context The environment of an activity. See also
process context, hardware context, and software
context.

counted string A data structure consisting of a
byte-sized length followed by the string. Although a
counted string is not used as a procedure argument,
it is a convenient representation in memory.

context indexing The ability to index through a
data structure automatically because the size of the
data type is known and used to determine the offset
factor.

current access mode The processor access
mode of the currently executing software. The Current Mode field of the Processor Status Longword
indicates the access mode of the currently executing
software.

context switching Interrupting the activity in progress and switching to another activity. Context
switching occurs as one process after another is
scheduled for execution. The operating system
saves the interrupted process' hardware context in
its hardware process control block (PCB) using the
Save Process Context instruction, and loads another
process' hardware PCB into the hardware context
using the Load Process Context instruction, scheduling that process for execution.

cylinder The tracks at the same radius on all recording surfaces of a disk.
D_floating (point) datum A floating point datum
consisting of B contiguous bytes (64 bits) starting on
an arbitrary byte boundary. The value of the
D_floating datum is in the approximate range ( + or
- ) 0.29 x 1o- 3 a to 1. 7 x 10 38 . The precision is approximately one part in 2 55 or typically sixteen decimal digits.

continuation character A hyphen at the end of a
command line signifying that the command string
continues on to the next command line.
G-4

data base 1. All the occurrences of data described
by a data base management system. 2. A collection
of related data structures.

device name The field in a file specification that
identifies the device unit on which a file is stored.
Device names also include the mnemonics that identify an 1/0 peripheral device in a data transfer request. A device name consists of a mnemonic followed by a controller identification letter (if
applicable), followed by a unit number (if applicable) , and ends with a colon(:).

data structure Any table, list, array, queue, or tree
whose format and access conventions are well-defined for reference by one or more images.
data type In general, the way in which bits are
grouped and interpreted. In reference to the processor instructions, the data type of an operand identifies the size of the operand and the significance of
the bits in the operand. Operand data types include:
byte, word, longword and quadword integer, floating
and double floating, character string, packed decimal string , and variable-length bit field .

device queue

See spool queue.

device register A location in device controller logic used to request device functions (such as 1/0
transfers) and/or report status.
device unit One drive, and its controlling logic, of
a mass storage device system. A mass storage system can have several drives connected to it.

deferred echo Refers to the fact that terminal
echoing does not occur until a process is ready to
accept input entered by type ahead.

diagnostic A program that tests logic and reports
any faults it detects .

delta time A time value expressing an offset from
the current date and time. Delta times are always
expressed in the system as negative numbers whose
absolute value is used as an offset from the current
time.

direct 1/0 An 1/0 operation in which the system
locks the pages containing the associated buffer in
memory for the duration of the 1/0 operation. The
110 transfer takes place directly from the process
buffer. Contrast with system buffered 1/0.

demand zero page A page, typically of an image
stack or buffer area, that is initialized to contain all
zeros when dynamically created in memory as a result of a page fault. This feature eliminates the waste
of disk space that would otherwise be required to
store blocks (pages) that contain only zeros.

direct mapping cache A cache organization in
which only one address comparison is needed to
locate any data in the cache because any block of
main memory data can be placed in only one possible position in the cache. Contrast with fully associative cache.

descriptor A data structure used in calling sequences for passing argument types, addresses and
other optional information. See character string descriptor.

directory A file used to locate files on a volume
that contains a list of file names (including type and
version number) and their unique internal identifications.

detached process A process that has no owner.
The parent process of a tree of subprocesses. Detached processes are created by the job controller
when a user logs on the system or when a batch job
is initiated . The job controller does not own the user
processes it creates; these processes are therefore
detached .

directory name The field in a file specification that
identifies the directory file in which a file is listed. The
directory name begins with a left bracket ([or <)and
ends with a right bracket(] or>).
displacement deferred indexed mode An indexed addressing mode in which the base operand
specifier uses displacement deferred mode
addressing.

device The general name for any physical
term inus or link connected to the processor that is
capable of receiving, storing, or transmitting data.
Card readers, line printers, and terminals are examples of record-oriented devices. Magnetic tape devices and disk devices are examples of mass storage devices. Terminal line interfaces and interprocessor links are examples of communications
devices.

displacement deferred mode In displacement
deferred mode addressing, the specifier extension is
a byte, word , or longword displacement. The displacement is sign extended to 32 bits and added to a
base address obtained from the specified register.
The result is the address of a longword which contains the address of the actual operand. If the PC is
used as the register, the updated contents of the PC
are used as the base address. The base address is
the address of the first byte beyond the specifier
extension .

device interrupt An interrupt received on interrupt
priority level 16 through 23. Device interrupts can be
requested only by devices, controllers, and memories.
G-5

event A change in process status or an indication
of the occurrence of some activity that concerns an
individual process or cooperating processes. An incident reported to the scheduler that affects a process' ability to execute. Events can be synchronous
with the process' execution (a wait request), or they
can be asynchronous (1/0 completion). Some other
events include: swapping; wake request; page fault.

displacement indexed mode An indexed addressing mode in which the base operand specifier
uses displacement mode addressing.
displacement mode In displacement mode addressing, the specifier extension is a byte, word, or
longword displacement. The displacement is sign
extended to 32 bits and added to a base address
obtained from the specified register. The result is the
address of the actual operand. If the PC is used as
the register, the updated contents of the PC are used
as the base address. The base address is the address of the first byte beyond the specifier extension.

event flag A bit in an event flag cluster that can be
set or cleared to indicate the occurrence of the event
associated with that flag. Event flags are used to synchronize activities in a process or among many
processes.

drive The electro-mechanical unit of a mass storage device system on which a recording medium
(disk cartridge, disk pack, or magnetic tape reel) is
mounted.

event flag cluster A set of 32 event flags which are
used for event posting. Four clusters are defined for
each process: two process-local clusters, and two
common event flag clusters. Of the process-local
flags, eight are reserved for system use.

driver The set of code that handles physical 1/0 to
a device.

exception An event detected by the hardware
(other than an interrupt or jump, branch, case, or call
instruction) that changes the normal flow of instruction execution. An exception is always caused by the
execution of an instruction or set of instructions
(whereas an interrupt is caused by an activity in the
system independent of the current instruction).
There are three types of hardware exceptions: traps,
faults, and aborts. Examples are: attempts to execute a privileged or reserved instruction, trace faults,
compatibility mode faults, breakpoint instruction execution, and arithmetic faults such as floating point
overflow, underflow, and divide by zero.

dynamic access A technique in which a program
switches from one record access mode to another
while processing a file.
echo A terminal handling characteristic in which
the characters typed by the terminal user on the keyboard are also displayed on the screen or printer.
effective address The address obtained after indirect or indexing modifications are calculated.
entry mask A word whose bits represent the
registers to be saved or restored on a subroutine or
procedure call using the call and return instructions.
entry point A location that can be specified as the
object of a call. It contains an entry mask and exception enables known as the entry point mask.

exception condition A hardware- or software-detected event other than an interrupt or jump, branch,
case, or call instruction that changes the normal flow
of instruction execution.

equivalence name The string associated with a
logical name in a logical name table. An equivalence
name can be, for example, a device name, another
logical name, or a logical name concatenated with a
portion of a file specification.

exception dispatcher An operating system procedure that searches for a condition handler when
an exception condition occurs. If no exception
handler is found for the exception or condition, the
image that incurred the exception is terminated .

error logger A system process that empties the
error log buffers and writes the error messages into
the error file. Errors logged by the system include
memory system errors, device errors and timeouts,
and interrupts with invalid vector addresses.

exception enables
exception vector

See trap enables.
See vector.

executable image An image that is capable of being run in a process. When run, an executable image
is read from a file for execution in a process.

escape sequence An escape is a transition from
the normal mode of operation to a mode outside the
normal mode. An escape character is the code that
indicates the transition from normal to escape mode.
An escape sequence refers to the set of character
combinations starting with an escape character that
the terminal transmits without interpretation to the
software set up to handle escape sequences.

executive The generic name for the collection of
procedures included in the operating system software that provides the basic control and monitoring
functions of the operating system.
executive mode The second most privileged
processor access mode (mode 1). The record management services (RMS) and many of the operating
G-6

system 's programmed service procedures execute
in executive mode.

file header A block in the index file describing a
file on a Files-11 disk structure. The file header identifies the locations of the file's extents. There is a file
header for every file on the disk.

exit An image exit is a rundown activity that occurs
when image execution terminates either normally or
abnormally. Image rundown activities include deassigning 1/0 channels and disassociation of common
event flag clusters. Any user- or system-specified
exit handlers are called .

file name The field preceding a file type in a file
specification that contains a 1- to 9-character logical
name for a file.
filename extension

exit handler A procedure executed when an image exits. An exit handler enables a procedure that
is not on the call stack to gain control and clean up
procedure-own data bases before the actual image
exit occurs.

See file type.

file organization The physical arrangement of data in the file. You select the specific organization
from those offered by V AX-11 RMS, based on your
individual needs for efficient data storage and retrieval. See indexed file organization, relative file organization, and sequential file organization.

extended attribute block (XAB) An RMS user data structure that contains additional file attributes
beyond those expressed in the file access block
(FAB) , such as boundary types (aligned on cylinder,
logical block number, virtual block number) and file
protection information.

Files-11 The name of the on-disk structure used
by the RSX-11, IAS and VAX/VMS operating systems. Volumes created under this structure are
transportable between these operating systems.
file specification A unique name for a file on a
mass storage medium. It identifies the node, the device, the directory name, the file name, the file type,
and the version number under which a file is stored.

extension The amount of space to allocate at the
end of a file each time a sequential write exceeds the
allocated length of the file .
extent The contiguous area on a disk containing a
file or a portion of a file . Consists of one or more
clusters.

file structure The way in which the blocks forming
a file are distributed on a disk or magnetic tape to
provide a physical accessing technique suitable for
the way in which the data in the file is processed.

F_floating (point) datum A floating point datum
consisting of 4 contiguous bytes (32 bits) starting on
an arbitrary byte boundary. The value of the
F_floating datum is in the approximate range
(+ or -) 0.29 x 10-3s to 1. 7 x 10 38 . The precision is
approximately one part in 223 or typically seven decimal digits.

file system A method of recording, cataloging,
and accessing files on a volume.
file type The field in a file specification that is
preceded by a period or dot (.) and consists of a
zero- to three-character type identification. By convention, the type identifies a generic class of files
that have the same use or characteristics, such as
ASCII text files, binary object files, etc.

failure exception mode A mode of execution selected by a process indicating that it wants an exception condition declared if an error occurs as the result of a system service call. The normal mode is for
the system service to return an error status code for
which the process must test.

fixed control area An area associated with a variable length record available for controlling or assisting record access operations. Typical uses include
line numbers and printer format control information.

fault A hardware exception condition that occurs
in the middle of an instruction and that leaves the
registers and memory in a consistent state, such that
elimination of the fault and restarting the instruction
will give correct results.

fixed-length record format Property of a file in
which all records are of the same size. This format
provides simplicity in determining the exact location
of a record in the file and eliminates the need to
prefix a record size field to each record.

field 1. See variable-length bit field. 2. A set of
contiguous bytes in a logical record .

floating (point) datum A numeric data type in
which the number is represented by a fraction (less
than 1 and greater than or equal to 112) multiplied by 2
raised to a power. There are four floating point data
types: F_floating (4 bytes), D_floating (8 bytes),
G_floating (8 bytes), and H_floating (16 bytes)

file An organized collection of related items (records) maintained in an accessible storage area, such
as disk or tape.
file access block (FAB) An RMS user data structure that represents a request for data operations
related to the file as a whole, such as OPEN, CLOSE,
or CREATE.

foreign volume Any volume other than a Files-11
formatted volume which may or may not be file
structured.
G-7

fork process A dynamically created system process such as a process that executes device driver
code or the timer process. Fork processes have minimal context. Fork processes are scheduled by the
hardware rather than by the software. The timer
process is dispatched directly by software interrupt.
1/0 driver processes are dispatched by a fork dispatcher. Fork processes execute at software interrupt levels and are dispatched for execution immediately. Fork processes remain resident until they terminate.

global symbol table (GST) In a library, an index of
strongly defined global symbols used to access the
modules defining the global symbols. The linker will
also put global symbol tables into an image. For example, the linker appends a global symbol table to
executable images that are intended to run under
the symbolic debugger, and it appends a global
symbol table to all shareable images.
group 1. A set of users who have special access
privileges to each other's directories and files within
those directories (unless protected otherwise), as in
the context " system, owner, group, world ," where
group refers to all members of a particular owner's
group. 2. A set of jobs (processes and their subprocesses) who have access privileges to a group's
common event flags and logical name tables, and
may have mutual process controlling privileges,
such as scheduling, hibernation, etc.
group number The first number in a User Identification Code (UIC).

frame pointer General register 13 (R13) . By
convention, FP contains the base address of the
most recent call frame on the stack.
fully associative cache A cache organization in
which any block of data ·from main memory can be
placed anywhere in the cache. Address comparision
must take place against each block in the cache to
find any particular block . Contrast with direct mapping cache.

H_floating (point) datum A floating point datum
consisting of 16 contiguous bytes (128 bits) starting
on an arbitrary byte boundary. The value of the
H_floating datum is in the approximate range
(+ or - ) 0.84 x 10 4932 to 0.59 x 104932 • The precision
is approximately one part in 2 112 or typically 33 decimal digits.

G_floating (point) datum A floating point datum
consisting of 8 contiguous bytes (64 bits) starting on
an arbitrary byte boundary. The value of the
G_floating datum is in the approximate range
(+ or - ) 0.56 x 10-3os to 9 x 10308 • The precision is
approximately one part in 252 or typically fifteen.
general register Any of the sixteen 32-bit registers
used as the primary operands of the native mode
instructions. The general registers include 12 general purpose registers which can be used as accumulators, as counters, and as pointers to locations in
main memory, and the Frame Pointer (FP). Argument Pointer (AP), Stack Pointer (SP), and Program
Counter (PC) registers.

hardware context The values contained in the following registers while a process is executing: the
Program Counter (PC); the Processor Status Longword (PSL); the 14 general registers (RO through
R13) ; the four processor registers (POBR , POL~ .
P1 BR and P1 LR) that describe the process virtual
address space; the Stack Pointer (SP) for the current
access mode in which the processor is executing ;
plus the contents to be loaded in the Stack Pointer
for every access mode other than the current access
mode. While a process is executing, its hardware
context is continually being updated by the processor. While a process is not executing, its hardware
context is stored in its hardware PCB.

generic device name A device name that identifies the type of device but not a particular unit; a
device name in which the specific controller and/or
unit number is omitted.
giga Metric term used to represent the number 1
followed by nine zeros.

hardware process control block (PCB) A data
structure known to the processor that contains the
hardware context when a process is not executing. A
process' hardware PCB resides in its process head er.

global page table The page table containing the
master page table entries for global sections.
global section A data structure (e.g., FORTRAN
global common) or shareable image section
potentially available to all processes in the system.
Access is protected by privilege and/or group number of the UIC.

hibernation A state in which a process is inactive,
but known to the system with all of its current status.
A hibernating process becomes active again when a
wake request is issued. It can schedule a wake request before hibernating, or another process can issue its wake request. A hibernating process also becomes active for the time sufficient to service any
AST it may receive while it is hibernating . Contrast
with suspension .

global symbol A symbol defined in a module that
is potentially available for reference by another module. The linker resolves (matches references with
definitions) global symbols. Contrast with local symbol.
G-8

home block A block in the index file that contains
the volume identification , such as volume label and
protection.

ing the result to the base operand address. The
addressing modes resulting from index mode addressing are formed by adding the suffix "indexed"
to the addressing mode of the base operand specifier: register deferred indexed, autoincrement indexed, autoincrement deferred indexed (or absolute
indexed), autodecrement indexed, displacement indexed, and displacement deferred indexed .

image An image consists of procedures and data
that have been bound together by the linker. There
are three types of images: executable, shareable,
and system.
image activator A set of system procedures that
prepare an image for execution. The image activator
establishes the memory management data structures required both to map the image's virtual pages
to physical pages and to perform paging .
image exit

indexed file organization A file organization
which allows random retrieval of records by key values and sequential retrieval of records in sorted order by key value. See key (indexed files).
indirect command file

See exit.

input stream The source of commands and data.
One of: the user's terminal, the batch stream, or an
indirect command file.

image 1/0 segment That portion of the control region that contains the RMS internal file access
blocks (IFAB) and 1/0 buffers for the image currently
being executed by a process.

instruction buffer A buffer in the processor used
to contain bytes of the instruction currently being
decoded and to pre-fetch instructions in the instruction stream. The control logic continously fetches
data from memory to keep the buffer full.

image name The file name of the file in which an
image is stored .
image privileges The privileges assigned to an
image when it is linked. See process privileges.

interleaving Assigning consecutive physical
memory addresses alternately between two memory
controllers.

image section (isect) A group of program sections (psects) with the same attributes (such as readonly access, read/write access, absolute, relocatable, etc.) that is the unit of virtual memory allocation
for an image.

interlocked The property of a read followed by a
write to the same datum with no possibility of an
intervening reference by a second processor or 1/0
device. Examples are the Branch on Bit Interlocked
and Add Aligned Word Interlocked instructions.

immediate mode In immediate mode addressing,
the PC is used as the register in autoincrement mode
addressing.

interprocess communication facility A common
event flag, mailbox, or global section used to pass
information between two or more processes.

index The structure which allows retrieval of records in an indexed file by key value . See key (indexed files) .

interrecord gap A blank space deliberately placed
between data records on the recording surface of a
magnetic tape.

index file The file on a Files-11 volume that
contains the access information for all files on the
volume and enables the operating system to identify
and access the volume.

interrupt An event other than an exception or
branch, jump, case, or call instruction that changes
the normal flow of instruction execution. Interrupts
are generally external to the process executing when
the interrupt occurs. See also device interrupt, software interrupt, and urgent interrupt.

index file bit map A table in the index file of a
Files-11 volume that indicates which file headers are
in use.
index register
dress offset.

See command procedure.

A register used to contain an ad-

interrupt priority level (IPL) The interrupt level at
which the processor executes when an interrupt is
generated. There are 31 possible interrupt priority
levels. IPL 1 is lowest, 31 highest. The levels arbitrate
contention for processor service. For example, a device cannot interrupt the processor if the processor
is currently executing at an interrupt priority level
greater than the interrupt priority level of the device's
interrupt service routine.

indexed addressing mode In indexed mode addressing, two registers are used to determine the
actual instruction operand: an index register and a
base operand specifier. The contents of the index
register are used as an index (offset) into a table or
array. The base operand specifier supplies the base
address of the array (the base operand address or
BOA). The address of the actual operand is calculated by multiplying the contents of the index register
by the size (in bytes) of the actual operand and add-

interrupt service routine The routine executed
when a device interrupt occurs.
G-9

LOGIN image for the job, maintains the accounting
record for the job, manages symbionts, and termi nates a process and its subprocesses.

interrupt stack The system-wide stack used when
executing in interrupt service context. At any time,
the processor is either in a process context executing in user, supervisor, executive or kernel mode, or
in system-wide interrupt service context operating
with kernel privileges, as indicated by the interrupt
stack and current mode bits in the PSL. The interrupt stack is not context switched.

job queue A list of files that a process has suppl ied
for processing by a specific device, for example, a
line printer.
kernel mode The most privileged processor
access mode (mode 0). The operating system 's most
privileged services, such as 1/0 drivers and the pager, run in kernel mode.

interrupt stack pointer The stack pointer for the
interrupt stack. Unlike the stack pointers for process
context stacks, which are stored in the hardware
PCB, the interrupt stack pointer is stored in an internal register.
interrupt vector
1/0 driver

key
indexed files: A character string, a packed decimal
number, a 2- or 4-byte unsigned binary number, or a
2- or 4-byte signed integer within each data record in
an indexed file. You define the length and location
within the records; VAX-11 RMS uses the key to
build an index. See primary key, alternate key, and
random access by key value.

See vector.

See driver.

1/0 function An 1/0 operation that is interpreted by
the operating system and typically results in one or
more physical 1/0 operations.

relative tiles : The relative record number of each data record in a data file; VAX-11 RMS uses the relative
record numbers to identify and access data records
in a relative file in random access mode. See relative
record number.

1/0 function code A 6-bit value specified in a
Queue 1/0 Request system service that describes
the particular 1/0 operation to be performed (e .g.,
read , write, rewind).
1/0 function modifier A 10-bit value specified in a
Queue 1/0 Request system service that modifies an
1/0 function code (e.g ., read terminal input no echo).

lexical function A command language construct
that the command interpreter evaluates and substitutes before it performs expression analysis on a
command string. Lexical functions return information about the current process, such as UIC or default directory; and about character strings, such as
length or substring locations.

1/0 lockdown The state of a page such that it cannot be paged or swapped out of memory until any
1/0 in progress to that page is completed.
1/0 rundown An operating system function in
which the system cleans up any 1/0 in progress
when an image exits.

librarian A program that allows the user to create,
update, modify, list, and maintain object library, image library, and assembler macro library files.

1/0 space The region of physical address space
that contains the configuration registers, and device
control/status and data registers.
1/0 status block A data structure associated with
the Queue 1/0 Request system service. This service
optionally returns a status code, number of bytes
transferred, and device- and function-dependent information in an 1/0 status block. It is not returned
from the service call, but filled in when the 1/0 request completes.
job 1. A job is the accounting unit equivalent to a
process and the collection of all the subprocesses, if
any, that it and its subprocesses create. Jobs are
classified as batch and interactive. For example, the
job controller creates an interactive job to handle a
user's requests when the user logs onto the system
and it creates a batch job when the symbiont manager passes a command input file to it. 2. A print job.
job controller The system process that establishes
a job's process context, starts a process running the
G- 10

library file A direct access file containing one or
more modules of the same module type.
limit The size or number of given items requ iri ng
system resources (such as mailboxes, locked pages,
1/0 requests, open files, etc.) that a job is allowed to
have at any one time during execution, as specified
by the system manager in the user authorization file .
See also quota.
line number A number used to identify a line of
text in a file processed by a text editor.
linker A program that reads one or more object
files created by language processors and produces
an executable image file, a shareable image file , or a
system image file .
linking The resolution of external references
between object modules used to create an image,
the acquisition of referenced library routines, service
entry points, and data for the image, and the assignment of virtual addresses to components of an image.

literal mode In literal mode addressing , the instructi on operand is a constant whose value is expressed in a 6-bit field of the instruction. If the
operand data type is byte, word , longword , quadword , or octaword , the operand is zero-extended
and can express values in the range 0 through 63
(decimal). If the operand data type is F_, D_, G_, or
H_floating, the 6-bit field is composed of two 3-bit
fields , one for the exponent and the other for the
fraction. The operand is extended to F_, D_, G_, or
H_floating format.
locality

See program locality.

local symbol A symbol meaningful only to the
module that defines it. Symbols not identified to a
language processor as global symbols are considered to be local symbols. A language processor resolves (matches references with definitions) local
symbols. They are not known to the linker and cannot be made available to another object module.
They can, however, be passed through the linker to
the symbolic debugger. Contrast with global symbol.
locate mode Technique used for a record input
operation in which the data records are not copied
from the 1/0 buffer. See move mode.

particular process, a particular group, or the system.

logical 1/0 functions A set of 1/0 functions that
allow restr icted direct access to device level 1/0 operations.
logical record
a unit.

A group of related fields treated as

longword Four contiguous bytes (32 bits) starting
on an addressable byte boundary. Bits are numbered from right to left, 0 through 31. The address of
the longword is the address of the byte containing bit
0. When interpreted arithmetically, a longword is a
2's complement integer with significance increasing
from bit 0 to bit 30. When interpreted as a signed
integer, bit 31 is the sign bit. The value of the signed
integer is in the range -2,147,483 ,648to2 ,147,483,647. When interpreted as an unsigned integer, significance increases from bit 0 to bit 31 . The value of
the unsigned integer is in the range 0 through
4,294,967 ,295 .
macro A statement that requests a language processor to generate a predefined set of instructions.
mailbox A software data structure that is treated
as a record-oriented device for general interprocess
communication. Communication using a mailbox is
similar to other forms of device-independent 1/0.
Senders perform a write to a mailbox, the receiver
performs a read from that mailbox. Some systemwide mailboxes are defined: the error logger and
OPCOM read from system-wide mailboxes.

locking a page in memory Mak ing a page in an
image ineligible for either paging or swapping . A
page stays locked in memory until it is specifically
unlocked .
locking a page in the working set Making a page
in an image ineligible for paging out of the working
set for the image. The page can be swapped when
the process is swapped. A page stays locked in a
work ing set until it is specifically unlocked .

main memory

See physical memory.

mapping window A subset of the retrieval information for a file that is used to translate virtual block
numbers to log ical block numbers.

logical block number A number used to identify a
block on a mass storage device. The number is a
volume-relative address rather than its physical (device-oriented) address or its virtual (file-relative) ad dress. The blocks that constitute the volume are labeled sequentially starting with logical block 0.

mass storage device A device capable of reading
and writing data on mass storage media such as a
disk pack or a magnetic tape reel.
member number The second number in a user
identification code that uniquely identifies that code.

logical 1/0 function A set of 1/0 operations (e.g.,
read and write logical block) that allow restricted
direct access to device level 1/0 operations using
logical block addresses.

memory management The system funct ions that
include the hardware's page mapping and protection and the operating system's image activator and
pager.

logical name A user-specified name for any portion or all of a file specification. For example, the
logical name INPUT can be assigned to a terminal
device from which a program reads data entered by
a user. Logical name assignments are maintained in
logical name tables for each process, each group,
and the system. A logical name can be created and
assigned a value permanently or dynamically.

Memory Mapping Enable (MME) A bit in a processor register that governs address translation.
modify access type The specified operand of an
instruction or procedure is read, and is potentially
modified and wr itten , during that instruction 's or
procedure's execution .

logical name table A table that contains a set of
logical names and their equivalence names for a
G-11

module 1. A portion of a program or program library, as in a source module, object module , or im-

age module. 2. A board, usually made of plastic covered with an electrical conductor, on which logic devices (such as transistors, resistors, and memory
chips) are mounted, and circuits connecting these
devices are etched, as in a logic module.

Monitor Console Routine (MCR)
interpreter in an RSX-11 system.

which is used as input to the linker.

object time system (OTS)
cedure Library.

See Run Time Pro-

octaword Sixteen contiguous bytes (128 bits)
starting on an addressable byte boundary. Bits are
numbered from right to left, 0 to 127. An octaword is
identified by the address of the byte containing the
low-order bit (bit 0).

The command

mount a volume 1. To logically associate a volume
with the physical unit on which it is loaded (an activity
accomplished by system software at the request of
an operator). 2. To load or place a magnetic tape or
disk pack on a drive and place the drive online (an
activity accomplished by a system operator).

offset A fixed displacement from the beginning of
a data structure. System offsets for items within a
data structure normally have an associated symbolic
name used instead of the numeric displacement.
Where symbols are defined, programmers always
reference the symbolic names for items in a data
structure instead of using the numeric displacement.

move mode Technique used for a record transfer
in which the data records are copied between the 1/0
buffer and your program buffer for calculations or
operations on the record. See locate mode.

opcode The pattern of bits within an instruction
that specify the operation to be performed.

mutex A semaphore that is used to control exclusive access to a region of code that can share a data
structure or other resource. The mutex (mutual exclusion) semaphore ensures that only one process at
a time has access to the region of code.
name block (NAM) An RMS user data structure
that contains supplementary information used in
parsing file specifications.
native image An image whose instructions are executed in native mode.
native mode The processor's primary execution
mode in which the programmed instructions are interpreted as byte-aligned, variable-length instructions that operate on byte, word, longword, quadword, and octaword integer, F_, D_, G_ and
H_floating format, character string, packed decimal,
and variable-length bit field data. The instruction execution mode other than compatibility mode.

operand specifier The pattern of bits in an instruction that indicate the addressing mode, a register and/or displacement, which, taken together,
identify an instruction operand.
operand specifier type The access type and data
type of an instruction's operand(s). For example, the
test instructions are of read access type, since they
only read the value of the operand. The operand can
be of byte, word, or longword data type, depending
on whether the opcode is for the TSTB (test byte) ,
TSTW (test word), or TSTL (test longword) instruction.
Operator Communication Manager (OPCOM) A
system process that is always active. OPCOM receives input from a process that wants to inform an
operator of a particular status or condition, passes a
message to the operator, and tracks the message.
operator's console Any terminal identified as a
terminal attended by a system operator.

network A collection of interconnected individual
computer systems.
nibble
byte.

The low-order or high-order four bits of a

node

An individual computer system in a network.

owner In the context "system, owner, group,
world," an owner is the particular member (of a
group) to which a file, global section, mailbox, or
event flag cluster belongs.
owner process The process (with the exception of
the job controller) or subprocess that created a subprocess.

null process A small system process that is the
lowest priority process in the system and takes one
entire priority class. One function of the null process
is to accumulate idle processor time.
numeric string A contiguous sequence of bytes
representing up to 31 decimal digits (one per byte)
and possibly a sign. The numeric string is specified
by its lowest addressed location, its length, and its
sign representation.
object module The binary output of a language
processor such as the assembler or a compiler,
G-12

packed decimal A method of representing a decimal number by storing a pair of decimal digits in one
byte, taking advantage of the fact that only four bits
are required to represent the numbers 0 through 9.
packed decimal string A contiguous sequence of
up to 16 bytes interpreted as a string of nibbles.
Each nibble represents a digit except the low-order
nibble of the highest addressed byte, which repre-

sents the sign . The packed decimal string is specified by its lowest addressed location and the number
of dig its.

physical address The address used by hardware
to identify a location in physical memory or on
directly addressable secondary storage devices
such as a disk . A physical memory address consists
of a page frame number and the number of a byte
within the page. A physical disk block address consists of a cylinder or track and sector number.

page 1. A set of 512 contiguous byte locations
used as the unit of memory mapping and protection.
2. The data between the beginning of file and a page
marker, between two markers, or between a marker
and the end of a file .

physical address space The set of all possible 30bit physical addresses that can be used to refer to
locations in memory (memory space) or device
registers (1/0 space) .

page fault An exception generated by a reference
to a page which is not mapped into a working set.
page fault cluster size
in on a page fault.

The number of pages read

physical block A block on a mass storage device
referred to by its physical (device-oriented) address
rather than a log ical (volume-relative) or virtual (filerelative) address.

page frame number (PFN) The address of the first
byte of a page in physical memory. The high-order
21 bits of the physical address of the base of a page.

physical 1/0 functions A set of 1/0 functions that
allow access to all device level 1/0 operations except
maintenance mode.

page marker A character or characters (generally
a form feed) that separates pages in a file that is
processed by a text editor.

physical memory The memory modules connected to the SBI that are used to store: 1) instructions
that the processor can directly fetch and execute,
and 2) any other data that a processor is instructed
to manipulate. Also called main memory.

pager A set of kernel mode procedures that executes as the result of a page fault. The pager makes
the page fo r which the fault occurred available in
physical memory so that the image can continue execution. The pager and the image activator provide
the operating system 's memory management functions.

position-dependent code Code that can execute
properly only in the locations in virtual address
space that are assigned to it by the linker.

page table entry (PTE) The data structure that
identifies the location and status of a page of virtual
address space. When a virtual page is in memory,
the PTE contains the page frame number needed to
map the virtual page to a physical page. When it is
not in memory, the page table entry contains the
information needed to locate the page on secondary
storage (disk) .

position-independent code Code that can execute properly without modification wherever it is located in virtual address space, even if its location is
changed after it has been linked . Generally, this
code uses addressing modes that form an effective
address relative to the PC.
primary key The mandatory key within the data
records of an indexed file; used by VAX-11 RMS to
determine the placement of records within the file
and to bu ild the primary index. See key (indexed
files) and alternate key.

paging The action of bringing pages of an executing process into physical memory when referenced.
When a process executes , all of its pages are said to
reside in virtual memory. Only the actively used
pages, however, need to reside in physical memory.
The remaining pages can reside on disk until they
are needed in physical memory. In this system, a
process is paged only when it references more
pages than it is allowed to have in its working set.
When the process refers to a page not in its working
set, a page fault occurs. This causes the operating
system's pager to read in the referenced page if it is
on disk (and , optionally, other related pages depend ing on a cluster factor), replacing the least recently faulted pages as needed . A process pages
only against itself.
parameter

primary vector A location that contains the starting address of a condition handler to be executed
when an exception condition occurs. If a primary
vector is declared , that condition handler is the first
handler to be executed .
private section An image section of a process that
is not shareable among processes. See also global
section .
privilege See process privilege, user privilege,
and image privilege.
privileged instructions In general, any instructions intended for use by the operating system or
privileged system programs. In particular, instructions that the processor will not execute unless the

See command parameter.

per-process address space
space.

See process address

G- 13

Processor Status Word (PSW) The low-order
word of the Processor Status Longword . Processor
status information includes: the condition codes
(carry, overflow, zero, negative), the arithmetic trap
enable bits (integer overflow, dec imal overflow ,
floating underflow), and the trace enable bit.

current access mode is kernel mode (e.g., HALT,
SVPCTX, LDPCTX, MTPR, and MFPR).

procedure 1. A routine entered via a Call instruction . 2. See command procedure.
process The basic entity scheduled by the system
software that provides the context in which an image
executes. A process consists of an address space
and both hardware and software context.
process address space

process page tables The page tables used to describe process virtual memory.
process priority The priority assigned to a process for scheduling purposes. The operating system
recognizes 32 levels of process priority, where 0 is
low and 31 high. Levels 16 through 31 are used for
time-critical processes. The system does not modify
the priority of a time-critical process (although the
system manager or process itself may). Levels O
through 15 are used for normal processes. The system may temporarily increase the priority of a normal process based on the activity of the process.

See process space.

process context The hardware and software contexts of a process.
process control block (PCB) A data structure
used to contain process context. The hardware PCB
contains the hardware context. The software PCB
contains the software context, which includes a
pointer to the hardware PCB.
process header A data structure that contains the
hardware PCB, accounting and quota information ,
process section table, working set list, and the page
tables defining the virtual layout of the process.

process privileges The privileges granted to a
process by the system, which are a combination of
user privileges and image privileges. They include,
for example, the privilege to: affect other processes
associated with the same group as the user's group,
affect any process in the system regardless of UIC,
set process swap mode, create permanent event flag
clusters, create another process, create a mailbox,
and perform direct 1/0 to a file-structured device.

process header slots That portion of the system
address space in which the system stores the process headers for the processes in the balance set.
The number of process header slots in the system
determines the number of processes that can be in
the balance set at any one time.

process section

process identification (PIO) The operating system's unique 32-bit binary value assigned to a process.

See private section.

process space The lowest-addressed half of virtual address space, where per-process instructions
and data reside . Process space is divided into a program region and a control region.

process 1/0 segment That portion of a process
control region that contains the process permanent
RMS internal file access block for each open file, and
the 1/0 buffers, including the command interpreter's
command buffer and command descriptors.

Program Counter (PC) General register 15 (R15) .
At the beginning of an instruction's execution, the PC
normally contains the address of a location in memory from which the processor will fetch the next instruction it will execute.

process name A 1- to 15-character ASCII string
that can be used to identify processes executing under the same group number.

program locality A characteristic of a program
that indicates how close or far apart the references
to locations in virtual memory are over time. A program with a high degree of locality does not refer to
many widely scattered virtual addresses in a short
period of time.

processor register A part of the processor used
by the operating system software to control the
execution states of the computer system. They include the system base and length registers, the program and control region base and length registers,
the system control block base register, the software
interrupt request register, and many more.

programmer number

Processor Status Longword (PSL) A system programmed processor register consisting of a word of
privileged processor status and the PSW. The privileged processor status information includes: the
current IPL (interrupt priority level), the previous access mode, the current access mode, the interrupt
stack bit, the trace fault pending bit, and the compatibility mode bit.
G-14

See member number.

program region The lowest-addressed half of
process address space (PO space). The program region contains the image currently being executed by
the process and other user code called by the image.
Program region Base Register (POBR) The processor register, or its equivalent in a hardware process control block, that contains the base virtual address of the page table entry for virtual page number
0 in a process program region.

Program region Length Register (POLR) The
processor register, or its equivalent in a hardware
process control block, that contains the number of
entries in the page table for a process program region.

random access by relative record number Retrieval of a record by its relative record
number. See relative record number. For relative
files, random access by relative record number is
synonymous with random access by key. See random access by key (relative files only) .

program section (psect) A portion of a program
with a given protection and set of storage management attributes. Program sections that have the
same attributes are gathered together by the linker
to form an image section.

read access type An instruction or procedure operand attribute indicating that the specified operand
is only read during instruction or procedure execution .

project number
number.

record A set of related data that your program
treats as a unit.

pure code

See group number or account

See re-entrant code.

record access block (RAB) An RMS user data
structure that represents a request for a record access stream . A RAB relates to operations on the records within a file, such as UPDATE, DELETE, or GET.

quadword Eight contiguous bytes (64 bits) starting
on an addressable byte boundary. Bits are numbered from right to left, 0 to 63. A quadword is identified by the address of the byte containing the loworder bit (bit 0). When interpreted arithmetically, a
quadword is a 2's complement integer with significance increasing from bit 0 to bit 62. Bit 63 is used as
the sign bit. The value of the integer is in the range
-2 63 to 2 63 - 1.

record access mode The method used in RMS for
retrieving and storing records in a file. One of three
methods: sequential , random, and record's file address.
record blocking The technique of grouping multiple rcords into a single block. On magnetic tape, an
IRG is placed after the block rather than after each
record. This technique reduces the number of 1/0
transfers required to read or write the data; and, in
addition {for magnetic tape), increases the amount
of usable storage area. Record blocking also applies
to disk files.

qualifier A portion of a command string that modifies a command verb or command parameter by
selecting one of several options. A qualifier, if present, follows the command verb or parameter to
which it applies and is in the format: " /qualifier:option. " For example, in the command string "PRINT
filename/COPIES :3," the COPIES qualifier indicates
that the user wants three copies of a given file printed.

record cell A fixed-length area in a relative file that
can contain a record. The concept of fixed-length
record cells lets VAX-11 RMS directly calculate the
record 's actual position in the file.

queue 1. n. A circular, doubly-linked list. See system queues. v. To make an entry in a list or table,
perhaps using the INSQUE instruction. 2. See job
queue.

record format The way a record physically
appears on the recording surface of the storage medium . The record format defines the method for determining record length.

queue priority The priority assigned to a job
placed in a spooler queue or a batch queue.

record length The size of a record; that is, the
number of bytes in a record.

quota The total amount of a system resource , such
as CPU time, that a job is allowed to use in an accounting period , as specified by the system manager
in the user authorization file. See also limit.

record locking A facility that prevents access to a
record by more than one record stream or process
until the initiating record stream or process releases
the record.

random access by key Indexed files only: Retrieval of a data record in an indexed file by either a
primary or alternate key within the data record . See
key {indexed files) .

Record Management Services A set of operating
system procedures that are called by programs to
process files and records within files. RMS allows
programs to issue READ and WRITE requests at the
record level (record 1/0) as well as read and write
blocks {block 1/0). RMS is an integral part of the
system software. RMS procedures run in executive
mode.

random access by record's file address The retrieval of a record by its unique address, which is
provided to the program by RMS. This method of
access is the only means of randomly accessing a
sequentially organized file containing variable length
records.

record-oriented device A device such as a terminal, line printer, or card reader, on which the largest
G-15

Run Time Procedure Library The collection of
procedures available to native mode images at run
time. These library procedures (such as trigonometric functions, etc.) are common to all native mode
images, regardless of the language processor used
to compile or assemble the program.

unit of data a program can access in one 1/0 operation is the device's physical record.
record's file address The unique address of a
record in a file, which is returned by RMS whenever
a record is accessed, that allows records in disk files
to be access randomly regardless of file organization. This address is valid only for the life of the file. If
an indexed file is reorganized, then the RFA of each
record will typically change.

scatter/gather The ability to transfer in one 1/0
operation data from discontiguous pages in memory
to contiguous blocks on disk, or data from contiguous blocks on disk to discontiguous pages in memory .

re-entrant code Code that is never modified during execution. It is possible to let many users share
the same copy of a procedure or program written as
re-entrant code.

secondary storage
age.

register A storage location in hardware logic other
than main memory. See also general register, processor register, and device register.

secondary vector A location that identifies the
starting address of a condition handler to be executed when a condition occurs and the primary vector
contains zero or the handler to which the primary
vector points chooses not to handle the condition .

register deferred indexed mode An indexed addressing mode in which the base operand specifier
uses register deferred mode addressing.

section A portion of process virtual memory that
has common memory management attributes (protection, access, cluster factor, etc.). It is created from
an image section, a disk file , or as the result of a
Create Virtual Address Space system service. See
global section, private section, image section, and
program section.

register deferred mode In register deferred mode
addressing, the contents of the specified register are
used as the address of the actual instruction
operand.
register mode In register mode addressing, the
contents of the specified register are used as the
actual instruction operand.

self-relative queue A circularly linked list whose
forward and backward links use the address of the
entry in which they occur as the base address for the
link displacement to the linked entry. Contrast with
absolute addresses used to link a queue.

relative file organization The arrangement of records in a file where each record occupies a cell of
equal length within a bucket. Each cell is assigned a
successive number, called a relative record number,
which represents the cell's position relative to the
beginning of the file.

sequential file organization A file organization in
which records appear in the order in which they were
originally written. The records can be fixed length or
variable length.

relative record number An identification number
used to specify the position of a record cell relative
to the beg inning of the file; used as the key during
random access by key mode to relative files .

sequential record access mode Record storage
or retrieval which starts at a designated point in the
file and continues in one-after-the-other fashion
through the file. That is, records are accessed in the
order in which they physically appear in the file.

resource A physical part of the computer system
such as a device or memory, or an interlocked data
structure such as a mutex. Quotas and limits control
the use of physical resources.
resource wait mode An execution state in which a
process indicates that it will wait until a system resource becomes available when it issues a service
request requiring a resource. If a process wants notification when a resource is not available, it can disable resource wait mode during program execution .
return status code

Random access mass stor-

See status code.

shareable image An image that has all of its internal references resolved, but which must be linked
with an object module(s) to produce an executable
image. A sharable image cannot be executed . A
shareable image file can be used to contain a library
of routines. A shareable image can be used to create
a global section by the system manager.
shell process A predefined process that the job
initiator copies to create the minimum context
necessary to establish a process.

RMS-11 A set of routines which is linked with
compatibility mode programs, and provides similar
functional capabilities to VAX-11 RMS. The file organizations and record formats used by RMS-11 are
identical to those of VAX-11 RMS .

signal 1. An electrical impulse conveying information . 2. The software mechanism used to indicate
that an exception condition was detected.
G-16

slave terminal A terminal from which it is not possible to issue commands to the command interpreter. A terminal assigned to application software.
small process A system process that has no control region in its virtual address space and has an
abbreviated context. Examples are the working set
swapper and the null process. A small process is
scheduled in the same manner as user processes,
but must remain resident during its execution.
software context The context maintained by the
operating system that describes a process. See
software process control block (PCB) .
software interrupt An interrupt generated on interrupt priority level 1 through 15, which can be
requested only by software.

state queue A list of processes in a particular
processing state. The scheduler uses state queues
to keep track of processes' eligibility to execute.
They include: processes waiting for a common event
flag, suspended processes, and executable
processes.
status code A longword value that indicates the
success or failure of a specific function. For example, system services always return a status code in
RO upon completion.

software process control block (PCB) The data
structure used to contain a process' software context. The operating system defines a software PCB
for every process when the process is created. The
software PCB includes the following kinds of information about the process: current state; storage address if it is swapped out of memory; unique identification of the process, and address of the process
header (which contains the hardware PCB) . The
software PCB resides in system region virtual address space. It is not swapped with a process.
software priority
priority.

stack pointer General register 14 (R14). SP contains the address of the top (lowest address) of the
processor-defined stack. Reference to SP will access one of the five possible stack pointers (kernel ,
executive, supervisor, user, or interrupt) depending
on the value in the current mode and interrupt stack
bits in the Processor Status Longword (PSL) .

store through

See write through.

strong definition Definition of a global symbol that
is explicitly available for reference by modules linked
with the module in which the definition occurs. The
linker always lists a global symbol with a strong definition in the symbol portion of the map. The librarian
always includes a global symbol with a strong definition in the global symbol table of a library.
strong reference A reference to a global symbol
in an object module that requests the linker to report
an error if it does not find a definition for the symbol
during linking . If a library contains the definition , the
linker incorporates the library module defining the
global symbol into the image containing the strong
reference.

See process priority and queue

spooling output spooling: The method by which
output to a low-speed peripheral device (such as a
line printer) is placed into queues maintained on a
high-speed device (such as disk) to await transmission to the low-speed device. Input spooling: the
method by which input from a low-speed peripheral
(such as the card reader) is placed into queues
maintained on a high-speed device (such as disk) to
await transmission to a job processing that input.

subprocess A subsidiary process created by
another process. The process that creates a subprocess is its owner. A subprocess receives resource
quotas and limits from its owner. When an owner
process is removed from the system, all its subprocesses (and their subprocesses) are also removed.

spool queue The list of files supplied by processes
that are to be processed by a symbiont. For example, a line printer queue is a list of files to be printed
on the line printer.

supervisor mode The third most privileged processor access mode (mode 2). The operating
system's command interpreter runs in supervisor
mode.

stack An area of memory set aside for temporary
storage, or for procedure and interrupt service linkages . A stack uses the last-in, first-out concept. As
items are added to ("pushed on" ) the stack, the
stack pointer decrements. As items are retrieved
from (" popped off" ) the stack, the stack pointer increments.

suspension A state in which a process is inactive,
but known to the system. A suspended process becomes active again only when another process requests the operating system to resume it. Contrast
with hibernation .

stack frame A standard data structure built on the
stack during a procedure call, starting from the location addressed by the FP to lower addresses, and
popped off during a return from procedure. Also
called call frame.

swap mode A process execution state that determines the eligibility of a process to be swapped out
of the balance set. If process swap mode is disabled,
the process working set is locked in the balance set.
G-17

system dynamic memory Memory reserved for
the operating system to allocate as needed for temporary storage. For example, when an image issues
an 1/0 request, system dynamic memory is used to
contain the 1/0 request packet. Each process has a
limit on the amount of system dynamic memory that
can be allocated for its use at one time.

swapping The method for sharing memory resources among several processes by writing an
entire working set to secondary storage (swap out)
and reading another working set into memory (swap
in). For example, a process' working set can be written to secondary storage while the process is waiting
for 1/0 completion on a slow device. It is brought
back into the balance set when 1/0 completes. Contrast with paging.
switch

System Identification Register A processor
register which contains the processor type and serial
number.

See (command) qualifier.

symbiont A full process that transfers record-oriented data to or from a mass storage device. For
example, an input symbiont transfers data from card
readers to disks. An output symbiont transfers data
from disks to line printers.

system image The image that is read into memory
from secondary storage when the system is started
up.
System Length Register (SLR) A processor register containing the length of the system page table in
longwords, that is, the number of page table entries
in the system region page table.

symbiont manager The function (in the system
process called the job controller) that maintains
spool queues, and dynamically creates symbiont
processes to perform the necessary 1/0 operations.

System Page Table (SPT) The data structure that
maps the system region virtual addresses, including
the addresses used to refer to the process page tables. The System Page Table (SPT) contains one
Page Table Entry (PTE) for each page of system region virtual memory. The physical base address of
the SPT is contained in a register called the SBA.

symbol See local symbol, global symbol, and universal global symbol.
Synchronous Backplane Interconnect (SBI) The
part of the hardware that interconnects the processor, memory controllers, MASSBUS adapters, and
the UNIBUS adapter.

system process A process that provides systemlevel functions. Any process that is part of the operating system. See also small process, fork process.

synchronous record operation A mode of record
processing in which a user program issues a record
read or write request and then waits until that request is fulfilled before continuing to execute.

system programmer A person who designs
and/or writes operating systems, or who designs
and writes procedures or programs that provide
general purpose services for an application system.

system In the context "system, owner, group,
world," the system refers to the group numbers that
are used by operating system and its controlling
users, the system operators and system manager.
system address space
system region.

system queue A queue used and maintained by
operating system procedures. See also state
queues.

See system space and

system buffered 1/0 An 1/0 operation, such as
terminal or mailbox 1/0, in which an intermediate
buffer from the system buffer pool is used instead of
a process-specified buffer. Contrast with direct 1/0.

system region The third quarter of virtual address
space. The lowest-addressed half of system space.
Virtual addresses in the system region are shareable
between processes. Some of the data structures
mapped by system region virtual addresses are: system entry vectors, the System Control Block (SCB),
the System Page Table (SPT), and process page tables.

System Control Block (SCB) The data structure in
system space that contains all the interrupt and exception vectors known to the system.

system services Procedures provided by the operating system that can be called by user processes.

System Base Register (SBR) A processor register
containing the physical address of the base of the
system page table.

system space The highest-addressed half of
virtual address space. See also system region.

System Control Block Base register (SCBB) A
processor register containing the base address of
the system control block.
system device The random access mass storage
device unit on which the volume containing the operating system software resides.
G-18

system virtual address A virtual address identifying a location mapped by an address in system
space.
system virtual space

See system space.

task An RSX-11 /IAS term for a process and image
bound together.

mand. The commands typed ahead are not echoed
on the terminal until the command processor is
ready to process them. They are held in a type ahead
buffer.

terminal The general name for those peripheral
devices that have keyboards and video screens or
printers. Under program control, a terminal enables
people to type commands and data on the keyboard
and receive messages on the video screen or printer. Examples of terminals are the LA36 DECwriter
hard-copy terminal and VT100 video display terminal.

unit record device
or line printer.

A device such as a card reader

universal global symbol A global symbol in a
shareable image that can be used by modules linked
with that shareable image. Universal global symbols
are typically a subset of all the global symbols in a
shareable image. When creating a shareable image,
the linker ensures that universal global symbols remain available for reference after symbols have
been resolved.

time-critical process A process assigned to a
software priority level between 16 and 31, inclusive.
The scheduling priority assigned to a time-critical
process is never modified by the scheduler, although it can be modified by the system manager or
process itself.

unwind the call stack To remove call frames from
the stack by tracing back through nested procedure
calls using the current contents of the FP register
and the FP register contents stored on the stack for
each call frame .

timer A system fork process that maintains the
time of day and the date. It also scans for device
timeouts and performs time-dependent scheduling
upon request.

urgent interrupt An interrupt received on interrupt
priority levels 24 through 31. These can be generated only by the processor for the interval clock, serious errors, and power fail.

track A collection of blocks at a single radius on
one recording surface of a disk.
transfer address The address of the location containing a program entry point (the first instruction to
execute).

user authorization file A file containing an entry
for every user that the system manager authorizes to
gain access to the system . Each entry identifies the
user name, password, default account, User Identification Code (UIC), quotas, limits, and privileges assigned to individuals who use the system.

translation buffer An internal processor cache
containing translations for recently used virtual addresses.
trap An exception condition that occurs at the end
of the instruction that caused the exception. The PC
saved on the stack is the address of the next instruction that would normally have been executed. All
software can enable and disable some of the trap
conditions with a single instruction.

user environment test package (UETP) A
collection of routines that verify that the hardware
and software systems are complete, properly installed, and ready to be used.
User File Directory (UFO)

See directory.

User Identification Code (UIC) The pair of numbers assigned to users and to files, global sections,
common event flag clusters, and mailboxes that
specifies the type of access (read and/or write access, and in the case of files, execute and/or delete
access) available to the owners, group, world, and
system. It consists of a group number and a member
number separated by a comma.

trap enables Three bits in the Processor Status
Word that control the processor's action on certain
arithmetic exceptions.
two's complement A binary representation for integers in which a negative number is one greater
than the bit complement of the positive number.
two-way associative cache A cache organization
which has two groups of directly mapped blocks.
Each group contains several blocks for each index
position in the cache. A block of data from main
memory can go into any group at its proper index
position. A two-way associative cache is a compromise between the extremes of fully associative
and direct mapping cache organizations that takes
advantage of the features of both.

user mode The least privileged processor access
mode (mode 3). User processes and the Run Time
Library procedures run in user mode.
user name The name that a person types on a
terminal to log on to the system .
user number

See member number.

user privileges The privileges granted a user by
the system manager. See process privileges.

type ahead A terminal handling technique in
which the user can enter commands and data while
the software is processing a previously entered com-

utility
G-19

A program that provides a set of related

management hardware translates a virtual address
to a physical address. The term "virtual address"
may also refer to the address used to identify a virtual block on a mass storage device.

general purpose functions, such as a program development utility (an editor, a linker, etc.), a file management utility (file copy or file format translation
program), or operations management utility (disk
backup/restore, diagnostic program, etc.).

virtual address space The set of all possible virtual addresses that an image executing in the context
of a process can use to identify the location of an
instruction or data. The virtual address space seen
by the programmer is a linear array of 4,294,967,296
(232) byte addresses.

value return registers The general registers RO
and R1 used by convention to return function values.
These registers are not preserved by any called procedures. They are available as temporary registers
to any called procedure. All other registers (R2,
R3, ... ,R11, AP, FP, SP, PC) are preserved across
procedure calls.

virtual block A block on a mass storage device
referred to by its file-relative address rather than its
logical (volume-oriented) or physical (device-oriented) address. The first block in a file is always virtual
block 1.

variable-length bit field A set of 0 to 32 contiguous bits located arbitrarily with respect to byte
boundaries. A variable bit field is specified by four
attributes: 1) the address A of a byte, 2) the bit position P of the starting location of the bit field with
respect to bit 0 of the byte at address A, 3) the size,
in bits, of the bit field, and 4) whether the field is
signed or unsigned.

virtual 1/0 functions A set of 1/0 functions that
must be interpreted by an ancillary control process.
virtual memory The set of storage locations in
physical memory and on disk that are referred to by
virtual
addresses.
From
the
programmer's
viewpoint, the secondary storage locations appear to
be locations in physical memory. The size of virtual
memory in any system depends on the amount of
physical memory available and the amount of disk
storage used for non-resident virtual memory.

variable-length record format A file format in
which records are not necessarily the same length.
variable with fixed-length control record format Property of a file in which records of variablelength contain an additional fixed control area capable of storing data that may have no bearing on the
other contents of the record. Variable with fixedlength control record format is not applicable to indexed files .

virtual page number
of virtual memory.

The virtual address of a page

volume
Disks: An ordered set of 512-byte blocks. The basic
medium that carries a Files-11 structure.

VAX-11 Record Management Services (VAX-11
RMS) The file and record access subsystem of the
VAX/VMS operating system for VAX. VAX-11 RMS
helps your application program process records
within files, thereby allowing interaction between
your application program and its data.

Magnetic tape: A reel of magnetic tape, which may
contain a part of a file, a complete file, or more than
one file.

volume set

vector 1. A interrupt or exception vector is a storage location known to the system that contains the
starting address of a procedure to be executed when
a given interrupt or exception occurs. The system
defines separate vectors for each interrupting device
controller and for classes of exceptions. Each system vector is a longword. 2. For exception handling,
users can declare up to two software exception vectors (primary and secondary) for each of the four
access modes. Each vector contains the address of
a condition handler. 3. A one-dimensional array.

A collection of related volumes.

wait To become inactive. A process enters a process wait state when the process suspends itself, hibernates, or declares that it needs to wait for an
event, resource, mutex, etc.
wake To activate a hibernating process. A hibernating process can be awakened by another process
or by the timer process, if the hibernating process or
another process scheduled a wake-up call .
weak definition Definition of a global symbol that
is not explicitly available for reference by modules
linked with the module in which the definition occurs.
The librarian does not include a global symbol with a
weak definition in the global symbol table of a library. Weak definitions are often used when creating
libraries to identify those global symbols that are
needed only if the module containing them is otherwise linked with a program.

version number 1. The field following the file type
in a file specification. It begins with a period(.) and is
followed by a number which generally identifies it as
the latest file created of all files having the identical
file specification but for version number. 2. The number used to identify the revision level of program .
virtual address A 32-bit integer identifying a byte
"location" in virtual address space. The memory
G-20

weak reference A reference to a global symbol
that requests the linker not to report an error or to
search the default library's global symbol table to
resolve the reference if the definition is not in the
modules explicitly supplied to the linker. Weak references are often used when creating object modules
to identify those global symbols that may not be
needed at run time.

memory for the process to execute. The remaining
pages of that process, if any, are either in memory
and not in the process working set or they are on
secondary storage.

working set swapper A system process that
brings process working sets into the balance set and
removes them from the balance set.
world In the context "system, owner, group,
world," world refers to all users, including the system
operators, the system manager, and users both in an
owner's group and in any other group.

wild card A symbol, such as an asterisk, that is
used in place of a file name, file type, directory
name, or version number in a file specification to
indicate "all" for the given field.
window

write access type The specified operand of an instruction or procedure is written only during that instruction's or procedure's execution.

See mapping window.

word Two contiguous bytes (16 bits) starting on an
addressable byte boundary. Bits are numbered from
the right, 0 through 15. A word is identified by the
address of the byte containing bit 0. When interpreted arithmetically, a word is a 2's complement integer
with significance increasing from bit O to bit 14. If
interpreted as a signed integer, bit 15 is the sign bit.
The value of the integer is in the range -32,768 to
32,767. When interpreted as an unsigned integer,
significance increases from bit 0 through bit 15 and
the value of the unsigned integer is in the range 0
through 65,535.

write allocate A cache management technique in
which cache is allocated on a write miss as well as on
the usual read miss.
write back A cache management technique in
which data from a write operation to cache are copied into main memory only when the data in cache
must be overwritten. This results in temporary inconsistencies between cache and main memory. Contrast with write through .
write through A cache management technique in
which data from a write operation are copied in both
cache and main memory. Cache and main memory
data are always consistent. Contrast with write back.

working set The set of pages in process space to
which an executing process can refer without incurring a page fault. The working set must be resident in

G-21

Index
4-7

Balance Set

6-19

2-1, 4-17, 4-18

Bandwidth
memory

4-30

Absolute mode addressing
Access modes

Accounting statistics

3-8

ACP (ancillary control process)
6-20, 6-21
Address
manipulation instructions
4-13
mapping registers 4-17
physical 4-17 to 4-22
Address
virtual

4-10,

Base priority
BASIC
PDP-11
VAX-11
7-21

6-18, 6-19

2-1,2-2, 7-1, 7-34, 7-35
1-1,2-1, 2-2, 7-1, 7-14to

Batch processing
3-9,8-1,8-3,8-4
Battery backup

4-17 to 4-22

Addressing modes
Address sort

Bit field

4-1 , 4-5 to 4-7

9-14

Address space
physical 4-18 to 4-20
virtual 4-1 , 4-4, 4-5, 4-18 to
4-20,6-5
Address translation buffer

4-29

Ancillary control process (ACP)
6-20,6-21
Application programming
3-5, 3-10 to 3-14

3-1 to

Arbitration
Memory Interconnect

4-28, 4-29

Argument Pointer (AP)

4-5, 4-7, 4-8

Arguments
definitions 4-7
passing 4-4 , 4-7, 4-8
Arithmetic exceptions

4-8

Communication
between processes 2-2, 3-6, 3-7,
6-12 to 6-14
interprocessor 5-6
network 10-1

2-1 , 2-5, 3-3, 3-6,
2-4, 4-29

5-1 , 5-2 , 6-10, 9-10, 9-11

4-10, 4-13 to

4-29, 4-32

BUS
MASSBUS 2-2, 4-29 , 4-30, 4-33,
5-1, 5-2
Memory Interconnect 2-1, 2-2 ,
4-28 to 4-29
UNIBUS 2-2, 4-27, 4-30, 4-33,
5-1 , 5-2 , 5-5
Byte

4-1 to 4-4
2-1, 4-29, 4-32

Assembler (see also VAX-11 ,
MACRO; (instruction set))
MACR0- 11 (PDP-11) 7-36
VAX-11 MACRO 3-4, 7-33, 7-34

CALL
facility 7-1, 7-5, 7-11, 7-18, 7-23
instructions 4-10, 4-15

Asynchronous system trap
processing services 6-8, 6-11

Card reader

Case instruction

4-10, 4-13, 4-14

Autodecrement addressing
mode 4-5, 4-6

Character data

4-2, 4-4

Autoincrement addresssing
mode 4-5, 4-6
Autoincrement Deferred addressing
mode 4-5, 4-6
Automatic recovery
power failure 3-9
Automatic restart
Backup disks

2-4

Backup files

9-4

Bad block locator
Bad blocks

2-4

Call frame

4-8
5-3

Character string instructions
4- 12
Clocks

2-3, 5-1, 5-2

4-9,

4-1 , 6-2, 6-3

Context switching

COBOL 1- 1, 2-1 , 2-2, 3-4, 7-1 , 7-7
to 7-14, 9-9
Command language
3-4

2-2, 2-5, 3-1 to

Command procedures
3-3, 8- 1, 8-8, 8-9

2-2, 2-5, 3-1

5-3, 5-4

Commercial system
example 3-10, 3-11 , 3-13

4-16

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

Data Communications Facilities 10-1
to 10-8
Data
integrity 2-3
managment facilities 9-1 to 9-17
protection 6-4, 6-5
throughput 4-30, 5-5, 5-6
types
VAX-11 4-1 to 4-4
VAX-11 BASIC 7-15
VAX-11 COBOL 7-9
VAX-11 FORTRAN 7-2
9-12to9-14

DDCMP (DIGITAL Data
Communications Message
Protocol) 5-6, 10-2
Debugger

2-2, 3-6, 7-1 , 8-1, 8-6, 8-7

Debugger
BLISS-32 7-31
COBOL 7-11
FORTRAN 7-7
Decimal data
DECnet

4-2, 4-3

10-1 to 10-7

Default device name

2-1

Command terminal
9-4

Context

DATATRIEVE

Cache memory

6-11

2-1, 4-25, 4-28, 4-31 , 4-32,

CR11 card reader

2-3, 2-4

Branch instructions
4-15

4-8, 6-11, 6-12

Console
5-6, 5-7

Control region

Blocks
bad 2-3, 5-1, 5-2

Buffers

4-4 , 4-8

Connect-to-interrupt

4-4

Bootstrap

Condition codes

Condition handlers

BLISS-32 1-1, 2-1, 2-2, 3-4, 3-5,
6-11, 7-29 to 7-32
Block 1/0

Compatibility mode 2-5 , 3-4, 3-5,
4-1,4-15,4- 16,6-22,6-23, 7-1 , 7-34
to 7-36

Definitions

3-7

G-1 to G-21

Detached process

6-3

Device
drivers 3-13, 4-23, 4-24, 5-1, 6-2 ,
6-11, 6-22, 9-1
independence 3-6, 6-20
name 6-11
Diagnostics
on-line 3-9 , 3-10

peripherals 2-3, 2-4
remote 2-4 , 3-10
DIGIT AL Data Communications
Message Protocol (DDCMP) 5-6,
10-2
DIGIT AL Network Architecture
(DNA) 10-2
Direct memory access (OMA)
devices 5-1 to 5-3, 5-5
Di rectory

3-7, 9-1, 9-2

Disks
backup 2-4, 9-4
supported 5-1 , 5-2
Disk Save and Compress (DSC)
utility 9-4
Displacement addressing mode
4-5,4-6
Displacement Deferred addressing
mode 4-6
Distributed computing network
10-1 , 10-2
OMA (direct memory access)
devices 5-1 to 5-3, 5-5
DMC11 communications link

5-6

2-5, 3-10, 10-2

DR11-B direct memory access digital
interface 5-5

Drivers
device 3-13, 4-23, 4-24, 5-1, 6-2,
6-11 , 6-22, 9-1
DSC (Disk Save and Compress)
utility 9-4
9-7

DZ11 terminal line interface
Edit instruction
Ed itors

5-5

4-9, 4-11, 4-12

2-1 , 8-1 to 8-4

Exceptions
6-12

4-4, 4-8, 4-16, 6-11,

Eception vectors
Fault detection

4-22
2-3

Files
backup 9-4
comparing 9-4

Interprocessor communications
link 5-5, 5-6, 6-14
Interrupts

4- 16, 4- 23

Interrupt vectors

4-22

1/0
controller interfaces 4-29 to 4- 33
device drivers 3- 13, 4-23 , 4-24,
5-1,6-2,6-11,6-22,9-1
processing 6-19 to 6-22
RMS 9-9 to 9-11
space 4-23
system services 6-7, 6-10 to 6-12,
6-20

9-8, 9-9

Known image

6-4

4-10 , 4-13 to 4-15

LA 11 line printer

5-3

6-10

LA 120 hard copy terminal

5-4

FORTRAN
PDP-11 FORTRAN IV IV AX to
RSX 2-1, 2-2, 3-4, 3-5, 7-1, 7-36
VAX-11 FORTRAN 3-4, 7-1 , 7-2to
7-7

LA36 hard copy terminal

5-4

Frame Pointer (FP)

4-5, 4-7 , 4-8

General registers
Global sections
Glossary

Librarian

2-1 , 8-1, 8-7

Libraries

3-6 , 7-5, 7-11 , 7-31 , 7-34

Linker

5-3

2-1 , 8-1 , 8-4 , 8-5

Literal mode addressing

6-14

Local event flag

G-1 to G-21

Hard copy terminal

Languages (see also names of
specific languages) 1-1 , 2-1 , 2-2,
3-4, 3-5, 7-1 to 7-36

Line printers

4- 1, 4-5

Hardware
context 4-16, 4-25
process control block

Logical names
Longword
4-23

Host development mode

4-2, 4-3

LP11 line printers

Immediate mode addressing

MACRO assembler
7-1 , 7-33, 7-34

Index sort

4-7
4-6

7-2,7-9,7-10, 9-Sto

Index instruction
Index register

4-10, 4-12

4-6
4-29, 4-32

Instruction set
compatibilitymode 4-1 , 4-15,
4-16
native mode 4-1 , 4-8 to 4-15
Integer instructions

4-9, 4-11

Interactive terminals

5-3 to 5-5

Interactive text editor

8-1 to 8-3

Interleaving

4-29

5-3
2-1 , 2-2, 3-4,

Macros
BLISS-32 7-31
MACRO 7-34
Magnetic tape

5-2, 5-3

Mailbox 2-2, 3-5, 3-6, 3-11, 6-10,
6-11 , 6-13 , 6-14

9- 14

Instruction buffer

4-10,

LPA11-K direct memory access
controller 5-5

7- 1

4-1 , 6-2, 6-14

lndexedfiles
9-10

6- 11

6- 11 , 9-3, 9-4

Loop control instructions
4-13, 4-14

High performance interface
(DR780) 5-5, 5-6

4-6

6-11

Local event flag clusters

5-4

Indexed addressing mode

Exception condition handling
services 6-8, 6-11, 6-12

2-2,

Floating point
accelerator 2-3, 4-30
data 4-2 , 4-3
instructions 4-9, 4-11

Error Correcting Code (ECC) MOS
memory 2-1 , 2-3, 4-27, 4-32, 5-2
Event flag
clusters 6-11 , 6-13
common 6-13
local 6-11
system services 6-8

Interprocess communication
3-6, 3-7, 6-12 to 6-14

Key
indexed files

Image

6-18

10-7,

Jump instruction

Error
logging 2-4, 3-9
read 5-2

Event

lnternets (protocol emulators)
10-8

Flight simulation
example 3-10, 3-12, 3-13

General register manipulation
instructions 4-10, 4-13

DR780 (High Performance
Interface) 5-5, 5-6

Dynamic access

Files-11 On-Disk Structure Level 2
(ODS-2) 2-2

Foreign volume

DNA (DIGITAL Network
Architecture) 10-2
Down-line loading

directories 9-1
logical names 9-3, 9-4
management 2-2 , 9-1 to 9-4
manipulation
DECnet-VAX 10-2 to 10-4
PDP-11 BASIC-PLUS-2/V AX
7-35
protection 3-6, 3-7
RMS-11 2-5, 6-22, 6-23
sorting 9-4, 9-14to 9-16
specificaions 9-1, 9-2
VAX-11 BASIC 7-17
VAX-11 COBOL 7-9, 7-10
VAX-11 FORTRAN 7-2
VAX-11 RMs 2-2 , 3-5, 6-20 , 7- 1,
7-2, 9-1to9-12

Main memory (see also Memory)
2-1 to 2-4, 4-27 to 4-29, 4-32
Manager
system

3-6 to 3-8

Map-to-1 / 0 page
Mapping

4-18 to 4-22, 6-14, 6-1 5

Mapping registers
MASSBUS
5-1 , 5-2

6-11
4-17

2-2, 4-29, 4-30, 4-33,

Mass storage devices
5-3, 6-10
Mathematics library

2-2, 5-1 to
2-3, 8-6

Memory
bandwidth 4-30
battery backup 2-4, 4-29
interleaving 4-29
main 2-1 to 2-4, 4-29, 4-32
management 2-1 , 4-18 to 4-22,
6-2, 6-14 to 6-18
Management control system
services 6-9, 6-10, 6-17, 6-18
mapping 4-18 to 4-22, 6-14, 6-15
multiported 5-6, 6-14
physical 2-1 to 2-4, 4-29 , 4-32
protection 2-1 , 4-17, 4-18
shared areas 5-6, 6-14
virtual (see also Virtual addressing ;
Virtual memory) 4-17
Modified pages

6-16, 6-17

MOS (Main) memory
4-29, 4-32

2-1 to 2-4,

Multiprogramming
4-16

2-1 , 3-5, 3-6,

Native mode
instruction set 4- 1 , 4-8 to 4-15
programming environment 7-1 to
7-36, 8-1to8-10
Network Ancillary Control Process
(NETACP) 10-3
Network services
to 10-8

2-1 , 2-2, 2-5, 10-1

Network Services Protocol
(NSP) 10-2
Non-processor request (NPR)
devices 5-1 to 5-3
Non-transparent interprocess
communication 10-3
macro 10-6
Octaword

4- 2, 4-3

On-line diagnostics

3-9, 3-10

Operating system
compatibility mode 6-22, 6-23
interprocess communication and
control 6-12 to 6-14
110 processing 6-19 to 6-22
memory management 6-2, 6-14
to 6-18
overview 2-1 to 2-5 , 6-1 to 6-5
process scheduling 6-2, 6-18,
6-19
system services 6-5 to 6-14
user environment 6-5 to 6-14
Operator
system

3-8 to 3-10

Optimizations
VAX-11 FORTRAN
Owner process

7-5 to 7-7

6-3

Packed decimal
data 4-2
instructions 4-9, 4-11
Page
description

2-1 , 6-15 to 6-17

region

fault 6-16
mapping 4-20 to 4-22
tables 4-20, 4-22, 6-14
Paging

2-1 , 3-5, 3-6, 6-15 to 6-17

PDP-11
BASIC-PLUS-2/V AX 2-1 , 2-2,
3-4, 3-5, 7-1 , 7-34, 7-35
Compatibility 2-5, 3-4, 3-5, 4-1 ,
4-15, 4-16, 6-22, 6-23, 7-1 7-34 to
7-36
DATATRIEVE 9-12to9-14
FORTRANIV/VAXtoRSX 2-1 ,
2-2 , 3-4, 3-5, 7-1 , 7-36
instructions 4-1, 4-16
MACR0-11 2-1, 2-2, 3-4, 3-5, 7-1 ,
7-36
Performance

Performance analysis statistics

Per-process space

3-8

2-2, 5-1 to 5-7
4-4, 4-5, 6-6

Physical address

4-18 to 4-20

Physical memory
4-32

2-1to2-4, 4-29,

Position independent code
Power failure

4-7

3-9

Priority 2-1 , 2-2, 2-5, 4-16, 4-17,
4-22, 6-18 , 6-19
Private Pages
Privilege

2-3

3-7, 6-4

Privileged instructions

4-10, 4-18

Procedure
control instructions
definition 4-4

4-10, 4-15

Process
communication 2-2, 3-6, 6-12 to
6-14 , 10-3, 10-4
control blocks 4-23 to 4-25
control system services 6-8, 6-9
context 4- 1
description 3-5, 4-1, 4-16, 6-2 to
6-4
mapping 6-14, 6-15
page table 4-20, 4-22, 4-23
scheduling 6-18, 6-19
virtual address space 4-1, 4-4,
4-5
virtual memory 6-15
Processor
access modes 4-17 , 4-18
description 2-1, 4-1 to 4-33

Programming
interfaces 2-5
languages 1-1 , 2-1 , 2-2, 3-4 , 3-5,
7-1 to 7-36
2-1 , 2- 3, 4-17, 4-18

Protection

Protection code

3-6

Protocol Emulators (lnternets)
10-8

4-2, 4-3

Queue control

3-8, 3-9

Queue instructions

4-10, 4-13

Queue 1/0 Request processing
6-21 , 6-22
Quota

3-7 , 3-8

Random record access mode
Read error

Real-time clock

2-1

Real-time flight simulation
example 3-10, 3-12, 3-13
Real-time Interface Extensions
connect-to-interrupt 6-11
map-to-1/0 page 6-11
Real-time processes
memory management 3-6
priority 2-1, 2-2, 2-5 , 4-16, 4-17
resource allocation 3-7, 3-8, 6-1 ,
6-18,6-19
Record
access modes
RMS 9-6, 9-7
attributes 9-7 to 9-9
processing
RMS 9-9to9-12
Record 1/0

9-9, 9-10

Record Management Services
(RMS) 2-2, 2-5, 3-5, 6-20, 7-1, 7-2,
7-9 , 7-17, 7-23, 9-1to9-12
Record 's File Address (RFA access
mode) 9-7, 9-9
Record sort
Regions

9-14

4-5

Processor Status Longword
(PSL) 4-17

Processor Status Word

Process control system services
6-8, 6-9, 6-12 to 6-14

Registers

Program Counter (PC)

Relative Files

Program
application 3-1 Oto 3-13
development tools 2-2, 3-6, 8-1 to
8-10

9-7

5-2

4-1, 4-6, 4-7

10-7,

PSL (Processor Status
Longword) 4-17

Process-oriented paging (see also
Paging) 2-1

4-8

2-1

Programmed interrupt request
devices 5-1

Quadword

2-3

Peripheral devices

4-5, 6-5, 6-6

Programmable real-time clock

4-1, 4-5
9-5 to 9-7, 9-9, 9-10

Reliability features

2-3, 2-4

Remote diagnostics

2-4, 3-1 O

Resource
allocation 6-4
quota 3-7, 3-8
Resource-sharing network
Restart

10-1

2-4

RFA (Record 's File Address) access
mode 9-7, 9-9
RK06 disk drive

5-2

RK07 disk drive

5-1, 5-2

RL02 disk drive

5-1, 5-2

RM03 disk drive

5-1 , 5-2

RM05 disk drive

5-1 , 5-2

RMS-11

System Control Block
Tag sort
Terminals

RMS (Record Management
Services) 2-2, 2-5, 3-5 , 6-20 , 7-1,
7-2 , 7-9, 7-17, 7-23, 9-1to9-12
5-1, 5-2

RX01 Floppy disk

5-6, 5-7
5-1, 5-2

Scatter/gather transfers

5-2, 5-3

5-3 to 5-5

Terms
definitions

G-1 to G-21

Text editors 8-1
Interactive
EDT 8-1 to 8-3
sos 8-1 ,8-2
Batch
SLP 8-1 , 8-3, 8-4

RSX-11 M (see also PDP-11,
Compatibility) 6-22, 6-23
RX02 disk drive

4-22, 4-24

9-14

TE16 tape storage system

2-5, 6-22, 6-23

RP06 disk drive

System
automatic recovery
power failure 2-4, 3-9
event 6-18
manager 3-6 to 3-8
operator 3-8 to 3-1 O
page table 4-20, 4-21
programming 4-17 to 4-26
region 4-5, 6-5 , 6-6
services 6-5 to 6-14
space 4-5

4-29

Time-of-year clock

2-1

Scheduling 2-1 , 2-2, 2-5, 3-5, 3-6,
6-2, 6-18 , 6-19

Traceback
8-6, 8-7

2-2, 3-6, 7-1 , 8-1 ,

Sequential files

Trace faults

4-8

9-5 to 9-7

Sequential record access mode
9-6, 9-7,9-9
Serial line multiplexer

5-5

Sharable image

8-5

SLP text ed itor

8-1, 8-3, 8-4

Traps

9-14to 9-16

SOS interactive text editor

8-1, 8-2

Special function instructions
4-15

4-10 ,

Spooling
Stack

2-3, 3-8, 3-9
4-4, 4-8

Stack Pointer (SP)

String handling
BASIC-PLUS-2/VAX
Subprocess

Subroutine control instructions
4-10, 4-14, 4-15
Swapping

6-19

Symbolic debugger
8-1 , 8-6, 8-7

2-2, 3-6, 7-1,

Symbolic Traceback Facility
7-7, 7-11 , 7-12

7-1,

4-25 to 4-28

TU58 tape storage system

5-7

TU77 tape storage system

5-2

VAX/VMS
command language 2-2, 2-5, 3-1
to 3-4
DEBUG program 2-2, 3-6, 7-1 ,
8-1 , 8-6,8-7
operating system (see also
Operating system) 2-1to2-5, 6- 1
to 6-23

Type-ahead

5-3

UETP (User Environment Test
Package) 3-10
3-6 to

2-2, 4-27, 4-30, 4-31, 4-33

Unit record peripherals

5-3

User Environment Test Package
(UETP) 3-10

Variable-length bit field
instructions 4-9, 4-12, 4-13

4-22 , 4-24
5-4, 5-5

Virtual addressing
4-17 to 4-22, 6-5

3-1 O, 6-4, 6-5

User identification code (UIC)
3-8, 6-4, 6-5

Vectors

Video terminal

3-1 O

User authorization file

6-3

4-31 to 4-33

5-2

User authorization

7-35

V AX-11 /750

TU45 tape storage system

UNIBUS

4-5, 4-7 , 4-8

5-2

V AX-11 /780

UIC (user identification code)
3-8 , 6-4,6-5

4-4

Stack frame

4-8

TS11 tape storage subsystem

Software process control block
4-23
Sort/MERGE

Transparent interprocess
communication
10-3
macro 10-4

VAX-11
750 Processor 4-31 to 4-33
780 Processor 4-27 to 4-30
BASIC 1-1, 2-1 , 2-2, 3-4, 7-1, 7-14
to 7-21
BLISS-32 1-1 , 2-1 , 2-2, 3-4 , 3-5 ,
7-1 , 7-29 to 7-32
COBOL 1-1 , 2-1, 2-2, 3-4, 7-1 , 7-7
to 7-14
Common language
environment 7-1 , 7-2
CORAL 66 2-1, 2-2, 3-4 , 3-5, 7-1 ,
7-32, 7-33
data types 4-1 to 4-4
Forms Management System
(FMS) 9-16, 9- 17
FORTRAN 1-1 , 2-1 , 2-2, 3-4 , 7-1 ,
7-2 to 7-7
Interactive Symbolic debugger
2-2,3-6, 7-1 , 8-1,8-6,8-7
MACRO 2-1 , 2-2, 3-4, 7-1 , 7-33,
7-34
PASCAL 1-1, 2-1 , 2-2, 3-4, 3-5,
7-1 , 7-26 to 7-29
PL/I 1- 1, 2- 1, 2-2, 3-4, 7-1 , 7-21 to
7-26
Procedure Calling Standard 2-5,
7-1, 7-23
Processor architecture 4-1 to
4-26
Record Management System
(RMS) 2-2, 2-5, 3- 5, 6-20, 7-1,
7-2, 7-9, 7-17, 7-23,9-1 to9-12
RUNOFF 8-10
SORT/MERGE 9-14to9- 16

3-6 to

4-1 , 4-4, 4-5,

Virtual memory
operating system 2-1 to 2-5 , 6-1
to 6-23
process 2-1 to 2-3, 6-2, 6-3, 6-12
programming considerations
6-17 to 6-19
VT100 video terminal
Word

5-4, 5-5

4-2, 4-3

Working set

6-15, 6-16

DIGITAL EQUIPMENT CORPORATION, Corporate Headquarters: Maynard, MA 01754 , Tel. (617) 897-5111 - SALES AND SERVICE OFFICES ;
UNITED STATES - ALABAMA, Birmingham, Huntsville ARIZONA , Phoenix , Tucson ARKANSAS , Little Rock CALIFORNIA, Costa Mesa, El
Segundo , Los Angeles, Oakland, Sacramento, San Diego , San Francisco , Monrovia, Santa Barbara, Santa Clara , Sherman Oaks COLORADO ,
Colorado Springs , Denver CONNECTICUT , Fairfield , Meriden DELAWARE , Newark FLORIDA , Miami , Orlando , Pensacola , Tampa GEORGIA,
Atlanta HAWAII , Honolulu IDAHO , Boise ILLINOIS, Chicago , Peoria INDIANA, Indianapolis IOWA , Bettendorf KENTUCKY , Louisville LOUISIANA , New Orleans MARYLAND , Baltimore MASSACHUSETTS, Boston, Springfield, Waltham MICHIGAN , Detroit, Kalamazoo MINNESOTA,
Minneapolis MISSOURI , Kansas City, St. Louis NEBRASKA , Omaha NEW HAMPSHIRE, Manchester NEW JERSEY, Cherry Hill , Parslppany ,
Princeton , Somerset NEW MEXICO , Albuquerque , Los Alamos NEW YORK , Albany , Buffalo, Long Island, New York City , Rochester , Syracuse,
Westch:?ster NORTH CAROLINA , Chapel Hill, Charlotte OHIO , Cincinnati, Cleveland , Columbus , Dayton OKLAHOMA, Tulsa OREGON , Portland
PENNSYLVANIA, Harrisburg, Philadelphia, Pittsburgh RHODE ISLAND, Providence SOUTH CAROLINA , Columbia , Greenville TENNESSEE ,
Knoxville, Nashville TEXAS , Austin , Dallas , El Paso , Houston , San Antonio UTAH , Salt Lake City VERMONT , Burlington VIRGINIA , Fairfa x,
Richmond WASHINGTON , Seattle, Spokane WASHINGTON D.C. WEST VIRGINIA , Charleston WISCONSIN , Milwaukee INTERNATIONAL EUROPEAN AREA HEADQUARTERS: Geneva , Tel: [41) (22)-93-33-11 INTERNATIONAL AREA HEADQUARTERS : Acton , MA 01754 , U.S.A ., Tel :
(617) 263-6000 AUSTRALIA , Adelaide, Brisbane, Canberra, Hobart, Melbourne, Perth , Sydney, Townsville AUSTRIA , Vienna BELGIUM , Brussels BRAZIL, Rio de Janeiro, Sao Paulo CANADA , Calgary , Edmonton, Halifax, Kingston , London, Montreal , Ottawa , Quebec City , Regina ,
Toronto , Vancouver , Victoria , Winnipeg DENMARK , Copenhagen ENGLAND, Basingstoke, Birmingham , Bristol , Ealing , Epsom , Leeds , Leice ster , London , Manchester, Reading, Welwyn FINLAND, Helsinki FRANCE , Bordeaux , Lyon , Paris , Puteaux , Strasbourg HOLLAND , Amstelveen ,
Delft, Utrecht HONG KONG IRELAND, Dublin ISRAEL , Tel Aviv ITALY, Milan , Rome , Turin JAPAN , Osaka , Tokyo MEXICO , Mexico City ,
Monterrey NEW ZEALAND, Auckland , Christchurch , Wellington NORTHERN IRELAND, Belfast NORWAY , Oslo , PUERTO RICO , San Juan
SCOTLAND , Edinburgh REPUBLIC OF SINGAPORE SPAIN , Barcelona, Madrid SWEDEN, Gothenburg , Stockholm SWITZERLAND, Geneva,
Zurich , WEST GERMANY , Berlin , Cologne , Frankfurt , Hamburg , Hannover, Munich , Nurnberg, Stuttgart