Digital PDFs

EJ-31491-41

May 1988

63 pages

Original

2.9MB

Document:	FORTRAN Compatibility Guide
Order Number:	EJ-31491-41
Revision:	0
Pages:	63
Original Filename:

OCR Text

Compatibility
Guide

FORTRAN Compatibility Guide
Order Number: EJ-31491-41

MAY 1988
This manual documents the compatibility issues that exist between the DIGITAL FORTRAN IV,
FORTRAN-77, and VAX FORTRAN compilers for PRO, PDP-11, and VAX systems.

Revision/Update Information:

This is a new manual.

Software Version:

FORTRAN IV Version 2.6
FORTRAN-77 Version 5.2
V AX FORTRAN Version 4.7

digital equipment corporation
maynard, massachusetts

MAY 1988
The information in this document is subject to change without notice and should not be construed as a
commitment by Digital Equipment Corporation. Digital Equipment Corporation assumes no responsibility
for any errors that may appear in this document.
The software described in this document is furnished under a license and may be used or copied only
in accordance with the terms of such license.
No responsibility is assumed for the use or reliability of software on equipment that is not supplied by
Digital Equipment Corporation or its affiliated companies.
Copyright © 1988 by Digital Equipment Corporation
All Rights Reserved.
Printed in U.S.A.
The postpaid READER'S COMMENTS form on the last page of this document requests the user's critical
evaluation to assist in preparing future documentation.
The following are trademarks of Digital Equipment Corporation:

DEC
DEC/CMS
DEC/MMS
DECnet
DECsystem-10
DECSYSTEM-20
DECUS
DECwriter

DIBOL
EduSystem
lAS
MASSBUS
PDP
PDT
RSTS
RSX

UNIBUS
VAX
VAXcluster
VMS
VT
and the DIGITAL logo

ZK4701

Production Note
This book was produced with the VAX DOCUMENT electronic publishing system, a software
tool developed and sold by DIGITAL. In this system, writers use an ASCII text editor to create
source files containing text and English-like code; this code labels the structural elements of the
document, such as chapters, paragraphs, and tables. The VAX DOCUMENT software, which runs
on the VMS operating system, interprets the code to format the text, generate a table of contents
and index, and paginate the entire document. Writers can print the document on the terminal
or line printer, or they can use DIGITAL-supported devices, such as the LN03 laser printer and
PostScript<:fill printers (PrintServer 40 or LN03R ScriptPrinter), to produce a typeset-quality copy
containing integrated graphics.

<:fill PostScript is a trademark of Adobe Systems, Inc.

Contents
PREFACE

INTRODUCTION

1-1

1.1

LANGUAGE IMPLEMENTATION COMPARISONS

1-1

1.2

DEVELOPING FORTRAN PROGRAMS
1.2.1
Using EDT
1 .2.2
Using VAXTPU
1.2.3
Using VAXLSE
Using EDI
1 .2.4
1.2.5
Using TECO
1 .2.6
Using PROSE-PLUS

1-2
1-4
1-5
1-6
1-6
1-6
1-6

LANGUAGE COMPATIBILITY ISSUES

2-1

PRINCIPLES OF WRITING TRANSPORTABLE CODE
General Principles of Transportability
Speed of Program Execution

2-1
2-1

INCOMPATIBILITIES IN FORTRAN LANGUAGE ELEMENTS
Invoking the Compilers
Compiler Qualifiers and Switches
2.2.2.1
/ A Switch • 2-6
2.2.2.2
/ ANALYSIS_DATA Qualifier • 2-6
2.2.2.3
/CHECK Qualifier • 2-6
2.2.2.4
/CODE:arg Qualifier • 2-7
2.2.2.5
/CONTINUATIONS Qualifier • 2-8
2.2.2.6
/CROSS_REFERENCE Qualifier • 2-8
2.2.2.7
/DEBUG Qualifier • 2-8
2.2.2.8
/DI and /B Switches • 2-9
2.2.2.9
/D_UNES Qualifier • 2-9
2.2.2.10 /DUNES Qualifier • 2-9
2.2.2.11 /DMl Qualifier • 2-9
2.2.2.12 /EXTEND Qualifier • 2-10
2.2.2.13 /EXTEND_SOURCE Qualifier. 2-10
2.2.2.14 /FOR Qualifier • 2-10
2.2.2.15 /F77 Qualifier • 2-10
2.2.2.16 /G_FlOATING Qualifier. 2-10
2.2.2.17 /IDENTIFICATION Qualifier • 2-11
2.2.2.18 /14 Qualifier • 2-11
2.2.2.19 /lA Switch • 2-11
2.2.2.20 /LiBRARY Qualifier • 2-11
2.2.2.21 /UNE_NUMBERS Qualifier • 2-11
2.2.2.22 lUST Qualifier • 2-12
2.2.2.23 flO and /Q Switches • 2-13
2.2.2.24 /MACHINE_CODE Qualifier • 2-13
2.2.2.25 /MAP Qualifier • 2-13
2.2.2.26 /N Switch • 2-14
2.2.2.27 /0 Switch. 2-14

2-2
2-3

CHAPTER 1

CHAPTER 2
2.1

2.1 .1
2.1.2
2.2

2.2.1
2.2.2

2-2

2-4

Contents

2.2.3
2.2.4
2.2.5

2.2.6

2.2.7
2.2.8
2.2.9
2.2.10
2.3

2.2.2.28 /OBJECT Qualifier • 2-14
2.2.2.29 /OPTIMIZE Qualifier • 2-14
2.2.2.30 /R Switch • 2-14
2.2.2.31 /SHAREABLE Qualifier • 2-15
2.2.2;32 /SHOW Qualifier • 2-15
2.2.2.33 /SOURCE Qualifier· 2-16
2.2.2.34 /SP Switch • 2-16
2.2.2.35 /STANDARD Qualifier • 2-16
2.2.2.36 /TRACEBACK:arg Qualifier • 2-17
2.2.2.37 /U Switch • 2-18
2.2.2.38 /VECTORS Qualifier· 2-18
2.2.2.39 /WARNINGS Qualifier • 2-18
2.2.2.40 /WORK_FILES:n Qualifier • 2-19
2.2.2.41 /X Switch • 2-19
Compiler Control Statements
Syntax and Format
Statements
2.2.5.1
Assigned GO TO Label Lists • 2-21
2.2.5.2
DO Loop Minimum Iteration Count • 2-21
2.2.5.3
EXTERNAL Statement • 2-22
2.2.5.4
Blank Common Program Section (.$$$$.) • 2-22
2.2.5.5
X Format Edit Descriptor • 2-22
Subroutines
2.2.6.1
ASSIGN Subroutine • 2-24
2.2.6.2
CLOSE Subroutine • 2-24
2.2.6.3
ERRSET Subroutine • 2-25
2.2.6.4
ERRTST Subroutine • 2-26
2.2.6.5
FDBSET Subroutine • 2-26
2.2.6.6
IRAD50 Subroutine • 2-27
2.2.6.7
RANDU Subroutine • 2-28
2.2.6.8
R50ASC Subroutine. 2-29
2.2.6.9
USEREX Subroutine • 2-29
Functions
2.2.7.1
RAD50 Function • 2-30
2.2.7.2
RAN Function • 2-30
Data Definitions
2.2.8.1
Floating-Point Results • 2-31
2.2.8.2
Character and Hollerith Constants • 2-31
Expressions
Character Sets

SYSTEM DEPENDENCIES
2.3.1
I/O Differences and File Transfer
2.3.2
Optimization of I/O with Operating System-Specific Features
2.3.3
File Naming Conventions
2.3.4
Transportable File Specification Format
2.3.4.1
Using Logical Names • 2-36
2.3.4.2
File Specifications in String Variables • 2-36
File Support
2.3.5

2-19
2-20
2-20

2-23

2-29
2-31
2-32
2-34
2-34
2-34
2-35
2-35
2-35
2-37

Contents
2.3.6

2.3.7
2.3.8
2.3.9
2.4
2.5

Keywords
2.3.6.1
OPEN Statement BLANK Keyword Default • 2-37
2.3.6.2
OPEN Statement STATUS Keyword Default • 2-38
2.3.6.3
OPEN Statement INITIALSIZE Keyword • 2-38
2.3.6.4
DISPOSE = 'PRINT' Specification • 2-38
Record Management Services
Block I/O
Run-Time Libraries

2-39
2-39
2-39

PROGRAM SEGMENTATION

2-39

PROCEDURE CALLING
Calling Function Subprograms
Calling Subroutine Subprograms

2-40
2-40
2-40

ERRORS AND ERROR HANDLING
Run-Time Library Error Numbers
Error Handling and Reporting
2.6.2.1
Continuing After Errors • 2-42
2.6.2.2
I/O Errors with lOST AT or ERR Specified • 2-42
2.6.2.3
OPEN or CLOSE Statement Errors • 2-42

2-41
2-41
2-42

Commands for Program Development

1-3
1-4
2-2
2-33
2-36

2.5.1
2.5.2
2.6

2-37

2.6.1
2.6.2

INDEX
FIGURES
1-1
1-2
2-1
2-2
2-3
TABLES
2-1
2-2
2-3
2-4
2-5
2-6
2-7
2-8

T ext Editors
Layering of the Various FORTRAN Compilers
Logical Test Comparison
File Specification Logical Name Comparison

Invoking a FORTRAN Compiler
FORTRAN Command Qualifiers and Switches
FORTRAN Subroutines
Character Set Incompatibilities
Default Logical Unit Numbers
Logical Name Capabilities for File Specifications
Supported Files for DIGITAL FORTRAN Compilers
Incompatible Error Numbers

2-3
2-4
2-23
2-34
2-34
2-36
2-37
2-41

vii

Preface

This manual provides reference information for transporting FORTRAN programs from FORTRAN
IV to FORTRAN-77 and VAX FORTRAN, and from FORTRAN-77 to VAX FORTRAN. This
manual is intended to augment, not replace, the information contained in the various FORTRAN
compiler documentation sets.

I ntended Audience
This manual is intended for programmers who are transporting their programs from FORTRAN IV
to FORTRAN-77 and VAX FORTRAN, and from FORTRAN-77 to VAX FORTRAN.

Document Structure
This manual has two chapters. They are as follows:
Chapter 1, Introduction, describes the relationship of the various FORTRAN implementations
to the different FORTRAN standards. It also describes the process of developing FORTRAN
programs on VMS, RSX, RSTS/E, and P lOS systems.
Chapter 2, Language Compatibility Issues, describes in detail all the FORTRAN language features
that cause transportability problems across VMS, RSX, RSTS/E, and P lOS systems.

Associated Documents
For more information on topics discussed in this manual, refer to the appropriate manuals in the
FORTRAN documentation sets.

Preface

Conventions
Convention

Meaning
In examples, a key name (usually abbreviated) shown within a box
indicates that you press a key on the keyboard; in text, a key name
is not enclosed in a box. In this example, the key is the RETURN
key. (Note that the RETURN key is not usually shown in syntax
statements or in all examples; however, assume that you must press
the RETURN key after entering a command or responding to a
prompt.)

CTRL/C

A key combination, shown in uppercase with a slash separating two
key names, indicates that you hold down the first key while you
press the second key. For example, the key combination CTRL/C
indicates that you hold down the key labeled CTRL while you
press the key labeled C. In examples, a key combination is enclosed
in a box.

$ SHOW TIME

In examples, system output (what the system displays) is shown in
black. User input (what you enter) is shown in red.

05-JUN-19BB 11:55:22
$ TYPE MYFILE.DAT

In examples, a vertical series of periods, or ellipsis, means either
that not all the data that the system will display in response to a
command is shown or that not all the data you enter is shown.

input-file, ...

In examples, a horizontal ellipsis indicates that additional parameters, values, or other information can be entered, that preceding
items can be repeated one or more times, or that optional arguments in a statement have been omitted.

[logical-name]

Brackets indicate that the enclosed item is optional. (Brackets are
not, however, optional in the syntax of a directory name in a file
specification, or in the syntax of a substring specification in an
assignment statement.)

quotation marks
apostrophes

The term quotation marks is used to refer to double quotation
marks ("). The term apostrophe (') is used to refer to a single
quotation mark.

Chapter 1
Introduction

This chapter discusses the relationships of the various FORTRAN language implementations to
each other and to any existing standards, for example the American National Standards. The
DIGITAL FORTRAN implementations are designed to comply with the current FORTRAN ANSI
standard; however, each implementation also includes extensions. This chapter highlights these
extensions.
This chapter also discusses the tools and methods of program development on the following
systems: PRO, PDP-II, and VAX.

1.1

Language Implementation Comparisons
Both VAX FORTRAN and FORTRAN-77 are based on American National Standard FORTRAN-77
(ANSI X3.9-1978). This standard corresponds to the ISO standard for FORTRAN-77 (ISO 15391980 (E» and the FIPS 69-1 standard. FORTRAN IV conforms to the previous FORTRAN standard
(ANSI X3.9-1966); however, there is upward compatibility with the more recent standard.
VAX FORTRAN, FORTRAN-77, and FORTRAN IV all support some common extensions to the
standards on which they are based. For example, all three compilers support tab-format lines.
That is, instead of using a fixed format, you can specify the statement label field, the continuation
indicator field, and the statement field using tab formatting. All three compilers support Hollerith
constants and other additional data types.
VAX FORTRAN and FORTRAN-77 support the following additional extensions to the ANSI
X3.9-1978 standard:
•
•
•
•
•
•

Indexed file organizations
Relative file organization
Exclamation point (!) as a comment indicator for comments at the end of a line of source code
Letter D in column 1 to indicate a debugging statement (for use with the jD_LINES qualifier)
Arithmetic expressions as control parameters in computed GO TO statements
INCLUDE statements to incorporate statements from a separate file into a FORTRAN program
during compilation

VAX FORTRAN also supports the following extensions to the ANSI X3.9-1978 standard that are
not supported by FORTRAN-77:
•

Conformance with the VAX procedure-calling standard

Introduction

1-1

•
•
•
•
•
•

Records and structures
DO WHILE statement
Namelist-directed input and output
IMPLICIT NONE statement
Support for double-precision complex (D-floating and GJIoating), REAL*16 (H-floating),
and G -floating double-precision data types
Intrinsic functions to manipulate the bits in binary patterns that represent integers

Both VAX FORTRAN and FORTRAN-77 provide a ISTANDARD qualifier to the FORTRAN
compile command. This standard causes the compiler to flag any extensions to the ANSI
X3.9-1978 standard. For more information, see Chapter 2.

1 .2

Developing FORTRAN Programs
To develop FORTRAN programs on VMS, RSX-IIM-PLUS, RSTS/E, and P lOS systems, you
must invoke a text editor to create the source program and then compile, link, and run it.
Figure 1-1 shows this process.

1-2

Introduction

Figure 1-1:

Commands for Program Development

ACTION

INPUT jOUTPUT FILES

CREATE A SOURCE
PROGRAM

TEST.FTN
(FORTRAN-77 and
FORTRAN IV on RSX)
TEST.FOR
(V AX FORTRAN and
FORTRAN IV on RSTSjE)

COMPILE THE
SOURCE PROGRAM

TEST.OBJ

LINK THE OBJECT
MODULE

TEST.TSK
(FORTRAN IV and
FORTRAN-77)
TEST.EXE (VAX FORTRAN)

RUN THE
EXECUT ABLE IMAGE

ZK-6612-HC

Introduction

1-3

Different ·operating systems allow you to use different text editors. For compatibility, it is recommended that you use EDT, since EDT is available on most operating systems. Figure 1-2
illustrates all of the text editors that are available on PRO, PDP-II, and VAX systems. The following sections briefly describe these editors. For more information, see the appropriate operating
system compatibility guide.
Figure 1-2:

Text Editors

EVE

EDI

TEeD

PRDSEPLUS

LSE
ZK-6567-HC

1 .2.1

Using EDT

EDT is an interactive, general-purpose text editor that offers three editing modes: keypad, nokeypad, and line. With keypad mode, you issue commands by using the numeric keypad that appears
to the right of your main keyboard. With nokeypad mode, you issue commands on a command
line, which EDT processes when you press the RETURN key. Line mode focuses on the line as
the unit of text. With line mode, you issue commands at the line mode asterisk prompt ( * ).
Keypad mode and nokeypad mode continually display the contents of the file on your screen.
When you begin your editing session, editing in line mode is the default. Unlike keypad and
nokeypad mode, line mode is not suitable for editing text arranged in paragraphs.

1-4 Introduction

The following command line invokes the EDT editor and creates the file PROG1.FTN:
$ EDIT/EDT PROG1.FTN

To change from line mode to keypad mode, type the CHANGE command, or simply C, at the
asterisk prompt. To return to line mode from keypad mode, press CTRL/Z. To change from
line mode to nokeypad mode, type the SET NOKEYP AD command and then type the CHANGE
command at the asterisk prompt.
When you invoke EDT to create a file, a journal file is automatically created. You can use this
journal file to recover your edits if the system fails during an editing session. To recover your
edits, type the EDIT/RECOVER command specifying the name of the file you were editing
originally, not the name of the journal (.JOU) file.
EDT provides an online HELP facility that you can access during an editing session. In line
mode, type the HELP command. EDT displays general information on EDT as well as detailed
information on both line mode editing and nokeypad mode editing. In keypad mode, press the
HELP key or the PF2 key. EDT displays a keypad diagram on your screen and a list of keypad
editing keys. For help on a specific keypad function, press the key you want help on.
For more detailed information on how to use EDT, see the EDT documentation set for the appropriate operating system or the VAX EDT Reference Manual.

1.2.2

Using VAXTPU

The VAX Text Processing Utility (VAXTPU) is a high-performance, programmable utility that is
only available on VMS systems. VAXTPU provides the Extensible VAX Editor (EVE). You can also
create your own interfaces.
EVE is an interactive text editor that allows you to execute editing functions using either the EVE
keypad or the EVE command line. The following command line invokes the EVE editor and
creates the file PROG_1.FOR:
$ EDIT/TPU PROG_1.FOR

You can define a global symbol for the EDIT /TPU command by placing a symbol definition in
your LOGIN .COM file. For example:
$ EVE == "EDIT/TPU"

Once this command line is executed, you can type EVE at the DCL prompt followed by the name
of the file you want to modify or create.
For more information on using the advanced features of EVE, see the Guide to VMS Text Processing.
Like EDT, VAXTPU provides you with an online HELP, facility that you can access during your
editing session. For help on the keypad, press the PF2 key. For general help on VAXTPU commands, press the DO key and enter HELP. For help on a specific command, press the DO key and
enter the name of the command you want help on.
Like EDT, when you invoke VAXTPU to create a file, a journal file is automatically created. You
can use this journal file to recover your edits if the system fails during an editing session. To
recover your edits, include the /RECOVER qualifier on the command line that invokes the editor.
Unlike EDT, however, VAXTPU supports multiple windows and buffers. This feature allows you
to view two files on your screen at the same time, or remote parts of the same file.

Introduction

1-5

1.2.3

Using VAXLSE

The VAX Language-Sensitive Editor (VAXLSE) is available only with VMS and can be used
as a multiwindow, screen-oriented text editor. Users familiar with EDT or EVE can use the
corresponding keypads with VAXLSE.
One of the major features of VAXLSE is that it provides language templates that supply the correct
format and syntax for whatever programming language you are currently using. For example, by
typing the following line at the DCL command prompt, you initiate a VAXLSE editing session that
includes the appropriate template for VAX FORTRAN:
$ LSEDIT PROG_2.FOR

Within this editing session. you can code a syntactically correct VAX FORTRAN program without
knowing anything about FORTRAN syntax. VAXLSE allows you to compile your source programs
and review and correct any errors, all without ever leaving the editing session. In addition,
VAXLSE also allows you to create templates for languages that you define.

1 .2.4

Using EDI

The EDI text editor lets you work with text one line at a time. The editor has two control modes:
edit mode and input mode. In edit mode, you issue commands to change text already created. In
input mode, you type text that goes into the file.
If you use EDI on RSX, and want to move a line editor on VMS, your best choice is to move to
EDT and use the line editing mode there. Some of the commands are the same, but not all. For
example, both EDI and EDT have INSERT and DELETE commands that work the same. However,
the LOCATE command in EDI is the same as the FIND command in EDT. In general, EDT's line
editing mode offers more capabilities than ED!.

For more information on the EDI editor, see the RSX-llM-PLUS Utilities Manual.

1.2.5

Using TECO

TECO is a powerful text editing program that is supported on RSTSIE systems. You may use
TECO to edit any form of ASCII text such as source programs. Since TECO is a character-oriented
editor rather than a line editor, text edited with TECO does not have line numbers associated with
it, and it is not necessary to replace an entire line of text to change one character. Because TECO
is not supported on other systems, it is recommended that you not use it if compatibility is an
issue.

1.2.6

Using PROSE-PLUS

PROSE-PLUS is a compound document available only on P lOS systems. This editor enables you
to combine both text and graphics into a single document. PROSE-PLUS immediately displays the
effects of all text editing functions such as bold, underline, and right-margin 'justification. It also
provides the ability to display the combined text and graphics document prior to printing.
To create drawings, you can either use the picture mode provided by PROSE-PLUS, or you can
insert GIDIS format graphic files that were created by other applications.
PROSE-PLUS also features an interactive spelling checker with a base dictionary of over 60,000
words, along with a user dictionary for proper names and acronyms.

1-6

Introduction

Chapter 2
Language Compatibility Issues

This chapter discusses the compatibility issues that exist between VAX FORTRAN, PDP-II
FORTRAN IV, PDP-II FORTRAN-77, and PRO/Tool Kit FORTRAN-77. Compatibility issues
generally deal with features that are available only on a particular system. However, they can also
deal with inconsistencies in the behavior of certain FORTRAN language elements, depending on
which FORTRAN language implementation you are using.

2.1

Principles of Writing Transportable Code
Writing FORTRAN code that runs on PRO, PDP-II, and VAX systems is an effort well spent.
The techniques for writing transportable FORTRAN code coincide with many of the techniques
of good programming practice, so that your transportable code will be understandable and easily
maintained. Additionally, the effort involved in writing transportable code is small compared to
the effort required to convert a program designed without considering transportability.
The following sections describe some programming practices to use when designing transportable
FORTRAN programs. These practices may help you avoid some of the common pitfalls encountered when running FORTRAN programs on PRO, PDP-II, and VAX systems.

2.1 .1

General Principles of Transportability

There are two important principles to remember when writing transportable code:
1. Modularize the code
2. Isolate system dependencies
Modularizing the code is important because it gives you logical, functional program units that
are easy to create and maintain. Additionally, if you break your program into manageable units,
you can isolate system dependencies in a single module, so that only that module is affected
when you need to run your program on a different system. Although this makes maintenance
and transportability of system-specific code easier, it is still a good idea to avoid using any
system-specific features when writing transportable programs unless absolutely l)ecessary.

Language Compatibility Issues

2-1

2.1.2

Speed of Program Execution

Programs that use system-specific functions or capabilities generally execute faster than those
that do not, because system-specific capabilities are designed to make the best possible use of the
system hardware and software or both. However, in gaining execution speed, your program must
sacrifice transportability. You must weigh both of these factors when determining whether or not
to use system-specific capabilities in your applications.

2.2

Incompatibilities in FORTRAN Language Elements
PDP-II FORTRAN IV is a compatible subset of PRO/Tool Kit and PDP-II FORTRAN-77, which
is a compatible subset of VAX FORTRAN.
NOTE

The implementation of FORTRAN-77 for the PDP-II and the PRO are almost identical,
so throughout this chapter both of these compilers are referred to as FORTRAN-77.
Because of this layering of the various compilers, any FORTRAN-77 program not using superset
(extension) features will generally run correctly under FORTRAN IV, and any FORTRAN-77 or
FORTRAN IV program will run correctly under VAX FORTRAN, because VAX FORTRAN is an
extension of PRO and PDP-II FORTRAN. Similarly, any VAX FORTRAN program not using
any language extensions can run under either FORTRAN IV or FORTRAN-:-77. Remember that
FORTRAN IV is based on a previous standard, so any incompatibilities in the standards will cause
corresponding incompatibilities between FORTRAN IV, FORTRAN-77, and VAX FORTRAN.
Figure 2-1 illustrates the layering of the various FORTRAN compilers.
Figure 2-1 :

Layering of the Various FORTRAN Compilers

VAX FORTRAN
PRO and PDP-11 FORTRAN-77

PDP-11 FORTRAN IV

I
ZK-6569-HC

2-2

Language Compatibility Issues

To run FORTRAN IV or FORTRAN-77 on VMS systems, you must install VAX-II RSX. Under
VAX-II RSX, both FORTRAN IV and FORTRAN-77 support the transporting of object code to
RSX systems.
The following restrictions apply when you are running either PDP-II FORTRAN compiler on a
VMS system under VAX-II RSX:
•
•

FORTRAN IV only supports the linking and running of programs that do not use virtual
arrays.
FORTRAN-77 does not support the running of tasks.

Aside from the FORTRAN language extensions, the compatibility of your PRO, PDP-II, and VAX
FORTRAN programs may also be affected by the following:
•
•
•

2.2.1

Differences between the hardware architectures of PRO, PDP-II, and VAX systems
Differences between the P lOS, RSX-ll, RSTS/E, and VMS operating systems
Differences between the standards on which the various language implementations are based

Invoking the Compilers

The method used to invoke the various FORTRAN compilers varies based on the following factors:
•
•
•

The FORTRAN compiler itself
The operating system
The command line interface

Table 2-1 illustrates the different methods used to invoke the various FORTRAN compilers.
Table 2-1:
Compiler
FORTRAN IV

Invoking a FORTRAN Compiler
CLI

Command Line

RSTS/E

DeL

$ RUN $FORTRAN 1
FOR> [obj] [, list][jsw] = filel[jsw]

RSX

DeL

$ FORTRAN/FOR[jqual] file[jqual]

MeR

> FOR
FOR> [obj] [, list][jsw] = filel[jsw]

RSX, P lOS, and
RSTS/E

DeL

$ FORTRAN/F77[jqual] file[jqual]

RSX, RSTS /E, and
VAX-II RSX

MeR

> F77
F77> [obj] [, list][jsw] = filel[jsw]

VMS

DeL

$ FORTRAN[j qual] file

System

$ FORTRAN /FOR[j qual] file[j qual]2

FORTRAN-77

VAX FORTRAN

IThis command invokes the RSTSjE FORTRAN IV compiler as if under RT-ll.
2This command invokes the RSTSjE FORTRAN IV compiler as if under RSX.

Language Compatibility Issues

2-3

NOTE

Any notation used for FORTRAN-77 and FORTRAN IV on RSX systems can be used
under VAX-11 RSX.

2.2.2

Compiler Qualifiers and Switches

FORTRAN command qualifiers influence the way in which the compiler processes your file. In
many cases, the simplest form of the FORTRAN command is sufficient. However, you can select
appropriate optional qualifiers if you need special processing.
FORTRAN IV, FORTRAN-77, and VAX FORTRAN support many similar FORTRAN DCL command qualifiers. FORTRAN IV and FORTRAN-77 also support switches, many of which correspond to a particular DCL command qualifier. If you have VAX-11 RSX installed on a VMS
system, you can use these switches when compiling a FORTRAN IV or FORTRAN-77 program.
Table 2-2 lists the DCL qualifiers and the corresponding switches, and indicates their availability
for the various FORTRAN implementations.
NOTE

The switches listed in Table 2-2 and in the subsequent sections consist of MCR
switches. These switches are available when you invoke FORTRAN IV on RSX systems
or FORTRAN-77 using the MCR command line interface. RSTSjE systems do not
support the MCR command line interface; however, you can implement the specified
switches for interactive compilations by using the FORTRAN IV compiler on RSTS jE
systems. That is, once you have invoked the FORTRAN IV compiler on RSTSjE systems, you. can use the specified switches as if you were invoking the compiler using the
MCR command line interface.
Table 2-2:
Qualifier

FORTRAN Command Qualifiers and Switches
FORTRAN IV
FORTRAN-77
Switch
jA2

j ANALYSIS_DATA

VAX FORTRAN

Yes

jCHECK

jCK

Yes

jCODE

JCD 1, j12

Yes

jCONTINUAT1ONS

JCO

Yes

jCROSS-REFERENCE

Yes

jDEBUG

Yes

Yes3

Yes

JDB
jD1l, jB2

jD_LINES

1 For FORTRAN IV, this switch is available only on RSX systems.

2-4

Language Compatibility Issues

Table 2-2 (Cont.):

FORTRAN Command Qualifiers and Switches

Qualifier

Switch

FORTRAN IV

FORTRAN-77

VAX FORTRAN

jDUNES

jDEI, jD2

Yes

No 4

Yes

No 4

Yes

jDML
jEXTEND

JEX , jE2

jEXTEND_SOURCE

jFOR
jF77

jF77

jG_FLOATING

Yes

jlDENTIFICATION

jlDl

Yes

No6

jl4

jl41, T2

Yes

JLA

Yes

jUBRARY
jUNE-NUMBERS

jSN l , jS2,7

Yes

JUST

JUI, jL2

Yes

Yes 3

Yes

jMACHINE_CODE

JLOl, jQ2,7
jU l , jL2

Yes

Yes3

JUI, jL2
jN 2

Yes

No 4

Yes

j02

Yes

jMAP

jOBJECT
jOPTIMIZE

Yes

Yes s

Yess

Yes 4

jU l , jL2

Yes

jSpl

Yes

jSTANDARD

JST

Yes

jTRACEBACK

JTR

Yes

jSHAREABLE

JOP
jR2
l

JRO , jZ2

jSHOW
jSOURCE

1 For FORTRAN IV, this switch is available only on RSX systems.

2For FORTRAN IV, this switch is available only on RSTS/E systems.
3This qualifier exhibits some minor differences, depending on which FORTRAN language implementation you are using. Refer
to the appropriate qualifier section for more information.
4The effects of this qualifier can be implemented in VAX FORTRAN using another qualifier. Refer to the section describing
this qualifier for more information.
sThis qualifier is available only on the RSX-llM-PLUS operating system.
6This qualifier or switch is not provided on certain systems because the specified behavior is performed by default.
7This qualifier is not synonymous with the other associated qualifiers or switches. Instead, it inhibits the behavior specified by
the associated qualifiers.

Language Compatibility Issues 2-5

Table 2-2 (Cont.):

FORTRAN Command Qualifiers and Switches

Qualifier

Switch

FORTRAN IV

FORTRAN-77

VAX FORTRAN

Yes

jVECTORS

jU2
JVAI, jV 2,7
jWRI, jW2
jWF
jX2

jWARNINGS
jWORKJILES

Yes

Yes 3

Yes

I For FORTRAN IV, this switch is available only on RSX systems.

2For FORTRAN IV, this switch is available only on RSTSjE systems.
3This qualifier exhibits some minor differences, depending on which FORTRAN language implementation you are using. Refer
to the appropriate qualifier section for more information.
7This qualifier is not synonymous with the other associated qualifiers or switches. Instead, it inhibits the behavior specified by
the associated qualifiers.

The following sections discuss each qualifier or switch, which system or systems it applies to,
and any inconsistencies in the various interpretations. Typically, each section refers to the DCL
command qualifier but also discusses any corresponding switches. Switches that do not correspond
to a particular DCL qualifier are discussed in separate sections.
2.2.2.1

/A Switch

The I A switch is valid only for FORTRAN IV on RSTS IE systems. It causes the compiler to print
compilation statistics. The default behavior for the I A switch is determined when FORTRAN IV is
installed on your RSTS/E system.
2.2.2.2

/ANALYSIS_DATA Qualifier

The IANALYSIS_DATA qualifier is valid only for VAX FORTRAN. This qualifier produces a file
that contains analysis data about the source code being compiled. The default is
INOANALYSIS_DATA. The source-code analysis files created when you specify this qualifier can
be used with such products as the VAX Source Code Analyzer.
2.2.2.3

/CHECK Qualifier

The ICHECK qualifier is valid for FORTRAN-77 and VAX FORTRAN, but it is interpreted
differently depending on which FORTRAN compiler you are using. The corresponding switch is
ICK.
For FORTRAN-77, the ICHECK qualifier or ICK switch tells the compiler to generate code to
check that all your array references are within the array address bounds. The compiler does not
check individual subscripts against the dimension specifications.
The VAX FORTRAN ICHECK qualifier has the format ICHECK=arg, where arg can have anyone
of the following values:
•

2-6

BOUNDS

Language Compatibility Issues

The BOUNDS argument provides similar capabilities to the FORTRAN-77 jCHECK qualifier.
If you specify jCHECK=BOUNDS, the VAX FORTRAN compiler checks array and substring
references to ensure that they are within the address boundaries specified in the array or
character variable declaration.
For array bounds, only the address reference is checked. The system only checks to determine
whether or not you are in the same array; it does not check each individual dimension. Array
bounds checking is not performed for arrays that are dummy arguments in which the last
dimension bound is specified as an asterisk (*), or both upper and lower dimensions are 1.
•

OVERFLOW
The OVERFLOW argument specifies that the VAX FORTRAN compiler checks all BYTE,
INTEGER*2, and INTEGER*4 calculations for arithmetic overflow. Real and complex calculations are always checked for overflow and are not affected by jNOCHECK. Integer
exponentiation is performed by a routine in the mathematics library; this routine always
checks for overflow, even if jCHECK=NOOVERFLOW is specified.

•

UNDERFLOW
The UNDERFLOW argument tells the VAX FORTRAN compiler to check all real and complex
calculations for floating underflow.

•

ALL
The ALL argument specifies that all OVERFLOW, UNDERFLOW, and BOUNDS checks be
performed by the VAX FORTRAN compiler.

•

NONE
The NONE argument specifies that no checks be performed.

The default for FORTRAN-77 is jNOCHECK or jNOCK, meaning that no checks are performed.
In VAX FORTRAN, the default is jCHECK=(NOBOUNDS, OVERFLOW, NOUNDERFLOW).
2.2.2.4

/CODE:arg Qualifier

The jCODE:arg qualifier is valid only for FORTRAN IV. On RSX systems, the jCD:arg switch
provides the same capability as the DCL jCODE qualifier. On RSTSjE systems, use the jI:arg
switch.
These qualifiers and switches allow you to indicate that the code to be compiled includes some
special characteristics. This characteristic is indicated by arg, which can take anyone of the
following values:
•

EAE
The EAE argument specifies that the code uses the Extended Arithmetic Element.

•

EIS
The EIS argument specifies that the code uses the Extended Instruction Set.

•

FIS
The PIS argument specifies that the code uses the Floating Instruction Set, which also includes
the Extended Instruction Set.
THREADED (DCL) or THR (MCR)
The THREADED or THR argument specifies that the code is threaded.

•

The default DCL qualifier is jCODE:THREADED. The default switch on RSX systems is jCD:THR.
On RSTS jE systems, the jI:arg switch defaults to the value specified during installation.

Language Compatibility Issues

2-7

2.2.2.5

/CONTINUATIONS Qualifier

The ICONTINUATIONS qualifier is valid only for FORTRAN-77 and VAX FORTRAN, but
it has a different format depending on which FORTRAN compiler you are using. The corresponding switch is ICO. This qualifier specifies the number of continuation lines allowed in
a source program statement; it was initially designed for use with punch card systems. The
ICONTINUATIONS qualifier has the following form:
FORTRAN-77 qualifier:

/CONTINUATIONS:n

FORTRAN-77 switch:

/CO:n
/CONTINUATIONS:n
}
{ /CONTINUATIONS = n

VAX FORTRAN:

In all cases, n can accept values from 0 through 99. The default is 19.
2.2.2.6

/CROSS_REFERENCE Qualifier

The ICROSS-REFERENCE qualifier applies only to VAX FORTRAN. This qualifier specifies that
the storage map section of a listing file should include information about using symbolic names,
including the line numbers of the lines in which the symbols are defined and referenced. The
default is INOCROSS-REFERENCE.
2.2.2.7

/DEBUG Qualifier

The IDE BUG qualifier applies only to FORTRAN-77 and VAX FORTRAN, but it has a different
format depending on which FORTRAN compiler you are using. The corresponding switch is IDB.
For FORTRAN-77, IDEBUG or IDB specifies that the FORTRAN-77 compiler should provide
information for use by the PDP-II Symbolic Debugger.
For VAX FORTRAN, the IDE BUG qualifier tells the compiler to provide information to the VAX
Symbolic Debugger and the run-time error traceback mechanism. The VAX FORTRAN IDEBUG
qualifier has the form IDEBUG=arg, where arg can have anyone of the following values:

2-8

•

SYMBOLS
The SYMBOLS argument specifies that the VAX FORTRAN compiler provide the debugger with local symbol definitions for user-defined variables, arrays (including dimension
information), structures, and labels of executable statements.

•

TRACEBACK
The TRACEBACK argument specifies that the VAX FORTRAN compiler provide an address
correlation table so that the debugger and the run-time error traceback mechanism can
translate virtual addresses into source program routine names and compiler-generated line
numbers.

•

ALL
The ALL argument specifies that the VAX FORTRAN compiler provide both local symbol
definitions and an address correlation table. If you do not specify an argument to the IDEBUG
qualifier, IDE BUG = ALL is the default.

•

NONE
The NONE argument specifies that the VAX FORTRAN compiler not provide any debugging
information. This is the same as INODEBUG.

Language Compatibility Issues

The VAX FORTRAN default is /DEBUG = (NOSYMBOLS, TRACEBACK). On FORTRAN-77, the
default is /NODEBUG or /NODB.
NOTE
If you use the /DEBUG qualifier, it is strongly recommended that you also use the
/NOOPTIMIZE qualifier. Optimizations performed by the FORTRAN compilers can
cause unexpected behavior when using the appropriate symbolic debuggers.

2.2.2.8

/DI and /8 Switches

The /DI and /B switches are valid only for FORTRAN IV. Use /DI for FORTRAN IV on RSX systems, and use /B for FORTRAN IV on RSTS /E systems. Both of these switches enable expanded
listings of compiler internal diagnostic information. The default on RSX systems is /NODI. For
FORTRAN IV on RSTSjE systems, the default is determined when FORTRAN IV is installed.
2.2.2.9

/D_LlNES Qualifier

The /D_LINES qualifier is valid only for VAX FORTRAN, but the /DLINES qualifier provides
the same capability for FORTRAN-77. This qualifier specifies that lines with a D in column 1 are
compiled and not treated as comment lines. The default is /NOD_LINES, which indicates that
lines with a D in column 1 are treated as comments.
2.2.2.10

/DLINES Qualifier

The /DLINES qualifier is valid only for FORTRAN-77, but you can achieve the same behavior for
FORTRAN IV and FORTRAN-77 by using switches. (To achieve the same behavior using VAX
FORTRAN, use the /D_LINES qualifier.) For FORTRAN IV on RSX systems and FORTRAN-77,
the switch corresponding to the /DLINES qualifier is /DE. For FORTRAN IV on RSTS jE systems,
the corresponding switch is jD.
All of these command qualifiers and switches specify that lines with a D in column 1 be compiled
and not treated as comment lines. The default is /NODLINES, /NODE, or /NOD, all of which
indicate that lines with a D in column 1 are treated as comments.
2.2.2.11

/DML Qualifier

The /DML qualifier is valid only on VAX FORTRAN. It specifies that the FORTRAN Data
Manipulation Language (DML) preprocessor is to be invoked before the compiler. The preprocessor produces an intermediate file of FORTRAN source code in which FORTRAN DML commands
are expanded into FORTRAN statements. The compiler is then automatically invoked to compile
this intermediate file. You can use the VAX FORTRAN /SHOW=PREPROCESSOR qualifier along
with the /DML qualifier to cause the preprocessor-generated source code to be included in the
listing file. For more information on the /SHOW qualifier, see Section 2.2.2.32. The default is
/NODML.

Language Compatibility Issues

2-9

2.2.2.12

/EXTEND Qualifier

The jEXTEND qualifier applies only to the FORTRAN-77 compiler. However, the
jEXTEND_SOURCE qualifier provides the same capability for VAX FORTRAN, and the JEX and
jE switches provide the same capability for FORTRAN IV and FORTRAN-77. For FORTRAN
IV on RSX systems and FORTRAN-77, the corresponding switch is JEX. For FORTRAN IV on
RSTSjE systems, the corresponding switch is jE.
All of these command qualifiers and switches let you specify that the range of your source text
be extended from columns 1 through 72 to columns 1 through 132. If a line is longer than 132
characters, the FORTRAN IV and FORTRAN-77 compilers signal a fatal read error and your
compilation is immediately terminated. The default is jNOEXTEND, JNOEX, or jNOE.
2.2.2.13

/EXTEND_SOURCE Qualifier

The jEXTEND_SOURCE qualifier applies only to the VAX FORTRAN compiler, but it performs
the same function as the FORTRAN-77 jEXTEND qualifier and the JEX and jE switches. The
jEXTEND_SOURCE qualifier lets you specify that the range of your VAX FORTRAN source
text be extended from columns 1 through 72 to columns 1 through 132. If a line is longer than
132 characters, the VAX FORTRAN compiler signals a fatal read error and your compilation is
immediately terminated. The default is jNOEXTEND_SOURCE.
2.2.2.14

/FOR Qualifier

Use the jFOR qualifier at the PDP-II DCL command line to invoke the FORTRAN IV compiler.
On RSX systems, the default file type for a FORTRAN IV program is FTN. On RSTSjE systems,
the default file type is FOR.
2.2.2.15

/F77 Qualifier

The jF77 qualifier is valid only for FORTRAN-77 and VAX FORTRAN. The corresponding
switch is jF77. This qualifier specifies that the compiler use the FORTRAN-77 interpretation
rules for those statements with a meaning that is incompatible with FORTRAN IV-PLUS. If you
specify the jNOF77 qualifier, the compiler selects FORTRAN IV-PLUS interpretations in cases of
incompatibility. The default is jF77.
2.2.2.16

/G_FLOATING Qualifier

The jGJLOATING qualifier is valid only for the VAX FORTRAN compiler. It lets you control
how the VAX FORTRAN compiler implements the following data types:
•
•
•
•

REAL*8
COMPLEX*16
DOUBLE PRECISION
DOUBLE COMPLEX

The default for this qualifier is jNOG_FLOATING, which causes the compiler to implement
double-precision quantities using the VAX D_floating data type. If you specify jGJLOATING,
the compiler implements such quantities using the VAX G-Boating data type.

2-10

Language Compatibility Issues

2.2.2.17

jlDENTIFICATION Qualifier

The IIDENTIFICATION qualifier is valid only for FORTRAN-77, but the corresponding switch
liD is also valid for FORTRAN IV on RSX systems. This qualifier tells the compiler to print its
identification and version number on the user's terminal. The default is INOIDENTIFICATION or
INOID.

2.2.2.18

jl4 Qualifier

The 114 qualifier is valid on FORTRAN IV, FORTRAN-77, and VAX FORTRAN. There are also
corresponding switches: use 114 for FORTRAN-77 and for FORTRAN IV on RSX systems; use IT
for FORTRAN IV on RSTS/E systems.
All of these qualifiers and switches specify that the default allocation for integer variables be 2
words (4 bytes). This includes INTEGER and LOGICAL data types.
For FORTRAN IV and FORTRAN-77, the default is INOl4 or INOT, which means that integer
variables are interpreted as INTEGER*2 and LOGICAV"2. VAX FORTRAN defaults to 114,
meaning that integer variables are interpreted as INTEGER*4 and LOGICAL*4.
2.2.2.19

JLA Switch

The ILA switch is valid only for FORTRAN-77 in MCR interactive mode. This switch causes the
current switch settings to be retained (latched) for subsequent compilations in MCR interactive
mode. The default is INOLA.
2.2.2.20

jLlBRARY Qualifier

The ILIBRARY qualifier is valid only for VAX FORTRAN. This qualifier specifies that a file is a
text library file. You can specify one or more text library files in a list of files that are concatenated
by using plus signs. However, at least one of the files in the list must be a nonlibrary file. The
default file type for a VMS library file is TLB.
2.2.2.21

jLlNE_NUMBERS Qualifier

The ILINE-NUMBERS qualifier is valid only on FORTRAN IV. The corresponding switch is ISN
for FORTRAN IV on RSX systems. The ILINE_NUMBERS qualifier and ISN switch tell the
FORTRAN IV compiler to include Internal Sequence Numbers in your object code.
FORTRAN IV on RSTS/E systems does not supply a switch to specify this default behavior.
The IS switch is actually the equivalent of INOLINE-NUMBERS or INOSH, which causes
the FORTRAN IV compiler on RSTS IE systems to suppress the generation of Internal Sequence
Numbers.
Internal Sequence Numbers may take up space in the object code and may reduce program
execution speed, but they are useful in determining which line caused a run-time error when you
debug your program. ILINE_NUMBERS is the default DCL qualifier for FORTRAN IV. ISN
is the default switch for FORTRAN IV on RSX systems. The default on RSTS/E systems is to
generate Internal Sequence Numbers; you cannot specify this behavior using a switch.

Language Compatibility Issues 2-11

2.2.2.22

JUST Qualifier

The /LIST qualifier is valid for FORTRAN IV, FORTRAN-77, and VAX FORTRAN, but there
are some minor differences between the PRO and PDP-11 FORTRAN interpretation and the
VAX FORTRAN interpretation of this qualifier. The /LIST qualifier generally tells anyone of the
FORTRAN compilers to generate a listing file.
There is no corresponding switch for either FORTRAN IV or FORTRAN-77, because the effects of
the /LIST qualifier can be accomplished by specifying a file specification in the list-file field of the
appropriate command line. By including a listing file specification, you indicate that the compiler
should generate a listing file. See Section 2.2.1 for more information on the appropriate command
lines for FORTRAN IV and FORTRAN-77.
The /LI and /L switches let you indicate what information to include in the listing file. For
FORTRAN IV on RSX systems and FORTRAN-77, the switch is /LI:n. For FORTRAN IV on
RSTS/E systems, the switch is /L:n. For either switch, n can take the following values:

o or null:

List diagnostics only

1 or SRC:

List source program and diagnostics only

2 or MAP:

List storage map and diagnostics only

4 or COD:

List generated code and diagnostics only

You can specify any combination of the above list of options by adding together the numeric
argument values to compute the desired list options. For example, specifying 7 or ALL requests a
source listing, a storage mapi and a generated code listing. The default for both switches (/LI:n
and /L:n) is n = 3; that is, /LI:3 and /L:3. If you specify a listing file name in the command line
for FORTRAN IV or FORTRAN-77 and you do not specify either the /LI or /L switch, by default
the listing file will include a listing of the source program and storage map.
NOTE

For the PRO and PDP-11 FORTRAN compilers, the default file type of a listing file is
LST. For VAX FORTRAN, the default file type of a listing file is LIS.
When using the DeL /LIST qualifier, you can specify it as either a command qualifier or as a file
specification qualifier on PRO and PDP-11 systems. That is, for the. PRO and PDP-I1 FORTRAN
compilers, /LIST is a positional qualifier. The command qualifier and file specification qualifier
have the following form:
Command qualifier:

FORTRAN/LIST file-name.FTN

File specification qualifier:

FORTRAN file-name.FTN /LIST

If you specify /LIST as a command qualifier, the PRO and PDP-1I FORTRAN compilers print the
listing file and place a copy in your UFO. If you specify /LIST as a file specification qualifier, the
PDP-11 FORTRAN compiler places the file in your directory but does not automatically print the
file.

When using the VAX DeL /LIST qualifier, the position of the qualifier in the command line
does not have any effect on the printing of the file. If you want to print the file, you must do so
explicitly. For VAX FORTRAN, the default is /NOLIST for an interactive compilation and /LIST
for a compilation performed in batch mode.

2-12

Language Compatibility Issues

2.2.2.23

IlO and IQ Switches

The ILO and IQ switches are valid only for FORTRAN IV. The ILO switch tells the FORTRAN
IV compiler on RSX systems to print the names of program units (from PROGRAM, FUNCTION,
SUBROUTINE, and BLOCK DATA statements) on your terminal as they are compiled. The IQ
switch inhibits the printing of this information for the FORTRAN IV compiler on RSTS IE systems;
that is, the IQ switch is actually the equivalent of INOLO. The default on both systems is to print
this information.
2.2.2.24

IMACHINE_CODE Qualifier

The IMACHINE_CODE qualifier is valid for FORTRAN IV, FORTRAN-77, and VAX FORTRAN,
but there are some minor differences. FORTRAN IV and FORTRAN-77 both provide corresponding switches. For FORTRAN IV on RSX systems and FORTRAN-77, specify a file specification
for the listing file field and use the ILI:3 switch. For FORTRAN IV on RSTS IE systems, specify
a listing file and use the IL:3 switch. (For more information on the ILl and IL switches, see
Section 2.2.2.22.)
You can use anyone of the following methods to specify that the PDP-11 FORTRAN compiler
listing include binary machine code and diagnostics:
•
•
•

Specify the DCL IMACHINE_CODE qualifier.
Specify the ILI:3 switch to the listing file specification in the command line for either
FORTRAN-77 or FORTRAN IV on RSX systems.
Specify the IL:3 switch to the listing file specification in the command line for FORTRAN IV
on RSTS/E systems.

The DCL IMACHINE_CODE qualifier is a positional qualifier for PRO and PDP-11 FORTRAN,
which is similar to the ILIST positional qualifier. For more information on positional qualifiers,
see Section 2.2.2.22.
The IMACHINE_CODE qualifier for the VAX FORTRAN compiler does not imply ILIST. If you
do not request a listing file using ILIST, the VAX IMACHINE_CODE qualifier is ignored. This
qualifier tells the VAX FORTRAN compiler to include a symbolic representation of the object code
generated by the compiler. The format of this symbolic representation is very similar to a VAX
MACRO assembly listing. Do not try to compile this listing since it includes items that are not
supported by VAX MACRO. The default is INOMACHINE_CODE.
2.2.2.25

IMAP Qualifier

The IMAP qualifier is only available through the PRO and PDP-I1 FORTRAN compilers. The
ISHOW=MAP qualifier provides the same capability for VAX FORTRAN. Section 2.2.2.32 describes the ISHOW qualifier. Using a different command line, you can achieve the identical
behavior by including a file specification for the listing-file field of the appropriate command line
and specifying either ILI:2 for FORTRAN IV on RSX systems and FORTRAN-77, or IL:2 for
FORTRAN IV on RSTS IE systems.
You can use anyone of the following methods to specify that the compiler listing include a storage
map and diagnostics:
•
•

Specify the DCL IMAP qualifier.
Specify the ILI:2 switch to the listing file specification in the command line for either
FORTRAN IV on RSX systems or FORTRAN-77.

Language Compatibility Issues

2-13

•

Specify the /L:2 switch to the listing file specification in the command line for FORTRAN IV
on RSTS/E systems.

Like the DCL /MACHINE_CODE qualifier, /MAP implies the DCL /LIST qualifier on the
PRO and PDP-II FORTRAN compilers, and is a positional qualifier. For more information, see
Section 2.2.2.22.

2.2.2.26

IN Switch

The /N switch is valid only for FORTRAN IV on RSTS /E systems. It specifies the maximum
number of simultaneously open I/O channels allowed at run time. The format of this switch is
/N:n, where 1 < n < 15. The default behavior is specified when FORTRAN IV is installed on
your RSTS/E system.

2.2.2.27

10 Switch

The /0 switch is valid only for FORTRAN IV on RSTS/E systems. It specifies that the compiler
print an "Options-In-Effect" section before the listing. The default behavior is specified when
FORTRAN IV is installed on your RSTS /E system.

2.2.2.28

10BJE~T Qualifier

The /OBJECT qualifier is valid only for VAX FORTRAN, but you can request an object file using
the appropriate FORTRAN-77 or FORTRAN IV command line. You use the /OBJECT qualifier,
or the appropriate field in the command line, to specify a name for the object file generated by
the compiler. The VAX FORTRAN compiler specifies the /OBJECT qualifier by default so that an
object file is generated. If you want to suppress the object file, specify the /NOOBJECT qualifier.
An object file is not automatically generated by the PRO and PDP-II compilers. To request
that the FORTRAN IV or FORTRAN-77 compiler generate an object file, you must include a file
specification in the object-file field of the appropriate command line. For example:
FOR> objectfile.oBJ = inputfile.FTN
F77> objectfile.oBJ = inputfile.FTN

In this example, the object file to be generated is named OBJECTFILE.OBJ.

2.2.2.29

10PTIMIZE Qualifier

The /OPTIMIZE qualifier is valid only for FORTRAN-77 and VAX FORTRAN. The corresponding
switch is lOP. The /OPTIMIZE qualifier and lOP switch let you specify that the compiler is to
produce optimized code; /OPTIMIZE or lOP is the default. If you invoke the VAX or PDP-II
Symbolic Debugger, you can specify /NOOPTIMIZE or /NOOP to ensure that the debugger has
sufficient information to locate errors in the source program.

2.2.2.30

IR Switch

The /R switch is valid only for FORTRAN IV on RSTS /E systems. This qualifier specifies the
maximum record length (in bytes) on run-time I/O. This switch has the format /R:n, where n has
the range 4 < n < 4095. The default behavior is specified when FORTRAN IV is installed on
your RSTS /E system.

2-14

Language Compatibility Issues

2.2.2.31

/SHAREABLE Qualifier

The /SHAREABLE qualifier is available to the PRO and PDP-11 FORTRAN compilers. The
corresponding switch for FORTRAN-77 and FORTRAN IV on RSX systems is /RO. For FORTRAN
IV on RSTS/E systems, the corresponding switch is /Z.
These command qualifiers and switches tell the compiler to generate pure code and pure data
sections as read-only .. This might be useful if you want to take advantage of code sharing in
multiuser tasks on RSX-11M-PLUS systems. These qualifiers specify whether to place the data in
the $IDATA or $PDATA program section, so they can be used on RSTS/E systems as well. The
default is /NOSHAREABLE, /NORO, or /NOZ.
2.2.2.32

/SHOW Qualifier

The /SHOW qualifier is valid only for VAX FORTRAN. The /SHOW qualifier controls whether
or not optionally listed source lines (that is, text module source lines and preprocessor-generated
source lines) and a symbol map appear in the source listing.
You must specify the VAX FORTRAN /LIST qualifier in order for the /SHOW qualifier to take
effect. The /SHOW qualifier has the format /SHOW=arg, where arg can have anyone of the
following values:
•

DICTIONARY
The DICTIONARY argument specifies that FORTRAN source representations of any CDD
records referenced by DICTIONARY statements be included in the listing file.

•

INCLUDE
The INCLUDE argument specifies that the source lines from any files specified by INCLUDE
statements be included in the source listing.

•

MAP
The MAP argument specifies that the symbol map be included in the listing file. If the
/CROSS-REFERENCE qualifier is specified, MAP is ignored. This argument provides the
same feature for VAX FORTRAN as the /MAP qualifier does for the PRO and PDP-11
FORTRAN compilers.
PREPROCESSOR
The PREPROCESSOR argument specifies that preprocessor-generated source lines be included
in the listing file. The negative form, NOPREPROCESSOR, specifies that the source lines be
excluded from the source listing.

•

SINGLE
The SINGLE argument specifies that the symbolic names of parameter constants be included
in cross-reference listings, even if they are not referenced outside the PARAMETER statements
in which they are declared. The negative form, NOSINGLE, specifies that names of parameter
constants be suppressed if they are only declared and not referenced elsewhere. This is useful
for cross-reference listings of small programs that specify INCLUDE declarations but use only
a small number of the parameter constant names that have been declared.

•

ALL
The ALL argument specifies that all optionally listed source lines be included in the listing file.

•

NONE
The NONE argument specifies that no optionally listed source lines be included in the listing
file.

Language Compatibility Issues

2-1 5

The default arguments for the jSHOW qualifier are NODICTIONARY, NOINCLUDE, MAP,
NOPREPROCESSOR, and SINGLE. Specifying the jSHOW qualifier without any arguments
is equivalent to jSHOW=ALL. jNOSHOW without any arguments is equivalent to specifying
jSHOW=NONE.
2.2.2.33

/SOURCE Qualifier

The jSOURCE qualifier is valid only on the PRO and PDP-II FORTRAN compilers. Using a
different command line, you can achieve the same behavior by induding a file specification for the
listing-file field of the appropriate command line and specifying either jLI:2 for FORTRAN IV on
RSX systems and FORTRAN-77, or jL:2 for FORTRAN IV on RSTSjE systems.
You can use anyone of the following methods to specify that the compiler listing include source
code:
•
•
•

Specify the DCL jMAP qualifier.
Specify the jLI:2 switch to the listing file specification in the command line for either
FORTRAN-77 or FORTRAN IV on RSX systems.
Specify the jL:2 switch to the listing file specification in the command line for FORTRAN IV
on RSTSjE systems.
.

Both the DCL qualifier and the corresponding switches are very similar to the jMAP qualifier and
its corresponding switches. For more information, see Section 2.2.2.25.
2.2.2.34

/SP Switch

The jSP switch is valid only for FORTRAN-77 and FORTRAN IV on RSX systems. This switch
tells the compiler to automatically spool the listing file. The default is jSP.
2.2.2.35

/STANDARD Qualifier

The jSTANDARD qualifier is valid only for FORTRAN-77 and VAX FORTRAN because the
purpose of this qualifier is to tell the compiler to flag any extensions to the ANSI X3.9-1978
standard. The corresponding switch is JST. These qualifiers have the following formats:
Qualifier:

jSTANDARD=arg

Switch:

jST:arg

In either case, arg can have one of the following values:
•

•

2-16

SEMANTIC (VAX FORTRAN only)
The SEMANTIC argument specifies that an informational message be issued for syntax
extensions to the current ANSI standard - both for extensions that occur in individual
statements and for extensions that occur as a result of usage across statements in a program
unit. For example, an assignment statement that specifies a record field is not flagged as an
extension if SYNTAX checking is in effect. It is flagged as an extension if SEMANTIC checking
was in effect.
SOURCE (FORTRAN-77) or SOURCEJORM (VAX FORTRAN)
Either the SOURCE (FORTRAN-77) or SOURCEJORM (VAX FORTRAN) argument specifies
that an informational message be issued for statements using tab formatting or containing
lowercase characters.

Language Compatibility Issues

•

SYNTAX
The SYNTAX argument specifies that an informational message be issued for syntax extensions
to the current ANSI standard.

•

ALL
The ALL argument specifies that informational messages be issued for both the SYNTAX and
SOURCE form extensions to the current ANSI standard.

•

NONE
The NONE argument specifies that no informational messages be issued for extensions to the
current ANSI standard. The /STANDARD=NONE qualifier is invalid on RSTS/E systems, so
use /NOSTANDARD.

The default is /NOSTANDARD or /NOST, which is equivalent to /STANDARD=NONE. If
you specify the /NOWARNINGS qualifier, the /STANDARD qualifier is ignored. Specifying
/STANDARD without any arguments is equivalent to specifying /STANDARD=(SYNTAX,
NOSOURCE) for FORTRAN-77 or /STANDARD=(SYNTAX, NOSOURCEJ'ORM) for VAX
FORTRAN. Specifying the /ST switch without any arguments is equivalent to specifying
/ST:SYNTAX.
2.2.2.36

/TRACEBACK:arg Qualifier

The /TRACEBACK:arg qualifier is valid only for FORTRAN-77. The corresponding switch is
/TR:arg. The /TRACEBACK qualifier and /TR switch control the amount of extra code to be
included in the compiled output for use by the OTS during error traceback. This code is used to
produce diagnostic information and to identify faulty statements.
The /TRACEBACK qualifier and /TR switch accept anyone of the following values for arg:
•

BLOCKS
The BLOCKS argument specifies that traceback information be compiled for all subroutine and
function entries and initial statements in sequences called blocks.

•

LINES
The LINES argument is equivalent to the ALL argument.

•

NAMES
The NAMES argument specifies that traceback information be compiled only for subroutine
and function entries.

•

ALL
The ALL argument specifies that error traceback information be compiled for all source
statements and function and subroutine. entries.

•

NONE
The NONE argument specifies that no traceback information be produced.

The default is /TRACEBACK:BLOCKS or /TR:BLOCKS. If you do not specify an argument to
/TRACEBACK or /TR, the default is TRACEBACK:ALL or /TR:ALL.

Language Compatibility Issues

2-17

2.2.2.37

/U Switch

The jU switch is valid only for FORTRAN IV on RSTSjE systems. It disables USR swapping
at run time. The default behavior is specified when FORTRAN IV is installed on your RSTSjE
system.

2.2.2.38

/VECTORS Qualifier

The jVECTORS qualifier is valid only for the FORTRAN IV compiler running on an RSX system.
The corresponding switch is JVA for FORTRAN IV on RSX systems. These command qualifiers
tell the compiler to "vectorize" arrays. This increases the size of the object program but decreases
the overall execution time for your program. jVECTORS or JVA is the default behavior.
FORTRAN IV on RSTSjE systems does not supply a switch to specify this default behavior.
The jV switch is the equivalent of jNOVECTORS or jNOVA, which causes the FORTRAN IV
compiler on RSTSjE systems to suppress the vectorization of arrays.

2.2.2.39

/WARNINGS Qualifier

The jWARNINGS qualifier is valid for FORTRAN IV, FORTRAN-77, and VAX FORTRAN, but
there are some differences between the qualifiers for the various compilers. The corresponding
switch is jWR for FORTRAN IV on RSX systems and FORTRAN-77. For FORTRAN IV on
RSTSjE systems, the corresponding switch is jW.
For the PRO and PDP-II FORTRAN compilers, the jWARNINGS qualifier tells the compiler to
issue warning diagnostics. jWARNINGS is the default.
For VAX FORTRAN, the jWARNINGS qualifier specifies that the compiler generate informational
(I) and warning (W) diagnostic messages in response to informational and warning-level errors.
The VAX FORTRAN jWARNINGS qualifier has the format jWARNINGS=arg, where arg can be
anyone of the following:
•

DECLARATIONS
The DECLARATIONS argument causes the compiler to print warnings for any undeclared data
items used in the program. DECLARATIONS acts as an external VAX FORTRAN IMPLICIT
NONE declaration. The default is NODECLARATIONS.

•

GENERAL
The GENERAL argument causes the compiler to generate informational and warning diagnostic messages. An informational message indicates that a correct VAX FORTRAN statement
may have unexpected results, or that it contains a nonstandard syntax or source form. A
warning message indicates that the compiler has detected acceptable but nonstandard syntax
or has performed some corrective action. In either case, a warning message indicates that
unexpected results may occur. To suppress I and W diagnostic messages, specify the negative
form of this qualifier (NOGENERAL). The default is GENERAL.
ULTRIX (VAX FORTRAN only)
The ULTRIX argument causes the compiler to issue diagnostics for language features not
supported by VAX FORTRAN on ULTRIX systems. Using this option, you can develop VAX
FORTRAN programs on a VMS system and transport those programs to an UL TRIX system.
The default is NOUL TRIX.
VAXELN (VAX FORTRAN only)
The VAXELN argument causes the compiler to issue diagnostic messages for language features
not supported by VAX FORTRAN on a VAXELN system. The default is NOVAXELN.

•

2-18

Language Compatibility Issues

•

ALL
The ALL argument causes the compiler to print all informational and warning messages,
including warning messages for any undeclared data items.

•

NONE
The NONE argument suppresses all informational and warning messages.

For VAX FORTRAN, the default is WARNINGS = (NODECLARATIONS, GENERAL,
NOULTRIX, NOVAXELN).
2.2.2.40

IWORK_FILES:n Qualifier

The /WORKJILES:n qualifier is valid only for FORTRAN-77. The corresponding switch is
/WF:n. The /WORK_FILES:n qualifier and /WF:n switch specify the number of temporary ondisk files to be used during compilation. You can assign n a value between 1 and 3; the default is
2. If you increase n, this increases the maximum possible size of your program, but decreases the
overall time of compilation.
2.2.2.41

IX Switch

The /X switch is valid only for FORTRAN IV on RSTS/E systems. It indicates cross-compilation
for the target environment. This switch has the format /X:arg, where arg can have one of the
following values:
RT:

Selects RT -11

RST:

Selects RSTS IE

RSX:

Selects RSX-ll

The default behavior is determined when FORTRAN IV is installed on your RSTS/E system.

2.2.3

Compiler Control Statements

The FORTRAN-77 and VAX FORTRAN compilers provide the following compiler control statements (FORTRAN IV does not support any of these statements):
•

•

INCLUDE 'file-spec' or
INCLUDE 'ffile-spec](module-name),
The INCLUDE statement is valid for both FORTRAN-77 and VAX FORTRAN. It specifies that
the contents of a file or a text library module be incorporated into the FORTRAN compilation
directly following the INCLUDE statement. The file-spec argument is a character constant
that specifies the file to be included. The module-name argument is the name of a text module
located in a text library.
DICTIONARY' cdd-path'
The DICTIONARY statement is valid only for VAX FORTRAN. This statement incorporates
VAX Common Data Dictionary (CDD) data definitions into the current FORTRAN source file
during compilation. The cdd-path argument is the full or relative pathname of a CDD object.

Language Compatibility Issues

2-19

•

OPTIONS qualifier
The OPTIONS statement is valid only for VAX FORTRAN. You can use this statement to
override or confirm the FORTRAN command qualifiers in effect for a particular program unit.
The OPTIONS statement can take anyone of the following arguments:
/CHECK = arg
/EXTEND_SOURCE
/F77
/GJLOATING

/14
All of these arguments to the OPTIONS statement have corresponding FORTRAN command
qualifiers.

2.2.4

Syntax and Format

Because VAX FORTRAN is an extension of PRO and PDP-II FORTRAN, most of the language
syntax and format is identical. Coexistence between PRO, PDP-II, and VAX systems, or migration
from one to another, is made easier by the many similarities between the various FORTRAN
compilers. However, because one is an extension of the other, there are some differences.
Differences between the various FORTRAN implementations involve the following areas:
•
•
•
•
•
•

Logical tests
Floating-point results
Character and Hollerith constants
Logical unit numbers
Assigned GO TO label list
Effect of the DISPOSE = 'PRINT' specification

FORTRAN-77 and VAX FORTRAN are both based on the most recent American National
Standard FORTRAN-77 (X3.9-1978). FORTRAN IV is based on the previous standard, X3.9-1966.
Incompatibilities between the two standards exist in the following areas:
•
•
•
•
•

The minimum iteration count of DO loops
The EXTERNAL statement
The defaults for the OPEN statement's BLANK and STATUS keywords
The X format edit descriptor
The effect of attempting to open a connected unit

These differences are described in the following sections.

2.2.5

Statements

Despite the similarity of the various FORTRAN compilers, some differences do exist between the
statements available in each language, or the way these statements are handled by the compiler.
The following sections describe these differences.

2-20

Language Compatibility Issues

2.2.5.1

Assigned GO TO Label Lists

The labels that you specify in an assigned GO TO label list are checked by the various FORTRAN
compilers to ensure their validity in the program unit that you are compiling. However, both VAX
FORTRAN and FORTRAN IV do not perform a check at run time to ensure that a label that is
actually assigned is in the list. FORTRAN-77 does perform this check. For example:

500

ASSIGN 400 TO MEGO
GO TO MEGO, (150, 200, 350)
TYPE *,'Statement not found in Assigned GO TO'

400

In this example, the ASSIGN statement assigns a value of 400 to MEGO. This variable is then
used to indicate a statement label for the assigned GO TO statement; however, the label 400 is
not included in the label list provided in the assigned GO TO statement. If a statement labeled
400 does exist somewhere in the program unit, the inconsistency between the various compilers is
exhibited as shown in the following table:
Language

Action

FORTRAN IV

Executes the ASSIGN and GO TO statements and continues execution at
statement 400.

FORTRAN-77

Executes the ASSIGN statement and checks to see if label 400 is in the label
list provided in the GO TO statement. Because it is not, execution continues at
the next statement, in this case, statement 500.

VAX FORTRAN

Executes the ASSIGN and GO TO statements and continues execution at
statement 400.

2.2.5.2

DO Loop Minimum Iteration Count

In FORTRAN-77 and VAX FORTRAN, the body of a DO loop is not executed if the end condition
of the loop is satisfied when the DO statement is executed. In FORTRAN IV, however, the body
of a DO loop is always executed at least once. For example:
1=1
DO 10 J = 1,1

CONTINUE

FORTRAN IV performs a single iteration of this loop, even though the loop condition (J = I)
is satisfied. If you compile this loop using FORTRAN-77 or VAX FORTRAN, the loop is not
executed because the condition is already satisfied. If you specify the /NOF77 qualifier when
compiling the FORTRAN-77 or VAX FORTRAN program, the execution of this loop matches the
FORTRAN IV behavior.

Language Compatibility Issues

2-21

2.2.5.3

EXTERNAL Statement

If you specify a function using FORTRAN IV with the same name as a FORTRAN intrinsic
function or library function in an EXTERNAL statement, that function name is assumed to refer
to the named intrinsic or library function and not a user-defined function. If you want to use the
FORTRAN IV EXTERNAL statement to specify a user-defined function, you must precede the
function name with an asterisk.

In the following example, the first EXTERNAL statement, if compiled under FORTRAN IV,
indicates the FORTRAN intrinsic SIN function. However, since SIN is preceded by an asterisk
in the second EXTERNAL statement, FORTRAN IV assumes that this refers to a user-defined
function named SIN.
EXTERNAL SIN
EXTERNAL *SIN

If you specify a function using FORTRAN-77 or VAX FORTRAN with the same name as a

FORTRAN intrinsic function or library function in an EXTERNAL statement, that function name
is always assumed to refer to a user-defined function. If you precede a function name with an
asterisk using FORTRAN-~7 or VAX FORTRAN, the asterisk has no meaning and is marked as an
error. However, if you specify the jNOF77 switch when compiling your FORTRAN-77 or VAX
FORTRAN program, the function name in the EXTERNAL statement is interpreted as it would be
under FORTRAN IV, and the asterisk is valid.
To use the name of an intrinsic function as an argument to a subprogram in FORTRAN-77 or
VAX FORTRAN, you must use the INTRINSIC state:qlent. This statement provides the same
capabilities for FORTRAN-77 and VAX FORTRAN that the EXTERNAL statement (without an
asterisk) provides for FORTRAN IV.
2.2.5.4

Blank Common Program Section (.$$$$.)

Under FORTRAN-77, the blank common program section (.$$$$.) has the SAV attribute; it does
not have this attribute under FORTRAN IV. The SAV attribute on a program section lets you pull
that program section into the root segment of an overlay.
In FORTRAN-77, the jF77 compiler switch lets you control the default assignment of the SAV
attribute; under jF77, the blank common program section is assigned the SAV attribute by default.
2.2.5.5

X Format Edit Descriptor

The nX edit descriptor lets you specify that the transmission of the next character to or from
a record should occur at the position n characters to the right of the current position. In a
FORTRAN-77 or VAX FORTRAN output statement, the character positions that are skipped
are not modified, and the length of the output record is not affected. However, in a FORTRAN IV
statement, the X edit descriptor writes blanks and may extend the output record.
In the following example, the FORMAT statement specifies that transmission of the character string
'ABCDEF' begin a single space to the right of the current (first) position. The T4 indicates that
the current position be set to the fourth character position. At this location, FORTRAN-77 and
VAX FORTRAN skip two positions. FORTRAN IV writes two blanks and then the string '12345'
followed by three blanks. The different interpretations of the X edit descriptor are illustrated in
the output of the following statements:
10

2-22

WRITE (1.10)
FORMAT(1X. 'ABCDEF'. T4. 2X. '12345'. 3X)

Language Compatibility Issues

These statements produce the following results:
FORTRAN-77 and VAX FORTRAN

FORTRAN IV
#AB##12345###

#ABCD12345

The /NOF77 switch does not affect the interpretation of the X edit descriptor. If you want to
simulate the FORTRAN IV method of interpreting the X edit descriptor, change nX to n(' ').

2.2.6

Subroutines

FORTRAN-77 supplies a number of utility subroutines. These subroutines are described in the
PDP-ll FORTRAN-77 User's Guide.

Six of these subroutines are supplied as a standard part of VAX FORTRAN. Table 2-3 describes
these FORTRAN subroutines and the operations they perform.
Table 2-3:
Subroutine

FORTRAN Subroutines
Operation Performed

DATE

Returns a 9-byte string containing the ASCII representation of the current date.

ERRSNS

Returns information about the most recently detected error condition. If the error was
an I/O error, the additional arguments receive the following information:
•

The primary file system error code: FCS-ll F.ERR value or RMS-ll STS value

•

The secondary file system error code: FCS-11 F.ERR+1 value or RMS-11 STY value

•

The logical unit number

EXIT

Terminates the execution of a program and returns control to the operating system. If
you do not specify the exit status parameter, the terminator status is success.

IDATE

Returns three integer values representing the current month, day, and year.

SECNDS

Provides the system time of day, or elapsed time, as a floating-point value in seconds.
The value returned by SECNDS is accurate to the resolution of the system clock:
0.0166 ... seconds for a 60-cycle (60 Hertz) clock, and 0.02 seconds for a 50-cycle (50
Hertz) clock.

TIME

Returns an 8-byte string containing the ASCII representation of the current time in
hours, minutes, and seconds.

VAX FORTRAN supports the following subroutines and functions for the purposes of maintaining
compatibility with PDP-11 FORTRAN. These subroutines and functions are:
ASSIGN

ERRSET

ERRTST

FDBSET

IRAD50

RAD50

RAN

RANDU

R50ASC

USEREX

Language Compatibility Issues 2-23

Most of these subroutines have been superseded by features included in VAX FORTRAN, so if
you are migrating to VAX FORTRAN you may want to use the new VAX FORTRAN capabilities
instead of these subroutines. However, if you are planning coexistence between PRO, VAX, and
PDP-II systems, continue to use these subroutines to ensure transportability of your programs.
2.2.6.1

ASSIGN Subroutine

The ASSIGN subroutine lets you supply device or file name information to a logical unit. The
assignment you make remains in effect until your program terminates, or until you close the
logical unit using a CLOSE statement or the CLOSE subroutine. Since the ASSIGN subroutine
assigns device or file information to a logical unit, you must call this subroutine before issuing the
first I/O statement for that logical unit.
If you use the ASSIGN subroutine for a particular logical unit, you can also use the FDBSET

subroutine and the DEFINE FILE statement. However, you cannot use the OPEN or INQUIRE
statements for that same logical unit. There are two other ways that you can assign a device or a
file name to a logical unit number: specify the FILE keyword in an OPEN statement, or use the
ASSIGN system command.
A call to the ASSIGN subroutine has the following format:
CALL ASSIGN (n [,name] [,icnt])

The arguments for this subroutine have the following definitions:
n
is an integer value specifying the logical unit number. If you only specify the logical unit number
in your CALL statement, you nullify any previous associations pertaining to that unit, and the
file/device association takes on the default value. That is, by only passing the n parameter, you
reset any previous values of name and icnt.
name
is a variable,' array, array element, or character constant containing any standard file specification.
icnt
is an INTEGER*2 value that specifies the number of characters contained in the string name.
If you omit the icnt argument or specify it as zero, any file specification is read until the first
ASCII null character is encountered. If you specify the icnt argument, you must specify the name
argument.
2.2.6.2

CLOSE Subroutine

The CLOSE subroutine lets you close the file currently open on a particular logical unit. A call to
the CLOSE subroutine has the following format:
CALL CLOSE(n)

The definition of the argument n is as follows:

2-24

language Compatibility Issues

n
is an integer value specifying the logical unit.

After you close the file, the logical unit again assumes the default file name specification.
2.2.6.3

ERRSET Subroutine

The ERRSET subroutine lets you determine what action to take in response to a particular error
detected by the Run-Time Library. This error action is independent of other errors, so if you
specify certain behavior for a particular error, that behavior is not followed if some other error
occurs.
The VAX condition-handling facility provides a more general method for defining actions to be
taken when errors are detected. If you are planning to migrate your applications to VAX systems,
refer to the VMS Run-Time Library Routines Reference Manual for more information on condition
handlers.
A call to the ERRSET subroutine has the following format:
CALL ERRSEr (number [,contin ,count ,type ,log ,maxlim])

In this format, the arguments have the following definitions:
number
is an integer value specifying the error number. You cannot specify a null argument for number,
but null arguments are legal for all the other parameters to ERRSET.
contin
is a logical value that determines whether or not to continue executing your program after an error
is detected.
.TRUE.

Continue after error is detected .

.FALSE.

Exit after error is detected .

count
is a logical value that determines whether or not the error is counted against your maximum error
limit.
.TRUE .

Count the error against the maximum error limit.

.FALSE.

Do not count the error against the maximum error limit.

type
. is a logical value that determines where control is passed after an error is detected.
.TRUE.

Pass control to an ERR transfer label, if one is specified .

.FALSE.

Return control to the routine that detected the error for default error recovery .

Language Compatibility Issues 2-25

log
is a logical value that determines whether or not to produce an error message for the detected
error.
.TRUE.

Produce an error message for this error .

.FALSE.

Do not produce an error message for this error .

maxlim
is a positive INTEGER*2 value that specifies the maximum error limit. The default is set to 15 at
the time of program initialization.

If you specify a null argument for any other parameter, it has no effect on the current state of that
argument. A null argument does not reset the previous value of the parameter.

NOTE
An external reference to ERRSET or ERRTST causes a special PDP-II FORTRAN
compatibility error handler to be established before the main program is called. This
special error handler transforms the executing environment to resemble that of PDP-II
FORTRAN.
2.2.6.4

ERRTST Subroutine

The ERRTST subroutine lets you check for a specific error. You can use ERRTST to determine
what error has occurred and take appropriate actions in response. A call to the ERRTST subroutine
has the following format:
CALL ERRTST (i .j)

The parameters for this subroutine have the following definitions:
is an integer value specifying the error number.
j
is a variable used to receive the return value of the error check.
j = 1:

Error i has occurred.

j = 2:

Error i has not occurred.

2.2.6.5

FDBSET Subroutine

You can use the FDBSET subroutine to specify special I/O options. It is provided primarily for
compatibility with older FORTRAN implementations; similar and more extensive capabilities are
available through the OPEN statement.
Using the FDBSET subroutine for a particular logical unit allows you to use the ASSIGN subroutine and the DEFINE FILE statement. However, you cannot use the OPEN or INQUIRE statements
for the same logical unit. For VAX FORTRAN, the recommended method to specify I/O options
is to use the OPEN statement. If you are planning to migrate to VAX systems, use the OPEN
statement to assign a logical unit and specify any special options. If you do use the FDBSET
subroutine, it must only be called before issuing the first I/O statement for the specified logical
unit.

2-26

Language Compatibility Issues

A call to the FDBSET subroutine has the following format:
CALL FOBSET (unit [.acc .share .numbuf .initsz .extend])

The parameters in this subroutine call have the following definitions:
unit
is an integer value specifying the logical unit. This is the only required parameter; all others are
optional.
ace
is a character constant specifying the access mode to be used. The possible access modes are:
'READONLY':

Establish read-only access.

'NEW':

Create a new file.

'OLD':

Access an existing file.

'APPEND':

Extend an existing sequential file.

'UNKNOWN':

Try accessing an existing file first. If no such file exists, create a new one.

share
is a character constant 'SHARE' indicating that shared access is allowed.
numbuf
is an INTEGER*2 value specifying the number of buffers to be used for multibuffered I/O.
initsz
is an INTEGER*2 value specifying the number of blocks initially allocated for a new file.
extend
is an INTEGER*2 value specifying the number of blocks by which to extend a file.
2.2.6.6

IRAD50 Subroutine

You can use the IRAD50 subroutine to convert Hollerith data to Radix-50 form. You may call
IRAD50 as a function subprogram if the return value is desired (see the first format below), or
as a subroutine if the return value is not desired (see the second format below). The IRAD50
subroutine has the following formats:
n = lRA050 (icnt .input .output)
CALL lRA050 (icnt .input .output)

In the first format, n is an INTEGER*2 value containing the number of characters actually converted. In either format, the parameters have the following definitions:
icnt
is an INTEGER*2 value specifying the maximum number of characters to be converted.
input
is a Hollerith string to be converted to Radix-50. Note that the scanning of input characters
terminates on the first non-Radix-50 character in the input string.

Language Compatibility Issues

2-27

output
is a numeric variable or array element in which the Radix-50 results are stored. Three Hollerith
characters are packed into each output word. The number of output words is computed as
(icnt + 2)/3. For example, if a value of 4 is specified for icnt, two output words will result, even if
an input string of only one character is converted.
2.2.6.7

RANDU Subroutine

The RANDU subroutine computes a pseudo-random number by implementing a general random
number generator of the multiplicative congruential type. The random number computed by this
random number generator is a single-precision value uniformly distributed in the following range:

0.0 .LE. VALUE .AND. VALUE .LT. 1.0
A call to the RANDU subroutine has the following format:
CALL RANDU (i1 ,i2 ,x)

In this call format, the parameters have the following definitions:

i1
is an INTEGER*2 variable or array element containing the seed for computing the random number.
i2
is an INTEGER*2 variable or_ array element containing the seed for computing the random number.

x
is a real variable or array element in which the computed random number is stored.
The values of i1 and i2 are updated during the computation to contain the updated seed values.
The algorith~ for computing the random number value is:

For compatibility, another multiplicative congruential random number generator is provided as the
file LB:[1,l]F4PRAN.OBJ. The algorithm used by this random number generator is:
If i1 = 0, i2 = 0,
set the generator base
X{n + 1)

216 + 3

Otherwise
X{n + 1)

= (2 16 + 3) * X{n) mod 231

Store generator base X(n = 1) in il, i2.

The result of this algorithm is X(n+1) scaled to a real value Y(n = 1) for 0.0 ~ Y(N = 1)

2-28

Language Compatibility Issues

< 1.

2.2.6.8

R50ASC Subroutine

The R50ASC subprogram lets you convert Radix-50 values to Hollerith strings. A call to R50ASC
has the following format:
CALL R50ASC (icnt .input .output)

In this format, the parameters have the following definitions:
icnt
is an INTEGER*2 value specifying the number of ASCII characters to be produced.
input
is a numeric variable or array element containing the Radix-50 data. The number of words of
input is determined by (icnt + 2)/3.
output
is a numeric variable or array element where the Hollerith characters are to be stored.
If the undefined Radix-50 code is detected, or if the Radix-50 word exceeds 174777 (octal),
question marks are placed in the output location.

2.2.6.9

USEREX Subroutine

The USEREX subroutine lets you specify a routine that will be called as part of your program
termination process. This allows you to perform clean-up operations, for example, in nonFORTRAN routines.
On VAX systems, you can establish a termination handler directly by calling the system service
routine SYS$DCLEXH. For more information, see the VMS System Services Reference Manual.
A call to the USEREX subroutine has the following format:
CALL USEREX (name)

In this format, name has the following definition:
name
specifies the routine to be called. The name of this routine must appear in an EXTERNAL statement in the program unit.
NOTE

Do not try to perform FORTRAN I/O operations as part of an exit handler. The
FORTRAN I/O system provides its own exit-handling functions, such as file closing, so
user-specified I/O operations may not work as expected. All OTS error handling is also
disabled.

2.2.7

Functions

VAX FORTRAN also supports certain functions for compatibility with PDP-II FORTRAN. The
following sections discuss these functions. Some of these functions may have been superseded by
features included in VAX FORTRAN, so if you are planning to migrate to VAX systems, consider
using these new capabilities. To ensure the transportability of your programs, use these functions.

Language Compatibility Issues

2-29

2.2.7.1

RAD50 Function

The RADSO function subprogram provides a simplified way to encode six Hollerith characters as
two words of Radix-50 data. The entry point for this function has the following format:
RAD50 (name)

In this function entry point, the name parameter has the following definition:
name
is a numeric variable name or array element corresponding to a Hollerith string.
NOTE

The RADSO function is equivalent to calling the IRADSO subroutine with icnt specified
as 6 and output specified as RADSO. The following FORTRAN statements accomplish
the same task:
FUNCTION RAD50 (A)
CALL IRAD50 (6.A.RAD50)
RETURN
END

You may use the RADSO function as an argument to an RSX-ll system directive
subroutine. For example:
REAL*8 A
DATA A/ITASK All
CALL REQUES (RAD50(A) .... )

2.2.7.2

RAN Function

The RAN function subprogram returns a pseudo-random number as the function value. The entry
point for the RAN function is:
RAN (j)
or

RAN (i1 .i2)

In this entry point, the parameters have the following definitions:
j

is an INTEGER*4 variable or array element containing the seed for computing the random number.

i1
is an INTEGER*2 variable or array element containing the seed for computing the random number.
i2
is an INTEGER*2 variable or array element containing the seed for computing the random number.
The values of it and i2 are updated during the computation of the random number to contain the
updated seed values.

2-30

Language Compatibility Issues

NOTE

The algorithm for computing the random number value is identical to the first algorithm
specified in the RANDU subroutine (see Section 2.2.6.7). The RAN function is equivalent to calling the RANDU subroutine with RAN specified as the variable to receive the
random number. These calls to the RAN function and the RANDU subroutine have the
following formats:
TYPE *.RAN(il. i2)
CALL RANDU(il. i2. ran)
TYPE *.ran

This RAN function is distinguished from the single-argument VAX FORTRAN RAN function by
the number of arguments. The single-argument form uses a statistically better algorithm and is
recommended when compatibility with PDP-ll FORTRAN is not an issue.

2.2.8

Data Definitions

There are some differences in data definition between VAX FORTRAN, FORTRAN IV, and
FORTRAN-77. The following sections describe these differences.
2.2.8.1

Floating-Point Results

If you call a math library routine, you may receive different results depending on which FORTRAN
implementation you are using. The VAX implementations of the math library routines take
advantage of the VAX instruction set, which is not available to the PDP-ll FORTRAN compilers.
These VAX functions typically produce results with an accuracy equal to or greater than the
corresponding PDP-ll functions, but in some cases there may be differences.

In addition, VAX FORTRAN does not immediately convert floating-point constants without
exponents to REAL*4, as is the case in PRO and PDP-ll FORTRAN. This feature provides greater
accuracy when such constants are used in double-precision expressions.
2.2.8.2

Character and Hollerith Constants

VAX FORTRAN and FORTRAN-77 support both Hollerith and character constants, with the
following respective notations:
nHa .. a
'a .. a"

In FORTRAN IV, both of these notations are used for Hollerith constants. (Note that Hollerith
constants have no data type; Hollerith constants assume a data type consistent with their use.)
In most cases, the conflicting use of the 'a .. a' notation is not a problem; VAX FORTRAN and
FORTRAN-77 can determine from the context of your program if you intend to use a Hollerith
constant or a character constant. There is one case in which this is not true. In an actual argument
list for a CALL or function reference, where the subprogram has a dummy argument, a constant
in the' a .. a' notation is always passed as a character constant, never as a Hollerith constant.

Language Compatibility Issues

2-31

Consider the following code segment. If subroutine F expects a Hollerith constant to be passed as
argument 5, execution of this code fragment is not correct since the actual and dummy arguments
must agree in data type.
SUBROUTINE F(S)

CALL F(' ABCD')

To avoid this problem, use the actual Hollerith constant format for specifying the argument as
follows:
SUBROUTINE F(S)

CALL F(4HABCD)

2.2.9

Expressions

The logical constants .TRUE. and .FALSE. are defined, respectively, as all ones and all zeros
by both VAX FORTRAN and PDP-II FORTRAN. However, when testing these constants, each
compiler determines the value of the constant by checking different parts of the constant itself.
The following list details the methods used by the various FORTRAN compilers to test. TRUE.
and .FALSE. constants:
•

•

VAX FORTRAN tests the low-order bit (bit 0) of a logical value. This is the system-wide VAX
convention for testing logical values. VAX FORTRAN always evaluates any odd number as
TRUE, because it tests bit O.
FORTRAN-77 tests the sign bit of a logical value: bit 7 for LOGICALo"I, bit 15 for
LOGICAL*2, and bit 31 for LOGICAL*4. FORTRAN-77 always evaluates any negative
number as TRUE, because it tests the sign bit.
FORTRAN IV tests the low-order byte of a logical value; all zeros is a .FALSE. value, and any
nonzero bit pattern is a .TRUE. value.

Figure 2-2 illustrates the data tested by each of the FORTRAN compilers.

2-32

Language Compatibility Issues

Figure 2-2:

Logical Test Comparison

FORTRAN-77

2423

VAX FORTRAN

16 15

8 7

II
t

FORTRAN IV
ZK-6613-HC

In most cases, this difference does not have any effect on the logical test comparison results. It is
significant only for nonstandard FORTRAN programs that perform arithmetic operations on logical
values and then make logical tests on the result. For example:
LOGICAL*1 BA
BA = 3
IF (BA) GO TO 10

The LOGICAL*l variable BA has the following format:

ZK-6570-HC

The various compilers interpret this LOGICAL*l value as follows:
•
•
•

FORTRAN IV produces a value of .TRUE., since it interprets any nonzero bit pattern,in the
entire byte as a .TRUE. value.
FORTRAN-77 produces a value of .FALSE., since it checks the sign bit (in this case, bit 7) to
determine a .TRUE. or .FALSE. value.
VAX FORTRAN produces a value of .TRUE., since it tests bit 0 to determine a .TRUE. or
.FALSE. value.

Language Compatibility Issues

2-33

2.2.10

Character Sets

The character sets for the various FORTRAN compilers have minor incompatibilities. Table 2-4
lists the special characters that are only valid under VAX FORTRAN.
Table 2-4:
Symbol

Character Set Incompatibilities
Character
Underscore

Percent sign

Ampersand

By avoiding these special characters, you should not encounter any character set incompatibilities
when transporting your programs.

2.3

System Dependencies
Using a system-specific feature can improve program execution speed, but does so at the expense
of transportability. If you must use the system-dependent features of FORTRAN language elements that are incompatible across systems, try to isolate them in subprograms. This minimizes
the changes you must make to convert the programs because it keeps the program logic from
being tied to these system-specific features.
The following sections discuss differences in system dependencies.

2.3.1

I/O Differences and File Transfer

If you do not specify a logical unit number in an I/O statement, a default unit number is used.
Table 2-5 shows the differences in the defaults used by VAX FORTRAN and those used by
FORTRAN IV and FORTRAN-77.

Table 2-5:

Default Logical Unit Numbers

I/O Statement

FORTRAN IV /FORTRAN-77 Unit

VAX FORTRAN Unit

READ

-4

-1

TYPE

-2

-3

The FORTRAN IV and FORTRAN-77 compilers use the normal logical unit numbers as the
defaults; VAX FORTRAN uses unit numbers that are unavailable to users. This feature prevents
conflicts between I/O statements that use the default logical unit numbers and those that use
explicit logical unit numbers. This should have no visible effect on program execution.

2-34

language Compatibility Issues

If you try to OPEN a logical unit already connected to a file, FORTRAN IV generates an error. In
VAX FORTRAN and FORTRAN-77, the behavior is as follows:
•
•

If the file specification specified (or the default) matches that of the currently opened file, the
new value (if any) of the BLANK keyword is used and the new open request is ignored.
If the file specifications do not match, the currently open file is closed and the new file is
opened.

In either case, neither VAX FORTRAN nor FORTRAN-77 generates an error, and neither language
supports any method to achieve behavior that is compatible with FORTRAN IV.

2.3.2

Optimization of I/O with Operating System-Specific Features

If you want to optimize the I/O operations performed by your programs, you can use system
services such as $QIO that are available on your system. However, these system service calls
are not totally transportable. For more information, refer to the appropriate operating system
compatibility guide.

2.3.3

File Naming Conventions

There is a difference betw~en the file naming conventions of FORTRAN IV, FORTRAN-77, and
VAX FORTRAN. Specifically, FORTRAN IV and FORTRAN-77 source code files have a default
file type of FTN. VAX FORTRAN source code files have a default file type of FOR.
There are also some differences between file naming conventions on PDP-II, PRO, and VAX
systems. The most obvious difference is in the length of allowable file names. On PDP-II and
PRO systems, a file name can consist of from 1 to 9 alphanumeric characters. On VAX systems,
a file name can consist of from 1 to 39 characters. Also, PDP-II and PRO listing files have a
default file type of LST; VAX listing files have a default file type of LIS. VAX systems also provide
a default file type for text library files (TLB); these files are not supported on PDP-II and PRO
systems. For more information on default file types, refer to the compatibility guide for the
appropriate operating system.

2.3.4 Transportable File Specification Format
There is no way to include completely transportable file specifications in a FORTRAN program.
However, there are two techniques that can make it easier to support FORTRAN programs that
include file specifications. These techniques are:
1. Using logical names
2. Assigning file specifications to string variables
The following sections describe these techniques.

Language Compatibility Issues

2-35

2.3.4.1

Using Logical Names

Logical names let you assign a mnemonic name to a file specification. Depending on the system,
you can use a logical name to represent different fields of a file specification. Table 2-6 defines
what fields of a file specification can be assigned to a logical name for each system.
Table 2-6:

Logical Name Capabilities for File Specifications

System

Logical Name Capability

RSTS

Device and account

RSX

Device

PI OS

Device

VMS

Full file specification, including node, device, and directory

Figure 2-3 illustrates the fields of a file specification that can be assigned logical nantes.
Figure 2-3:

File Specification Logical Name Comparison

VMS

node::device:[directory]filename.type;version

LJ
RSX, PRO

RSTS
ZK-6568-HC

Regardless of the system, you have to assign logical names at the monitor command level, not
within a FORTRAN program.
The main advantage of using logical names is that you can redefine at least the device portion of a
file specification at the monitor command level, regardless of the system.
2.3.4.2

File Specifications in String Variables

Placing file specifications in string variables allows you to avoid some 'Of the common problems of
hard-coding device specifications, which makes your code more transportable.
For example, if you assign a file specification to a string variable and then specify that string
variable in all statements referencing that particular file, change the assignment of the string
variable to transport your program from one system to another.

2-36

Language Compatibility Issues

2.3.5

File Support

The types of files supported by FORTRAN IV, FORTRAN-77, and VAX FORTRAN differ according to which file services are supported by that compiler. Table 2-7 shows the different types of
files supported by the various FORTRAN compilers.
Table 2-7:

Supported Files for DIGITAL FORTRAN Compilers

Compiler

Supported File Structures

FORTRAN IV

Direct access or sequential, with fixed- or variable-length records

FORTRAN-77

Direct access, sequential, relative, or indexed sequential, with fixed- or variablelength records

VAX FORTRAN

Direct access, sequential, relative sequential or direct, or indexed sequential or keyed,
with fixed- or variable-length records

For more information on file services, see the appropriate operating system compatibility manual.

2.3.6

Keywords

There are minor differences in some of the values available for the OPEN statement keywords.
The following sections describe these differences.
2.3.6.1

OPEN Statement BLANK Keyword Default

In FORTRAN-77 and VAX FORTRAN, the BLANK keyword to the OPEN statement lets you
control the interpretation of blanks in numeric input fields.
When BLANK specifies 'NULL', all blanks in a numeric input field are ignored (except if the
field is comprised of all blanks, in which case it is treated as zero).
When BLANK specifies 'ZERO', all blanks other than leading blanks are treated as zeros.

•
•

For both FORTRAN-77 and VAX FORTRAN, the default is BLANK = 'NULL' .
.The FORTRAN IV OPEN statement does not support a BLANK keyword. FORTRAN IV interprets
blanks in a numeric input field as zeros, which is identical to specifying BLANK = 'ZERO' in the
OPEN statement. If you use the /NOF77 qualifier to compile a FORTRAN-77 or VAX FORTRAN
program, or you do not explicitly open a logical unit with an OPEN statement, both FORTRAN-77
and VAX FORTRAN default to BLANK = 'ZERO'.
The following example illustrates the differences between these interpretations:
Program:
10

OPEN (UNIT=l, STATUS='OLD') READ(l,10)I, J
FORMAT (215) END

Data Record:
#1#2####12

Language Compatibility Issues

2-37

These statements under FORTRAN IV (or under FORTRAN-77 or VAX FORTRAN with the
jNOF77 qualifier) have the following results:
I = 1020

J = 12

Under FORTRAN-77 or VAX FORTRAN, these same statements have the following results:
I

= 12

J = 12

You can also use the BZ edit descriptor in the FORMAT statement to indicate that blanks be
interpreted as zeros. However, this edit descriptor only applies to that particular FORMAT
statement; it does not specify a default interpretation of blanks for an entire file.
2.3.6.2

OPEN Statement STATUS Keyword Default

In FORTRAN-77 and VAX FORTRAN, you can use the STATUS keyword to the OPEN statement
to specify the initial status of a file ('OLD', 'NEW', 'SCRATCH', or 'UNKNOWN'). In FORTRAN
IV, use the TYPE keyword instead of STATUS. (The TYPE keyword is also supported on VAX
FORTRAN and FORTRAN-77 for compatibility with FORTRAN IV.)
By default, both FORTRAN-77 and VAX FORTRAN assume a STATUS of 'UNKNOWN'. An
OPEN statement compiled under either of these compilers first searches to see if a file of that
name already exists. If so, that file is opened; if not, a new one is created. FORTRAN IV assumes
a TYPE of 'NEW' so that, by default, an OPEN statement compiled by FORTRAN IV creates a
new file with the name specified instead of first searching for any existing file of that name.
If you specify the jNOF77 qualifier to either VAX FORTRAN or FORTRAN-77, the default status

of the file is 'NEW' as in FORTRAN IV.
2.3.6.3

OPEN Statement INITIALSIZE Keyword

For VAX FORTRAN, the space requested by the INITIALSIZE keyword is allocated contiguously,
if possible, on what is called a best-try basis. If you specify an INITIALSIZE value and there is
sufficient contiguous space available, the space allocated is contiguous. If there is not enough

contiguous space, the space allocated is noncontiguous.
For FORTRAN-77, allocation of contiguous or noncontiguous space depends on the sign of the
value specified for the INITIALSIZE and EXTENDSIZE keywords. To be compatible with PDP-II
FORTRAN, VAX FORTRAN uses the absolute value of the user-supplied value.
2.3.6.4

DISPOSE

='PRINT' Specification

On RSTSjE, RSX-IIM, RSX-IIM-PLUS, and VMS systems, a file printed under DISPOSE =
'PRINT' is always saved. To print and then delete a file, specify DISPOSE = 'PRINT jDELETE'.

2-38

Language Compatibility Issues

2.3.7

Record Management Services

FORTRAN-77 provides the following methods for controlling file and record access:
•
•

File Control Services (FCS)
Record Management Services (RMS)

These file-handling facilities differ in the types of files they support, but the calls to mutually
supported routines are identical. For example, indexed or relative files are not supported under
FCS, but are supported by RMS.
If you do not use any file structures that are not supported by FCS, you can still request that
FORTRAN-77 use RMS as your file-handling facility by specifying this option at link time. VAX
FORTRAN supports only RMS; FORTRAN IV supports only FCS.

2.3.8

Block I/O

If you are using VAX FORTRAN or FORTRAN-77 with RMS, you can perform block I/O. Block
I/O is an intermediate step between RMS record operations and the direct use of the $QIO
system service. Block I/O operations let you directly read or write the blocks of a file. For more
information on block I/O, refer to the appropriate operating system compatibility manual.

2.3.9

Run-Time Libraries

Run-time libraries are provided for all PRO, PDP-ll, and VAX systems. Although these library
routines may have some internal differences, they are designed to make these differences transparent to you. Therefore, if you have a program that calls run-time libraries, that program should
be transportable between PRO, PDP-ll, and VAX systems. For more information on run-time
libraries, see the appropriate operating system compatibility manual.

2.4

Program Segmentation
It is a good idea to modularize your code, even if you are not explicitly concerned with transporta-

bility. Modularizing your code means that you take a complex task, divide it into modules, and
code each module as a separate procedure. This has several advantages over coding a complex
program as a single module. For example:
•
•
•
•
•
•

You can use any modular procedure in any of your programs.
You can add a modular procedure to a library at any time.
You do not have to rewrite common algorithms every time you need them in a program.
You can divide a complex program into simpler procedures, which reduces development time
and complexity and increases reliability.
You can modify or replace a procedure without having to modify each program that calls it.
You can significantly reduce the time and effort involved in debugging an application.

Each module should contain a single procedure or a group of related procedures.

Language Compatibility Issues

2-39

2.5

Procedure Calling
After you modularize your code, you must pull in all the modules required by your main program.
These modules can be coded and invoked either as function or subroutine subprograms.

2.5.1

Calling Function Subprograms

A function subprogram is a program unit consisting of a FUNCTION statement followed by a
series of statements that define the computing procedure of that function. To transfer control to
a function subprogram, you use what is called a function reference. A function ,reference means
using the name of the function as a variable in an assignment statement. A RETURN or END
statement returns control from the function subprogram to the calling program.
A function subprogram returns a single value to the calling program unit by assigning that value
of the function's name.
The following example illustrates the definition of a FORTRAN function subprogram, and the
statement used to transfer control to that function:
FUNCTION ROOT(A)

x = 1.0
2

EX = EXP(X)
EMINX = 1./EX
ROOT = «EX + EMINX) * .5 + COS(X) -A)/«EX - EMINX) * .5 - SIN(X))
IF (ABS(X-ROOT) .LT. 1E-6) RETURN
X = ROOT
GO TO 2
END·

This function subprogram uses the Newton-Raphson iteration method to obtain the root of the
following function:
F(X) = cosh(X) + cos(X) - A = 0
The value of A is passed as an argument. This calculation is repeated until the difference between
the values is less than 1.0E-6. To invoke this function, use the following statement:
FUNCROOT = ROOT(A)

2.5.2

Calling Subroutine Subprograms

A subroutine subprogram is a program unit consisting of a SUBROUTINE statement followed by
a series of statements that define the computing procedure. To transfer control to a subroutine
subprogram, you must use the FORTRAN CALL statement. You must also include a RETURN or
END statement to return control from the subroutine subprogram to the calling program.
The following example illustrates a simple subroutine subprogram:
SUBROUTINE ADD(SUM,A,B)
SUM = A + B
RETURN
END

2-40

Language Compatibility Issues

To invoke this subroutine subprogram, include the following statement in your calling program:
FIRST = 5.0
SECOND = 9.5
TOTAL = 0.0
CALL ADD (TOTAL , FIRST, SECOND)

2.6

Errors and Error Handling
Differences in run-time support between VAX FORTRAN and PRO and PDP-ll FORTRAN are
reflected in run-time error numbers and in run-time error reporting. The following sections discuss
these differences.

2.6.1

Run-Time Library Error Numbers

If you have a program that uses the ERRSNS subroutine, that program may need to be modi-

fied because certain PRO and PDP-ll FORTRAN run-time errors were either deleted from, or
redefined in, the VAX Run-Time Library. Table 2-8 lists the error numbers that are affected.
Table 2-8: Incompatible Error Numbers
Error Number

Incompatibility

2 through 14

Deleted; these error numbers reported fatal PDP-ll
hardware conditions.

37 (INCONSISTENT RECORD LENGTH)

Redefined; continuation is not allowed.

65 (FORMAT TOO BIG FOR "FMTBUF")

. Deleted; this error cannot occur because space is
acquired dynamically for run-time formats.

72, 73, 82, 83, 84

Redefined; floating-point arithmetic errors and. math
library errors return -0.0 (a hardware reserved
operand) rather than +0.0.

75 (FPP FLOATING TO INTEGER CONVERSION
OVERFLOW)

Deleted; error number 70 is now reported. Error
number 70 indicates ARITHMETIC TRAP, INTEGER
OVERFLOW.

86 (INVALID ERROR NUMBER)

Deleted; error number 48 is now reported. Error
number 48 indicates INVALID ARGUMENT TO
FORTRAN RUN-TIME LIBRARY.

91 (COMPUTED GO TO OUT OF RANGE)

Deleted; no error is generated by the VAX hardware
when this condition occurs. Program execution
continues in line.

92 (ASSIGNED LABEL NOT IN LIST)

Deleted; VAX FORTRAN does not perform this check
at run time.

94 (ARRAY REFERENCE OUTSIDE ARRAY)

Deleted; error number 77 is now reported. Error
number 77 indicates TRAP, SUBSCRIPT OUT OF
. RANGE.

95 through 101

Deleted; these error numbers reported PDP-11
FORTRAN errors that cannot occur in VAX
FORTRAN.

Language Compatibility Issues

2-41

2.6.2

Error Handling and Reporting

VAX FORTRAN differs from FORTRAN IV and FORTRAN-77 in the way that it treats error
continuation, I/O errors, and OPEN or CLOSE statement errors. The following sections discuss
these differences.

2.6.2.1

Continuing After Errors

In PDP-11 FORTRAN, program execution after errors, such as floating-point overflows, normally
continues until 15 such errors occur. At that point, execution is terminated. VAX FORTRAN,
however, sets a limit of one such error; program execution normally terminates when the first
error occurs. To change this behavior, you can take one of the following steps:
•

Include a condition handler in your program to change the severity level of the error. Severity
levels of Warning and Error permit continuation.
Include the ERRSET subroutine. ERRSET alters the Run-Time library's default error processing tomatch the behavior of FORTRAN-77.

•

2.6.2.2

I/O Errors with 10STAT or ERR Specified

If an 10STAT or ERR specification is included in the I/O statement, VAX FORTRAN neither

generates an error message nor increments the image error count when an I/O error occurs.
Under these circumstances, FORTRAN IV and FORTRAN-77 both report the error and increment
the task error count.

2.6.2.3

OPEN or CLOSE Statement Errors

Unlike PRO and PDP-II FORTRAN, VAX FORTRAN reports only the first error encountered
in an OPEN or CLOSE statement. PRO and PDP-II FORTRAN report all errors detected in
processing these statements.

2-42

Language Compatibility Issues

Index

/DML qualifier - 2-9
DO loops - 2-21

/ ANAL YSIS_DA T A qualifier - 2-6
ANSI Standard
FORTRAN-77-1-1,2-20
Standard X3.9-1966-1-1, 2-20
Standard X3.9-1978-1-1, 2-20
ASSIGN subroutine - 2-24
/ A switch - 2-6

B
Block I/O - 2-39
/B switch - 2-9

c
CHANGE command - 1-5
Character constants - 2-31
Character sets - 2-34
incompatibilities - 2-34t
/CHECK qualifier - 2-6
CLOSE subroutine - 2-24
/CODE qualifier - 2-7
Command qualifiers and switches - 2-4 to 2-19
Common blocks
blank - 2-22
Compiler control statements - 2-19 to 2-20
Compilers
Invoking - 2-3
/CONTINUA TIONS qualifier - 2-8
/CROSS_REFERENCE qualifier - 2-8

D
/D_LlNES qualifier - 2-9
Data definitions - 2-31
/DEBUG qualifier - 2-8
Default logical unit number - 2-34
DISPOSE = 'PRINT' specification - 2-20, 2-38
/DI switch - 2-9
/DLINES qualifier - 2-9

EDI editor
using -1-6
Edit descriptors
XformM-2-22,2-23
EDT editor
invoking - 1-5
using - 1-4
Error handling - 2-41 to 2-42
Errors - 2-41 to 2-42
ERRSET subroutine - 2-25 to 2-26
ERRTST subroutine - 2-26
EVE (Extensible V AX Editor) interface - 1-5
Expressions - 2-32 to 2-33
/EXTEND qualifier - 2-10
/EXTEND_SOURCE qualifier - 2-10
Extensible V AX Editor
see EVE
EXTERNAL statement - 2-22

F
/F77 qualifier - 2-10
FDBSET subroutine - 2-26, 2-27
File naming conventions - 2-35
File support - 2-37
File transfer - 2-34
Floating-point results - 2-31
/FOR qualifier - 2-10
Functions - 2-29 to 2-31
Function subprograms
calling - 2-40

G
/G_FLOATING qualifier- 2-10
GO TO statement
assigned - 2-21

Index-1

H
Hollerith constants· 2-31

Program segmentation· 2-39
PROSE-PLUS
using· 1-6

Q
I/O
differences. 2-34
errors • 2-42
/14 qualifier· 2-11
/IDENTIFICA TION qualifier· 2-11
INTRINSIC statement. 2-22
IRAD50 subroutine· 2-27

K
Keypad mode· 1-4
Keywords· 2-37 to 2-38

L
label lists· 2-21
/lA switch· 2-11
/UBRARY qualifier· 2-11
Line mode· 1-4
/UNE_NUMBERS qualifier· 2-11
JUST qualifier· 2-12
logical names
using· 2-36
flO switch· 2-13

M
/MACHINE_CODE qualifier. 2-13
/MAP qualifier· 2-13

/0 switch • 2-13

R
R50ASC subroutine· 2-29
RAD50 function • 2-30
RANDU subroutine· 2-28
RAN function • 2-30, 2-31
Record Management Services· 2-39
/R switch· 2-14
Run-Time libraries· 2-39
error numbers • 2-41

5
/SHAREABlE qualifier. 2-15
/SHOW qualifier· 2-15
/SOURCE qualifier. 2-16
/SP switch • 2-16
/ST ANDARD qualifier· 2-16 to 2-17
Statements • 2-20 to 2-23
for compiler control· 2-19 to 2-20
String variables
file specifications in· 2-36
Subroutines • 2-23 to 2-29
Subroutine subprograms
calling • 2-40
System dependencies· 2-34 to 2-39

Nokeypad mode· 1-4
/N switch. 2-14

TECO editor
using· 1-6
/TRACEBACK qualifier· 2-17
Transportable code· 2~ 1 to 2-2

/OBJECT qualifier· 2-14
OPEN statement
BLANK keyword default. 2-37, 2-38
INITIAlSIZE keyword • 2-38
ST A TUS keyword default. 2-38
/OPTIMIZE qualifier· 2-14
/0 switch· 2-14

USEREX subroutine· 2-29
/U switch· 2-18

p
Procedure calling • 2-40

2-lndex

v
VAXlSE
using· 1-6
V AXTPU (V AX Text Processing Utility)
using· 1-5
/VECTORS qualifier· 2-18

w
/W ARNINGS qualifier· 2-18
/WORK_FILES qualifier· 2-19

x
/X switch. 2-19

Index-3

CII
~

CII
1::1

c::I

Prin~dinU.S .A~3149141~8M433QOMEMCopyri~tI9~88~D~i~W~·t~al~E~q~u~~~m~en~t~C~o~~~o~r~a~ti~o~n~
.~~I~n~·g~h~t~s~r~~~e~~~e~d~.~~~~~~~~~~~~~~~~~~~~~~