Pascal Usermanual

Compaq Pascal User Manual for OpenVMS Systems
Order Number: AA-PXSNC-TK June 1999 This manual contains information about selected programming tasks using the Compaq Pascal programming language. It supersedes DEC Pascal User Manual for OpenVMS Systems, order number AA-PXSNB-TK.
Revision/Update Information: Software Version:
This is an updated manual. Compaq Pascal Version 5.7
Compaq Computer Corporation Houston, Texas
June 1999 Digital Equipment Corporation makes no representations that the use of its products in the manner described in this publication will not infringe on existing or future patent rights, nor do the descriptions contained in this publication imply the granting of licenses to make, use, or sell equipment or software in accordance with the description. Possession, use, or copying of the software described in this publication is authorized only pursuant to a valid written license from Digital Equipment Corporation or an authorized sublicensor. Digital Equipment Corporation 1999. All Rights Reserved. Compaq, the Compaq logo and the DIGITAL logo are Registered in the U.S. Patent and Trademark Ofce. Alpha, AlphaServer, AlphaStation, Bookreader, DEC, DEC Pascal, DIGITAL, OpenVMS, Tru64 Unix, ULTRIX, VAX, VMS, and the DIGITAL logo are trademarks of Digital Equipment Corporation. The following are third-party trademarks: IEEE is a registered trademark of the Institute of Electrical and Electronics Engineers, Inc. Oracle Rdb, Oracle CODASYL DBMS, Oracle CDD/Repository, Oracle CDD/Administrator, Oracle RALLY, Oracle TRACE, Oracle Expert, Oracle InstantSQL, Oracle Graphical Schema Editor, Oracle RMU, Oracle RMUwin, Oracle TRACE Collector, Oracle SQL/Services, Oracle DBA Workcenter, and Oracle Module Language are trademarks of Oracle Corporation. OSF/1 is a registered trademark of The Open Group. PostScript is a registered trademark of Adobe Systems, Inc. UNIX is a registered trademark in the United States and other countries licensed exclusively through The Open Group. X/Open is a trademark of The Open Group. All other trademarks and registered trademarks are the property of their respective holders. Compaq conducts its business in a manner that conserves the environment and protects the safety and health of its employees, customers, and the community.
ZK6140 This manual is available on CDROM.
This document was prepared using VAX DOCUMENT Version 2.1.
Contents
Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 Compiling Compaq Pascal Programs
1.1 PASCAL Command . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.1.1 PASCAL Command Examples . . . . . . . . . . . . . . . . . . . . . 1.1.2 PASCAL Qualiers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.1.3 Contents of the Compilation Listing File . . . . . . . . . . . . . 1.1.3.1 Table of Contents (OpenVMS VAX systems only) . . . . . . . . . . . 1.1.3.2 Source Code . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.1.3.3 Cross-Reference Section . . . . . . . . . . . . . . . . . . . . . . . 1.1.3.4 Machine Code Section . . . . . . . . . . . . . . . . . . . . . . . . 1.1.3.5 Structured Layout Section . . . . . . . . . . . . . . . . . . . . . 1.1.3.6 Inline Summary (OpenVMS VAX systems only) . . . . . . . . . . . . 1.1.3.7 Compilation Statistics . . . . . . . . . . . . . . . . . . . . . . . . 1.1.4 Text Libraries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.1.4.1 Using the %INCLUDE Directive for Text Libraries . . 1.1.4.2 Specifying Text Libraries on the PASCAL Command Line . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.1.4.3 Dening Default Libraries . . . . . . . . . . . . . . . . . . . . . 1.2 LINK Command . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.2.1 LINK Command Examples . . . . . . . . . . . . . . . . . . . . . . . 1.2.2 LINK Qualiers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.2.3 Object Module Libraries . . . . . . . . . . . . . . . . . . . . . . . . . . 1.3 RUN Command . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.4 Error Messages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 13 13 124 125 125 125 125 126 126 126 127 127 128 129 129 130 131 134 135 136 xi
iii
2 Separate Compilation
2.1 2.2 2.3 2.4 The ENVIRONMENT, HIDDEN, and INHERIT Attributes . Interfaces and Implementations . . . . . . . . . . . . . . . . . . . . . . Data Models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Separate Compilation Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 28 212 216
3 Program Correctness, Optimization, and Efciency

3.1 Compiler Optimizations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1.1 Compile-Time Evaluation of Constants . . . . . . . . . . . . . . 3.1.2 Elimination of Common Subexpressions . . . . . . . . . . . . . 3.1.3 Elimination of Unreachable Code . . . . . . . . . . . . . . . . . . . 3.1.4 Code Hoisting from Structured Statements . . . . . . . . . . . 3.1.5 Inline Code Expansion for Predeclared Functions . . . . . . 3.1.6 Inline Code Expansion for User-Declared Routines . . . . . 3.1.7 Testing for Inline Expansion on OpenVMS VAX Systems 3.1.8 Operation Rearrangement . . . . . . . . . . . . . . . . . . . . . . . . 3.1.9 Partial Evaluation of Logical Expressions . . . . . . . . . . . . 3.1.10 Value Propagation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1.11 Strength Reduction (OpenVMS Alpha systems only) . . . . . . . . . . . 3.1.12 Split Lifetime Analysis (OpenVMS Alpha systems only) . . . . . . . . 3.1.13 Code Scheduling (OpenVMS Alpha systems only) . . . . . . . . . . . . . 3.1.14 Loop Unrolling (OpenVMS Alpha systems only) . . . . . . . . . . . . . . . 3.1.15 Alignment of Compiler-Generated Labels . . . . . . . . . . . . 3.1.16 Error Reduction Through Optimization . . . . . . . . . . . . . . 3.1.17 Software Pipelining (OpenVMS Alpha systems only) . . . . . . . . . . . 3.1.18 Processor Tuning (OpenVMS Alpha systems only) . . . . . . . . . . . . . 3.1.19 Compiling for Optimal Performance . . . . . . . . . . . . . . . . . 3.2 Programming Considerations . . . . . . . . . . . . . . . . . . . . . . . . . 3.3 Implementation-Dependent Behavior . . . . . . . . . . . . . . . . . . . 3.3.1 Subexpression Evaluation Order . . . . . . . . . . . . . . . . . . . 3.3.2 MAXINT and MAXINT64 Predeclared Constants . . . . . . 3.3.3 Pointer References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3.4 Variant Records . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3.5 Atomicity, Granularity, Volatility, and Write Ordering . . . 3.3.6 Debugging Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 33 34 35 35 36 36 36 37 38 38 39 39 310 310 311 311 312 313 313 314 315 316 316 317 318 318 320
iv
4 Programming Tools
4.1 4.2 4.2.1 4.2.2 4.2.3 4.3 4.3.1 4.3.2 Debugger Support for Compaq Pascal . . . . . . . . . . . . . . . . . . . Language-Sensitive Editor/Source Code Analyzer Support for Compaq Pascal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Programming Language Placeholders and Tokens . . . . . . . Placeholder and Design Comment Processing . . . . . . . . . . LSE and SCA Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . Accessing CDD/Repository from Compaq Pascal . . . . . . . . . . . Equivalent Compaq Pascal and CDDL Data Types . . . . . . CDD/Repository Example . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . 41 42 43 43 45 48 410 412
5 Calling Conventions
5.1 5.1.1 5.1.2 5.1.3 5.1.4 5.1.5 5.2 5.3 5.3.1 5.3.2 5.3.3 5.3.3.1 5.3.3.2 5.3.3.3 5.3.3.4 5.3.4 5.4 5.4.1 5.4.2 5.4.3 OpenVMS Calling Standard . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Parameter Lists . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Function Return Values . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Contents of the OpenVMS Alpha Call Stack . . . . . . . . . . . . . Contents of the OpenVMS VAX Call Stack . . . . . . . . . . . . . . Unbound Routines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Parameter-Passing Semantics . . . . . . . . . . . . . . . . . . . . . . . . . . . Parameter-Passing Mechanisms . . . . . . . . . . . . . . . . . . . . . . . . . By Immediate Value . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . By Reference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . By Descriptor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . CLASS_S Attribute . . . . . . . . . . . . . . . . . . . . . . . . . . . . . CLASS_A and CLASS_NCA Attributes . . . . . . . . . . . . . . %STDESCR Mechanism Specier . . . . . . . . . . . . . . . . . . %DESCR Mechanism Specier . . . . . . . . . . . . . . . . . . . . Summary of Passing Mechanisms and Passing Semantics . . . Passing Parameters between Compaq Pascal and Other Languages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Parameter Mechanisms Versus Parameter Semantics . . . . . . Passing Nonroutine Parameters between Compaq Pascal and Other Languages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Passing Routine Parameters between Compaq Pascal and Other Languages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51 52 52 53 54 54 55 55 57 57 57 511 512 512 513 514 515 515 516 518
6 Programming on OpenVMS Systems

6.1 6.2 6.2.1 6.2.2 6.2.3 6.2.4 6.2.5 6.2.6 6.3 6.4 6.5 6.6 6.7 6.7.1 6.7.2 Using System Denitions Files . . . . . . . . . . . . . . . . . . . . . . Declaring System Routines . . . . . . . . . . . . . . . . . . . . . . . . . Methods Used to Obtain OpenVMS Data Types . . . . . . . Methods Used to Obtain Access Methods . . . . . . . . . . . . Methods Used to Obtain Passing Mechanisms . . . . . . . . Data Structure Parameters . . . . . . . . . . . . . . . . . . . . . . Default Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . Arbitrary Length Parameter Lists . . . . . . . . . . . . . . . . . Calling System Routines . . . . . . . . . . . . . . . . . . . . . . . . . . . Using Attributes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Using Item Lists . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Using Foreign Mechanism Speciers on Actual Parameters Using 64-Bit Pointer Types on OpenVMS Alpha . . . . . . . . . Pascal Language Features Not Supported with 64-Bit Pointers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Using 64-Bit Pointers with System Denition Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61 66 67 67 68 610 611 613 613 614 615 617 618 618 620
.... ....
7 Input and Output Processing

Environment I/O Support . . . . . . . . . . . . . . . . . . . . . . . . . . . Indexed Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . OpenVMS Components and RMS Records . . . . . . . . . . . Count Fields for Variable-Length Components . . . . . . . . Variable-Length with Fixed-Length Control Field (VFC) Component Format . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.1.5 Random Access by Record File Address (RFA) . . . . . . . . 7.1.6 OPEN Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.1.6.1 OPEN Defaults . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.1.6.2 OPEN and RMS Data Structures . . . . . . . . . . . . . . . 7.1.7 Default Line Limits . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.2 User-Action Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.3 File Sharing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.4 Record Locking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.1 7.1.1 7.1.2 7.1.3 7.1.4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71 72 72 72 73 73 73 74 75 710 710 715 716
vi
8 Error Processing and Condition Handling

8.1 8.2 8.2.1 8.2.2 8.3 8.3.1 8.3.2 8.3.3 8.3.4 8.3.5 8.4 8.5 Condition Handling Terms . . . . . . . . . . . . . . . . . . . . . . . . . . . Overview of Condition Handling . . . . . . . . . . . . . . . . . . . . . . Condition Signals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Handler Responses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Writing Condition Handlers . . . . . . . . . . . . . . . . . . . . . . . . . . Establishing and Removing Handlers . . . . . . . . . . . . . . . Declaring Parameters for Condition Handlers . . . . . . . . . Handler Function Return Values . . . . . . . . . . . . . . . . . . . Condition Values and Symbols . . . . . . . . . . . . . . . . . . . . . Using Condition Handlers that Return SS$_CONTINUE Fault and Trap Handling (OpenVMS VAX systems only) . . . . . . . . . . . Examples of Condition Handlers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82 83 83 83 84 84 86 87 88 88 810 811
9 Migrating from OpenVMS VAX to OpenVMS Alpha

9.1 9.2 9.3 9.4 9.5 9.6 9.7 9.7.1 9.7.2 9.7.3 9.7.4 9.7.5 9.7.6 9.7.7 Sharing Environment Files Across Platforms . . . . . . . . . . . Default Size for Enumerated Types and Booleans . . . . . . . . Default Data Layout for Unpacked Arrays and Records . . . Overow Checking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bound Procedure Values . . . . . . . . . . . . . . . . . . . . . . . . . . . Different Descriptor Classes for Conformant Array Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Data Layout and Conversion . . . . . . . . . . . . . . . . . . . . . . . . Natural Alignment, VAX Alignment, and Enumeration Sizes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Compaq Pascal Features Affecting Data Alignment and Size . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Optimal Record Layout . . . . . . . . . . . . . . . . . . . . . . . . . Optimal Data Size . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Converting Existing Records . . . . . . . . . . . . . . . . . . . . . Applications with No External Data Dependencies . . . . Applications with External Data Dependencies . . . . . . . . . . . . . . . . . . . . . . . . . . . 91 91 92 92 92 93 93 94 95 95 97 98 98 910
.... .... .... . . . . . . . . . . . . . . . . . . . . . . . .
A Errors Returned by STATUS and STATUSV Functions
vii
B Entry Points to Compaq Pascal Utilities

B.1 B.2 B.3 B.4 PAS$FAB( f ) . . . . . . PAS$RAB( f ) . . . . . . PAS$MARK2( s ) . . . PAS$RELEASE2( p ) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B1 B2 B2 B3
C Diagnostic Messages
C.1 C.2 Compiler Diagnostics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Run-Time Diagnostics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C1 C65
Index Examples
21 22 23 41 42 43 61 62 71 An Interface Module for Graphics Objects and Routines . . . An Implementation Module for Graphics Objects and Routines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A Graphics Main Program . . . . . . . . . . . . . . . . . . . . . . . . . . Using LSE to Create a FOR Statement . . . . . . . . . . . . . . . . Using LSE Comments in Program Design (OpenVMS VAX systems only) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Using %DICTIONARY to Access a Oracle CDD/Repository Record Denition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Inheriting STARLET.PEN to Call SYS$HIBER . . . . . . . . . . Using $GETJPIW to Retrieve a Process Name . . . . . . . . . . User-Action Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 216 217 218 45 46 413 65 611 713
Figures
21 22 23 Cascading Inheritance of Environment Files . . . . . . . . . . . . . Inheritance Path of an Interface, an Implementation, and a Program . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cascading Using the Interface and Implementation Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 28 210
viii
Tables
1 11 12 13 14 15 16 17 18 19 110 111 112 113 114 41 42 43 51 52 53 54 61 62 63 71 72 73 74 75 Conventions Used in This Manual . . . . . . . . . . . . . . . . . . . . /ALIGN Qualier Options . . . . . . . . . . . . . . . . . . . . . . . . . . /ARCHITECTURE Qualier Options . . . . . . . . . . . . . . . . . . /CHECK Qualier Options . . . . . . . . . . . . . . . . . . . . . . . . . . /DEBUG Qualier Options . . . . . . . . . . . . . . . . . . . . . . . . . . /DESIGN Qualier Options . . . . . . . . . . . . . . . . . . . . . . . . . /ENUMERATION_SIZE Qualier Options . . . . . . . . . . . . . . /FLOAT Qualier Options . . . . . . . . . . . . . . . . . . . . . . . . . . /OPTIMIZE Qualier Options . . . . . . . . . . . . . . . . . . . . . . . /PLATFORMS Qualier Options . . . . . . . . . . . . . . . . . . . . . /SHOW Qualier Options . . . . . . . . . . . . . . . . . . . . . . . . . . . /STANDARD Qualier Options . . . . . . . . . . . . . . . . . . . . . . /TERMINAL Qualier Options . . . . . . . . . . . . . . . . . . . . . . . /USAGE Qualier Options . . . . . . . . . . . . . . . . . . . . . . . . . . Compilation Listing Contents and Qualiers . . . . . . . . . . . . Placeholders Within the Declaration Section . . . . . . . . . . . . Placeholders Within the Executable Section . . . . . . . . . . . . Equivalent CDD/Repository Language and Compaq Pascal Data Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Parameter-Passing Descriptions on OpenVMS Alpha Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Parameter-Passing Methods on Compaq Pascal . . . . . . . . . . Parameter Descriptors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Summary of Passing Mechanisms and Passing Semantics . . Compaq Pascal Denitions Files . . . . . . . . . . . . . . . . . . . . . Access Type Translations . . . . . . . . . . . . . . . . . . . . . . . . . . . Mechanism Type Translations . . . . . . . . . . . . . . . . . . . . . . . Default Values for OpenVMS File Specications . . . . . . . . . Setting of RMS File Access Block Fields by a Call to the OPEN Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Setting of RMS Record Access Block Fields by a Call to the OPEN Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Setting of Extended Attribute Block Fields by a Call to the OPEN Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Setting of Name Block Fields by a Call to the OPEN Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiv 14 15 17 110 111 111 113 116 117 118 119 120 121 124 44 44 411 56 56 58 514 62 67 68 74 75 78 79 79
ix
91 A1 C1
Unpacked Sizes of Fields and Components . . . . . . . . . . . . . . STATUS and STATUSV Return Values . . . . . . . . . . . . . . . . . Compiler Message Warning Levels . . . . . . . . . . . . . . . . . . . . .
95 A2 C1
Preface
This manual describes selected programming tasks using the Compaq Pascal programming language. It contains information on using some Compaq Pascal language elements in combination, and it provides examples of how to improve programming efciency. All references to VMS systems refer to OpenVMS Alpha operating systems and OpenVMS VAX systems unless otherwise specied. You can use the information in this manual to write programs or modules for the OpenVMS operating system. If you need to write Compaq Pascal programs or modules that must be compiled by another compiler, see the Compaq Pascal Language Reference Manual for a checklist of language extensions not included in the Pascal standard. The Compaq Pascal Language Reference Manual also provides information on the Pascal standard.
Intended Audience
This manual is intended for experienced applications programmers with a rm understanding of the Pascal language. Some familiarity with the operating system is helpful.
Document Structure
This manual consists of the following chapters and appendixes: Chapter 1 provides information on compiling programs, linking programs, running programs, and using text and object-module libraries. Chapter 2 describes the use of separately compiled modules. Chapter 3 describes programming techniques that improve the efciency of compilation and execution. Chapter 4 provides information on tools that you may want to use to develop Compaq Pascal programs.
xi
Chapter 5 provides information on the OpenVMS Calling Standard as applied to Compaq Pascal programs. Chapter 6 provides information on Compaq Pascal system denitions les, and declaring and calling system routines. Chapter 7 provides information on the relationship between Compaq Pascal input and output, and the underlying OpenVMS VAX Record Management Services (RMS) constructs. Chapter 8 provides information on error processing and writing condition handlers. Chapter 9 provides information on migrating from OpenVMS to OpenVMS Alpha systems. Appendix A provides a list of possible error values returned by the STATUS and STATUSV functions. Appendix B provides a list of entry points to utilities in the OpenVMS Run-Time Library that can be called as external routines by a Compaq Pascal program. Appendix C provides descriptions of diagnostic messages that can be generated by a Compaq Pascal program at compile time and run time.
Related Documents
The following manuals may also be useful when programming in Compaq Pascal: Compaq Pascal Language Reference ManualProvides information on the syntax and semantics of the Compaq Pascal programming language. In addition, this manual provides information about the alignment, allocation, and internal representation of data types supported by Compaq Pascal and descriptions of the error messages. Compaq Pascal Installation Guide for OpenVMS SystemsProvides information on how to install Compaq Pascal on your operating system.
Readers Comments
Compaq welcomes your comments. If you would like to comment on a Compaq Pascal manual, please send the manual title, order number, and your comments by one of the following methods: FAX: 6038840120 Attn: Languages Documentation, ZK023/K35
xii
A letter sent to the following address: Compaq Computer Corporation Languages Documentation, ZK023/K35 110 Spit Brook Road Nashua, NH 030622698 USA
Compaq Pascal Home Page

You can access the Compaq Pascal home page by going to: http://www.compaq.com/openvms Follow the Production Software link, then the Pascal link. Please visit the web site and follow the "Pascal Feedback" link. By registering with us, we can send you updates on future Compaq Pascal releasess.
How to Order Additional Documentation

Use the following World Wide Web page to order additional documentation: http://www.compaq.com/openvms
To reach the OpenVMS documentation web site, click the Documentation link. If you need help deciding which documentation best meets your needs, call: 800ATCOMPAQ
xiii
Conventions
Table 1 presents the conventions used in this manual. Table 1 Conventions Used in This Manual
Convention $ Meaning The default user prompt is your system name followed by a right angle bracket. This manual uses a dollar sign ( $ ) to represent this prompt. A vertical ellipsis in a gure or example means that not all of the statements are shown.
. . .
WRITELN
Uppercase letters and special symbols in syntax descriptions indicate reserved words and predeclared identiers. For example:
BEGIN END
Temp : INTEGER; PRED( n ) $_File: $ PASCAL MYPROG Lowercase letters represent user-dened identiers or elements that you must replace according to the description in the text. In interactive examples, prompts or displayed text appears in a monospace font. User input appears in bold monospace font. This typeface indicates the name of a command, ag, pathname, le name, directory path, or partition. This typeface is also used in examples of program code, interactive examples, and other screen displays. Large braces enclose a list from which you must choose one item. For example:
monospaced
8 9
STATUS DISPOSE DISP
[]
Square brackets enclose items that are optional. For example:
BLOCK DATA [nam]

(continued on next page)
xiv
Table 1 (Cont.) Conventions Used in This Manual

Convention { Source File: EXAMPLE.PAS } Meaning Many of the code examples in this manual have a suggested source-le name in the comments at the top of the program. Compaq Pascal ships some (but not all) of the examples in this manual on the kit; these les are named according to the source-le name presented in this book. For information on the location of these les, refer to the release notes. Refers to OpenVMS Alpha operating systems and OpenVMS VAX operating systems unless otherwise specied. This manual uses labels to indicate information that applies to one or more platforms.
OpenVMS systems
(OpenVMS Alpha systems only) (OpenVMS VAX systems only)
In this manual, complex examples have been divided into several lines to make them easy to read. Compaq Pascal does not require that you format your programs in any particular way.
xv
1
Compiling Compaq Pascal Programs
After you use a text editor to write code to a le, you use Digital Command Language (DCL) commands to compile modules and programs, and to link and run programs. This chapter covers the following topics: The PASCAL command (Section 1.1) The LINK command (Section 1.2) The RUN command (Section 1.3) Error messages (Section 1.4)
For information on DCL syntax, see HELP or the documentation set for your operating system.
1.1 PASCAL Command

The PASCAL command invokes the Compaq Pascal compiler, which veries program source statements, issues error messages, and generates and groups machine language instructions into an object module for the linker. The default for compiler output les (object modules) is the .OBJ le type. The Pascal command format is as follows: PASCAL [[{/command-qualier} . . . ]] {le-spec[[{/le-qualier} . . . ]]}
n+ o
,
...
8 le-spec > > identier > > > < (identier, . . . ) /qualier [[= >n > > > > : name=identier, . . .
directory, . . .
9 > > > > > = > > > > > ;
]]
Compiling Compaq Pascal Programs 11
/command-qualier The name of a qualier that indicates special processing to be performed by the compiler on all les listed. le-spec The name of one of the following: The input source le that contains the program or module to be compiled. If you separate multiple source le specications with commas, the programs are compiled separately. If you separate the le specications with plus signs, the les are concatenated and compiled as one program. The default le type for an input le is either .PAS (source le) or .TLB (text-library module). The output le used only with the /ANALYSIS_DATA, /ENVIRONMENT, /LIST, /OBJECT, or /DIAGNOSTICS qualiers.
/le-qualier The name of a qualier that indicates special processing to be performed by the compiler on the les to which the qualier is attached. identier The name of one or more options that modify the /ALIGN, /CHECK, /DEBUG, /DESIGN, /FLOAT, /OPTIMIZE, /SHOW, /STANDARD, /TERMINAL, and /USAGE qualiers. n The value of an integer constant. When used with the /ERROR_LIMIT qualier, this parameter indicates the maximum number of errors to be detected before compilation ceases. When used with the /OPTIMIZATION=LEVEL=n qualier, this parameter indicates a specic level of optimization. name=value The denition of a constant or name with the specied value when used with the /CONSTANT qualier. directory The input directory that contains the include le, environment le, or text library processed by a %INCLUDE directive or [INHERIT] attribute when used with the /INCLUDE qualier.
12 Compiling Compaq Pascal Programs
1.1.1 PASCAL Command Examples

This section contains examples of PASCAL command lines. $ PASCAL/LIST [DIR]M The source le M.PAS in directory [DIR] is compiled, producing an object le named M.OBJ and a listing le named M.LIS. The compiler places the object and listing les in the default directory. $ PASCAL/LIST A, B, C Source les A.PAS, B.PAS, and C.PAS are compiled as separate les, producing object les named A.OBJ, B.OBJ, and C.OBJ, and listing les named A.LIS, B.LIS, and C.LIS. $ PASCAL X + Y + Z Source les X.PAS, Y.PAS, and Z.PAS are concatenated and compiled as one le producing an object le named X.OBJ. By default, batch mode produces a listing le, which takes its name from the name of the rst le on the command line. In this example, the name of the listing le would be X.LIS. $ PASCAL A, B, C+D/LIST, F When a qualier follows the le specication, it applies only to the le immediately preceding it. Files concatenated with plus signs are considered one le. This command line produces four object les, A.OBJ, B.OBJ, C.OBJ, and F.OBJ, and one listing le, D.LIS. $ PASCAL A + CIRC/NOOBJECT + X This command completely suppresses the object le; that is, source les A.PAS, CIRC.PAS, and X.PAS are concatenated and compiled, but no object le is produced.
1.1.2 PASCAL Qualiers

This section describes the command and le qualiers that you can use when compiling code. /ALIGN=keyword Controls the default alignment rules. Note that specifying the ALIGN attribute overrides any value that was previously specied for the /ALIGN qualier.
Table 11 lists the values for keyword. Table 11 /ALIGN Qualier Options
Option Alpha_AXP Action Uses natural alignment when positioning record elds or array components. Natural alignment is when a record eld or an array component is positioned on a boundary based on its size. For example, 32-bit integers are aligned on the nearest 32-bit boundary. Uses byte alignment when positioning record elds or array components. Record elds or array components larger than 32 bits are positioned on the nearest byte boundary. Default Information Default on OpenVMS Alpha systems if /ALIGN is not specied.
VAX
Default on OpenVMS VAX systems if /ALIGN is not specied.
On OpenVMS VAX systems, when you specify a value of Alpha_AXP, automatic variables are aligned on longword boundaries instead of quadword boundaries. This occurs because the largest allowable alignment for the stack is longword alignment. /[NO]ANALYSIS_DATA /NOANALYSIS_DATA (default) Creates a le containing source code analysis information. If you omit the le specication, the analysis le defaults to the name of your source le with a .ANA le type. The source code analysis le is used with products such as the Language-Sensitive Editor/Source Code Analyzer (LSE/SCA). [/NO]ARCHITECTURE /ARCHITECTURE=GENERIC (default) (OpenVMS Alpha systems only) Species which version of the Alpha architecture for which to generate instructions. All Alpha processors implement a core set of instructions and, in some cases, the following extensions: BWX (byte- and word- manipulation instructions) and MAX (multimedia instructions). See the Alpha Architecture Reference Manual for additional information. Table 12 lists the available keywords:
Table 12 /ARCHITECTURE Qualier Options

Option GENERIC Action Generate instructions that are appropriate for all Alpha processors. Generate instructions for the processor on which the compiler is running (for example, EV56 instructions on an EV56 processor, and EV4 instructions on an EV4 processor). Generate instructions for the EV4 processor (21064, 20164A, 21066, and 21068 chips) and EV5 processor (some 21164 chips). (Note that the EV5 and EV56 processors both have the same chip number - 21164.) Generate instructions for EV56 processors (some 21164 chips). This option permits the compiler to generate any EV4 instruction, plus any instructions contained in the BWX extension. Applications compiled with this option may incur emulation overhead on EV4 and EV5 processors. PCA56 Generate instructions for PCA56 processors (21164PC chips). This option permits the compiler to generate any EV4 instruction, plus any instructions contained in the BWX and MAX extensions. However, Compaq Pascal does not generate any of the instructions in the MAX (multimedia) extension to the Alpha architecture. (continued on next page) Default Information This option is the default.
HOST
EV4, EV5
EV56
Table 12 (Cont.) /ARCHITECTURE Qualier Options

Option EV6 Action Generate instructions for EV6 processors (21264 chips). This option permits the compiler to generate any EV4 instruction, any instruction contained in the BWX and MAX extensions, plus any instructions added for the EV6 chip. These new instructions include a oatingpoint square root instruction (SQRT), integer/oating-point register transfer instructions, and additional instructions to identify extensions and processor groups. Applications compiled with this option may incur emulation overhead on EV4, EV5, EV56, and PCA56 processors. Default Information
Beginning with OpenVMS Alpha V7.1 and continuing with subsequent versions, the operating system includes an instruction emulator. This capability allows any Alpha chip to execute and produce correct results from Alpha instructions, regardless of whether some of the instructions are not implemented on the chip. Applications using emulated instructions will run correctly, but may incur signicant emulation overhead at run time. /[NO]CHECK /CHECK=(BOUNDS,DECLARATIONS) (default) Directs the compiler to generate code to perform run-time checks. A single identier or a list of identiers enclosed in parentheses can follow /CHECK; these identiers are the names of options that tell the compiler which aspects of the compilation unit to check. The system issues an error message and normally terminates execution if any of the conditions in the options list occur. Table 13 lists the available checking options, their corresponding actions, and their negations.
Table 13 /CHECK Qualier Options

Option ALL BOUNDS Action Generates checking code for all options. Veries that an index expression is within the bounds of an arrays index type, that character-string sizes are compatible with the operations being performed, and that schemata are compatible. Veries that the value of a case selector is contained in the corresponding case-label list. Veries that schema denitions yield valid types and that uses of GOTO from one block to an enclosing block are correct. Also controls whether the ASSERT statement is processed. Veries that the result of an integer computation does not exceed the machine representation. Veries that the value of a pointer variable is not NIL. Veries that values assigned to variables of subrange types are within the subrange; veries that a set expression is assignment compatible with a set variable; veries that MOD operates on positive numbers. Negation NONE NOBOUNDS
CASE_SELECTORS
NOCASE_SELECTORS
DECLARATIONS
NODECLARATIONS
OVERFLOW
NOOVERFLOW
POINTERS SUBRANGE
NOPOINTERS NOSUBRANGE
The BOUNDS and DECLARATIONS options are the only checking options enabled by default. The /CHECK qualier without options is equivalent to /CHECK=ALL. The negation /NOCHECK is equivalent to /CHECK=NONE. The CHECK attribute in the source program or module overrides the /CHECK qualier on the command line.
/CONSTANT=(name=value, . . . ) The /CONSTANT qualier allows a limited set of Pascal constants to be dened from the command line. This capability can be used to augment the conditional-compilation facility. name is the name of a Pascal constant to create. You cannot dene any predeclared Pascal name by the command line. value can be one of the following: integer-literal integer-literal TRUE FALSE "string-literal" string-literal
Nonbase-10 integer literals are not supported on the command line. For example: $ PASCAL/CONSTANT=(DEBUG=TRUE,MAXSIZE=10,OFFSET=-10,IDENT="V1.0") Note that the denition of /CONSTANT is such that DCL does not remove any double-quote characters used in the name=value clauses. All characters are literally passed to the compiler for processing. This behavior is slightly different from the usual behavior of putting items in double quotes on DCL commands to preserve the case, but not to pass the double quotes to the target application. This denition of /CONSTANT allows you to dene Pascal string literals with embedded single quotes and for DCL symbol substitutions. For example, $ PASCAL/CONSTANT=MSG="Special compile run for Monday" $ IDENT = "V1.0" $ PASCAL/CONSTANT=MSG="IDENT" Inserting double-quote characters and inserting adjacent single-quote characters can be accomplished by using the \ escape character allowed in Compaq Pascal double-quoted string constants. By using the \ single-quote escape character, you can insert adjacent single quotes without DCL interpreting it as a symbol substition, as shown in the following example: $ PASCAL/CONSTANT=MSG="String with 2 \\ single quote characters"
Do not use \" to insert a double-quote character into the string literal, as DCL will interpret the double quote as the end of the string. Instead, use the \x22 character literal (16#22 is the ASCII code for the double-quote character) to insert a double-quote character into the string literal without DCL interpreting it as the end of the string. This is shown in the following example: $ PASCAL/CONSTANT=MSG="String with a \x22 double-quote character" To use a single-quote string literal with /CONSTANT, enclose the entire name=value clause in double quotes to prevent DCL from trying to perform symbol substitution when it sees the single-quote character. For example: $ PASCAL/CONSTANT="MSG=Single-quoted literal" In this case, the double quotes are discarded by the compiler, and the singlequoted string literal is processed. However, using double-quoted literals with /CONSTANT is easier and more exible. The extended-string syntax for string literals is not supported on the command line. The extended-string syntax is as follows: {printing-string({constant-expression},...)}... {"printing-string"({constant-expression},...)}... These extended-string literals are constant expressions, not simple literals. To insert nonprintable characters into a string literal from the command line, you can use a double-quoted string literal and the \xnn escape sequence. /[NO]CROSS_REFERENCE /NOCROSS_REFERENCE (default) Produces a cross-reference section within the listing le. The compiler ignores this qualier if you do not also specify /LIST on the same command line. /[NO]DEBUG /DEBUG=TRACEBACK (default) Species that the compiler is to generate information for use by the debugger and the run-time error traceback mechanism. A single identier or a list of identiers enclosed in parentheses can follow /DEBUG; these identiers are the names of options that inform the compiler which type of information it should generate. Table 14 lists the available options, their corresponding actions, and their negations.
Table 14 /DEBUG Qualier Options

Option ALL Action Species that the compiler should include symbol and traceback information in the object module. Species that the compiler should include in the object module symbol denitions for all identiers in the compilation. Species that the compiler should include in the object module traceback information permitting virtual addresses to be translated into source program routine names and compilergenerated line numbers. Negation NONE
SYMBOLS
NOSYMBOLS
TRACEBACK
NOTRACEBACK
When debugging programs that contain schema, you must use the /NOOPTIMIZE qualier on the PASCAL DCL command. If you do not use /NOOPTIMIZE, you might receive incorrect debug information or an Internal Debug Error when manipulating schema. When you specify SYMBOLS without TRACEBACK, the table of compilergenerated line numbers is omitted from the debugger symbol table. You should consider using /NOOPTIMIZE when you are using /DEBUG. Allowing optimizations to occur can make debugging difcult and can obscure some sections of the compilation unit. The /DEBUG qualier without options is equivalent to /DEBUG=ALL. The negation /NODEBUG is equivalent to /DEBUG=NONE. /[NO]DESIGN /NODESIGN (default) Directs the compiler to accept design phase placeholders on all OpenVMS systems and comments on OpenVMS VAX systems as valid program elements within a Compaq Pascal program. Placeholders are produced by you or by LSE/SCA; design comments are intended for use with LSE/SCA. Table 15 lists the options, their corresponding action, and their negation.
Table 15 /DESIGN Qualier Options

Option PLACEHOLDERS Action Directs the compiler to accept placeholders as valid program elements. Directs the compiler to recognize design comments. Negation NOPLACEHOLDERS
COMMENTS
(OpenVMS VAX systems only)
NOCOMMENTS
The /DESIGN qualier without an option is equivalent to /DESIGN=(PLACEHOLDERS) on OpenVMS Alpha systems and /DESIGN=(PLACEHOLDERS, COMMENTS) on OpenVMS VAX systems. /[NO]DIAGNOSTICS /NODIAGNOSTICS (default) Creates a le containing compiler messages and diagnostic information. If you omit the le specication, the diagnostics le defaults to the name of your source le with a .DIA le type. The diagnostics le is used by products such as LSE/SCA.. /ENUMERATION_SIZE=keyword Controls the allocation of unpacked enumerated data types and Boolean data types, which are considered to be an enumerated type containing two elements. Note that specifying the ENUMERATION_SIZE attribute overrides any value you previously specied with this qualier. Table 16 lists the available values for keyword. Table 16 /ENUMERATION_SIZE Qualier Options
Option BYTE Action Allocates unpacked enumerated data types with up to 255 elements in a single byte. Otherwise, enumerated data types are allocated in a 16-bit word. Default Information Default on OpenVMS VAX systems.
Table 16 (Cont.) /ENUMERATION_SIZE Qualier Options

Option LONG Action Allocates all unpacked enumerated data types in a 32-bit longword. Default Information Default on OpenVMS Alpha systems.
/[NO]ENVIRONMENT determined by attributes (default) Produces an environment le in which declarations and denitions made at the outermost level of a compilation unit are saved. The default le name is the same as the source le name. The default le type is .PEN, an abbreviation for Pascal Environment. You can provide a different name for the environment le by including a le specication after the /ENVIRONMENT qualier, for example, /ENVIRONMENT=MASTER.PEN. The /ENVIRONMENT qualier on the command line overrides the ENVIRONMENT attribute in the source program or module. By default, the attributes of the source program or module determine whether an environment le is created; however, if the /ENVIRONMENT qualier is specied at compile time, an environment le will always be created. /[NO]ERROR_LIMIT /ERROR_LIMIT=30 (default) Terminates compilation after the occurrence of a specied number of error messages, excluding warning-level and information-level errors. If you specify /NOERROR_LIMIT, compilation continues until 500 errors have been detected. /FLOAT=oattype Selects the default format for REAL and DOUBLE data types. You must specify oattype if you use the /FLOAT qualier. Table 17 lists the available options, their corresponding actions, and default information. If the source program includes the [NO]G_FLOATING attribute, then the value of the /FLOAT qualier must be in agreement with the value of the attribute.
Table 17 /FLOAT Qualier Options

Option D_FLOAT Action REAL data type will be dened in the F_oating-point format; DOUBLE will be dened in the D_oating-point format. REAL data type will be dened in the F_oating-point format; DOUBLE will be dened in the G_oating-point format. REAL data type is dened in the IEEE S oating-point format; DOUBLE is dened in the IEEE T oating-point format. Default Information Default for OpenVMS VAX systems if /FLOAT or /NOG_FLOATING is not specied. Default for OpenVMS Alpha systems if /FLOAT or /NOG_FLOATING is not specied.
G_FLOAT
IEEE_FLOAT
Routines and compilation units between which double-precision quantities are passed should not mix oating-point formats. le-spec/LIBRARY none (default) Species that a le is a text library le. The text library le specication is required. The text library les in a list of source les must be concatenated by plus signs. The default le type is .TLB. /GRANULARITY=keyword (OpenVMS Alpha systems only) Directs the compiler to generate additional code to preserve the indicated granularity. Granularity refers to the amount of storage that can be modied when updating a variable. You can specify the following values for keyword: BYTE LONGWORD QUADWORD ( Default for OpenVMS Alpha systems )
To update a variable that is smaller than a longword on older Alpha systems, Compaq Pascal must issue multiple instructions to fetch the surrounding longword or quadword, update the memory inside to longword or quadword, and then write the longword or quadword back into memory. If multiple processes are writing into memory that is contained in the same longword or quadword, you might incur inaccurate results, unless /GRANULARITY=BYTE or some other synchronization mechanism is used.
on newer Alpha systems, the architecture has additional instructions that can modify byte and word-sized data directly. See the /ARCHITECTURE qualier for additional information. /INCLUDE=(directory, . . . ) Allows you to specify search locations for %INCLUDE directives and [INHERIT] attributes that specify le names without explicit disk or directory specications. The qualier takes a list of directories to search. The compiler applies the /INCLUDE information to the following Pascal constructs: %INCLUDE name or %INCLUDE name.ext [INHERIT(name)] or [INHERIT(name.ext)] %INCLUDE name(modname) or %INCLUDE name.ext(modname)
The compiler searches as follows: 1. The current directory with a default extension of .pas for %INCLUDE directives, .pen for [INHERIT] attributes, and .tlb for %INCLUDE from text libraries (just like before) 2. Any directories specied with the /INCLUDE qualifer (in the order specied) with the appropriate default extension 3. SYS$LIBRARY: with the appropriate default extension /[NO]LIST /NOLIST (interactive default) /LIST=input_le_name.LIS (batch default) Produces a source listing le with a le type of .LIS. See the /SHOW qualier for more information on controlling the contents of the source listing le. /[NO]MACHINE_CODE /NOMACHINE_CODE (default) Produces a machine code section within the listing le. If the compiler detects errors in the source code, the compiler does not generate this section. The compiler ignores this qualier if you do not also specify /LIST on the same command line. /MATH_LIBRARY=keyword (OpenVMS Alpha systems only) Determines whether the compiler uses alternate math library routines that boost performance, but sacrice accuracy. You can specify the following values for keyword: ACCURATE ( Default )
FAST
/[NO]OBJECT /OBJECT=input_le_name.OBJ (default) Species the name of the object le. If the compiler detects errors in the source code, the compiler writes no representation of object code to the listing le. /[NO]OPTIMIZE /OPTIMIZE (default) Directs the compiler to optimize the code for the program or module being compiled so that the compiler generates more efcient code. A single identier or a list of identiers enclosed in parentheses can follow /OPTIMIZE; these identiers are the names of options that tell the compiler which aspects of the compilation unit to optimize. Table 18 lists the available options, their corresponding actions, and their negations.
Table 18 /OPTIMIZE Qualier Options

Option ALL INLINE=keyword LEVEL=num Action Enables all optimization components. Enables inline expansion of user-dened routines.
systems only).
Negation NONE NOINLINE

Alpha
Controls the optimization level (OpenVMS Values for num are:
0 1 2
Disables all optimizations. Identical in function to /NOOPTIMIZE. Enables local optimizations and recognition of common subexpressions. Enables all level 1 optimizations and some global optimizations, including the following: code motion, strength reduction and test replacement, split lifetime analysis, and code scheduling. Enables all level 2 optimizations and some additional global optimizations that improve speed at the cost of extra code size. These optimizations include integer multiplication and division expansion (using shifts), loop unrolling, and code replication to eliminate branches. Identical in function to /OPTIMIZE=NOINLINE. Enables all level 3 optimizations and inline expansion of procedures and functions. Identical in function to /OPTIMIZE. Enables software pipelining and additional software dependancy analysis, which in certain cases improves run-time performance.
UNROLL=num
Controls number of times loops are unrolled. The default is 4. /UNROLL=0 disables loop unrolling. Loop unrolling is only enabled above optimzation level 2. (OpenVMS Alpha systems only) Tune the object code to run best on the processor chosen. The default is Generic. Values for the processor are EV4, EV5, EV56, EV6, Generic, and Host. (OpenVMS Alpha systems only)
TUNE=processor
The /OPTIMIZE qualier without options is equivalent to /OPTIMIZE=ALL. The negation /NOOPTIMIZE is equivalent to /OPTIMIZE=NONE. The OPTIMIZE and NOOPTIMIZE attributes in the source program or module override the /OPTIMIZE and /NOOPTIMIZE qualiers on the command line. For More Information: On compiler optimizations (Section 3.1)
The /NOOPTIMIZE qualier guarantees full evaluation of both operands of the AND and OR Boolean operators to aid in diagnosing all potential programming errors. If you wish to have short-circuit evaluation even with the /NOOPTIMIZE qualier, use the AND_THEN and OR_ELSE Boolean operators. On OpenVMS Alpha systems, you can also specify an optimization level. Optimization levels from level 2 and higher include all optimizations from lower levels. /[NO]PLATFORMS /NOPLATFORMS (default) Displays informational messages about nonportable language features for the specied platform. Table 19 lists the supported platforms. Table 19 /PLATFORMS Qualier Options
Option COMMON OpenVMS_AXP OpenVMS_VAX OSF1_AXP Action Displays informational messages for all platforms. Displays informational messages for the OpenVMS Alpha platform. Displays informational messages for the OpenVMS VAX platform. Displays informational messages for the Compaq Tru64 UNIX platform.
/PSECT_MODEL=[NO]MULTILANGUAGE (OpenVMS Alpha systems only) /PSECT_MODEL=NOMULTILANGUAGE (default) This qualier controls whether the compiler pads the size of overlaid PSECTs, so as to ensure compatibility when the PSECT is shared by code created by other Compaq compilers. When a PSECT generated with a [COMMON] attribute is overlaid with a PSECT consisting of a C struct or a Fortran COMMON block, linker error messages can result due to the inconsistent sizes of the PSECTs; some languages pad the size of PSECTS, while other do not.
/[NO]SHOW /SHOW=(DICTIONARY,HEADER,INCLUDE,SOURCE,STATISTICS, TABLE_OF_CONTENTS) (default) Species a list of items to be included in the listing le. A single identier or a list of identiers enclosed in parentheses can follow /SHOW; these identiers are the names of options that inform the compiler which type of information it should generate. Table 110 lists the available options, their corresponding actions, and their negations. Table 110 /SHOW Qualier Options
Option ALL DICTIONARY HEADER INCLUDE INLINE
Action Enables listing of all options.
Negation NONE
Enables listing of %DICTIONARY NODICTIONARY records. Enables page headers. Enables listing of %INCLUDE les. Enables listing of inline summary. Enables listing of Compaq Pascal source code. Enables listing of compilation statistics. Enables listing of the sizes, record eld offsets, and comments about nonoptimal performance for variables and types in your program. Enables listing of a table of contents only if the %TITLE or %SUBTITLE directive was specied in the source code. NOHEADER NOINCLUDE NOINLINE NOSOURCE NOSTATISTICS NOSTRUCTURE_LAYOUT
SOURCE STATISTICS STRUCTURE_ LAYOUT
TABLE_OF_ CONTENTS
NOTABLE_OF_ CONTENTS
On OpenVMS VAX systems, the inline summary, enabled by the /SHOW=INLINE qualier, shows only the names of routines that were expanded inline in the compilation. If you want to know why routines were not expanded inline, you must specify an additional qualier, either /OPTIMIZE=INLINE or /OPTIMIZE=ALL. Although /OPTIMIZE defaults to /OPTIMIZE=ALL, you must explicitly specify the ALL option to generate these reasons.
The compiler ignores the /SHOW qualier if you do not also specify the /LIST qualier on the same command line. The negation /NOSHOW is equivalent to /SHOW=NONE; /SHOW is equivalent to /SHOW=ALL. /[NO]STANDARD /NOSTANDARD (default) Causes the compiler to generate messages wherever the compilation unit uses Compaq Pascal language extensions, which are nonstandard Pascal features. Within the Compaq Pascal documentation set, these standards are collectively referred to as the Pascal standard. Table 111 lists the available options, their corresponding actions, and their negations. Table 111 /STANDARD Qualier Options
Option NONE ANSI ISO EXTENDED VALIDATION Action Disables standards checking. Uses the rules of the ANSI standard. Uses the rules of the ISO standard. Uses the rules of the Extended standard. Performs validation for the given standard. Negation N.A. N.A. N.A. N.A. NOVALIDATION
The /STANDARD qualier allows you to use only two options. The rst option selects the standard to be used (ANSI, ISO or EXTENDED). The second option determines whether the strict validation rules are to be enforced ([NO]VALIDATION). /STANDARD=(ANSI, ISO, VALIDATION) is not allowed because both ANSI and ISO are specied. By default, these information-level messages are written to the error le SYS$ERROR. Using the VALIDATION option changes all nonstandard information-level messages to error-level messages. The /STANDARD qualier without options is equivalent to /STANDARD=(ANSI, NOVALIDATION). /STANDARD=VALIDATION is equivalent to /STANDARD=(ANSI, VALIDATION). The negation /NOSTANDARD is equivalent to /STANDARD=NONE.
/[NO]SYNCHRONOUS_EXCEPTIONS (OpenVMS Alpha systems only) /NOSYNCHRONOUS_EXCEPTIONS (default) Species that the compiler should generate code to insure that exceptions are reported as near as possible to the instruction that generated the exception. This can avoid confusion in tracing the source of an exception, however, there is a performance penalty for using this qualier. /[NO]TERMINAL /NOTERMINAL (default) Species a list of items to be displayed on the terminal. A single identier or a list of identiers enclosed in parentheses can follow the /TERMINAL qualier; these identiers are options that inform the compiler which type of information to display. Table 112 lists the available options, their corresponding actions, and their negations. Table 112 /TERMINAL Qualier Options
Option ALL FILE_NAME Action Displays all options. Displays le names on Pascal command line as they are being processed. Displays routine names as code is generated. Displays compiler statistics. Negation NONE NOFILE_NAME
ROUTINE_NAME
NOROUTINE_NAME
STATISTICS
NOSTATISTICS
The /TERMINAL qualier without options is equivalent to /TERMINAL=ALL. The negation /NOTERMINAL is equivalent to /TERMINAL=NONE. /[NO]TIE (OpenVMS Alpha systems only) /NOTIE(default) Species that the generated code can call images translated by the VAX Environment Software Translator (VEST) utility, which translates OpenVMS VAX system images into functionally equivalent OpenVMS Alpha system images. The Translated Image Environment (TIE) allows translated images to execute as if on an OpenVMS VAX system.
/[NO]USAGE (/USAGE=EMPTY_RECORDS, NONGRNACC, PACKED_ACTUALS, UNSUPPORTED_ CDD, UNINITIALIZED, VOLATILE) (defaults) Directs the compiler to perform compile-time checks indicated by the chosen options. A single identier or a list of identiers enclosed in parentheses can follow /USAGE; these identiers are options that tell the compiler which checks to perform. Table 113 lists the available options, their corresponding actions, and their negations. Table 113 /USAGE Qualier Options
Option ALL EMPTY_RECORDS Action Enables checking of all options. Checks for fetching records with no elds. Such elds are usually created by the %DICTIONARY directive for unsupported data types. Species that the compiler should issue warning messages for code sequences that might not match your granularity request from the /GRANULARITY qualier. When the compiler cannot guarantee that the generated code matches the granularity setting, a warning message is issued. You should examine your code to make sure that the variable being accessed is quadword-aligned and is a multiple of quadwords in size. In this case, the resulting code will be correct, although the compiler might not be able to determine that at compile time. Such cases involve pointer dereferences or VAR parameters. These messages are enabled by default by the compiler. (continued on next page) Negation NONE NOEMPTY_RECORDS
NONGRNACC
(OpenVMS Alpha systems only)
NONONGRNACC
Table 113 (Cont.) /USAGE Qualier Options

Option PACKED_ACTUALS Action Checks for passing components of packed structures to VAR parameters. Checks for variables and record elds that are poorly sized or aligned on inefcient boundaries. This provides the same information that is found in the /SHOW=STRUCTURE_ LAYOUT listing section. Species whether the compiler should issue informational messages for routines that are declared but never called. Checks for variables that can be uninitialized depending on program ow. Checks for variables that are known to be uninitialized. Negation NOPACKED_ACTUALS
PERFORMANCE
NOPERFORMANCE
UNCALLABLE
NOUNCALLABLE
UNCERTAIN
NOUNCERTAIN
UNINITIALIZED UNSUPPORTED_CDD
NOUNINITIALIZED
Checks for usage of CDD/RepositoryNOUNSUPPORTED_ constructs that do not correspond CDD to Compaq Pascal data types. Checks for variables that are declared but never referenced. Checks for VOLATILE variables that are not aligned properly. On the Alpha platform, certain unaligned VOLATILE variables cannot be updated in an atomic fashion. NOUNUSED NOVOLATILE
UNUSED VOLATILE
The following types of variables are not checked for uninitialization: Variables that have a le component Predeclared INPUT or OUTPUT identiers Variables that have global, external, or inherited visibility Variables declared with the AT attribute Variables declared with the COMMON attribute Variables declared with the READONLY attribute
Variables declared with the VOLATILE attribute Variables used as parameters Variables used as function identiers
The /USAGE qualier without options is equivalent to /USAGE=ALL. The negation /NOUSAGE is equivalent to /USAGE=NONE. The Compaq Pascal compiler can detect when some variables are uninitialized; however, it cannot detect that an uplevel variable is uninitialized at the point at which it was referenced. This is because at the time the routine is lexically scanned, the compiler has not seen any of the calls to that routine. /[NO]WARNINGS /WARNINGS (default) Directs the compiler to generate diagnostic messages in response to warninglevel or informational-level errors. By default, these messages are written to the error le SYS$ERROR. A warning or informational diagnostic message indicates that the compiler has detected acceptable but unorthodox syntax or has performed some corrective action; in either case, unexpected results can occur. Note that informational messages generated when you specify the /STANDARD qualier do not appear if /NOWARNINGS is enabled. /[NO]ZERO_HEAP (OpenVMS Alpha systems only) /ZERO_HEAP (default) Species that heap memory should be zeroed after allocation. By default, the Pascal RTL will return zero-lled memory for each call to the NEW built-in. Using the /NOZERO_HEAP qualier can increase runtime performance. For More Information: On debugging (Section 4.1) On text libraries (Section 1.1.4) On LSE and SCA information (Section 4.2) On error messages (Section 1.4) On the contents of a compiler listing (Section 1.1.3) On Pascal standards (Compaq Pascal Language Reference Manual) On using environment les (Section 2.1)
On the AND_THEN and OR_ELSE Boolean operators (Compaq Pascal Language Reference Manual)
1.1.3 Contents of the Compilation Listing File

You control the contents of a compilation listing by appending qualiers to the PASCAL command. Table 114 lists the parts of a complete compilation listing and the qualiers that cause them to be generated. Table 114 Compilation Listing Contents and Qualiers
Section Table of contents
VAX systems only) (OpenVMS
Generated With /LIST /SHOW=TABLE_OF_CONTENTS /LIST /LIST /CROSS_REFERENCE /LIST /MACHINE_CODE

(OpenVMS VAX
Source code Cross-reference Machine code Inline summary

systems only)
/LIST /SHOW=INLINE /LIST /SHOW=STATISTICS /LIST /SHOW=STRUCTURE_LAYOUT
Compilation statistics Structure layout
A compilation listing le usually contains source code because the /SHOW=SOURCE qualier is enabled by default. The /LIST qualier does not initiate the printing of the listing le. To obtain a line printer copy of your listing le, use the PRINT command. You can control the number of lines that appear on a listing page by dening the SYS$LP_LINES logical name before invoking the compiler. For example: $ DEFINE SYS$LP_LINES 100 $ PASCAL/LIST [DIR]M This set of commands creates a printed page size of 94 lines (the compiler subtracts six lines for margins). The following sections describe the contents of each part of the listing le. For More Information: On the PASCAL command qualiers (Section 1.1.2) On the SYS$LP_LINES logical (OpenVMS Programming Concepts Manual)
1.1.3.1 Table of Contents (OpenVMS VAX systems only) The table of contents lists the line number, listing page number, and source le page number of each section of the source code. These sections are delineated by %TITLE or %SUBTITLE directives that indicate the name by which the section is known; for example, Main Program Body. The compiler ignores the /SHOW=TABLE_OF_CONTENTS qualier if the compilation unit does not contain a %TITLE or %SUBTITLE directive. 1.1.3.2 Source Code The source code part of a listing le includes source code line numbers (LINE column); a notation identifying %INCLUDE directive code, %DICTIONARY directive code, and comments (IDC column); a procedure nesting level (PL column); a statement nesting level (SL column); source code; and diagnostic messages. The following example shows several lines of a procedure accessed with an %INCLUDE directive on an OpenVMS VAX system. On OpenVMS Alpha systems, the format can be slightly different. (LINE) (IDC) (PL) 00021 I 1 00022 I 1 00023 I 1 (SL) 0 PROCEDURE PRINT (Arr : Arrtype); 0 VAR I := INTEGER; 1 BEGIN
1.1.3.3 Cross-Reference Section The cross-reference part of a listing le contains a list of all identiers and labels used within the source code. This list includes the name of the identier or label, the program element it represents, the source code line numbers where it appears, and, where applicable, the attributes, declaring block, and function result type associated with it. 1.1.3.4 Machine Code Section The machine code part of a listing le contains a representation of the object code generated by the compiler. Information is organized by program section and, within each program section, by executable block. For each program section, the compiler generates the program section name and properties, hexadecimal representation of the code, computer-generated labels, symbolic opcode, and symbolic operands (if needed). Note that the listing format is similar to, but not exactly like, MACRO; that is, if the section is edited to remove the hexadecimal notation on the left side, it will not assemble using the MACRO-64 assembler, or the MACRO-32 assemblers for VAX or Alpha systems.
1.1.3.5 Structured Layout Section This listing section gives the sizes, record eld offsets, and comments about nonoptional performance for variables and types in your program. This section is useful to aid in restructuring data types for optimal performance. 1.1.3.6 Inline Summary (OpenVMS VAX systems only) The inline summary part of a listing le contains a summary indicating which routine calls of user-dened routines were or were not expanded inline. This summary includes the name of the calling routine, program, or module; the call and line number of the call; and a notation indicating whether expansion occurred. 1.1.3.7 Compilation Statistics The compilation statistics part of a listing le contains the following categories of summary information: Psect Summary, listing the program section name, number of bytes, and attributes of all program sections created during compilation. Environment Statistics, listing the names of all environment les inherited by the compilation and symbol information. This information includes the total number of symbols in the environment le, the number of symbols actually used by the compilation, and the percentage of used symbols versus dened symbols. Note that the Compaq Pascal compiler denes symbol in terms of internal representation. This denition can not reect the complexity of the environment source; that is, the number of symbols shown loaded can not reect the number of symbols in your program. Command Qualiers and Options List, containing the exact command line passed by DCL to the Compaq Pascal compiler (on OpenVMS VAX systems only), and the qualier options in effect during compilation. Compiler Internal Timing Statistics, noting the number of page faults and amount of elapsed time and CPU time required for each phase of the compilation. Compilation Statistics, listing the total number of messages generated at each levelinformational, warning, error, and fatal; the time and speed of compilation; and the number of page faults that occurred. The last line is a message indicating that the compilation of the source code is complete.
1.1.4 Text Libraries

A text library contains modules of source text that you can incorporate in a program by using the %INCLUDE directive. This directive indicates the module and, optionally, the text library in which the module can be found. Text library names can be specied in the following ways: In the %INCLUDE directive On the PASCAL command line In a DEFINE default library command (DCL)
1.1.4.1 Using the %INCLUDE Directive for Text Libraries The %INCLUDE directive has the following form: %INCLUDE [[le-spec]] (module-name) [[/[[NO]]LIST ]] le-spec The name of the text library containing a module to be included in the compilation. module-name The name of a text module, located in a text library, that is to be included in the source le. The name of the module must be enclosed in parentheses. The module names can include any printable character except a space, horizontal tab, comma, or exclamation point. The maximum length of the module name is determined when the text library is created. Module names are also case insensitive. /[NO]LIST Indicates that the included module should be printed in the listing of the program if a listing is being generated. If not specied, the default is determined by the [NO]INCLUDE option on the /SHOW qualier. The INCLUDE option enables the listing of %INCLUDE les and is enabled by default. For example, the following %INCLUDE directive species both the text library DATAB.TLB and the module External_Declarations: %INCLUDE DATAB.TLB (External_Declarations) If the text library is not specied in the %INCLUDE directive, its name must appear on the PASCAL command line or it must be specied by a DCL DEFINE command.
For More Information: On /LIST and /SHOW qualiers (Section 1.1.2) On default libraries (Section 1.1.4.3)
1.1.4.2 Specifying Text Libraries on the PASCAL Command Line The /LIBRARY qualier identies text libraries specied on the PASCAL command line. When you compile a source le that includes a module from a text library, concatenate the name of the text library to the name of the source le and append the /LIBRARY qualier. You specify concatenation with a plus sign. For example: $ PASCAL APPLIC+DATAB/LIBRARY This command instructs the compiler to search the DATAB text library each time it encounters an %INCLUDE directive within the APPLIC source le. If more than one library is specied, the compiler searches the libraries in the order they appear on the command line. For example: $ PASCAL APPLIC+DATAB/LIBRARY+DATAC/LIBRARY+DATAD/LIBRARY If you request multiple compilations, the /LIBRARY qualier must appear after each compilation in which it is needed. For example: $ PASCAL METRIC+DATAB/LIBRARY, APPLIC+DATAB/LIBRARY If you are concatenating source les, the /LIBRARY qualier can appear only after the last source le. For example: $ PASCAL METRIC.PAS+APPLIC.PAS+DATAB/LIBRARY Any Pascal output qualiers that appear after the /LIBRARY qualier, such as /OBJECT or /LISTING, apply to the last source le name that you specied. For example, the following PASCAL command creates APPLIC.OBJ: $ PASCAL METRIC+APPLIC+DATAB/LIBRARY/OBJECT For More Information: On the PASCAL command and qualiers (Section 1.1)
1.1.4.3 Dening Default Libraries You can dene one of your private text libraries as a default text library for the Pascal compiler to search. The Compaq Pascal compiler searches the default library after it searches libraries specied in the PASCAL command. To establish a default library, dene the logical name PASCAL$LIBRARY, as in the following example of the DEFINE command: $ DEFINE PASCAL$LIBRARY DISK$:[LIB]DATAB While this assignment is in effect, the compiler automatically searches the library DISK$:[LIB]DATAB.TLB for any included modules that it cannot locate in libraries explicitly specied on the PASCAL command. The Compaq Pascal compiler uses PASCAL$LIBRARY as the le name for the default text library; the location and search order of the logical name tables are controlled by Record Management Services (RMS). If PASCAL$LIBRARY is dened as a search list, the compiler opens the rst item specied in the list. If the include module is not found there, the search is terminated and an error message is issued. For More Information: On the DCL command DEFINE (OpenVMS DCL Dictionary)
1.2 LINK Command

The LINK command invokes the OpenVMS Linker, which combines object modules into one executable image, which can then be executed by the OpenVMS operating system. The linker uses the name of the input le that you specied rst on the command line for the name of the output le. The default for linker output les (executable images) is the .EXE le type. The LINK command format is as follows: LINK [[{/command-qualier} . . . ]] {le-spec[[{/le-qualier} . . . ]]} , . . .
( le-spec
/qualier [[= library-module (library-module, . . . )
)
]]
/command-qualier The name of a qualier that indicates special processing to be performed by the linker on all les listed. le-spec The name of one of the following: The input le (which can be the name of an object module library) that contains the object code to be linked. The options le, used only with the /OPTIONS qualier. The output le, used only with the /EXECUTABLE and /MAP qualiers.
/le-qualier The name of a qualier (the /INCLUDE, /LIBRARY, or /OPTIONS qualier) that indicates special processing to be performed by the linker on the les to which the qualier is attached. library-module The name of one or more object modules or shareable image libraries that you can only specify using the /INCLUDE or /LIBRARY qualiers. A source program or module cannot run on the system until it is linked. If you are using .PEN (Pascal Environment) les that include variables, procedures, or functions, make sure you link the object le into the .EXE le. When you execute the LINK command, the OpenVMS Linker performs the following functions: Resolves local and global symbolic references in the object code Assigns values to the global symbolic references Signals an error message for any unresolved symbolic reference
1.2.1 LINK Command Examples

This section contains examples of LINK command lines. $ LINK DANCE.OBJ, CHACHA.OBJ, SWING.OBJ This command links the object les DANCE.OBJ, CHACHA.OBJ, and SWING.OBJ to produce one executable image called DANCE.EXE. $ LINK/EXECUTABLE=TEST CIRCLE This command links CIRCLE.OBJ and then causes the executable image generated by the linker to be named TEST.EXE.
$ LINK SCHEDULE,COURSES/INCLUDE=(HISTORY,ALGEBRA,PHILOSOPHY) This example shows the use of the /INCLUDE qualier with a library named COURSES. The linker extracts the modules HISTORY, ALGEBRA, and PHILOSOPHY from the library COURSES and includes them in the executable image SCHEDULE.EXE. $ LINK SCHEDULE,COURSES/LIBRARY/INCLUDE=(HISTORY,ALGEBRA, PHILOSOPHY) This example also causes the linker to include the modules HISTORY, ALGEBRA, and PHILOSOPHY in the image le SCHEDULE.EXE. However, the /LIBRARY qualier causes the linker to search the rest of the library COURSES and link in any other modules needed to resolve symbolic references in SCHEDULE, HISTORY, ALGEBRA, and PHILOSOPHY. $ LINK UPDATE/EXE=[PROJECT.EXE]/MAP=[PROJECT.MAP] This command produces these les: [PROJECT.EXE]UPDATE.EXE [PROJECT.MAP]UPDATE.MAP
For More Information: On debugging (Section 4.1) On error messages (Section 1.4) On including modules from object module libraries (Section 1.2.3) On messages generated by the linker (OpenVMS Linker Utility Manual) On DCL syntax (HELP) On the OpenVMS Linker (OpenVMS Linker Utility Manual)
1.2.2 LINK Qualiers

The following are command and le qualiers that you can use when linking object modules: /[NO]DEBUG /NODEBUG (default) Indicates that the VMS Debugger is to be included in the executable image and that a symbol table is to be generated. If you specify LINK/DEBUG, the program links and then executes under the control of the debugger.
/[NO]EXECUTABLE /EXECUTABLE (default) Produces an executable image. A le specication can follow /EXECUTABLE to designate a name for the image le. The /NOEXECUTABLE qualier, which suppresses production of the image le, is useful when you want to verify the results of linking an object le before the image is produced. /INCLUDE none (default) Species that the input le is an object module or a shareable image library, and that the modules named are the only ones in the library to be explicitly included as input. In the case of shareable image libraries, the module is the shareable image name. You must specify at least one module name with the /INCLUDE qualier. The default for library modules is the .OLB le type. This qualier is a le qualier and cannot be used directly on the LINK command. /LIBRARY none (default) Species that the input le is an object module or shareable image library, which the linker must search to resolve undened symbols within other input modules specied on the same command line. You can use the /LIBRARY qualier with the /INCLUDE qualier to modify the same input le specication. In that case, the same library is searched for unresolved references. This qualier is a le qualier and cannot be used directly on the LINK command. The default for the le to which this qualier is applied is the .OLB le type. /[NO]MAP /NOMAP (interactive default) /MAP/NOCROSS_REFERENCE (batch default) Controls the generation of a map le and its contents. The /MAP qualier produces a map le, which you can name by including a le specication. The map le is stored on the default device in the default directory. If you do not include a le specication with /MAP, the map le is given the name of the rst input le and a .MAP le type. With the /MAP qualier, you can use the qualiers /BRIEF, /FULL, and /CROSS_REFERENCE to dene the type of information included in the map le.
lename/OPTIONS none (default) Species that the input le is a linker options le, which can contain input le specications as well as special instructions recognized only by the linker. You can also use options les to create shareable images. /[NO]SHAREABLE /NOSHAREABLE (default) Creates a shareable image. A shareable image has all of its internal references resolved, but must be linked with one or more object modules to produce an executable image. For example, a shareable image can contain a library of routines or can be used by the system manager to create a global section for all users. To include a shareable image as input to the linker, you can insert the shareable image into a shareable image library and specify the library as input to the LINK command. By default, the linker automatically searches the system-supplied shareable image library SYS$LIBRARY:IMAGELIB.OLB after searching any libraries you specify on the LINK command line. You can also include a shareable image by using a linker options le. The /NOSHAREABLE qualier species that the image produced cannot be linked with other images. /[NO]TRACEBACK /TRACEBACK (default) Causes the generation of error messages to be accompanied by symbolic traceback information. This information shows the sequence of calls that transferred control to the program in which the error occurred. /NOTRACEBACK suppresses production of traceback information. The traceback capability is automatically included with the /DEBUG qualier; therefore, if you specify both /DEBUG and /NOTRACEBACK, /NOTRACEBACK has no effect. For More Information: On debugging (Section 4.1) On object-module libraries (Section 1.2.3) On shareable images, options les, and contents of map les (OpenVMS Linker Utility Manual)
1.2.3 Object Module Libraries

An object module library contains modules of program text that have been successfully compiled. To link modules contained in a object module library, use the /INCLUDE qualier and specify the modules you want to link. For example: $ LINK GARDEN,VEGETABLES/INCLUDE=(EGGPLANT,TOMATO, BROCCOLI,ONION) This example directs the linker to link the subprogram modules EGGPLANT, TOMATO, BROCCOLI, and ONION with the main program module GARDEN. Besides program modules, an object module library can also contain a symbol table with the names of each global symbol in the library, and the name of the module in which they are dened. You specify the name of the object module library containing symbol denitions with the /LIBRARY qualier. When you use the /LIBRARY qualier during a link operation, the linker searches the specied library for all unresolved references found in the included modules during compilation. Also, by default, the linker automatically searches the system-supplied shareable image library SYS$LIBRARY:IMAGELIB.OLB after searching any libraries you specify on the LINK command. In the following example, the linker uses the library RACQUETS to resolve undened symbols in BADMINTON, TENNIS, and RACQUETBALL. $ LINK BADMINTON, TENNIS, RACQUETBALL, RACQUETS/LIBRARY You can dene an object module library to be your default library by using the DCL command DEFINE. The linker searches default user libraries for unresolved references after it searches modules and libraries specied in the LINK command. For More Information: On the LINK command and qualiers (Section 1.2) On the OpenVMS Linker (OpenVMS Linker Utility Manual) On the DCL command DEFINE (OpenVMS DCL Dictionary)
1.3 RUN Command

The RUN command executes programs that have been linked into an executable image by the OpenVMS Linker. This command has the following format: RUN [/command-qualier] le-spec /command-qualier The name of a qualier that indicates special processing to be performed by the linker on all les listed. le-spec The name of the executable image you want to run. The default le type for executable images is .EXE. The image activator accepts one command qualier, as follows: /[NO]DEBUG depends on linking (default) The /[NO]DEBUG qualier is optional. Specify the /DEBUG qualier to request the debugger, if the image was not linked with it. You cannot use /DEBUG on images linked with the /NOTRACEBACK qualier. If the image was linked with the /DEBUG qualier and you do not want the debugger to prompt, use the /NODEBUG qualier. The default action depends on whether you specied /DEBUG on the LINK command line. Consider the following examples: $ RUN PROG This example executes the image PROG.EXE. If you specied /DEBUG to the linker while creating PROG.EXE, the image activator passes control to the debugger upon execution. If you did not specify /DEBUG to the linker while creating PROG.EXE, the image activator executes the program. $ RUN/NODEBUG PROG This example executes the image PROG.EXE without invoking the debugger. For More Information: On debugging (Section 4.1) On messages generated by the image activator (Section 1.4) On the DCL command RUN (OpenVMS DCL Dictionary)
1.4 Error Messages

During program development, you can have to respond to messages regarding possible syntax or logic errors in your program. These messages have the following form: %SOURCE-CLASS-MNEMONIC, message_text SOURCE A code that identies the origin of the message. For example, the PASCAL code identies the Compaq Pascal compiler, the LINK code identies the OpenVMS Alpha Linker Utility, and the PAS code identies the Compaq Pascal run-time system. CLASS A single character that determines message severity. The four classes of error messages are: Informational ( I ), Warning ( W ), Error ( E ), and Fatal ( F ). The denition for each class depends on the source of the message, but execution of your request does not continue when E- or F-level errors occur. MNEMONIC A name that is unique to that message. message_text Explains the event that caused the message to be generated. For example, a common linker error occurs when you omit required le or library names from the command line, and the linker cannot locate the denition for a specied global symbol reference. The following error messages appear when a main program in OCEAN.OBJ calls a subprogram in SEAWEED.OBJ that is not specied in the LINK command: %LINK-W-NUDFSYMS, 1 undefined symbol %LINK-I-UDFSYMS, SEAWEED %LINK-W-USEUNDEF, module "OCEAN" references undefined symbol "SEAWEED" %LINK-W-DIAGISUED, completed but with diagnostics For More Information: On the complete list of Compaq Pascal compile-time and run-time errors (Appendix C) On the complete list of linker messages (OpenVMS Linker Utility Manual)
2
Separate Compilation
Pascal allows you to divide your application into subprograms by creating procedures and functions. Compaq Pascal allows you further modularity by allowing you to create compilation units, called programs and modules, that can be compiled separately. This chapter discusses the following topics about separate compilation: Sharing data with the ENVIRONMENT and INHERIT attributes (Section 2.1) Isolating data with the HIDDEN attribute (Section 2.1) Interfaces and implementations (Section 2.2) Data models (Section 2.3) Examples (Section 2.4) Note The sections at the beginning of this chapter use code fragments from the examples in this chapter and in the online example directory, which by default is PASCAL$EXAMPLES.
For More Information: On the ENVIRONMENT, HIDDEN, and INHERIT attributes (Compaq Pascal Language Reference Manual) On compiling and executing programs and modules (Chapter 1)
Separate Compilation 21
2.1 The ENVIRONMENT, HIDDEN, and INHERIT Attributes

To divide your program into a program and a series of modules, you need to decide, according to the needs of your application, which data types, constants, variables, and routines need to be shared either by other modules or by the program. To share data, create an environment le by using the ENVIRONMENT attribute in a module. Consider the following example: { Source File: share_data.pas This program initializes data to be shared with another compilation unit. } [ENVIRONMENT( share_data )] Module Share_Data; CONST Rate_For_Q1 = 0.1211; Rate_For_Q2 = 0.1156; Rate_For_Q3 = 0.1097; Rate_For_Q4 = 0.11243; TYPE Initialized_Type = ARRAY[1..10] OF INTEGER VALUE [1..5: 67; 6,9: 105; OTHERWISE 33]; END. If you do not specify a le name, Compaq Pascal creates an environment le using the le name of the source le and a default extension of .PEN. Another compilation unit can access the types and constants in the previous example by inheriting the environment le, as follows: { Source File: program.pas This code inherits data declarations and uses them in a program. } [INHERIT( share_data )] PROGRAM Use_Data( OUTPUT ); VAR a, b, c : Initialized_Type; Total : REAL VALUE 0.0; BEGIN Total := Total + ( Total * Rate_For_Q3 ); WRITELN( b[7] ); {b is of an initialized type} END. To build and run the application made up of the code in the previous examples, use the following commands:
$ $ $ $
PASCAL SHARE_DATA PASCAL PROGRAM LINK PROGRAM RUN PROGRAM 33
If a module contains variable declarations, routine declarations, schema types, or module initialization or nalization sections, you must link the program with the module that created the environment le to resolve external references. To prevent errors, you may wish to link programs with modules of inherited environment les as standard programming practice. For example, if SHARE_DATA contained a variable declaration, you must enter the following to resolve the external reference: $ $ $ $ PASCAL SHARE_DATA PASCAL PROGRAM LINK PROGRAM, SHARE_DATA RUN PROGRAM 33
For many applications, it is a good idea to place all globally accessible data into one module, create a single environment le, and inherit that module in other compilation units that need to make use of that data. Using environment les in this way reduces the difculties in maintaining the data (it is easier to maintain one le) and it eliminates problems that can occur when you cascade environment les. If compilation unit A inherits an environment le from compilation unit B, and if unit B inherits a le from unit C, then inheritance is cascading. Figure 21 shows a cascading inheritance path and a noncascading inheritance path.
Figure 21 Cascading Inheritance of Environment Files Problematic: C.U.1 Efficient: C.U.1 C.U.2 C.U.3
C.U.2
C.U.3
C.U.4
Key:
C.U.4 Inherited by
ZK1469AGE
Cascading is not always undesirable; it depends on your application and on the nature of the environment les. For example, if cascading occurs for a series of constant and type denitions that are not likely to change, cascading may require very little recompiling and relinking. However, if the constant and type denitions change often or if environment les contain routines and variables, you may nd it easier to redesign the inheritance paths of environment les due to the recompiling and relinking involved. Also, the inheritance path labeled Efcient in Figure 21 is not immune to misuse. That inheritance path, although it avoids the problems of cascading, may still involve multiply declared identiers (identical identiers contained in several of the compilation units whose environment les are inherited by compilation unit 4). In many instances, Compaq Pascal does not allow multiply declared identiers in one application. For example, a compilation unit cannot inherit two environment les that declare the same identier; also, a compilation unit usually cannot inherit an environment le that contains an identier that is identical to an identier in the outermost level of the unit (one exception, for example, is the redeclaration of a redenable reserved word or of an identier predeclared by Compaq Pascal). Also, Compaq Pascal allows the following exceptions to the rules concerning multiply declared identiers: A variable identier can be multiply declared if all declarations of the variable have the same type and attributes, and if all but one declaration at most are external. A procedure identier can be multiply declared if all declarations of the procedure have congruent parameter lists and if all but one declaration at most are external.
A function identier can be multiply declared if all declarations of the function have congruent parameter lists and identical result types, and if all but one declaration at most are external.
If a compilation unit creates an environment le and if it contains data that you do not want to share with other compilation units, you can use the HIDDEN attribute. Consider the following example: [ENVIRONMENT] MODULE Example; TYPE Array_Template( Upper : INTEGER ) = [HIDDEN] ARRAY[1..Upper] OF INTEGER; Global_Type : Array_Template( 10 ); VAR i : [HIDDEN] INTEGER; {Used for local incrementing} PROCEDURE x; BEGIN i := i + 1; END; PROCEDURE y; BEGIN FOR i := i + 1; END; END. The code in the previous example hides the schema type, preventing the schema type from being used in inheriting modules. (Whether to hide the type depends on the requirements of a given application.) Also, Compaq Pascal does not include the variable i in the environment le; this allows inheriting modules to declare the identier variable i as an incrementing variable without being concerned about generating errors for a multiply dened identier. Compaq Pascal performs compile-time and link-time checks to ensure that all compilations that inherit environment les actually used the same environment le denition. Information is placed in the object le such that the OpenVMS Linker performs the same check between each object le that inherited environment les. By default, compilation units that inherit an environment le compare the embedded compilation time inside the environment le: Uses found in any other environment les that are also inherited.
If the times are different, a compile-time message is displayed.
This checking can be disabled or modied by using the PEN_CHECKING_ STYLE attribute in the Pascal source le that created the environment le. Once the environment le exists, its selected checking style will be performed at each use. The PEN_CHECKING_STYLE attribute is valid at the beginning of a MODULE that creates an environment. The syntax is: PEN_CHECKING_STYLE(keyword) In this format "keyword" is: COMPILATION_TIME Uses the compilation time of the environment le in all subsequent compile-time checking for users of this environment le. This is the default. IDENT_STRING Uses the [IDENT( )] string of the environment le in all subsequent compile-time checking for users of this environment le. NONE Disables all compile-time checking for users of this environment le. When Compaq Pascal compiles a module with the /ENVIRONMENT qualier or [ENVIRONMENT] attribute, it generates an environment le and an object le. This is also true when compiling a or program, although it is not customary to generate an environment from a program. The PEN le contains compressed symbol table information and is used by subsequent Pascal compilations with the INHERIT attribute. Neither the OpenVMS Linker or ANALYZE/OBJECT reads environment les. Only the Compaq Pascal compiler uses these les. The OBJ le contains the following: Storage for variables declared at the outermost level of the module Code for procedures/functions contained in the module Linker timestamp verication for all inherited environment les and for the environment le being created. Compiler-generated variables and routines to support schema types declared or discriminated at the outermost level of the module Code for the TO BEGIN DO or TO END DO sections (the module will have an invisible TO BEGIN DO section if it inherits an environment whose creating module also had a TO BEGIN DO or TO END DO section)
It is possible for the OBJ le to contain just the linker timestamp record for the environment being created. Since the linker does not require the timestamp record for correct behavior, you may not need the OBJ le created when generating a PEN le however, if the module is used inherited variables or routines, then the OBJ le is needed at link time. This applies not only to the environments inherited by the program, but also to any environments inherited by modules which create subsequent environments. With the addition of schema types and TO BEGIN DO and TO END DO sections, the decision on which OBJ les to include on the LINK command becomes more complicated. This is due to: The compiler-generated variables and routines for schema types Any module initialization or nalization sections (TO BEGIN DO and TO END DO sections) The fact that currently PROGRAMs that inherit modules with initialization routines do not call the modules initialization routines directly but rely on the OpenVMS LIB$INITIALIZE feature to activate them.
The rule is still basically "Use any OBJ whose module contained variables or routines" but now you must consider compiler-generated variables or routines, as well as user-dened variables or routines. Depending on the modules involved, missing OBJ les might not be detected by the linker, since PROGRAMs do not directly call their inherited initialization routines. Compiling with the /DEBUG qualier might require more OBJ les to be included on the LINK command. Normally, the compiler only uses the inherited compiler-generated variables and routines if the corresponding schema types are used in certain Pascal constructs. However, the Debug information generated to describe the schema types always requires the compiler-generated variables and routines. This is because a user might ask the Debugger to perform some operation on the variable that did not appear in the source code. Therefore, if you omitted an OBJ le on the LINK command and did not get a linker error, you might get an error if the /DEBUG qualier was used. All OBJ les from modules creating environments must be included on the LINK command line if they contain variables or routines (either user-generated or compiler-generated). To receive the full benet of the linkers timestamp verication, you should include all OBJ les on the LINK command.
2.2 Interfaces and Implementations

If your application requires, you can use a method of creating and inheriting environment les that minimizes the number of times you have to recompile compilation units. This method involves the division of module declarations into two separate modules: an interface module and an implementation module. The interface module contains data that is not likely to change: constant denitions, variable declarations, and external routine declarations. The implementation module contains data that may change: bodies of the routines declared in the interface module, and private types, variables, routines, and so forth. The interface module creates the environment le that is inherited by both the implementation module and by the program. Figure 22 shows the inheritance process. Figure 22 Inheritance Path of an Interface, an Implementation, and a Program
Interface Module
Program
Implementation Module
means "is inherited by"

ZK1491AGE
Consider this code fragment from the interface module in Example 21 (see Section 2.4): [ENVIRONMENT( interface )] MODULE Graphics_Interface( OUTPUT ); {Globally accessible type} {Provide routines that manipulate the shapes:} PROCEDURE Draw( s : Shape ); EXTERNAL; PROCEDURE Rotate( s : Shape ); EXTERNAL; PROCEDURE Scale( s : Shape ); EXTERNAL; PROCEDURE Delete( s : Shape ); EXTERNAL; {Module initialization section} END. The code contained in the interface is not likely to change often. The implementation code can change without requiring recompilation of the other modules in the application. Consider this code fragment from the implementation module in Example 22 (see Section 2.4): [INHERIT( Interface )] {Predeclared graphics types and routines} MODULE Graphics_Implementation( OUTPUT ); [GLOBAL] PROCEDURE Rotate( s : Shape ); BEGIN WRITELN( Rotating the shape :, s.t ); END; To compile, link, and run the code in Examples 21, 22, and 23 (the main program), use the following commands: $ PASCAL GRAPHICS_INTERFACE $ PASCAL GRAPHICS_IMPLEMENTATION $ PASCAL GRAPHICS_MAIN_PROGRAM $ LINK GRAPHICS_MAIN_PROGRAM, GRAPHICS_IMPLEMENTATION,_$ GRAPHICS_INTERFACE $ RUN GRAPHICS_MAIN_PROGRAM If you need to change the code contained in any of the routine bodies declared in the implementation module, you do not have to recompile the program to reect the changes. For example, if you have to edit the implementation module, you can regenerate the application with the following commands: $ EDIT GRAPHICS_IMPLEMENTATION $ PASCAL GRAPHICS_IMPLEMENTATION $ LINK GRAPHICS_MAIN_PROGRAM, GRAPHICS_IMPLEMENTATION,_$ GRAPHICS_INTERFACE $ RUN GRAPHICS_MAIN_PROGRAM
In this manner, interfaces and implementations can save you maintenance time and effort. In addition, the interface and implementation design allows you to better predict when cascading inheritance may provide maintenance problems. Figure 23 shows two forms of cascading. Figure 23 Cascading Using the Interface and Implementation Design
Interface and Implementation
Interface
Interface and Implementation
Interface
Interface and Implementation A
Interface
B
ZK1492AGE
If the compilation units creating environment les are designed to contain both interface and implementation declarations, the cascading in column A may lead to more recompiling, more relinking, and more multiply declared identiers. The design shown in column B does not always provide easy maintenance, but it is more likely to do so. For example, if each interface provided a different kind of constant or type (as determined by your application) and if the constants and types are not derived from one another, the inheritance path in column B may be quite efcient and orderly, and may require little recompiling and relinking.
Do not place the following in an implementation module: Nonstatic types and variables at the module level A module initialization section (TO BEGIN DO) A module nalization section (TO END DO)
These restrictions are necessary because Compaq Pascal cannot determine the order of activation of initialization and nalization sections that do not directly follow an environment-le inheritance path. Since implementation modules do not create environment les, the initialization and nalization sections in those modules are effectively outside of any inheritance path. Also, if you use the previously listed objects in implementation modules, there may be attempts to access data that has not yet been declared. Consider the following example: {In one file:} [ENVIRONMENT( interface )] MODULE Interface; PROCEDURE x; EXTERNAL; END. {In another file:} [INHERIT( interface )] MODULE Implementation( OUTPUT ); VAR My_String : STRING( 10 ); [GLOBAL] PROCEDURE x; BEGIN WRITELN( My_String ); END; TO BEGIN DO My_String := Okay; END. In the previous example, it is possible for you to call procedure x ( in some other module that also inherits INTERFACE.PEN ) before the creation and initialization of the variable My_String. You can circumvent this problem by using a routine call to initialize the variable and by moving the code to the interface module, as shown in the next example: {In one file:} [ENVIRONMENT( interface )] MODULE Interface; VAR My_String : STRING( 10 ); PROCEDURE x; EXTERNAL; PROCEDURE Initialize; EXTERNAL;
TO BEGIN DO Initialize; END. {In another file:} [INHERIT( interface )] MODULE Implementation( OUTPUT ); [GLOBAL] PROCEDURE x; BEGIN WRITELN( My_String ); END; [GLOBAL] PROCEDURE Initialize; BEGIN My_String := Okay; END; END.
2.3 Data Models

Using separate compilation and a few other features of Compaq Pascal (including initial states, constructors, the HIDDEN attribute, and TO BEGIN DO and TO END DO sections), you can construct models for creating, distributing, isolating, and restricting data in an application. Of course, the design of the data model depends on the needs of a particular application. However, to show some of the power of Compaq Pascal features used in conjunction, Examples 21, 22, and 23 in Section 2.4 create a generic graphics application. Consider the following code fragment from Example 21: TYPE Shape_Types = ( Rectangle, Circle ); {Types of graphics objects} Shape( t : Shape_Types ) = RECORD {Starting coordinate points} Coordinate_X, Coordinate_Y : REAL VALUE 50.0; CASE t OF {Shape-specific values} Rectangle : ( Height, Width : REAL VALUE 10.0 ); Circle : ( Radius : REAL VALUE 5.0 ); END; {Provide routines that manipulate the shapes:} PROCEDURE Draw( s : Shape ); EXTERNAL; PROCEDURE Rotate( s : Shape ); EXTERNAL; PROCEDURE Scale( s : Shape ); EXTERNAL; PROCEDURE Delete( s : Shape ); EXTERNAL;
The interface module provides an interface to the rest of the application. This module contains types and external procedure declarations that the data model chooses to make available to other compilation units in the application; other units can access these types and routines by inheriting the generated environment le. The type Shape_Types denes two legal graphical objects for this application: a circle and a rectangle. The type Shape can be used by other units to create circles and rectangles of specied dimensions. This code uses a variant record to specify the different kinds of data needed for a circle (a radius value) and a rectangle (height and width values). Since the type has initial-state values, any variable declared to be of this type receives these values upon declaration. Providing initial states for types that are included in environment les can prevent errors when other compilation units try to access uninitialized data. The initial states in this code are specied for the individual record values. You can also provide an initial state for this type using a constructor, as follows: Shape( t : Shape_Types ) = RECORD Coordinate_X, Coordinate_Y : REAL; CASE t OF Square : ( Height, Width : REAL ); Circle : ( Radius : REAL ); END VALUE [ Coordinate_X : 50.0; Coordinate_Y : 50.0; CASE Circle OF [ Radius : 5.0 ] ]; If you use constructors for variant records, you can only specify an initial state for one of the variant values. If you need to specify initial states for all variant values, you must specify the initial states on the individual variants, as shown in Example 21. The interface module also declares routines that can draw, rotate, scale, and delete an object of type Shape. The bodies of these routines are located in the implementation module. The interface module also contains a TO BEGIN DO section, as shown in the following code fragment: [HIDDEN] PROCEDURE Draw_Logo; EXTERNAL; { Before program execution, display a logo to which the main program has no access. } TO BEGIN DO Draw_Logo;
As with the other routines, the body of Draw_Logo is located in the implementation module. The HIDDEN attribute prevents compilation units that inherit the interface environment le from calling the Draw_Logo routine. This ensures that the application only calls Draw_Logo once at the beginning of the application. Using this design, the interface module can provide graphical data and tools to be used by other compilation units without the other units having to worry about implementation details. The actual details are contained in one implementation module. For example, the routine bodies are contained in the implementation module. Consider the following code fragment from Example 22: {Declare routine bodies for declarations in the interface} [GLOBAL] PROCEDURE Draw( s : Shape ); BEGIN CASE s.t OF Circle : WRITELN( Code that draws a circle ); Rectangle : WRITELN( Code that draws a rectangle ); END; END; {Procedure Draw} The routine bodies of the external routines declared in the interface module are located in the implementation module. The code in each of the routines uses the actual discriminant of parameter s to determine if the shape is a circle or a rectangle and draws the shape. If this code needs to change, it does not require that you recompile the code in Examples 21 or 23 in Section 2.4. Example 22 also contains code that is isolated and hidden from other compilation units that inherit the interface environment le. Consider the following code fragment from the interface module: [GLOBAL] PROCEDURE Draw_Logo; VAR Initial_Shape : Shape( Circle ) {Declare object} VALUE [ Coordinate_X : 50.0; Coordinate_Y : 50.0; CASE Circle OF [Radius : 15.75;]]; BEGIN WRITELN( Drawing a company logo ); Draw( Initial_Shape ); {Code pauses for 30 seconds as the user looks at the logo.} Delete( Initial_Shape ); WRITELN; {Ready for the rest of the graphics program to begin.} END;
In the graphical data model, you may wish to dene a company logo, and you may wish to display that logo on the screen before any other graphical data is drawn or displayed. This code declares the variable Initial_Shape. Since this variable is declared locally to Draw_Logo and since Draw_Logo is contained in a module that does not produce an environment le, other modules that may have access to the interface environment le do not have access to this variable. In this application, you may not wish to give other compilation units the power to alter the company logo. The code in the interfaces TO BEGIN DO section, which executes before any program code, displays the company logo and deletes it to begin the application. Consider again the compilation process for interfaces, implementations, and programs: $ PASCAL GRAPHICS_INTERFACE $ PASCAL GRAPHICS_IMPLEMENTATION $ PASCAL GRAPHICS_MAIN_PROGRAM $ LINK GRAPHICS_MAIN_PROGRAM, GRAPHICS_IMPLEMENTATION,_$ GRAPHICS_INTERFACE $ RUN GRAPHICS_MAIN_PROGRAM Compaq Pascal executes the TO BEGIN DO section according to the inheritance order of environment les. Remember that Compaq Pascal cannot determine the order of execution for TO BEGIN DO sections contained in implementation modules, so do not use them there. Using this design, you can allow different sites that run the graphics application to access global data through the interface module. One location can maintain and control the contents of the implementation module, shipping the implementations object module for use at other sites. You can use this method for other types of sensitive data or data that needs to be maintained locally.
2.4 Separate Compilation Examples

Example 21 shows an interface module that creates the environment le INTERFACE.PEN. This environment le is inherited in Examples 22 and in 23. Example 21 An Interface Module for Graphics Objects and Routines { Source File: graphics_interface.pas This module creates an interface to graphical data and routines. } [ENVIRONMENT( interface )] MODULE Graphics_Interface; TYPE Shape_Types = ( Rectangle, Circle ); {Types of graphics objects} Shape( t : Shape_Types ) = RECORD {Starting coordinate points:} Coordinate_X, Coordinate_Y : REAL VALUE 50.0; CASE t OF {Shape-specific values} Rectangle : ( Height, Width : REAL VALUE 10.0 ); Circle : ( Radius : REAL VALUE 5.0 ); END; {Provide routines that manipulate the shapes:} PROCEDURE Draw( s : Shape ); EXTERNAL; PROCEDURE Rotate( s : Shape ); EXTERNAL; PROCEDURE Scale( s : Shape ); EXTERNAL; PROCEDURE Delete( s : Shape ); EXTERNAL; [HIDDEN] PROCEDURE Draw_Logo; EXTERNAL; { Before program execution, display a logo to which the main program has no access. } TO BEGIN DO Draw_Logo; END.
Example 22 shows the implementation of the routines declared in Example 21. Example 22 An Implementation Module for Graphics Objects and Routines { Source File: graphics_implementation.pas This module implements the graphics routines and data declarations made global by the interface module. } [INHERIT( Interface )] {Predeclared graphics types and routines} MODULE Graphics_Implementation( OUTPUT ); {Declare routine bodies for declarations in the interface:} [GLOBAL] PROCEDURE Draw( s : Shape ); BEGIN CASE s.t OF Circle : WRITELN( Code that draws a circle ); Rectangle : WRITELN( Code that draws a rectangle ); END; END; {Procedure Draw} [GLOBAL] PROCEDURE Rotate( s : Shape ); BEGIN WRITELN( Rotating the shape :, s.t ); END; [GLOBAL] PROCEDURE Scale( s : Shape ); BEGIN WRITELN( Scaling the shape :, s.t ); END; [GLOBAL] PROCEDURE Delete( s : Shape ); BEGIN WRITELN( Deleting the shape :, s.t ); END;
Example 22 (Cont.) An Implementation Module for Graphics Objects and Routines [GLOBAL] PROCEDURE Draw_Logo; VAR Initial_Shape : Shape( Circle ) {Declare object} VALUE [ Coordinate_X : 50.0; Coordinate_Y : 50.0; CASE Circle OF [Radius : 15.75;]]; BEGIN WRITELN( Drawing a company logo ); Draw( Initial_Shape ); {Code pauses for 30 seconds as the user looks at the logo.} Delete( Initial_Shape ); WRITELN; {Ready for the rest of the graphics program to begin.} END; END. Example 23 shows a main program and its use of the types and routines provided by the interface module. Example 23 A Graphics Main Program { Source File: graphics_main_program.pas This program inherits the interface environment file, which gives it access to the implementations declarations. } [INHERIT( Interface )] {Types and routines in interface module} PROGRAM Graphics_Main_Program( OUTPUT ); VAR My_Shape : Shape( Rectangle ) VALUE [ Coordinate_X : 25.0; Coordinate_Y : 25.0; CASE Rectangle OF [Height : 12.50; Width : 25.63]]; BEGIN { You cannot access the variable Initial_Shape, because it is in the implementation module, and that module does not create an environment file.
Example 23 (Cont.) A Graphics Main Program You can work with My_Shape. If you did not provide initial values in this declaration section, the module Graphics_Interface provided initial values for the schema type Shape. } Draw( My_Shape ); Scale( My_Shape ); Rotate( My_Shape ); Delete( My_Shape ); END. To compile, link, and run the code in Examples 21, 22, and 23, enter the following: $ PASCAL GRAPHICS_INTERFACE $ PASCAL GRAPHICS_IMPLEMENTATION $ PASCAL GRAPHICS_MAIN_PROGRAM $ LINK GRAPHICS_MAIN_PROGRAM, GRAPHICS_IMPLEMENTATION,_$ GRAPHICS_INTERFACE $ RUN GRAPHICS_MAIN_PROGRAM Drawing a company logo Code that draws a circle Deleting the shape : CIRCLE Code that draws a rectangle Scaling the shape : RECTANGLE Rotating the shape : RECTANGLE Deleting the shape : RECTANGLE
3
Program Correctness, Optimization, and Efciency
The objective of optimization is to produce source and object programs that achieve the greatest amount of processing with the least amount of time and memory. Realizing this objective requires programs that are carefully designed and written, and compilation techniques, such as those used by Compaq Pascal, that take advantage of the operating system and machine architecture environment. (The benets of portable code and program efciency depend on the requirements of your application.) This chapter discusses the following topics: Compiler optimizations (Section 3.1) Programming considerations (Section 3.2) Implementation-dependent behavior (Section 3.3)
3.1 Compiler Optimizations

By default, programs compiled with the Compaq Pascal compiler undergo optimization. An optimizing compiler automatically attempts to remove repetitious instructions and redundant computations by making assumptions about the values of certain variables. This, in turn, reduces the size of the object code, allowing a program written in a high-level language to execute at a speed comparable to that of a well-written assembly language program. Optimization can increase the amount of time required to compile a program, but the result is a program that may execute faster and more efciently than a nonoptimized program. The language elements you use in the source program directly affect the compilers ability to optimize the object program. Therefore, you should be aware of the ways in which you can assist compiler optimization. In addition, this awareness often makes it easier for you to track down the source of a problem when your program exhibits unexpected behavior.
Program Correctness, Optimization, and Efciency 31
The compiler performs the following optimizations: Compile-time evaluation of constant expressions Elimination of some common subexpressions Partial elimination of unreachable code Code hoisting from structured statements, including the removal of invariant computations from loops Inline code expansion for many predeclared functions Inline code expansion for user-declared routines Rearrangement of unary minus and NOT operations to eliminate unary negation and complement operations Partial evaluation of logical expressions Propagation of compile-time known values Strength reduction
(OpenVMS Alpha systems only) (OpenVMS Alpha systems only)
Split lifetime analysis Code scheduling Loop unrolling
(OpenVMS Alpha systems only) (OpenVMS Alpha systems only)
Software pipelining
These optimizations are described in the following sections. In addition, the compiler performs the following optimizations, which can be detected only by a careful examination of the machine code produced by the compiler: Global assignment of variables to registers If possible, reduce the number of memory references needed by assigning frequently referenced variables to registers. Reordering the evaluation of expressions This minimizes the number of temporary values required. Peephole optimization of instruction sequences The compiler examines code a few instructions at a time to nd operations that can be replaced by shorter and faster equivalent operations. For More Information: On Compaq Pascal language elements (Compaq Pascal Language Reference Manual)
3.1.1 Compile-Time Evaluation of Constants

The compiler performs the following computations on constant expressions at compile time: Negation of constants The value of a constant preceded by unary minus signs is negated at compile time. For example: x := -10.0; Type conversion of constants The value of a lower-ranked constant is converted to its equivalent in the data type of the higher-ranked operand at compile time. For example: x := 10 * y; If x and y are both real numbers, then this operation is compiled as follows: x := 10.0 * y; Arithmetic on integer and real constants An expression that involves +, , *, or / operators is evaluated at compile time. For example: CONST nn = 27; {In the executable section:} i := 2 * nn + j; This is compiled as follows: i := 54 + j; Array address calculations involving constant indexes These are simplied at compile time whenever possible. For example: VAR i : ARRAY[1..10, 1..10] OF INTEGER; {In the executable section:} i[1,2] := i[4,5]; Evaluation of constant functions and operators Arithmetic, ordinal, transfer, unsigned, allocation size, CARD, EXPO, and ODD functions involving constants, concatenation of string constants, and logical and relational operations on constants, are evaluated at compile time.
For More Information: On the complete list of compile-time operations and routines (Compaq Pascal Language Reference Manual)
3.1.2 Elimination of Common Subexpressions

The same subexpression often appears in more than one computation within a program. For example: a := b * c + e * f; h := a + g - b * c; IF ((b * c) - h) <> 0 THEN ... In this code sequence, the subexpression b * c appears three times. If the values of the operands b and c do not change between computations, the value b * c can be computed once and the result can be used in place of the subexpression. The previous sequence is compiled as follows: t := b * c; a := t + e * f; h := a + g - t; IF ((t) - h) <> 0 THEN ... Two computations of b * c have been eliminated. In this case, you could have modied the source program itself for greater program optimization. The following example shows a more signicant application of this kind of compiler optimization, in which you could not reasonably modify the source code to achieve the same effect: VAR a, b : ARRAY[1..25, 1..25] OF REAL; {In the executable section:} a[i,j] := b[i,j]; Without optimization, this source program would be compiled as follows: t1 := (j - 1) * 25 + i; t2 := (j - 1) * 25 + i; a[t1] := b[t2]; Variables t1 and t2 represent equivalent expressions. The compiler eliminates this redundancy by producing the following optimization: t = (j - 1) * 25 + i; a[t] := b[t];
3.1.3 Elimination of Unreachable Code

The compiler can determine which lines of code, if any, are never executed and eliminates that code from the object module being produced. For example, consider the following lines from a program: CONST Debug_Switch = FALSE; {In the executable section:} IF Debug_Switch THEN WRITELN( Error found here ); The IF statement is designed to write an error message if the value of the symbolic constant Debug_Switch is TRUE. Suppose that the error has been removed, and you change the denition of Debug_Switch to give it the value FALSE. When the program is recompiled, the compiler can determine that the THEN clause will never be executed because the IF condition is always FALSE; no machine code is generated for this clause. You need not remove the IF statement from the source program. Code that is otherwise unreachable, but contains one or more labels, is not eliminated unless the GOTO statement and the label itself are located in the same block.
3.1.4 Code Hoisting from Structured Statements

The compiler can improve the execution speed and size of programs by removing invariant computations from structured statements. For example: FOR j := 1 TO i + 23 DO BEGIN IF Selector THEN a[i + 23, j - 14] := 0 ELSE b[i + 23, j - 14] := 1; END; If the compiler detected this IF statement, it would recognize that, regardless of the Boolean value of Selector, a value is stored in the array component denoted by [i + 23, j 14]. The compiler would change the sequence to the following: t := i + 23; FOR j := 1 TO t DO BEGIN u := j - 14; IF Selector THEN a[t,u] := 0 ELSE b[t,u] := 1; END; This removes the calculation of j 14 from the IF statement, and the calculation of i + 23 from both the IF statement and the loop.
3.1.5 Inline Code Expansion for Predeclared Functions

The compiler can often replace calls to predeclared routines with the actual algorithm for performing the calculation. For example: Square := SQR( a ); The compiler replaces this function call with the following, and generates machine code based on the expanded call: Square := a * a; The program executes faster because the algorithm for the SQR function has already been included in the machine code.
3.1.6 Inline Code Expansion for User-Declared Routines

Inline code expansion for user-declared routines performs in the same manner as inline code expansion for predeclared functions: the compiler can often replace calls to user-declared routines with an inline expansion of the routines executable code. Inline code expansion is useful on routines that are called only a few times. The overhead of an actual procedure call is avoided, which increases program execution. The size of the program, however, may increase due to the routines expansion. To determine whether or not it is desirable to inline expand a routine, compilers use a complex algorithm. Section 3.1.7 describes the algorithm for Compaq Pascal on OpenVMS VAX systems; Compaq Pascal on OpenVMS Alpha systems uses a similar algorithm to make the determination.
3.1.7 Testing for Inline Expansion on OpenVMS VAX Systems

The rst part of the algorithm performs tests for cases that always prohibit the routine from being inlined. A failure of one of these tests can be thought of as a hard failure. These hard failure tests verify that the following are false; if any one of these tests is true, the routine is not inlined: The called routine is an external or inherited routine. Either the calling routine or the called routine does not have inlining optimization enabled. Note that optimization is enabled by default. The called routine establishes an exception handler, or is used as an exception handler. The called function result is a structured result type. The calling routine and the called routine do not have the same checking options enabled.
The calling routine and the called routine do not use the same program section. The called routine declares a routine parameter or is itself a routine parameter. The called routines parameter list contains a LIST or TRUNCATE parameter, a read-only VAR parameter, or a conformant parameter. The called routine declares local le variables or contains any nonlocal GOTO operations. The called routine references automatic variables in an enclosing scope. The called routine uses or declares nonstatic types.
The second part of the algorithm performs tests to determine how desirable it is to inline the routine at a particular call point. A failure to one of these tests can be thought of as a soft failure. These tests check for the number of formal parameters, number of local variables, whether the called routine is directly recursive, the number of direct calls to the routine, and the size of both the calling and the called routine. If an explicit [OPTIMIZE(INLINE)] attribute is specied on the routine declaration, the hard failure tests are still performed; however, the soft failure tests are not. So if the routine passes the hard failure tests, that routine is inlined at all call points. Specifying this attribute provides you with more power in deciding which routines should be inlined. Note There is no stack frame for an inline user-declared routine and no debugger symbol table information for the expanded routine. Debugging the execution of an inline routine is difcult and is not recommended.
3.1.8 Operation Rearrangement

The compiler can produce more efcient machine code by rearranging operations to avoid having to negate and then calculate the complement of the values involved. For example: (-c) * (b - a)
If a program includes this operation, the compiler rearranges the operation to read as follows: c * (a - b) These two operations produce the same result, but because the compiler has eliminated negation or complement operations, the machine code produced is more efcient.
3.1.9 Partial Evaluation of Logical Expressions

The Pascal language does not specify the order in which the components of an expression must be evaluated. If the value of an expression can be determined by partial evaluation, then some subexpressions may not be evaluated at all. This situation occurs most frequently in the evaluation of logical expressions. For example: WHILE ( i < 10 ) AND ( a[i] <> 0 ) DO BEGIN a[i] := a[i] + 1; i := i + 1; END; In this WHILE statement, the order in which the two subexpressions ( i < 10 ) and ( a[i] <> 0 ) are evaluated is not specied; in fact, the compiler may evaluate them simultaneously. Regardless of which subexpression is evaluated rst, if its value is FALSE the condition being tested in the WHILE statement is also FALSE. The other subexpression need not be evaluated at all. In this case, the body of the loop is never executed. To force the compiler to evaluate expressions in left-to-right order with short circuiting, you can use the AND_THEN operator, as shown in the following example: WHILE ( i < 10 ) AND_THEN ( a[i] <> 0 ) DO BEGIN a[i] := a[i] + 1; i := i + 1; END;
3.1.10 Value Propagation

The compiler keeps track of the values assigned to variables and traces the values to most of the places that they are used. If it is more efcient to use the value rather than a reference to the variable, the compiler makes this change. This optimization is called value propagation. Value propagation causes the object code to be smaller, and may also improve run-time speed.
Value propagation performs the following actions: It allows run-time operations to be replaced with compile-time operations. For example: Pi := 3.14; Pi_Over_2 := Pi/2; In a program that includes these assignments, the compiler recognizes the fact that Pis value did not change between the time of Pis assignment and its use. So, the compiler would use Pis value instead of a reference to Pi and perform the division at compile time. The compiler treats the assignments as if they were as follows: Pi := 3.14; Pi_Over_2 := 1.57; This process is repeated, allowing for further constant propagation to occur. It allows comparisons and branches to be avoided at run time. For example: x := 3; IF x <> 3 THEN y := 30 ELSE y := 20; In a program that includes these operations, the compiler recognizes that the value of x is 3 and the THEN statement cannot be reached. The compiler will generate code as if the statements were written as follows: x := 3; y := 20;
3.1.11 Strength Reduction
Strength reduction speeds computations by replacing a multiply operation with a more efcient add instruction when computing array addresses each time around a loop.
3.1.12 Split Lifetime Analysis
Split lifetime analysis improves register usage by determining if the lifetime of a variable can be broken into multiple, independent sections. If so, the variable may be stored in different registers for each section. The registers can then be reused for other purposes between sections. Therefore, there may be times when the value of the variable does not exist anywhere in the registers. For example:
v:= 3.0 *q; . . . x:= SIN(y)*v: . . . v:= PI*x: . . . y:= COS(y)*v; This example shows that the variable v has two disjoint usage sections. The value of v in the rst section does not affect the value of v in the second section. The compiler may use different registers for each section.
3.1.13 Code Scheduling
Code scheduling is a technique for reordering machine instructions to maximize the amount of overlap of the multiple execution units inside the CPU. The exact scheduling algorithms depend on the implementation of the target architecture.
3.1.14 Loop Unrolling
Loop unrolling is a technique for increasing the amount of code between branch instructions and labels by replicating the body of a loop. Increasing the code optimizes instruction scheduling. The following code shows such a transformation: Original Code FOR i:= 1 to 12 DO a[i]:= b[i] + c[i] Unrolled Loop Code i:= 1 WHILE i < 12 DO BEGIN a[i]:= b[i] + c[i]; a[i+1]:= b[i+1] + c[i+1]; a[i+2]:= b[i+2] + c[i+2]; a[i+3]:= b[i+3] + c[i+3]; i:= i+4; END; In this example, the loop body was replicated four times, allowing the instruction scheduler to overlap the fetching of array elements with the addition of other array elements.
By default, loop unrolling makes 4 copies of an unrolled loop. You can change the number of copies from 1 to 16. This is controlled by: /OPTIMIZE=UNROLL="number" Numbers larger than 4 may improve performance at a cost of additional code size. However, larger numbers may decrease performance due to cache requirements, register conicts, and other factors.
3.1.15 Alignment of Compiler-Generated Labels

The compiler aligns the labels it generates for the top of loops, the beginnings of ELSE branches, and others, on machine-specic boundaries by lling in unused bytes with NO-OP instructions. A branch to a longword-aligned address is faster than a branch to an unaligned address. This optimization may increase the size of the generated code; however, it increases run-time speed.
3.1.16 Error Reduction Through Optimization

An optimized program produces results and run-time diagnostic messages identical to those produced by an equivalent unoptimized program. An optimized program may produce fewer run-time diagnostics, however, and the diagnostics may occur at different statements in the source program. For example:
Unoptimized Code a := x/y; b := x/y; FOR i := 1 TO 10 DO c[i] := c[i] * x/y; Optimized Code t := x/y; a := t; b := t; FOR i := 1 TO 10 DO c[i] := c[i] * t;
If the value of y is 0.0, the unoptimized program would produce 12 divideby-zero errors at run time; the optimized program produces only one. (The variable t is a temporary variable created by the compiler.) Eliminating redundant calculations and removing invariant calculations from loops can affect the detection of such arithmetic errors. You should keep this in mind when you include error-detection routines in your program.
3.1.17 Software Pipelining
Software pipelining and additional software dependency analysis are enabled using the /OPTIMIZE=LEVEL=5 command line option, which in certain cases improves run-time performance. /OPTIMIZE=LEVEL=5 is not the default; /OPTIMIZE=LEVEL=4 remains the default. As compared to regular loop unrolling (enabled at optimization level 3 or above), software pipelining uses instruction scheduling to eliminate instruction stalls within loops, rearranging instructions between different unrolled loop iterations to improve performance. For instance, if software dependency anaylsis of data ow reveals that certain calculations can be done before or after that iteration of the unrolled loop, software pipelining reschedules those instructions ahead or behind that loop iteration at places where their execution can prevent instruction stalls or otherwise improve performance. For this version of Compaq Pascal, loops chosen for software pipelining: Are always innermost loops (those executed the most). Do not contain branches or procedure calls.
By modifying the unrolled loop and inserting instructions as needed before and/or after the unrolled loop, software pipelining generally improves runtime performance, except for cases where the loops contain a large number of instructions with many existing overlapped operations. In this case, software pipelining may not have enough registers available to effectively improve execution performance and using optimization level 5 instead of optimization level 4 may not improve run-time performance. To determine whether using optimization level 5 benets your particular program, time program execution for the same program compiled at level 4 and 5. For programs that contain loops that exhaust available registers, longer execution times may result with optimization level 5. In cases where performance does not improve, consider compiling using /OPTIMIZE=(UNROLL=1,LEVEL=5) to (possibly) improve the effects of software pipelining.
3.1.18 Processor Tuning
Compaq Pascal provides support for tuning the generated code for specic Alpha processors. The two models supported are the EV4 (also known as the 21064[A]/21066[A]) and the EV5 (also known as the 21164). The differences between the two models can be seen in the instruction scheduling phase. Code tuned for EV4 will generally run quite well on EV5, though it will not exploit the EV5 to the maximum potential (particularly in the oating-point area.) This is because EV5 integer code is basically a "relaxed" set of issue rules relative to EV4. Code tuned for EV5 will generally be a little worse on an EV4 because taking advantage of the shorter latencies and relaxed issue pairing leads to cases where the EV4 stalls (whereas the EV5 running EV4 tuned code basically just misses opportunities to issue things earlier in the instruction stream, which may or may not matter.) The default is "generic" with the current meaning being EV4. The denition of "generic" is projected to change over time as the mix of the installedbase changes. The default should be satisfactory unless you need to exploit a particular processor. For oating-point code, the difference tends to be single-digit percentages.
3.1.19 Compiling for Optimal Performance

The following command lines will result in producing the fastest code from the compiler. Depending on the system, use one of the following: For OpenVMS Alpha systems use: PASCAL /NOZERO_HEAP /MATH_LIBRARY=FAST /OPT=LEVEL=5 /NOCHECK /ARCH=HOST Note Under certain circumstances an optimization level of 5 might be slower than the default optimization level of 4.
For OpenVMS VAX systems use: PASCAL /OPTIMIZE /NOCHECK In both cases, you may also want to use the performance agger to identify datatypes that could be modied for additional performance. For More Information: On performance agger (Section 1.1.2)
3.2 Programming Considerations

The language elements that you use in a source program directly affect the compilers ability to optimize the resulting object program. Therefore, you should be aware of the following ways in which you can assist compiler optimization and obtain a more efcient program: Dene constant identiers to represent values that do not change during your program. The use of constant identiers generally makes a program easier to read, understand, and later modify. In addition, the resulting object code is more efcient because symbolic constants are evaluated only once, at compile time, while variables must be reevaluated whenever they are assigned new values. Whenever possible, use the structured control statements CASE, FOR, IF-THEN-ELSE, REPEAT, WHILE, and WITH rather than the GOTO statement. You can use the GOTO statement to exit from a loop, but careless use of it interferes with both optimization and the straightforward analysis of program ow. Enclose in parentheses any subexpression that occurs frequently in your program. The compiler checks whether any assignments have affected the subexpressions value since its last occurrence. If the value has not changed, the compiler recognizes that a subexpression enclosed in parentheses has already been evaluated and does not repeat the evaluation. For example: x := SIN( u + (b - c) ); y := COS( v + (b - c) ); The compiler evaluates the subexpression (b 0 c) as a result of performing the SIN function. When it is encountered again, the compiler checks to see whether new values have been assigned to either b or c since they were last used. If their values have not changed, the compiler does not reevaluate (b 0 c).
Once your program has been completely debugged, disable all checking with [CHECK(NONE)] or with the appropriate compilation switch. Recall that Compaq Pascal enables bounds and declaration checking by default. When no checking code is generated, more optimizations can occur, and the program executes faster. Integer overow checking is disabled by default. If you are sure that your program is not in danger of integer overow, you should not enable overow checking. Because overow checking precludes certain optimizations, you can achieve a more efcient program by leaving it disabled.
When a variable is accessed by a program block other than the one in which it was declared, the variable should have static rather than automatic allocation. An automatically allocated variable has a varying location in memory; accessing it in another block is time-consuming and less efcient than accessing a static variable. On OpenVMS VAX systems, avoid using the same temporary variable many times in the course of a program. Instead, use a new variable every time your program needs a temporary variable. Because variables stored in registers are the easiest to access, your program is most efcient when as many variables as possible can be allocated in registers. If you use several different temporary variables, the lifetime of each one is greatly reduced; thus, there is a greater chance that storage for them can be allocated in registers rather than at memory locations. When creating schema records (or records with nonstatic elds), place the elds with run-time size at the end of the record. The generated code has to compute the offset of all record elds after a eld with run-time size, and this change minimizes the overhead.
For More Information: On Compaq Pascal language elements and on attributes (Compaq Pascal Language Reference Manual) On compilation switches (Chapter 1)
3.3 Implementation-Dependent Behavior

The Pascal language has several implementation-dependent behaviors that a program must not rely upon. Relying on these behaviors for correct behavior is illegal and is not portable to other platforms or other compiler versions. Refer to the Compaq Pascal Language Reference Manual for a list of the implementation-dependent behaviors.
For More Information: On attributes and on static and automatic variables (Compaq Pascal Language Reference Manual) On compilation switches (Chapter 1)
3.3.1 Subexpression Evaluation Order

The compiler can evaluate subexpressions in any order and may even choose not to evaluate some of them. Consider the following subexpressions that involve a function with side effects: IF f( a ) AND f( b ) THEN ... This IF statement contains two designators for function f with the same parameter a. If f has side effects, the compiler does not guarantee the order in which the side effects will be produced. In fact, if one call to f returns FALSE, the other call to f might never be executed, and the side effects that result from that call would never be produced. For example: q := f( a ) + f( a ); The Pascal standard allows a compiler to optimize the code as follows: Q := 2 * f( a ) If the compiler does so, and function f has side effects, the side effects would occur only once because the compiler has generated code that evaluates f( a ) only once. If you wish to ensure left-to-right evaluation with short circuiting, use the AND_THEN and OR_ELSE Boolean operators. For More Information: On the order of expression evaluation, see the description of the NOOPTIMIZE attribute (Compaq Pascal Language Reference Manual)
3.3.2 MAXINT and MAXINT64 Predeclared Constants

The smallest possible value of the INTEGER type is represented by the predeclared constant 0MAXINT. The largest possible value of the INTEGER type is represented by the predeclared constant MAXINT. However, the VAX and Alpha architectures support an additional integer value, which is (0MAXINT 01). If your program contains a subexpression with this value, the programs evaluation might result in an integer overow trap. Therefore, a computation involving the value (0MAXINT 01) might not produce the expected result. To evaluate expressions that include (0MAXINT 01), you should disable either optimization or integer overow checking.
Similarly, on OpenVMS Alpha of (0MAXINT64 expected results.
01) might not produce the
3.3.3 Pointer References

The compiler assumes that the value of a pointer variable is either the constant identier NIL or a reference to a variable allocated in heap storage by the NEW procedure. A variable allocated in heap storage is not declared in a VAR section and has no identier of its own; you can refer to it only by the name of a pointer variable followed by a circumex ( ^ ). Consider the following example: VAR x : INTEGER; p : ÎNTEGER; {In the executable section:} NEW( p ); p^ := 0; x := 0; IF p^ = x THEN p^ := p^ + 1; If a pointer variable in your program must refer to a variable with an explicit name, that variable must be declared VOLATILE or READONLY. The compiler makes no assumptions about the value of volatile variables and therefore performs no optimizations on them. Use of the ADDRESS function, which creates a pointer to a variable, can result in a warning message because of optimization characteristics. By passing a nonread-only or nonvolatile static or automatic variable as the parameter to the ADDRESS function, you indicate to the compiler that the variable was not allocated by NEW but was declared with its own identier. Because the compilers assumptions are incorrect, a warning message occurs. You can also use IADDRESS, which functions similarly to the ADDRESS function except that IADDRESS returns an INTEGER_ADDRESS value and does not generate any warning messages. Use caution when using IADDRESS. Similarly, when the parameter to ADDRESS is a formal VAR parameter or a component of a formal VAR parameter, the compiler issues a warning message that not all dynamic variables allocated by NEW may be passed to the function. For More Information: On attributes and on predeclared routines (Compaq Pascal Language Reference Manual)
3.3.4 Variant Records

Because all the variants of a record variable are stored in the same memory location, a program can use several different eld identiers to refer to the same storage space. However, only one variant is valid at a given time; all other variants are undened. You must store a value in a eld of a particular variant before you attempt to use it. For example: VAR x : INTEGER; a : RECORD CASE t : BOOLEAN OF TRUE : ( b : INTEGER ); FALSE : ( c : REAL ); END; {In the executable section:} x := a.b + 5; a.c := 3.0; x := a.b + 5; Record a has two variants, b and c, which are located at the same storage address. When the assignment a.c := 3.0 is executed, the value of a.b becomes undened because TRUE is no longer the currently valid variant. When the statement x := a.b + 5 is executed for the second time, the value of a.b is unknown. The compiler may choose not to evaluate a.b a second time because it has retained the elds previous value. To eliminate any misinterpretations caused by this assumption, variable a should be associated with the VOLATILE attribute. The compiler makes no assumptions about the value of VOLATILE objects. For More Information: On variant records or on the VOLATILE attribute (Compaq Pascal Language Reference Manual)
3.3.5 Atomicity, Granularity, Volatility, and Write Ordering

When data is shared by multiple code streams (either multiple processes, multiple threads, or asynchronous events such as AST routines or condition handlers), you need to be aware of certain issues to guarantee correct sharing of data. You must inform the compiler that the data being shared may change in an asychronous fashion. By default, the compiler assumes that data is only modied by assignment statements, routine calls, etc. If the data is being changed in a way that the compiler does not know about you must use the VOLATILE attribute to tell the compiler that it must fetch the data in an
atomic fashion from memory at each reference and the compiler must store the data in an atomic fashion back into memory at each assignment. To accomplish atomic access on the VAX for volatile objects 32 bits or smaller, fetches and stores are done with the normal MOVB/MOVW/MOVL/INSV/EXTV instructions. In a single CPU environment, the alignment of the objects is not relevant. However, in a multiple CPU SMP system, the data being accessed must reside in a single 32-bit longword otherwise the underlying memory system may return incorrect data if two CPUs are updating the same longwords at the same time. Larger objects that are manipulated with the MOVC3/MOVC5 instructions are not atomic as those instructions may be interrupted. To accomplish atomic access on the Alpha for volatile objects 64 bits or smaller, fetches and stores are done with the STx_L/LDx_C instruction sequence. This pair of instructions ensures that the volatile data is accessed in an atomic fashion. Without the VOLATILE attribute, you will not get this special instruction sequence and the data might become corrupted if two writers are trying to store to the shared data at the same time. Only aligned data objects are guaranteed to be accessed atomically. Larger objects that are manipulated with run-time routines are not atomic, as those routines may be interrupted. Granularity is a term on Alpha machines to describe the situation where two threads update nearby data at the same time. Because the compiler on the older Alpha must fetch the surrounding longword or quadword, modify it, and store it back, the two threads could possibly overwrite each others data. For these situations, the nearby data should be moved to separate quadwords or use the /GRANULARITY qualier to tell the compiler that you want longword or byte granularity at the expense of additional STx_L/LDx_C sequences. (See the /ARCHITECTURE qualier for additional information on the byte and word instructions available on newer Alpha systems.) Besides atomic accesses, many programs want to perform atomic operations on shared data. To facilitate this, Compaq Pascal provides the following built-in routines: ADD_INTERLOCKED(expr,variable) This routine adds the expression to the aligned word variable and returns -1 if the new value is negative, 0 if it is zero, or 1 if it is positive. On VAX, it generates the VAX ADAWI instruction. On Alpha, it uses the STx_L /LDx_C instructions. CLEAR_INTERLOCKED(Boolean-variable) SET_INTERLOCKED(Boolean-variable)
These routines clear or set a Boolean variable, respectively, and return the original value. On VAX, they generate the BBCCI and BBSSI instructions, respectively. On Alpha, they use the STx_L/LDx_C instructions. ADD_ATOMIC(expr,variable) AND_ATOMIC(expr,variable) OR_ATOMIC(expr,variable) These Alpha-only routines generate STx_L/LDx_C instruction sequences to add/and/or the value of the expression with the variable and return the original value of the variable. On the VAX, write operations to independent memory locations are completed in the order of the instructions. However, on Alpha, the architecture does not guarantee that independent writes will complete in the order they were issued. The Alpha architecture provides a special instruction to serialize write operations. Compaq Pascal provides the BARRIER built-in routine on Alpha systems to generate the MB instruction to preserve write ordering. If your code uses a higher-level synchronization scheme to guard critical regions (such as a lock manager or a semaphore package), then using the VOLATILE attribute, the GRANULARITY qualier and the INTERLOCKED /ATOMIC built-ins may not be necessary; you have already ensured that there are only single readers/writers in the critical section.
3.3.6 Debugging Considerations

Some of the effects of optimized programs on debugging are as follows: Use of registers When the compiler determines that the value of an expression does not change between two given occurrences, it may save the value in a register. In such a case, it does not recompute the value for the next occurrence, but assumes that the value saved in the register is valid. If, while debugging the program, you attempt to change the value of the variable in the expression, then the value of that variable is changed, but the corresponding value stored in the register is not. When execution continues, the value in the register may be used instead of the changed value in the expression, causing unexpected results. When the value of a variable is being held in a register, its value in memory is generally invalid; therefore, a spurious value may be displayed if you try to examine a variable under these circumstances. Coding order
Some of the compiler optimizations cause code to be generated in a order different from the way it appears in the source. Sometimes code is eliminated altogether. This causes unexpected behavior when you try to step by line, use source display features, or examine or deposit variables. Use of condition codes
This optimization technique takes advantage of the way in which the VAX processor condition codes are set. For example, consider the following source code: x := x + 2.5; IF x < 0 THEN ... Rather than test the new value of x to determine whether to branch, the optimized object code bases its decision on the condition code settings after 2.5 is added to x. If you attempt to set a debugging breakpoint at the second line and deposit a different value into x, you cannot achieve the intended result because the condition codes no longer reect the value of x. In other words, the decision to branch is being made without regard to the deposited value of the variable. Inline code expansion on user-declared routines There is no stack frame for an inline user-declared routine and no debugger symbol table information for the expanded routine. Debugging the execution of an inline user-declared routine is difcult and is not recommended. To prevent conicts between optimization and debugging, you should always compile your program with a compilation switch that deactivates optimization until it is thoroughly debugged. Then you can recompile the program (which by default is optimized) to produce efcient code. For More Information: On debugging tools (Chapter 4) On compilation switches (Chapter 1)
4
Programming Tools
This chapter describes some Pascal-specic assistance provided in the set of Compaq CASE tools. For general information on each tool, see the documentation for the tool. This chapter provides information on: Debugger support for Compaq Pascal (Section 4.1) Language-Sensitive Editor/Source Code Analyzer (LSE/SCA) support for Compaq Pascal (Section 4.2) Accessing Oracle CDD/Repository from Compaq Pascal source programs (Section 4.3) Equivalent CDD/Repository Language and Compaq Pascal data types (Section 4.3.1)
4.1 Debugger Support for Compaq Pascal

In general, the debugger supports the data types and operators of Compaq Pascal and the other debugger-supported languages. However, there are important language-specic limitations. (To get information on the supported data types and operators of any of the languages, enter the HELP LANGUAGE command at the DBG> prompt.) In general, you can examine, evaluate, and deposit into variables, record elds, and array components. An exception to this occurs under the following circumstances: if a variable is not referenced in a program, the Compaq Pascal compiler may not allocate the variable. If the variable is not allocated and you try to examine it or deposit into it, you will receive an error message. When depositing data into variables, the debugger truncates the high-order bits if the value being deposited is larger than the variable; it lls the highorder bits with zeros if the value being deposited is smaller than the variable. If the deposit violates the rules of assignment compatibility, the debugger displays an informational message.
Programming Tools 41
Automatic variables (within any active block) can be examined and can have values deposited into them; however, since automatic variables are allocated in stack storage and are contained in registers, their values are considered undened until the variables are initialized or assigned a value. For example: DBG> EXAMINE X MAINP\X: 2147287308 In this example, the value of variable X should be considered undened until after a value has been assigned to X. In addition, you may examine a VARYING OF CHAR string, but it is not possible to examine the LENGTH eld. For example, the following is not supported: DBG> EXAMINE VARY_STRING.LENGTH Because the current LENGTH of a VARYING string is the rst word, you should do the following to examine the LENGTH: DBG> EXAMINE/WORD VARY_STRING It should also be noted that the type cast operator (::) is not permitted when evaluating Compaq Pascal expressions. Pointers to undiscriminated schema cannot be correctly described to the debugger at this time since the type of the pointer is dependent upon the value pointed to by the pointer. They are described as pointers to UNSIGNED integers. For example, TYPE S(I:INTEGER) = ARRAY [1..I] OF INTEGER; VAR P : ^S; BEGIN NEW(P,expression); END;
4.2 Language-Sensitive Editor/Source Code Analyzer Support for Compaq Pascal

This section describes Compaq Pascal specic information for the following Language-Sensitive Editor/Source Code Analyzer ( LSE/SCA ) features: Programming language placeholders and tokens Placeholder processing Design comment processing
42 Programming Tools
4.2.1 Programming Language Placeholders and Tokens

LSE accepts keywords, or tokens, for all languages with LSE support, but the specic tokens themselves are language-dened. For example, you can expand the %INCLUDE token only when using Compaq Pascal. Likewise, LSE provides placeholders, or prompt markers, for all languages with LSE support, but the specic text or choices these markers call for are language-dened. For example, you see the %{environ_name_string}% placeholder only when using Compaq Pascal. Some Compaq Pascal keywords, like TYPE, VAR, IF, and FOR, can be placeholders as well as tokens. LSE supplies language constructs for these keywords when they appear on your screen as placeholders. You can also type the keywords into the buffer yourself, enter the EXPAND command, and see the same language constructs appear on your screen. You can use the SHOW TOKEN and SHOW PLACEHOLDER commands to display a list of all Compaq Pascal tokens and placeholders, or a particular token or placeholder. For example: LSE> SHOW TOKEN IF LSE> SHOW TOKEN {lists the token IF} {lists all tokens }
To copy the listed information into a separate le, rst enter the appropriate SHOW command to put the list into the $SHOW buffer. Then enter the following command: LSE> GOTO BUFFER $SHOW LSE> SAVE FILE filename.filetype
4.2.2 Placeholder and Design Comment Processing

While all languages with LSE support provide placeholder processing, each language denes specic contexts in which placeholders can be accepted as valid program code. Compaq Pascal denes contexts for declaration section placeholders and executable section placeholders. Table 41 lists the valid contexts within a Compaq Pascal declaration section where you can insert placeholders.
Table 41 Placeholders Within the Declaration Section

Can Replace PROGRAM or MODULE identier Program parameter Identier Data type Value Complete variant within the variant part of record Cannot Replace Directive Attribute Declaration-begin reserved word Complete declaration
Table 42 lists valid contexts within a Compaq Pascal executable section where you can insert placeholders. Table 42 Placeholders Within the Executable Section
Can Replace Statement Variable Expression Case label Complete case expression Iteration variable within a FOR statement Cannot Replace LABEL identier TO | DOWNTO within a FOR statement
Compaq Pascal support for placeholder and design comment processing includes the following language-specic stipulations: Pseudocode placeholders are designated with double left- and right-angle brackets (<< >>) or the 8 bit format ( ). The compiler produces an empty object le when it encounters pseudocode or LSE placeholders within a source program. Comment processing is limited to the declaration section.
4.2.3 LSE and SCA Examples

Example 41 shows how you can use LSE tokens and placeholders to create a FOR statement within a Compaq Pascal program. The callout numbers identify the steps in this process, which are detailed in the notes appearing after the example. Example 41 Using LSE to Create a FOR Statement BEGIN %[statement_list]%... END. . . . BEGIN 2 FOR %{control_var}% %{iteration_clause}% DO %{statement}%; %[statement_list]%... END. . . . BEGIN FOR INDEX := 1 TO MAX DO 3 %{statement}%; %[statement_list]%... END. . . . BEGIN FOR INDEX := 1 TO MAX DO 4 %{variable | func_id}% := %{value_expr}%; %[statement_list]%... END. . . . BEGIN 5 FOR INDEX := 1 TO MAX DO ARR[INDEX] := 0; %[statement_list]%... END.
1 1
As you begin the executable section of your program, the cursor rests on the placeholder %[statement_list]%. Type the token FOR over this placeholder and expand FOR.
LSE provides the FOR statement template. Select a FOR variable option from the menu. Expand the %{iteration_clause}% placeholder and expand the %{statement}% placeholder. LSE displays a menu, from which you can select the %{simple_statement}% option. A further menu appears, from which you select the ASSIGNMENT statement option. LSE provides the assignment statement template. Type an appropriate identifer or value expression over each placeholder. The completed FOR statement appears in your buffer.
Example 42 shows some contexts in which LSE placeholders and design comments might appear in the design of a Compaq Pascal program. Placeholder contexts are self-explanatory; the callout numbers identify types of comments, which are detailed in the notes following the example. Example 42 Using LSE Comments in Program Design PROGRAM Semester_Grades ( input,output ) ;
1
{ Author : P. Knox { Creation Date : 03/03/89 { Functional Description : This program calculates the numerical semester grade and determines the corresponding alphabetic grade for each student in a class. { Nonlocal References : None { Included Files : None { Keywords : Grade array procedures, semester grade file CONST number of students = %{compile_time_exp}%; number of semester grades = %{compile_time_exp}%; TYPE
} }
} } } }
grade array = ARRAY [ 1..number of semester grades] OF integer or real? determine later; grade range = 0 .. 100;
Example 42 (Cont.) Using LSE Comments in Program Design

systems only)
(OpenVMS VAX
VAR A_Grade_Array A_Grade Total %{variable}% : : grade array ; { var for : grade range ; { var for : grade range ; { var for INTEGER; { control array of grades individual grade semester grade var, FOR loops } } } }
PROCEDURE compute semester grade ( VAR Grd_Arry : grade array ; VAR Sum : grade range ) ;
3
{ Parameters : Grd_Arry : value parameter, array of semester grades for one student: Sum : variable parameter, returns the semester grade } BEGIN FOR %{control_var}% %{iteration_clause}% DO sum the grades in Grd_Arry Sum gets Sum divided by number of grades in array write students semester grade END; PROCEDURE assign letter grade ( Semester_Grade : grade range ) ; BEGIN CASE %{case_selector}% OF %{case_labels}%... : %{statement}%; %{case_labels}%... : %{statement}%; %{case_labels}%... : %{statement}%; %{case_labels}%... : %{statement}% OTHERWISE %{statement_list}%... END END;
Example 42 (Cont.) Using LSE Comments in Program Design

systems only)
(OpenVMS VAX
BEGIN FOR number of students DO BEGIN enter semester grades for student ; FOR %{control_var}% := %{value_expr}% TO number of semester grades DO BEGIN read grades and load into array END; Total := 0; compute routine ( Grade_Array, Total ); assign_letter routine ( Total ) END END.
These comments, which could be placed at the beginning of the program, are tagged comments. The comment begins with a predened term called a tag. The tag is followed by a tag terminator symbol (:) and free text. These comments, which do not contain tags, are called remark comments. A remark comment consists of free text. This structured comment contains both a tag (the identier parameters) and subtags (the identiers Grd_Arry and Sum). A tag terminator and a blank comment line separate the two subtags from each other.
4.3 Accessing CDD/Repository from Compaq Pascal

The Oracle CDD/Repository (CDD/Repository) must be purchased separately. The CDD/Repository allows language-independent structure declarations that can be shared by many OpenVMS layered products. Compaq Pascal support of the CDD/Repository allows Compaq Pascal programmers to share common record and data denitions with other Compaq languages and data management products. A system manager or data administrator creates the CDD/Repositorys directory hierarchies, history lists, and access control lists with the Dictionary Management Utility (DMU). Once record paths are established, you can enter data denitions into and extract them from the CDD/Repository.
To enter data denitions into the CDD/Repository, you rst create CDD/Repository source les written in the Common Data Dictionary Language (CDDL). The CDDL compiler converts the denitions to an internal formmaking them independent of the language used to access them. To extract data denitions from the CDD/Repository, include the %DICTIONARY directive in your Compaq Pascal source program. If the data attributes of the data denitions are consistent with Compaq Pascal requirements, the data denitions are included in the Compaq Pascal program during compilation. The %DICTIONARY directive incorporates CDD/Repository data denitions into the current Compaq Pascal source le during compilation. This directive can appear only in the TYPE section of a Compaq Pascal program, not in the executable section. For example: PROGRAM SAMPLE1; TYPE %DICTIONARY Pascal_SALESMAN_RECORD/LIST . . . A /LIST option in the %DICTIONARY directive (or the /SHOW= DICTIONARY qualier on the Pascal command line) includes the translated record in the programs listing. For example: TYPE %DICTIONARY PASCAL_SALESMAN_RECORD/LIST { CDD Path Name => PASCAL_SALESMAN_RECORD } PAYROLL_RECORD = PACKED RECORD SALESMAN : PACKED RECORD NAME : PACKED ARRAY [1..30] OF CHAR; ADDRESS : PACKED ARRAY [1..40] OF CHAR; SALESMAN_ID : [BYTE(5)] RECORD END; { numeric string, unsigned } END; { record salesman } END; { record payroll_record } The option (/LIST or /NOLIST) overrides the qualier (/SHOW=NODICTIONARY or /SHOW=DICTIONARY).
For More Information: On CDD/Repository (CDD/Repository Reference Manual) On CDDL (CDD/Repository Data Denition Language Reference Manual) On CDDL data types (CDD/Administrator Users Guide and CDD/Repository Data Denition Language Reference Manual) On the Compaq Pascal %DICTIONARY directive (Compaq Pascal Language Reference Manual)
4.3.1 Equivalent Compaq Pascal and CDDL Data Types

The CDD/Repository supports some data types that are not native to Compaq Pascal. If a data denition contains a eld declared with an unsupported data type, Compaq Pascal replaces the eld with one declared as a [BYTE(n)] RECORD END, where n is the appropriate length in bytes. By making the data addressable in this way, you are able to manipulate the data either by passing it to external routines as variables or by using the Compaq Pascal type casting capabilities to perform an assignment. However, because these empty records do not have elds, the size of the record is 0 bits. They should not be used in expressions or passed to formal value parameters. Recall that a size attribute used on a type denition has no effect on fetches. When fetching from these records, the compiler will fetch the actual size of the record, 0 bits. Table 43 summarizes the mapping between CDDL data types and the corresponding Compaq Pascal data types. NOTE Although this practice is discouraged, you can use both D_oating and G_oating data types in the same compilation unit; however, both types cannot be handled in the same expression. Not all processors support the G_oating and H_oating types.
Table 43 Equivalent CDD/Repository Language and Compaq Pascal Data Types

CDDL Data Type Unspecied Byte logical Word logical Longword logical Quadword logical Octaword logical Byte integer Word integer Longword integer Quadword integer Octaword integer F_oating D_oating G_oating H_oating
Compaq Pascal Data Type [BYTE(n)] RECORD END [BYTE] 0..255 [WORD] 0..65535 UNSIGNED [BYTE(8)] RECORD END [BYTE(16)] RECORD END [BYTE] 128..127 [WORD] 32768..32767 INTEGER [BYTE(8)] RECORD END [BYTE(16)] RECORD END SINGLE DOUBLE (/NOG_FLOATING) DOUBLE (/G_FLOATING) QUADRUPLE [BYTE (8)] RECORD END [BYTE(16)] RECORD END [BYTE(16)] RECORD END [BYTE(32)] RECORD END PACKED ARRAY [l..u] OF CHAR VARYING [u] OF CHAR [BYTE(n)] RECORD END [BYTE(n)] RECORD END [BYTE(n)] RECORD END [BYTE(n)] RECORD END (continued on next page)
F_oating complex D_oating complex G_oating complex H_oating complex Text Varying text Numeric string, unsigned Numeric string, left separate Numeric string, left overpunch Numeric string, right separate
Table 43 (Cont.) Equivalent CDD/Repository Language and Compaq Pascal Data Types
CDDL Data Type Numeric string, right overpunch Numeric string, zoned sign Bit Compaq Pascal Data Type [BYTE(n)] RECORD END [BYTE(n)] RECORD END [BIT(n)] 0..((2n )1) or [BIT(32)]UNSIGNED or [BIT(N)] RECORD END or ignored [BIT(n), POS(x)] 0..((2n )n1) or [BIT(32), POS(x)] UNSIGNED or [BIT(n), POS(x)] RECORD END or ignored [BYTE(n)] RECORD END [BYTE(n)] RECORD END Ignored VARYING [u] OF CHAR Variant record Pointer type
Bit unaligned
Date and time Date Virtual eld Varying string Overlay Pointer
4.3.2 CDD/Repository Example

In Example 43, the %DICTIONARY directive is used to access the CDD/Repository record denition Mail_Order_Info. With this denition, the Compaq Pascal program Show_Keys performs ISAM le manipulation on an existing indexed le, CUSTOMERS.DAT. Assume that CUSTOMERS.DAT has the primary key Order_Num and a eld name called Zip_Code. Note Oracle CDD/Repository has no equivalent for the Compaq Pascal KEY attribute, which is required to create new indexed les. You can use Oracle CDD/Repository data denitions to open existing indexed les (as in this example) but not new indexed les.
Example 43 Using %DICTIONARY to Access a Oracle CDD/Repository Record Denition Program Show_Keys(OUTPUT); TYPE %DICTIONARY Mail_Order_Info/LIST VAR Old_Customer_File Order_Rec Continue : FILE OF Mail_Order; : Mail_Order; : BOOLEAN;
BEGIN OPEN( File_Variable := Old_Customer_File, File_Name := Customers.Dat, History := OLD, Organization := Indexed, Access_Method := Keyed ); FINDK(Old_Customer_File, 1, 1000, NXTEQL); Continue := TRUE; WHILE Continue and NOT UFB(Old_Customer_File) DO BEGIN READ(Old_Customer_File, Order_Rec); IF Order_Rec.Zip_Code < 5000 THEN WRITELN(Order number, Order_Rec.Order_Num, has zip code, Order_Rec.Zip_Code) ELSE Continue := False; END; END. During the compilation of Show_Keys, the record denition Mail_Order_Info is extracted from the CDD/Repository. Show_Keys prints the order number and zip code of each le component that has a zip code greater than or equal to 1000 but less than 5000.
5
Calling Conventions
This chapter describes how Compaq Pascal passes parameters and calls routines. It discusses the following topics: OpenVMS Calling Standard (Section 5.1) Parameter-passing semantics (Section 5.2) Parameter-passing mechanisms (Section 5.3) Passing parameters from non-Pascal routines to Compaq Pascal routines (Section 5.4)
For More Information: On declaring and calling Compaq Pascal routines (Compaq Pascal Language Reference Manual) On procedure-calling and argument-passing mechanisms (OpenVMS Calling Standard)
5.1 OpenVMS Calling Standard

Programs compiled by the Compaq Pascal compiler conform to the OpenVMS calling standard. This standard describes how parameters are passed, how function values are returned, and how routines receive and return control. Because Compaq Pascal conforms to the calling standard, you can call and pass parameters to routines written in other Compaq languages from Compaq Pascal programs. For More Information: See the (OpenVMS Calling Standard)
Calling Conventions 51
5.1.1 Parameter Lists

Each time a routine is called, the Compaq Pascal compiler constructs a parameter list. On OpenVMS Alpha systems, the parameters are a sequence of quadword (8-byte) entries. The rst 6 integer parameters are located in integer registers designated as R16 to R21; the rst 6 oating-point parameters are located in oating-point registers designated as F16 to F21. Information about the parameter list is passed in the argument information register (R25). The rst byte of the register species the parameter count. On OpenVMS VAX systems, the parameters are a sequence of longword (4-byte) entries pointed to by the array pointer (AP). The rst byte of the rst entry in the list is a parameter count, which indicates how many parameters follow in the list. The form in which the parameters in the list are represented is determined by the passing mechanisms you specify in the formal parameter list and the values you pass in the actual parameter list. The parameter list contains the actual parameters passed to the routine.
5.1.2 Function Return Values

In Compaq Pascal, a function returns to the calling block the value that was assigned to its identier during execution. Compaq Pascal chooses one of three methods for returning this value. The method chosen depends on the amount of storage required for values of the type returned, as follows: On OpenVMS Alpha Systems: An nonoating-point scalar type, a schematic subrange, an array, a record, or set with size less than 64 bits, is returned in the rst integer register, designated as R0. If the value is less than 64 bits, R0 is sign-extended or zero-extended depending on the tape. A oating-point value that can be represented in 64 bits of storage is returned in the rst oating-point register, designated as F0. If the value is too large to be represented in 64 bits, if its type is a string type (PACKED ARRAY OF CHAR, VARYING OF CHAR, or STRING), or if the type is nonstatic, the calling routine allocates the required storage. An extra parameter (a pointer to the location where the function result will be stored) is added to the beginning of the calling routines actual parameter list.
On OpenVMS VAX Systems: If the value can be represented in 32 bits of storage, it is returned in register R0. If the value is less than 32 bits, the upper bits of R0 are undened. If the value requires from 33 to 64 bits, the low-order bits of the result are returned in register R0 and the high-order bits are returned in register R1. If the value is too large to be represented in 64 bits, if its type is a string type (PACKED ARRAY OF CHAR, VARYING OF CHAR, or STRING), or if the type is nonstatic, the calling routine allocates the required storage. An extra parameter (a pointer to the location where the function result will be stored) is added to the beginning of the calling routines actual parameter list.
Note that functions that require the use of an extra parameter can have no more than 254 parameters; functions that store their results in registers can have 255 parameters.
5.1.3 Contents of the OpenVMS Alpha Call Stack

The OpenVMS Alpha system conventions dene three types of procedures. The calling process does not need to know what type it is calling; the compiler chooses which type to generate based on the requirements of the procedure. The types of procedures are: Stack frame procedures, in which the calling context is placed on the stack Register frame procedures, in which the calling context is in registers No frame procedures, for which the compiler does not establish a context and which, therefore, execute in the context of the caller
If a stack frame is required, it consists of a xed part (that is known at compile time) and an optional variable part. The compiler determines the exact contents of the stack frame but all stack frames have common characteristics: Fixed temporary locations: an optional section that contains languagespecic locations required by the procedure context of some languages Register save area: a set of consecutive quadwords for storing registers saved and restored by the current procedure
Argument home area: if allocated, a region of memory used by the called process to assemble the arguments passed in registers adjacent to the arguments passed in memory. This allows all arguments to be addressed as a contiguous array. The argument home area is also used to store arguments passed in registers if an address for such an argument is required. Arguments passed in memory
5.1.4 Contents of the OpenVMS VAX Call Stack

Each time a routine is called by a Compaq Pascal program on an OpenVMS VAX system, the hardware creates a call frame structure on the call stack. The call frame for each active routine contains the following: A pointer to the call frame of the previous routine call. This pointer is called the saved frame pointer (FP). The saved argument pointer (AP) of the previous routine call. The storage address of the point at which the routine was called; that is, the address of the instruction following the call to the current routine. This address is called the saved program counter (PC). The saved contents of other general registers. Based on a mask specied in the control information, the system restores the saved contents of these registers to the calling routine when control returns to it.
When execution of a routine ceases, the system uses the frame pointer in the call frame of the current routine to locate the frame of the previous routine. The system then removes the call frame of the current routine from the stack.
5.1.5 Unbound Routines

The frame pointer of calling routines is stored in an implementation-dened register. If, however, you declare a routine with the UNBOUND attribute, the system does not assume that the frame pointer of the calling routine is stored in a register and there is no link between the calling and the called routines. As a result, an unbound routine has the following restrictions: It cannot access automatic variables declared in enclosing blocks. It cannot call bound routines declared in enclosing blocks. It cannot use a GOTO statement to transfer control to enclosing blocks other than the main program block.
By default, routines declared at program or module level and all other routines declared with the INITIALIZE, GLOBAL, or EXTERNAL attributes have the characteristics of unbound routines. Routines passed by the immediate value mechanism must be UNBOUND. Asynchronous system trap routines (ASTs) and RMS completion routines must have both the ASYNCHRONOUS and UNBOUND attributes. Because they are asynchronous, such routines can access only volatile variables, predeclared routines, and other asynchronous routines. Note that the Compaq Pascal run-time system does not permit a program and an asynchronous routine (such as an AST) to access the same le simultaneously. For More Information: On attributes (Compaq Pascal Language Reference Manual) On the immediate value mechanism (Section 5.3.1)
5.2 Parameter-Passing Semantics

Parameter-passing semantics describe how parameters behave when passed between the calling and called routine. Compaq Pascal passes parameter values by the following methods: Value passing semantics (Standard) Variable passing semantics (Standard) Foreign passing semantics (Compaq Pascal extension)
By default, Compaq Pascal passes arguments using value semantics. For More Information: On value, variable, and foreign semantics (Compaq Pascal Language Reference Manual)
5.3 Parameter-Passing Mechanisms

The way in which an argument species how the actual data to be passed by the called routine is dened by the parameter-passing mechanism. In compliance with the OpenVMS calling standard, Compaq Pascal supports the basic parameter-passing mechanisms, shown in Table 51.
Table 51 Parameter-Passing Descriptions on OpenVMS Alpha Systems

Mechanism By immediate value By reference By descriptor Description The argument contains the value of the data item. The argument contains the address of the data to be used by the routine. The argument contains the address of a descriptor, which describes type of the data and its location.
By default, Compaq Pascal uses the by reference mechanism to pass all actual parameters except those that correspond to conformant parameters and undiscriminated schema parameters, in which case the by descriptor mechanism is used. Table 52 describes the method you use in Compaq Pascal to obtain the desired parameter-passing mechanism. Table 52 Parameter-Passing Methods on Compaq Pascal
Mechanism By immediate value By reference By descriptor Methods Used by Compaq Pascal %IMMED or [IMMEDIATE] Default for nonconformant and nonschema parameters or %REF Default for conformant and schema parameters or %DESCR, %STDESCR, [CLASS_S],[CLASS_A], or [CLASS_NCA]
A mechanism specier usually appears before the name of a formal parameter, or if a passing attribute is used it appears in the attribute list of the formal parameter. However, in Compaq Pascal, a mechanism specier can also appear before the name of an actual parameter. In the latter case, the specier overrides the type, passing semantics, passing mechanism, and the number of formal parameters specied in the formal parameter declaration. For More Information: On passing mechanisms and passing semantics (Section 5.3.4)
5.3.1 By Immediate Value

The by immediate value passing mechanism passes a copy of a value instead of the address. Compaq Pascal provides the %IMMED foreign passing mechanism and the IMMEDIATE attribute in order to pass a parameter by immediate value. You cannot use variable semantics with the by immediate value passing mechanism. On OpenVMS systems, values that are less than or equal to 64 bits in size can be passed by immediate value. On OpenVMS VAX systems, values that are less than or equal to 32 bits in size can be passed by immediate value.
5.3.2 By Reference
The by reference mechanism passes the address of the parameter to the called routine. This is the default parameter-passing mechanism for non-conformant and non-schematic parameters. When using the by reference mechanism, the type of passing semantics used depends on the use of the VAR keyword. If the formal parameter name is preceded by the reserved word VAR, variable semantics is used; otherwise, value semantics is used. In addition to using the defaults, the Compaq Pascal compiler provides the %REF foreign passing mechanism and the REFERENCE attribute, which has more than one interpretation for the passing semantics depending on the data item represented by the actual parameter. This allows you to have the called routine use either variable semantics or true foreign semantics. The mechanism specier appears before the name of a formal parameter. The parameter passing attribute appears in the attribute list of the formal parameter.
5.3.3 By Descriptor
There are several types of descriptors. Each descriptor contains a value that identies the descriptors type. The called routine then uses the information held in the descriptor to identify its type and size. This is the default parameter-passing mechanism for conformant and schematic parameters. When you use one of the Compaq Pascal by descriptor mechanisms, the compiler passes the address of a string, array, or scalar descriptor. The Compaq Pascal compiler generates the descriptor supplying the necessary information.
Compaq Pascal provides three attributes for the by descriptor passing mechanism: [CLASS_S], [CLASS_A], and [CLASS_NCA]. With these three attributes, the type of passing semantics used for the by descriptor argument depends on the use of the VAR keyword. If the formal parameter name is preceded by the reserved word VAR, variable semantics is used; otherwise, value semantics is used. The parameter-passing attribute appears in the attribute list of the formal parameters. Sometimes you may want to choose either variable semantics or true foreign semantics. In these cases, the Compaq Pascal compiler provides two foreign passing mechanism speciers, %DESCR and %STDESCR. These speciers have more than one interpretation for the passing semantics depending on the data type of the actual parameter. The mechanism specier appears before the name of a formal parameter. Table 53 lists the class and type of descriptor generated for parameters that can be passed using the by descriptor mechanism. Table 53 Parameter Descriptors
Parameter Type %DESCR Ordinal SINGLE DSC$K_CLASS_S1 DSC$K_CLASS_S, DSC$K_DTYPE_F, DSC$K_DTYPE_FS DSC$K_CLASS_S, DSC$K_DTYPE_D, DSC$K_DTYPE_G, DSC$K_DTYPE_FT DSC$K_CLASS_S DSC$K_DTYPE_H/_FX6
D_type depends on size of type.
Descriptor Class and Type %STDESCR Value or VAR Semantics
DOUBLE
QUADRUPLE RECORD
1 Descriptors 6 DTYPE_H
is used on OpenVMS VAX systems. DTYPE_FX is used on OpenVMS Alpha systems.
Table 53 (Cont.) Parameter Descriptors

Parameter Type %DESCR ARRAY ARRAY OF VARYING OF CHAR Conformant ARRAY Conformant ARRAY OF VARYING OF CHAR4 VARYING OF CHAR Conformant VARYING OF CHAR STRING Schema name Discriminated schema SET FILE Pointer DSC$K_CLASS_A/_NCA2;5 DSC$K_CLASS_VSA DSC$K_DTYPE_VT DSC$K_CLASS_A/_NCA2;5 DSC$K_CLASS_VSA DSC$K_DTYPE_VT Descriptor Class and Type %STDESCR DSC$K_CLASS_S DSC$K_DTYPE_T3 Value or VAR Semantics
DSC$K_CLASS_S DSC$K_DTYPE_T3
DSC$K_CLASS_A/_ NCA2;5 DSC$K_CLASS_VSA DSC$K_DTYPE_VT
DSC$K_CLASS_VS DSC$K_DTYPE_VT DSC$K_CLASS_VS DSC$K_DTYPE_VT DSC$K_CLASS_S DSC$K_DTYPE_Z DSC$K_CLASS_S DSC$K_DTYPE_Z DSC$K_CLASS_S DSC$K_DTYPE_LU
DSC$K_CLASS_VS DSC$K_DTYPE_VT DSC$K_CLASS_VS DSC$K_DTYPE_VT Internal Compaq Pascal descriptor
2 Descriptors 3 Only
D_type depends on component type. type can be a conformant VARYING OF CHAR.
if PACKED ARRAY OF CHAR. is used on OpenVMS VAX systems. CLASS_NCA is used on OpenVMS Alpha systems.
4 Component 5 CLASS_A

Parameter Type %DESCR PROCEDURE or FUNCTION DSC$K_CLASS_S DSC$K_DTYPE_BPV Descriptor Class and Type %STDESCR Value or VAR Semantics Bound procedure value by reference
CLASS_A Ordinal SINGLE
CLASS_NCA
CLASS_S DSC$K_CLASS_S1 DSC$K_CLASS_S, DSC$K_DTYPE_F, DSC$K_DTYPE_FS DSC$K_CLASS_S, DSC$K_DTYPE_D, DSC$K_DTYPE_G, DSC$K_DTYPE_FT DSC$K_CLASS_S DSC$K_DTYPE_H/_ FX6 DSC$K_CLASS_S DSC$K_DTYPE_T3
DOUBLE
QUADRUPLE
RECORD ARRAY ARRAY OF VARYING OF CHAR Conformant ARRAY Conformant ARRAY OF VARYING OF CHAR4
1 Descriptors 2 Descriptors 3 Only
DSC$K_CLASS_A
2
DSC$K_CLASS_ NCA2
DSC$K_CLASS_A2
DSC$K_CLASS_ NCA2
DSC$K_CLASS_S DSC$K_DTYPE_T3
D_type depends on size of type. D_type depends on component type. type can be a conformant VARYING OF CHAR.
if PACKED ARRAY OF CHAR. is used on OpenVMS VAX systems. DTYPE_FX is used on OpenVMS Alpha systems.
4 Component 6 DTYPE_H

Parameter Type CLASS_A VARYING OF CHAR Conformant VARYING OF CHAR STRING Schema name Discriminated schema SET FILE Pointer PROCEDURE or FUNCTION Descriptor Class and Type CLASS_NCA CLASS_S
DSC$K_CLASS_S DSC$K_DTYPE_Z DSC$K_CLASS_S DSC$K_DTYPE_Z DSC$K_CLASS_S DSC$K_DTYPE_LU
5.3.3.1 CLASS_S Attribute When the CLASS_S attribute is used on a formal parameter, the compiler generates a xed-length scalar descriptor of a variable and passes its address to the called routine. Only ordinal, real, set, pointer, and one-dimensional packed arrays (xed or conformant) that are of type OF CHAR can have the CLASS_S attribute on the formal parameter. With the CLASS_S attribute, the type of passing semantics used for the by descriptor argument depends on the use of the VAR keyword. If the formal parameter name is preceded by the reserved word VAR, variable semantics is used; otherwise, value semantics is used.
5.3.3.2 CLASS_A and CLASS_NCA Attributes When you use the CLASS_A or CLASS_NCA attribute on a formal parameter, the compiler generates an array descriptor and passes its address to the called routine. The type of the CLASS_A and CLASS_NCA parameter must be an array (packed or unpacked, xed or conformant) of an ordinal or real type. With the CLASS_A and CLASS_NCA attributes, the type of passing semantics used for the by descriptor argument depends on the use of the VAR keyword. If the formal parameter name is preceded by the reserved word VAR, variable semantics is used; otherwise, value semantics is used. 5.3.3.3 %STDESCR Mechanism Specier When you use the %STDESCR mechanism specier, the compiler generates a xed-length descriptor of a character-string variable and passes its address to the called routine. Only items of the following types can have the %STDESCR specier on the actual parameter: character-string constants, string expressions, packed character arrays with lower bounds of 1, and packed conformant arrays with indexes of an integer or integer subrange type. The passing semantics depend on the variable represented by the actual parameter as follows: If the actual parameter is a variable of type PACKED ARRAY OF CHAR, %STDESCR implies variable semantics within the called routine. If the actual parameter is either a variable enclosed in parentheses, an expression, or a VARYING OF CHAR variable, %STDESCR implies foreign semantics.
If the actual parameter is not modied by the called external routine, the corresponding formal parameter should be declared READONLY, saving the copy from having to be made. The following function declaration requires one xed-length string descriptor as a parameter: [ASYNCHRONOUS,EXTERNAL(SYS$SETDDIR)] FUNCTION $SETDDIR (%STDESCR New_Dir : PACKED ARRAY [$L1..$U1: INTEGER] OF CHAR; VAR Old_Dir_Len : $UWORD := %IMMED 0; VAR Old_Dir : [CLASS_S]PACKED ARRAY [$L2..$U2 : INTEGER] OF CHAR := %IMMED 0) : INTEGER; EXTERN; . . . Status := $SETDDIR([Compaq_Pascal]); The actual parameter [Compaq_Pascal] is passed by string descriptor with foreign semantics to the formal parameter New_Dir.
5.3.3.4 %DESCR Mechanism Specier When you use the %DESCR mechanism specier, the parameter generates a descriptor for an ordinal, real, or array variable and passes its address to the called routine. The type of %DESCR parameter can be any ordinal or real type, a VARYING OF CHAR string, or an array (packed or unpacked, xed or conformant) of an ordinal or real type. The passing semantics depend on the actual parameters data type as follows: If the actual parameter is a variable, the %DESCR formal parameter implies variable semantics within the called routine. If the actual parameter is an expression or a variable enclosed in parentheses, %DESCR implies foreign semantics.
If the actual parameter is not modied by the called external routine, the corresponding formal parameter should be declared READONLY, saving the copy from having to be made. The following function declaration requires a varying-length string descriptor as its parameter: TYPE Vary = VARYING [30] OF CHAR; VAR Obj_String : Vary; Times_Found : INTEGER; [EXTERNAL] FUNCTION Search_String( %DESCR String_Val : Vary ) : BOOLEAN; EXTERNAL; . . . IF Search_String( Obj_String ) THEN Times_Found := Times_Found + 1; The actual parameter Obj_String is passed by varying string descriptor with variable semantics to the formal parameter String_Val. For More Information: On descriptor classes and types (OpenVMS Calling Standard)
5.3.4 Summary of Passing Mechanisms and Passing Semantics

Table 54 summarizes the passing semantics used when various mechanisms are specied on either the formal or the actual parameter. For example, if a variable is passed to a formal parameter that was declared without the keyword VAR and either %REF or [REFERENCE] was specied, then variable passing semantics will be used. Likewise, if a variable is passed to a formal parameter which was declared with the keyword VAR and either %REF or [REFERENCE] was specied, then an error will occur. If an actual parameter is preceded by an %IMMED specier, regardless of what passing mechanism is used to declare the formal parameter, foreign semantics would be used, because a specier appearing on the actual parameter always overrides the semantics specied on the formal parameter. Table 54 Summary of Passing Mechanisms and Passing Semantics
Actual Parameter Variable Passing Mechanism By immediate value %IMMED or [IMMEDIATE] By reference Default for nonconformant or nonschema %REF or [REFERENCE] By descriptor Default for conformant and schema [CLASS_S] [CLASS_A] [CLASS_NCA] %STDESCR %DESCR
1 If
(Variable) or Expression No VAR on Formal Foreign VAR on Formal Error
No VAR on Formal Foreign
VAR on Formal Error
Value Variable
Variable Error
Value Foreign
Value1 Error
Value Value Value Value Variable Variable
Variable Variable Variable Variable Error Error
Value Value Value Value Foreign Foreign
Value1 Value1 Value1 Value1 Error Error
the formal parameter is declared with the READONLY attribute, then value-passing semantics is used; otherwise, it is an error.
5.4 Passing Parameters between Compaq Pascal and Other Languages

Passing parameters between Compaq Pascal and other languages on OpenVMS systems requires some additional knowledge about the semantics and mechanisms used by the Compaq Pascal and the other compilers involved.
5.4.1 Parameter Mechanisms Versus Parameter Semantics

The Pascal language provides three parameter semantics, "VAR parameters", "value parameters", and "routine parameters". These models dene what happens to the parameters, not how the compiler actually implements them. "VAR parameters" are parameters that represent the actual variable passed to the routine. Changes made to the VAR parameter are reected back to the actual variable passed in to the routine. "Value parameters" are parameters that are local copies of the expression passed into the routine. Changes made to the value parameter are not reected back to any actual parameter passed in to the routine. "Routine parameters" are parameters that represent entire routines that may be called from inside the called routine. The Compaq Pascal compiler provides three parameter mechanisms, "by immediate value", "by reference", and "by descriptor". These forms represent the actual implementation used by the compiler for the parameter. These forms are denoted by the [IMMEDIATE], [CLASS_S], [CLASS_A], and [CLASS_NCA] attributes (note, the [REFERENCE] attribute doesnt just specify a parameter mechanism, but also species a parameter semantic model). Compaq Pascal also provides a fourth parameter model called "foreign parameters". These parameters become either VAR or value parameters depending on the actual parameter. If the actual parameter is a variable, then the parameter is treated as a VAR parameter. If the actual parameter is an expression, then the parameter is treated as a value parameter. These parameters are denoted by the %REF, %DESCR, and %STDESCR foreign mechanism speciers and the [REFERENCE] attribute (identical in behavior to the %REF foreign parameter specier). Be careful not to confuse the term "value parameter" with the "by immediate value" mechanism. The "value" in "value parameter" describes the semantics of the parameter where changes made to the parameter inside the called routine are not reected back to the actual parameter. It is a common misconception that Compaq Pascal uses the "by immediate value" mechanism for "value parameters."
5.4.2 Passing Nonroutine Parameters between Compaq Pascal and Other Languages
By default, Compaq Pascal will use the "by reference" mechanism for the following VAR and value parameter types: Ordinal (integer, unsigned, char, Boolean, pointers, subranges, and enumerated types), Real (real, double, quadruple), Record, Array, Set, Varying, and File. If you want to pass a parameter with the "by immediate value" mechanism, you can place the [IMMEDIATE] attribute on the denition of the formal parameters denition or use the %IMMED foreign mechanism specier on the actual parameter to override the default mechanism of the formal parameter. Only ordinal and real types may be passed with the "by immediate value" mechanism. Only value parameters may use the "by immediate value" mechanism. If you want to accept a value parameter with the "by immediate value", you can place the [IMMEDIATE] attribute on the denition of the formal parameter. Only ordinal and real types may be accepted with the "by immediate value" mechanism. For example, to pass an integer with the "by immediate value" mechanisn to another routine, [external] procedure rtn( p : [immediate] integer ); external; begin rtn(3); rtn(some-integer-expression); end; If you want to pass a parameter with the "by descriptor" mechanism, you can place the [CLASS_A], [CLASS_S], or [CLASS_NCA] attributes on the formal parameters denition. You can also use the %DESCR and %STDESCR foreign mechanism speciers, but be aware that these also imply parameter semantics as well as the parameter-passing mechanism. When passing values to a subrange parameter in a Pascal routine, the argument must be large enough to hold a value of the subranges base-type even if the formal parameter contained a size attribute. When passing Boolean or enumerated-type values to a VAR parameter in a Pascal routine, the calling routine must ensure that the sizes of the Boolean or enumerated-type matches the setting of the /ENUMERATION_SIZE qualier or [ENUMERATION_SIZE] attribute used in the Pascal routine. For value parameters, you can pass the address of a longword as that will work for either setting.
When passing arrays or records to a Pascal routine, the calling routine must ensure that the array and record has the same layout (including any alignment holes) as chosen by the Compaq Pascal compiler. You may want to use the /SHOW=STRUCTURE_LAYOUT listing section to help you determine the layout chosen by the Compaq Pascal compiler. By default, Compaq Pascal will use the "by descriptor" mechanism for VAR and value conformant parameters. For conformant-varying parameters, Compaq Pascal uses a CLASS_VS desciptor. For conformant-array parameters, Compaq Pascal uses a CLASS_A descriptor on OpenVMS systems and a CLASS_NCA descriptor on OpenVMS Alpha systems. Using a conformant-varying parameter or STRING schema parameter with a routine not written in Pascal is not very common since the called routine does not know how to deal with these strings. If you just are passing a string expression to the non-Pascal routine, using a conformant PACKED ARRAY OF CHAR is the right solution. Since Compaq Pascal will use either a CLASS_A or CLASS_NCA descriptor for the conformant PACKED ARRAY OF CHAR, but other languages will expect either a CLASS_S descriptor or the string "by reference", you will need to use either the [CLASS_S] attribute or the %REF foreign mechanism specier. For example, to pass a string expression to Fortran (which expects a CLASS_S descriptor), [external] procedure fortrtn( p : [class_s] packed array [l..u:integer] of char); external; begin fortrtn(string); fortrtn(some-string-expression); end; To pass a string expression to C (which expects a "by-reference" parameter and a null-terminated string), [external] procedure crtn( %ref p : packed array [l..u:integer] of char); external; begin crtn(string(0)); crtn(some-string-expression+0); end; Compaq Pascal on OpenVMS Alpha systems has additional support to deal with null-terminated strings.
For More Information Compaq Pascal Language Reference Manual
When passing strings to a Compaq Pascal routine from another language, you must use a descriptor if the Pascal formal parameter is a conformant parameter. Compaq Pascal cannot accept a conformant parameter with the "by reference" mechanism. For conformant array parameters, you must generate a CLASS_A descriptor on OpenVMS or a CLASS_NCA descriptor on OpenVMS Alpha unless you select another descriptor class using the [CLASS_S], [CLASS_A], or [CLASS_NCA] attributes. For conformant varying parameters, you must generate a CLASS_VS descriptor on both platforms. If you wish to use the "by reference" mechanism to pass strings into a Pascal routine, you must dene a Pascal datatype that represents a xed-length string (or varying-string with a maximum size) and use that datatype in the formal parameter denition. The Compaq Pascal schema type STRING is passed by CLASS_VS descriptor. Other Compaq Pascal schema types use private data structures when passed between routines and cannot be accessed from routines written in other languages.
5.4.3 Passing Routine Parameters between Compaq Pascal and Other Languages
By default, Compaq Pascal on OpenVMS VAX will pass the address of a BPV (Bound Procedure Value) for PROCEDURE or FUNCTION parameters. The BPV is a 2 longword structure with the 1st longword being the address of the entry mask of the routine being passed; the 2nd longword is either a 0 or a frame-pointer for any uplevel references that the called routine may attempt to perform. If you use the [UNBOUND] attribute on the PROCEDURE or FUNCTION parameter, the compiler still passes the address of a BPV, but the 2nd longword will always be zero. If you use the %IMMED foreign mechanism specier along with the [UNBOUND] attribute, Compaq Pascal will pass the address of the routine with the "by immediate" mechanism.
For example, if you want to pass a Pascal routine to an external routine that just expects the address of the routine (like a C routine or system service), [external] procedure crtn( %immed [unbound] procedure p); external; procedure x; begin ... end; begin crtn(x); end; Most system routines expect to be called in this fashion. A few routines like EDT$EDIT actually expect BPVs and not just entry point addresses. On OpenVMS VAX, Compaq Pascal only accepts a routine parameter by BPV. To pass a routine from a non-Pascal routine to a Pascal routine, you must use the [UNBOUND] attribute on the formal routine parameter and you must pass the address of a longword containing the address of the routines entry point. By default, Compaq Pascal on OpenVMS Alpha passes the address of a procedure descriptor for PROCEDURE or FUNCTION parameters. The presence of the [UNBOUND] attribute or the %IMMED foreign mechanism specier has no effect over the generated code since the procedure descriptors in the OpenVMS Alpha Calling Standard allow any combination of bound and unbound routines to be passed around and invoked. On OpenVMS Alpha, Compaq Pascal expects the address of a procedure descriptor for routine parameters. In all Alpha languages, asking for the address of a routine, yields the address of its procedure descriptor since the actual address of the instructions is not useful by itself.
6
Programming on OpenVMS Systems
To eliminate duplication of programming and debugging efforts, OpenVMS systems provide many routines to perform common programming tasks. These routines are collectively known as system routines. They include routines in the run-time library to assist you in such areas as mathematics, screen management, and string manipulation. Also included are OpenVMS Record Management Services (RMS), which are used to access les and their records. There are also system services that perform tasks such as resource allocation, information sharing, and input/output coordination. This chapter discusses the following topics: Using system denitions les (Section 6.1) Declaring system routines (Section 6.2) Calling system routines (Section 6.3) Using attributes (Section 6.4) Using item lists (Section 6.5) Using foreign mechanism speciers on actual parameters (Section 6.6)
Because all OpenVMS system routines adhere to the OpenVMS Calling Standard, you can declare any system routine as an external routine and then call the routine from a Compaq Pascal program.
6.1 Using System Denitions Files

To access system entry points, data structures, symbol denitions, and messages, Compaq Pascal provides les that you can inherit (.PEN) or include (.PAS) in your application. Table 61 summarizes the source and environment les that Compaq Pascal makes available to you in the directory SYS$LIBRARY (for instance, SYS$LIBRARY:STARLET.PEN).
Programming on OpenVMS Systems 61
Table 61 Compaq Pascal Denitions Files

File System Services Denitions File: STARLET.PAS/.PEN Contains OpenVMS system service denitions, LIB$ messages, MTH$ messages, OTS$ messages, SMG$ data structures and termtable, STR$ messages, RMS routine declarations, system symbolic names, and RMS data structures. Description
Run-Time Library Denitions Files: PASCAL$ACLEDIT_ROUTINES.PAS/.PEN PASCAL$C_ROUTINES.PAS/.PEN PASCAL$CLI_ROUTINES.PAS/.PEN PASCAL$CMA_ROUTINES.PAS/.PEN2 Contains ACLEDIT$ routine entry points. Contains C$ routine entry points. Contains CLI$ routine entry points. Contains routine entry points, data structures, and messages for DECthreads. For more information on DECthreads, see the Guide to DECthreads. Contains CONV$ routine entry points. Contains CVT$ routine entry points. Contains DCX$ routine entry points. Contains DTK$ routine entry points, data structures, and messages. Contains EDT$ routine entry points. Contains FDL$ routine entry points. Contains LBR$ routine entry points. Contains LIB$ routine entry points. Contains MAIL$ routine entry points. Contains MTH$ routine entry points. Contains NCS$ routine entry points.
PASCAL$CONV_ROUTINES.PAS/.PEN PASCAL$CVT_ROUTINES.PAS/.PEN PASCAL$DCX_ROUTINES.PAS/.PEN PASCAL$DTK_ROUTINES.PAS/.PEN PASCAL$EDT_ROUTINES.PAS/.PEN PASCAL$FDL_ROUTINES.PAS/.PEN PASCAL$LBR_ROUTINES.PAS/.PEN PASCAL$LIB_ROUTINES.PAS/.PEN PASCAL$MAIL_ROUTINES.PAS/.PEN PASCAL$MTH_ROUTINES.PAS/.PEN PASCAL$NCS_ROUTINES.PAS/.PEN
2 These
les are only provided on OpenVMS Alpha systems.
Table 61 (Cont.) Compaq Pascal Denitions Files

File Run-Time Library Denitions Files: PASCAL$OTS_ROUTINES.PAS/.PEN PASCAL$PPL_ROUTINES.PAS/.PEN PASCAL$PSM_ROUTINES.PAS/.PEN PASCAL$SMB_ROUTINES.PAS/.PEN PASCAL$SMG_ROUTINES.PAS/.PEN PASCAL$SOR_ROUTINES.PAS/.PEN PASCAL$STR_ROUTINES.PAS/.PEN PASCAL$TPU_ROUTINES.PAS/.PEN Symbol Denitions Files:1 LIBDEF.PAS Contains denitions for all condition symbols from the general utility run-time library routines. Contains denitions for all condition symbols from the mathematical routines library. Contains miscellaneous symbol denitions used in condition handlers. These denitions are also included in STARLET.PEN. Contains OTS$ routine entry points. Contains PPL$ routine entry points, data structures, and messages. Contains PSM$ routine entry points. Contains SMB$ routine entry points. Contain SMG$ routine entry points and messages. Contains SOR$ routine entry points and messages. Contains STR$ routine entry points. Contains TPU$ routine entry points. Description
MTHDEF.PAS
SIGDEF.PAS
Compaq Pascal run-time library Files: PASDEF.PAS Contains denitions for all condition symbols from the Compaq Pascal run-time library routines.
les are retained for compatibility with older versions of this product and do not contain symbol denitions for subsequent releases of the product. (For denitions that are complete for the latest release of OpenVMS, use the individual PASCAL$ les or STARLET). To access these les, use the %INCLUDE directive in the CONST declaration section of your program.
1 These
Table 61 (Cont.) Compaq Pascal Denitions Files

File Compaq Pascal run-time library Files: PASSTATUS.PAS Contains denitions for all values returned by the STATUS and STATUSV routines. Description
For instance, the external routine declarations in STARLET dene new identiers by which you can refer to the routines. Example 61 shows that you can refer to SYS$HIBER as $HIBER if you use STARLET.
Example 61 Inheriting STARLET.PEN to Call SYS$HIBER [INHERIT(SYS$LIBRARY:STARLET)] PROGRAM Suspend (INPUT,OUTPUT); TYPE Sys_Time = RECORD I,J : INTEGER; END; Unsigned_Word = [WORD] 0..65535; VAR Current_Time : PACKED ARRAY[1..80] OF CHAR; Length : Unsigned_Word; Job_Name : VARYING[15] OF CHAR; Ascii_Time : VARYING[80] OF CHAR; Binary_Time : Sys_Time; BEGIN { Print current date and time } $ASCTIM (TIMLEN := Length, TIMBUF := Current_Time); WRITELN (The current time is , SUBSTR(Current_Time, 1, Length); { Get name of process to suspend } WRITE (Enter name of process to suspend: ); READLN (Job_Name); { Get time to wake process } WRITE (Enter time to wake process: ); READLN (Ascii_Time); { Convert time to binary } IF NOT ODD ($BINTIM (Ascii_Time, Binary_Time)) THEN BEGIN WRITELN (Illegal format for time string); HALT; END; { Suspend process } IF NOT ODD ($SUSPND (PRCNAM := Job_Name)) THEN BEGIN WRITELN (Cannot suspend process); HALT; END; { Schedule wakeup request for self } IF ODD ($SCHDWK (DAYTIME := Binary_Time)) THEN $HIBER { Put self to sleep } ELSE BEGIN WRITELN (Cannot schedule wakeup); WRITELN (Process will resume immediately); END;
(continued on next page) Programming on OpenVMS Systems 65
Example 61 (Cont.) Inheriting STARLET.PEN to Call SYS$HIBER { Resume process } IF NOT ODD ($RESUME (PRCNAM := Job_Name)) THEN BEGIN WRITELN (Cannot resume process); HALT; END; END.
6.2 Declaring System Routines

Before calling a routine, you must declare it. System routine names conform to one of the two following conventions: [ [facility-code] ]$procedure-name For example, LIB$PUT_OUTPUT is the run-time library routine used to write a record to the current output device and $ASCTIM is a system service routine used to convert binary time to ASCII time. Because system routines are often called from condition handlers or asynchronous trap (AST) routines, you should declare system routines with the ASYNCHRONOUS attribute. Each system routine is documented with a structured format in the appropriate OpenVMS reference manual. The documentation for each routine describes the routines purpose, the declaration format, the return value, and any required or optional parameters. Detailed information about each parameter is listed in the description. The following format is used to describe each parameter: parameter-name OpenVMS Usage : type : access : mechanism : OpenVMS data type parameter data type parameter access parameter-passing mechanism
Using this information, you must determine the parameters data type (type), the parameters passing semantics (access), the mechanism used to pass the parameter (mechanism), and whether the parameter is required or optional from the call format. The following sections describe the methods available in Compaq Pascal to obtain the various data types, access methods, and passing mechanisms.
6.2.1 Methods Used to Obtain OpenVMS Data Types

The data specied by a parameter has a data type. Several OpenVMS standard data types exist. A system routine parameter must use one of these data types. For More Information: On OpenVMS data types and equivalent Compaq Pascal declarations (OpenVMS Calling Standard)
6.2.2 Methods Used to Obtain Access Methods

The access method describes the way in which the called routine accesses the data specied by the parameter. The following three methods of access are the most common: Read onlydata must be read by the called routine. The called routine does not write the input data. Thus, input data supplied by a variable is preserved when the called routine completes execution. Write onlydata that the called routine returns to the calling routine must be written into a location where the calling routine can access it. Such data can be thought of as output data. The called routine does not read the contents either before or after it writes into the location. Modifya parameter species data that is both read and written by the called routine. In this case, the called routine reads the input data, which it uses in its operations, and then overwrites the input data with the results. Thus, when the called routine completes execution, the input data specied by the argument is lost.
Table 62 lists all access methods that may appear under the access entry in a parameter description, as well as the Compaq Pascal translation. Table 62 Access Type Translations
Access Entry Call after stack unwind Function call (before return) Jump after unwind Method Used in Compaq Pascal Procedure or function parameter passed by immediate value Function parameter Not available (continued on next page)
Table 62 (Cont.) Access Type Translations

Access Entry Modify Read only Call without stack unwind Write only
1 It
Method Used in Compaq Pascal Variable semantics1 Value or foreign semantics Procedure parameter Variable semantics1
is possible to obtain variable semantics by either specifying the VAR keyword on the formal parameter or by passing a variable as an actual parameter using %REF, %DESCR, or %STDESCR.
6.2.3 Methods Used to Obtain Passing Mechanisms

The way in which an argument species the actual data to be used by the called routine is dened in terms of the parameter-passing mechanism. Table 63 lists all passing mechanisms that may appear under the mechanism entry in an argument description and the method used in Compaq Pascal to obtain the passing mechanism. Table 63 Mechanism Type Translations
Mechanism Entry By value By reference By descriptor Fixed-length Method Used in Compaq Pascal %IMMED or [IMMEDIATE] VAR, %REF or [REFERENCE] or default %STDESCR parameter of type PACKED ARRAY OF CHAR or [CLASS_S] %STDESCR parameter of type PACKED ARRAY OF CHAR or [CLASS_S] Array type, conformant array parameter, or [CLASS_A] N.A. N.A. (continued on next page)
Dynamic-string
Array Procedure Decimal-string
Table 63 (Cont.) Mechanism Type Translations

Mechanism Entry Noncontiguousarray Varying-string Method Used in Compaq Pascal %DESCR or [CLASS_NCA]
Value, VAR, or %DESCR conformant parameter of type VARYING OF CHAR, or %DESCR parameter of type VARYING OF CHAR Value, VAR, or %DESCR conformant parameter of type array of VARYING OF CHAR, or %DESCR parameter of type array of VARYING OF CHAR N.A.
Varying-stringarray
Unaligned-bitstring Unaligned-bit-array String-with-bounds Unaligned-bitstring-withbounds
N.A. N.A. N.A.
Parameters passed by reference and used solely as input to a system service should be declared with Compaq Pascal value semantics; this allows actual parameters to be compile-time and run-time expressions. When a system service requires a formal parameter with a mechanism specier, you should declare the formal parameter with the READONLY attribute to specify value semantics. Other parameters passed by reference should be declared with Compaq Pascal variable semantics to ensure that the output data is interpreted correctly. In some cases, by reference parameters are used for both input and output and should also be declared with variable semantics. The following example shows the declaration of the Convert ASCII String to Binary Time (SYS$BINTIM) system service and a corresponding function designator:
TYPE $QUAD = [QUAD,UNSAFE] RECORD L0 : UNSIGNED; L1 : INTEGER; END; VAR Ascii_Time : VARYING[80] OF CHAR; Binary_Time : $QUAD; [ASYNCHRONOUS,EXTERNAL(SYS$BINTIM)] FUNCTION $BINTIM (TIMBUF : [CLASS_S] PACKED ARRAY [$l1..$u1:INTEGER] OF CHAR; VAR TIMADR : [VOLATILE] $QUAD) : INTEGER; EXTERNAL; {In the executable section:} IF NOT ODD ($BINTIM(Ascii_Time, Binary_Time)) THEN BEGIN WRITELN (Illegal format for time string); HALT; END; The rst formal parameter requires the address of a character-string descriptor with value semantics; the second requires an address and uses variable semantics to manipulate the parameter within the service. Because you can call $BINTIM from a condition handler or AST routine, you should declare it with the ASYNCHRONOUS attribute. Also, because you may want to pass a volatile variable to the TIMADR parameter, you should use the VOLATILE attribute to indicate that the argument is allowed to be volatile.
6.2.4 Data Structure Parameters

Some system services require a parameter to be the address of a data structure that indicates a function to be performed or that holds information to be returned. Such a structure can be described as a list, a control block, or a vector. The size and POS attributes provide an efcient method of laying out these data structures. The size attributes ensure that the elds of the data structure are of the size required by the system service, and the POS attribute allows you to position the elds correctly. For example, the Get Job/Process Information (SYS$GETJPIW) system service requires an item list consisting of an array of records of 12 bytes, where all but the last array cell requests one piece of data and the last array cell represents the item list terminator. By packing the record, you can guarantee that the elds of each record are allocated contiguously. Example 62 uses the system service routine $GETJPIW to retrieve the processs name as a 12-byte string.
Example 62 Using $GETJPIW to Retrieve a Process Name [INHERIT(SYS$LIBRARY:STARLET)] PROGRAM Userid( OUTPUT ); TYPE Uword = [WORD] 0..65535; Itmlst_Cell = PACKED RECORD CASE INTEGER OF 1 : (Buf_Len : Uword; Item_Code : Uword; Buf_Addr : INTEGER; Len_Addr : INTEGER); 2 : (Term : INTEGER); END; VAR Username_String : [VOLATILE] VARYING [12] OF CHAR; Itmlst : ARRAY [1..2] OF Itmlst_Cell := ZERO; BEGIN Itmlst[1].Buf_Len := 12; { 12 bytes returned } Itmlst[1].Item_Code := JPI$_USERNAME; { return user name } Itmlst[1].Buf_Addr := { store returned name here } IADDRESS(Username_String.BODY); Itmlst[1].Len_Addr := { store returned length here } IADDRESS(Username_String.LENGTH); Itmlst[2].Term := 0; { terminate item list } IF NOT ODD( $GETJPIW(,,,Itmlst) ) THEN WRITELN(error) ELSE WRITELN(user name is ,Username_String); END. For More Information: On size attributes ( Compaq Pascal Language Reference Manual )
6.2.5 Default Parameters

In some cases, you do not have to supply actual parameters to correspond to all the formal parameters of a system service. In Compaq Pascal, you can supply default values for such optional parameters when you declare the service. You can then omit the corresponding actual parameters from the routine call. If you choose not to supply an optional parameter, you should initialize the formal parameter with the appropriate value using the by immediate value (%IMMED) mechanism. The correct default value is usually 0.
For example, the Cancel Timer (SYS$CANTIM) system service has two optional parameters. If you do not specify values for them in the actual parameter list, you must initialize them with zeros when they are declared. The following example is the routine declaration for SYS$CANTIM: [ASYNCHRONOUS,EXTERNAL(SYS$CANTIM)] FUNCTION $CANTIM ( %IMMED REQIDT : UNSIGNED := %IMMED 0; %IMMED ACMODE : UNSIGNED := %IMMED 0) : INTEGER; EXTERNAL; A call to $CANTIM must indicate the position of omitted parameters with a comma, unless they all occur at the end of the parameter list. For example, the following are legal calls to $CANTIM using the previous external declaration: $CANTIM (, PSL$C_USER); $CANTIM (I); $CANTIM; PSL$C_USER is a symbolic constant that represents the value of a user access mode, and I is an integer that identies the timer request being canceled. If you call $CANTIM with both of its default parameters, you can omit the actual parameter list completely. When it is possible for the parameter list to be truncated, you can also specify the TRUNCATE attribute on the formal parameter declaration of the optional parameter. The TRUNCATE attribute indicates that an actual parameter list for a routine can be truncated at the point that the attribute was specied. However, once one optional parameter is omitted in the actual parameter list, it is not possible to specify any optional parameter following that. For example: [ASYNCHRONOUS] FUNCTION LIB$GET_FOREIGN ( VAR Resultant_String : [CLASS_S,VOLATILE] PACKED ARRAY [$l1..$u1:INTEGER] OF Prompt_String : [CLASS_S,TRUNCATE] PACKED ARRAY [$l2..$u2:INTEGER] OF VAR Resultant_Length : [VOLATILE,TRUNCATE] VAR Flags : [VOLATILE,TRUNCATE] : INTEGER; EXTERNAL;
CHAR; CHAR := %IMMED 0; $UWORD := %IMMED 0; UNSIGNED := %IMMED 0)
With this declaration, it is possible to specify values for Resultant_String and Prompt_String and truncate the call list at that point. In this case, two parameters would be passed in the CALL instruction. You may want to use a combination of the %IMMED 0 and TRUNCATE methods. This combination would allow you to skip the specication of intermediate optional parameters, as well as allow you to truncate the call list once all desired parameters have been specied.
Note that OpenVMS system services require a value to be passed by parameters, including optional parameters; therefore, you should not use the TRUNCATE attribute when dening optional parameters to a system service. Instead, you should specify default values on the formal parameter declaration. The TRUNCATE attribute is useful when calling routines for which the optional parameter is truly optional, for example, when calling run-time library routines. For More Information: On the TRUNCATE attribute (Compaq Pascal Language Reference Manual)
6.2.6 Arbitrary Length Parameter Lists

Some run-time library routines require a variable number of parameters. For example, there is no xed limit on the number of values that can be passed to functions that return the minimum or maximum value from a list of input parameters. The LIST attribute supplied by Compaq Pascal allows you to indicate the mechanism by which excess actual parameters are to be passed. For example: [ASYNCHRONOUS] FUNCTION MTH$DMIN1 ( D_FLOATING : DOUBLE; EXTRA_PARAMS : [LIST] DOUBLE) : DOUBLE; EXTERNAL; Because the function MTH$DMIN1 returns the D_oating minimum of an arbitrary number of D_oating parameters, the formal parameter EXTRA_ PARAMS is declared with the LIST attribute. All actual parameters must be double-precision real numbers passed by reference with value semantics. For More Information: On the LIST attribute (Compaq Pascal Language Reference Manual)
6.3 Calling System Routines

All system routines are functions that return an integer condition value; this value indicates whether the function executed successfully. An odd-numbered condition value indicates successful completion; an even-numbered condition value indicates a warning message or failure. Your program can use the Compaq Pascal predeclared function ODD to test the function return value for success or failure. For example:
IF NOT ODD ($BINTIM(Ascii_Time,Binary_Time)) THEN BEGIN WRITELN(Illegal format for time string); HALT; END; In addition, run-time library routines return one or two values: the result of a computation or the routines completion status, or both. When the routine returns a completion status, you should verify the return status before checking the result of a computation. You can use the function ODD to test for success or failure or you can check for a particular return status by comparing the return status to one of the status codes dened by the system. For example: VAR Seed_Value : INTEGER; Rand_Result : REAL; [ASYNCHRONOUS] FUNCTION MTH$RANDOM ( VAR seed : [VOLATILE] UNSIGNED) : SINGLE; EXTERNAL; {In the executable section:} Rand_Result := MTH$RANDOM (Seed_Value); When the routines completion status is irrelevant, your program can treat the function as though it were an external procedure and ignore the return value. For example, your program can declare the Hibernate (SYS$HIBER) system service as a function but call it as though it were a procedure: [ASYNCHRONOUS,EXTERNAL(SYS$HIBER)] FUNCTION $HIBER : INTEGER; EXTERNAL; {In the executable section:} $HIBER; { Put process to sleep } Because SYS$HIBER is expected to execute successfully, the program will ignore the integer condition value that is returned.
6.4 Using Attributes

When writing programs that use OpenVMS System Services and run-time library routines, it is common to use several Compaq Pascal attributes. The VOLATILE attribute indicates that a variable is written or read indirectly without explicit program action. The most common occurrence for this is with item lists. In that case, the address of the variable is placed in the item list (most likely using the IADDRESS routine). This address is then used later when the entire item list is passed to a system service. Without the VOLATILE attribute, the compiler does not realize that the call to the system service or run-time library routine uses the variable.
The UNBOUND attribute designates a routine that does not have a static link available to it. Without a static link, a routine can only access local variables, parameters, or statically allocated variables. System services that require AST or action routines want the address of an UNBOUND routine. Routines at the outer level of a PROGRAM or MODULE are UNBOUND by default. The ASYNCHRONOUS attribute designates a routine that might be called asynchronously of any program action. This allows the compiler to verify that the asynchronous routine only accesses local variables, parameters, and VOLATILE variables declared at outer levels. Without the assurance that only VOLATILE variables are used, the asynchronous routine might access incorrect data, or data written by the routine will not be available to the main program.
6.5 Using Item Lists

Many OpenVMS system services use item lists. Item lists are sequences of control structures that provide input to the system service and that describe where the service should place its output. These item lists can have an arbitrary number of cells and are terminated with a longword of value 0. Since different programs need a different number of item list cells, you can use a schema type to dene a generic item list data type. This schema type can then be discriminated with the appropriate number of cells. Consider the following example: TYPE Item_List_Cell = RECORD CASE INTEGER OF 1: ( { Normal Cell } Buffer_Length : [WORD] 0..65535; Item_Code : [WORD] 0..65535; Buffer_Addr : UNSIGNED; Return_Addr : UNSIGNED ); 2: ( { Terminator } Terminator : UNSIGNED ); END; Item_List_Template( Count : INTEGER ) = ARRAY [1..Count] OF Item_List_Cell; The Item_List_Cell data type species what a single cell looks like. The Buffer_Addr and Return_Addr elds are declared as UNSIGNED since most applications use the IADDRESS predeclared routine to ll them in. The Item_List_Template schema type denes an array of item list cells with a upper bound to be lled in by an actual discriminant.
To use this schema type, rst determine the number of item list cells required including one cell for the terminator. After the number of cells has been determined, declare a variable discriminating the schema. Consider the following example: VAR Item_List : Item_List_Template( 2 ); Additionally, since actual discriminants to schema can be run-time expressions, you can write a routine that can have item lists with a number of cells that is determined at run time. After the item list variable has been declared, each cell must be lled in according to the system service and operation requested. Consider the following example using the SYS$TRNLNM system service: VAR Item_List : Item_List_Template( 2 ); Translated_Name : [VOLATILE] VARYING [132] OF CHAR; {Specify the buffer to return the translation:} Item_List[1].Buffer_Length := SIZE( Translated_Name.BODY ); Item_List[1].Item_Code := LNM$_String; Item_List[1].Buffer_Addr := IADDRESS( Translated_Name.BODY ); Item_List[1].Return_Addr := IADDRESS( Translated_Name.LENGTH); { Terminate the item list:} Item_List[2].Terminator := 0; The VAR section declares an item list with two cells. It also declares an output buffer for the system service. The VOLATILE attribute is used since the call to SYS$TRNLNM indirectly writes into the variable. The rst cell is lled in with the operation desired, the size of the output buffer, the location to write the result, and the location to write the size of the result. Using the SIZE predeclared function prevents the code from having to be modied if the output buffer ever changes size. Using the BODY and LENGTH predeclared elds of the VARYING string allows the system service to construct a valid VARYING OF CHAR string. Finally, the second cell of the item list is initialized. Since the second cell is the last cell, the terminator eld must be lled in with a value of 0.
6.6 Using Foreign Mechanism Speciers on Actual Parameters

The denition les provided by Compaq Pascal (SYS$LIBRARY:STARLET.PAS and so forth) are created from a generic description language used by the OpenVMS operating system. Since this description language does not contain all the features found in Compaq Pascal, some of the translations do not take advantage of Compaq Pascal features. Also, since several of the system services are generic in nature, it is impossible to provide a denitive denition for every situation. If a formal parameter denition does not reect the current usage, you can use a foreign mechanism specier to direct the compiler to use a different passing mechanism or different descriptor type than the default for that parameter. Consider the following: The ASTADR parameter Many system services dene this parameter to be a procedure parameter with no formal parameters. This is because the format of the arguments passed to the AST routine vary with the system service. If you specify a routine with parameters as the actual parameter to an ASTADR parameter, you will receive a compile-time error saying that the formal parameter and actual parameter have different parameter lists. To solve this problem, you can specify the %IMMED foreign mechanism specier on the actual parameter. This causes the compiler to pass the address of the routine without verifying that the parameter lists are identical. The ASTPRM parameter Many system services dene this parameter to be an UNSIGNED parameter passed by immediate value. Since the parameter to an AST routine is dependent on the application, it is often desired to pass the address of a variable instead of its contents. To solve this problem, you can specify the %REF foreign mechanism specier on the actual parameter. This causes the compiler to pass the address of the variable instead of the contents of the variable. The P1..Pn parameters The P1 through P6 parameters of the $QIO and $QIOW system services and the P1 through P20 parameters of the $FAO system services are also dened to be UNSIGNED parameters passed by immediate value. If the actual parameter is not UNSIGNED or requires a different passing mechanism, you can specify the %REF foreign mechanism specier on
the actual parameter. This causes the compiler to pass the address of the variable instead of the contents of the variable. The RESULTANT_FILESPEC parameter of the LIB$FIND_FILE run-time library routine This parameter is declared to be a VAR conformant PACKED ARRAY OF CHAR parameter and is passed by a CLASS_A descriptor. However, the LIB$FIND_FILE routine can also accept CLASS_VS descriptors of VARYING OF CHAR variables. To cause the compiler to build a CLASS_ VS descriptor instead of the default CLASS_A descriptor, you can specify the %DESCR foreign mechanism specier on the actual VARYING OF CHAR parameter.
6.7 Using 64-Bit Pointer Types on OpenVMS Alpha

Compaq Pascal on OpenVMS Alpha includes limited support for 64-bit pointers. 64-bit pointers can be declared by using the [QUAD] attribute on a pointer variable declaration. When [QUAD] is used, the generated code will use all 64 bits of the pointer.
6.7.1 Pascal Language Features Not Supported with 64-Bit Pointers

Several Pascal features are not supported with 64-bit pointers. These features are as follows: Base types of 64-bit pointers cannot contain le types or schema types. The NEW and DISPOSE built-in routines do not work with 64-bit pointer types. You must call OpenVMS system services directly (or a C language routine) to allocate memory in the P2 address space and assign the address into the 64-bit pointer variable directly. The READ built-in routine cannot read into variables accessed with 64-bit pointers. For example, the following code fragment will be rejected by the compiler: var quad_ptr : [quad] înteger; begin quad_ptr := my_alloc_routine(size(integer)); read(quad_ptr^); end
Compaq Pascal understands 32-bit descriptors as dened by the OpenVMS calling standard. Therefore, any Compaq Pascal construct that relies on descriptors is not supported for variables accessed with 64-bit pointers. The features rejected for 64-bit pointers are: The use of %DESCR or %STDESCR on actual parameter values accessed with 64-bit pointers. For example, you cannot do the following: type s32 = packed array [1..32] of char; var qp : [quad] ^s; begin qp := my_alloc_routine(size(s32)); some_routine( %stdescr qp^ ); end; Passing variables accessed with 64-bit pointers to formal parameters declared with %DESCR or %STDESCR foreign mechanism speciers. Passing variables accessed with 64-bit pointers to conformant array or conformant varying parameters. Passing variables accessed with 64-bit pointers to STRING parameters. At run time, the compiler will generate incorrect code when passing a VAR parameter that is accessed with a 64-bit pointer to a parameter that requires a descriptor. The generated code will build the descriptor with the lower 32-bits of the 64-bit address. For example: type s32 = packed array [1..32] of char; var qp : [quad] ^s; procedure a( p : packed array [l..u:integer] of char ); begin writeln(a); end; procedure b( var p : s32 ); begin a(p); { This will generate a bad descriptor } end; begin qp := my_alloc_routine(size(s32)); b(qp^); end;
6.7.2 Using 64-Bit Pointers with System Denition Files

For routines that have parameters that are 64-bit pointers, the Pascal denition uses a 64-bit record type. The denition les do not support either the INTEGER64 datatype or 64-bit pointers. You can override the formal denition inside of denition les by using a foreign mechanism specier (that is, %IMMED, %REF, %STDESCR, and %DESCR) on an actual parameter. For example, the following is an example of calling lib$get_vm_64 using %ref to override the denition from PASCAL$LIB_ROUTINES.PEN: [inherit(sys$library:pascal$lib_routines)] program p64(input,output); const arr_size = (8192 * 10) div 4; ! Make each array be 10 pages type arr = array [1..arr_size] of integer; arrptr = [quad] ârr; var ptr : arrptr; ptrarr : array [1..10] of arrptr; i,j,stat : integer; sum : integer64; ! PASCAL$LIB_ROUTINES.PAS on a V7.1 system contains ! the following definitions for LIB$GET_VM_64 ! !type ! $QUAD = [QUAD,UNSAFE] RECORD ! L0:UNSIGNED; L1:INTEGER; END; ! $UQUAD = [QUAD,UNSAFE] RECORD ! L0,L1:UNSIGNED; END; ! lib$routines$$typ4 = ^$QUAD; ! ![ASYNCHRONOUS] FUNCTION lib$get_vm_64 ( ! number_of_bytes : $QUAD; ! VAR base_address : [VOLATILE] lib$routines$$typ4; ! zone_id : $UQUAD := %IMMED 0) : INTEGER; EXTERNAL; ! ! Note that the BASE_ADDRESS parameter is a 64-bit pointer ! that will be returned by LIB$GET_VM_64. The definition ! incorrectly declared it as a pointer to a record that is ! quadword sized. ! begin
! ! ! ! ! ! ! ! !
Allocate memory with lib$get_vm_64. The definition of lib$get_vm_64 declares the return address parameter as a quadword-sized record since it doesnt have sufficient information to generate a INTEGER64 or other type. Use an explicit %ref foreign mechanism specifier to override the formal parameters type definition and pass our pointer to lib$get_vm_64.
writeln(arr_size = ,arr_size:1); for i := 1 to 10 do begin stat := lib$get_vm_64( size(arr), %ref ptrarr[i] ); if not odd(stat) then begin writeln(Error from lib$get_vm_64: ,hex(stat)); lib$signal(stat); end; writeln(ptrarr[,i:1,] = ,hex(ptrarr[i])); end; ! Read/write all the memory locations to get some page faults ! writeln(Initialize all memory); for i := 1 to 10 do for j := 1 to arr_size do ptrarr[i]^[j] := i + j; sum := 0; writeln(Add up all memory in reverse direction); for i := 10 downto 1 do for j := arr_size downto 1 do sum := sum + ptrarr[i]^[j]; writeln(Sum of array contents = ,sum:1); end. On OpenVMS Alpha, the compiler allows the LONG and QUAD attributes to be specied on pointer types, as shown in the following example: var long_ptr : înteger; quad_ptr : [quad] înteger; Both pointers point to integers, but long_ptr is 32 bits while quad_ptr is 64 bits.
7
Input and Output Processing
This chapter provides details on the input/output ( I/O ) support provided for OpenVMS systems and discusses the following topics: Environment I/O support (Section 7.1) User-action functions (Section 7.2) File sharing (Section 7.3) Record locking (Section 7.4)
7.1 Environment I/O Support

Compaq Pascal uses the Record Management Services (RMS) to perform I/O tasks at the system level. In this environment, all of the Compaq Pascal I/O model is supported; the model is based on RMS concepts. If these sections contain no information on a concept or element in the Compaq Pascal I/O model, then this environment supports the concept or element exactly as it is described in the Compaq Pascal Language Reference Manual. You can use RMS features through Compaq Pascal when you call the OPEN procedure. For instance, when you call this procedure, you can specify the le organization, the component format, and the access method. If you choose to use additional features of RMS that are not available in the Compaq Pascal I/O model, you can write a user-action function that manipulates the RMS control blocks: the le access block (FAB), the record access block (RAB), and the extended attribute block (XAB). Once you write the user-action function, you pass the function name as a parameter to the OPEN procedure.
Input and Output Processing 71
For More Information: On user-action functions (Section 7.2) On OPEN defaults (Section 7.1.6.1) On OPEN and the Compaq Pascal I/O model (Compaq Pascal Language Reference Manual) On RMS concepts (Guide to OpenVMS File Applications) On the user interface to RMS (OpenVMS Record Management Services Reference Manual)
7.1.1 Indexed Files

The Compaq Pascal I/O model allows you to use most of the features of RMS indexed les. However, if you wish to use segmented or null keys, you must write a user-action function. When an existing indexed le is opened, the run-time library compares the keys in the le against the KEY attributes specied in the program. If no KEY attribute was specied for the corresponding key in the indexed le, then the comparison is bypassed and the open continues. The run-time library compares the position and the data type of the les keys against the KEY attributes specied. If the KEY attribute explicitly species a collating sequence (ASCENDING or DESCENDING), then the specied sequence must match that of the key in the le. If no sequence is specied, either sequence is allowed. The CHANGES and DUPLICATES options are not checked. For More Information: On user-action functions (Section 7.2) On the OPEN procedure (Section 7.1.6) On indexed le organization and the KEY attribute (Compaq Pascal Language Reference Manual)
7.1.2 OpenVMS Components and RMS Records

In the Compaq Pascal I/O model, data items in a le are called components. In RMS, these items are called records.
7.1.3 Count Fields for Variable-Length Components

Each variable-length component contains a count eld as a prex. This count eld contains the number of bytes in the rest of the component. For les on tape, this count eld is 4 bytes in length; for les on disk, this count eld is 2 bytes in length.
7.1.4 Variable-Length with Fixed-Length Control Field (VFC) Component Format

The Compaq Pascal I/O model does not provide a direct means to create les of variable-length components with xed-length control elds (VFC). If you open a le of this component format, Compaq Pascal treats the le like a le of variable-length components. If you want to create les of this component format, you must write a user-action function. For More Information: On user-action functions (Section 7.2) On VFC components (Guide to OpenVMS File Applications)
7.1.5 Random Access by Record File Address (RFA)

The Compaq Pascal I/O model does not allow random access by record le address. If you want to use this type of access, you must write a user-action function. RMS supports random access by Record File Address (RFA) for relative and indexed les, and for sequential les only on disk. The RFA is a unique number supplied for les on disk. The RFA remains constant as long as the record is in the le. RMS makes the RFA available to your program every time the record is stored or retrieved. Your program can either ignore the RFA or it can keep it as a random-access pointer to the record for subsequent accesses. If your disk le is sequential with variable-length records, the RFA provides the only method for randomly accessing records of that le. For More Information: On user-action functions (Section 7.2) On RFA (Guide to OpenVMS File Applications)
7.1.6 OPEN Procedure

When you use the OPEN procedure, RMS applies default values for VMS le specications, and assigns values to FAB, RAB, XAB, and Name Block data structures.
7.1.6.1 OPEN Defaults When you use OPEN to open a le, RMS applies certain defaults when attempting to locate the physical le. Table 71 presents these defaults. Table 71 Default Values for OpenVMS File Specications
Element Node Device Directory File name File type Version number (history) Default Local computer Current user device Current user directory Compaq Pascal le variable name or its logical name translation .DAT OLD: highest current number NEW: highest current number + 1
The OPEN procedure includes a default le-name parameter. Using this parameter, you can access the RMS default le-name parameter to set lespecication defaults. Consider the following example: VAR My_File : VARYING [20] OF CHAR; My_File_Var : TEXT; BEGIN My_File := foo.bar; OPEN( FILE_NAME := My_File, FILE_VARIABLE := My_File_Var, DEFAULT := [another.dir] ); The OPEN statement in the previous example opens the le called [ANOTHER.DIR]FOO.BAR. RMS applies the defaults in Table 71 to determine the node, the device, and the version number of the le. For More Information: On le specications (OpenVMS Users Manual) On the OPEN procedure (Compaq Pascal Language Reference Manual)
7.1.6.2 OPEN and RMS Data Structures Table 72 presents the status of RMS FAB elds when you call the OPEN procedure. If a eld is not included in the following tables, it is initialized to zero. Table 72 Setting of RMS File Access Block Fields by a Call to the OPEN Procedure
Field FAB$L_CTX FAB$L_DEV FAB$L_DNA FAB$L_DNS FAB$B_FAC FAB$V_DEL FAB$V_GET FAB$V_PUT FAB$V_TRN FAB$V_UPD FAB$L_FNA FAB$B_FNS FAB$L_FOP FAB$V_CIF FAB$V_DFW FAB$V_DLT FAB$V_NAM FAB$V_SFC FAB$V_SQO Name Context Device characteristics Default le specication string address Default le specication string size File access options Allow deletions Allow reads Allow writes Allow truncations Allow updates File specication string address File specication string size File processing options Create if nonexistent Deferred write Delete on close service Name block inputs Submit command le (when closed) Sequential only 1, if HISTORY := UNKNOWN. 1 Set when le is closed, depends on DISPOSITION. 1, if terminal le reopened to enable prompting. Set when le is closed, depends on DISPOSITION. 1, if ACCESS_METHOD:=SEQUENTIAL (default). (continued on next page) 1, if not HISTORY:=READONLY. 1 1, if not HISTORY:=READONLY. 1, if not HISTORY:=READONLY. 1, if not HISTORY:=READONLY. FILE_NAME if specied, name of le variable if external le, else 0. Set to length of le name string. OPEN Parameters and Value Reserved to Compaq. Returned by RMS. DEFAULT parameter value, if specied; else, .DAT. Set to length of default le name string.
Table 72 (Cont.) Setting of RMS File Access Block Fields by a Call to the OPEN Procedure
Field FAB$V_TEF Name Truncate at end of le OPEN Parameters and Value Initialized to 0, set to 1 after REWRITE or TRUNCATE of a sequential organization le. 1, if nonexternal le with no FILE_NAME specied and DISPOSITION:=DELETE specied or implied. 2, if terminal le enabled for prompting. Returned by RMS. RECORD_LENGTH if specied; le component size if ORGANIZATION is not SEQUENTIAL or if RECORD_ TYPE:=FIXED. Set to address of name block (the expanded and resultant string areas are set up, but the related le name string is not). FAB$C_REL if ORGANIZATION:=RELATIVE; FAB$C_IDX if ORGANIZATION:=INDEXED; FAB$C_SEQ in all other cases. FAB$B_RAT FAB$V_FTN FAB$V_CR FAB$V_PRN
1 After
FAB$V_TMD
Temporary (marked for deletion) Fixed control area size Internal le identier Maximum record size
FAB$B_FSZ FAB$W_IFI FAB$W_MRS
FAB$L_NAM1
Name block address
FAB$B_ORG
File organization
Record attributes FORTRAN carriage control Add LF and CR Print le format 1, if CARRIAGE_CONTROL:=FORTRAN. 1, if CARRIAGE_CONTROL:=LIST (default for TEXT and VARYING OF CHAR les). 1, if terminal le enabled for prompting.
the call to OPEN, FAB$L_NAM must contain the same value it had before the call.
Table 72 (Cont.) Setting of RMS File Access Block Fields by a Call to the OPEN Procedure
Field FAB$B_RFM Name Record format OPEN Parameters and Value FAB$C_FIX if RECORD_TYPE:=FIXED or if le component is of xed size; FAB$C_VAR if RECORD_TYPE:=VARIABLE or le is VARYING or TEXT; FAB$C_STM if RECORD_TYPE:=STREAM; FAB$C_STMCR if RECORD_TYPE:=STREAM_ CR; FAB$C_STMLF if RECORD_TYPE:=STREAM_ LF; FAB$C_VFC if a terminal le enabled for prompting. Returned by RMS. The XAB chain always has a File Header Characteristics (FHC) extended attribute block in order to get the longest record length (XAB$W_LRL). If ACCESS_ METHOD:=KEYED, key index denition blocks are also present. Compaq may add additional XABs in the future. Your useraction function may insert XABs anywhere in the chain. This eld is only valid during execution of user-action functions; Compaq Pascal places 0 in this eld after the call to OPEN. 1, if SHARING:=READWRITE. 1, if SHARING is not NONE (default if HISTORY:=READONLY). 1, if SHARING:=READWRITE. 1, if SHARING:=READWRITE. 1, if SHARING:=NONE (default if HISTORY is not READONLY).
FAB$L_SDC FAB$L_XAB2
Spooling device characteristics Extended attribute block address
FAB$B_SHR FAB$V_SHRPUT FAB$V_SHRGET FAB$V_SHRDEL FAB$V_SHRUPD FAB$V_NIL

2 You
File sharing Allow other PUTs Allow other GETs Allow other DELETEs Allow other UPDATEs Allow no other operations
cannot change XABs provided by Compaq, but you can add and delete XABs that you insert using a user-action function.
Table 73 presents the status of RMS RAB elds when you call the OPEN procedure. If a eld is not included in the following table, it is initialized to zero.
Table 73 Setting of RMS Record Access Block Fields by a Call to the OPEN Procedure
Field RAB$L_CTX RAB$L_FAB RAB$W_ISI RAB$L_KBF RAB$B_KRF RAB$B_KSZ RAB$B_RAC RAB$L_RBF RAB$L_RHB
1
Name Context FAB address Internal stream identier Key buffer address Key of reference Key size Record access mode Record address Record header buffer
OPEN Parameters and Value Reserved to Compaq. Set to address of FAB (allocated by Compaq Pascal RTL). Returned by RMS. May be modied for individual le operations after the le is opened. May be modied for individual le operations after the le is opened. May be modied for individual le operations after the le is opened. May be modied for individual le operations after the le is opened. May be modied for individual le operations after the le is opened. Set to address of 2-byte carriage-control information for terminal les enabled for prompting. May be modied for individual le operations after the le is opened. 1 May be modied for individual le operations after the le is opened. 1, if ACCESS:=DIRECT. 1 May be modied for individual le operations after the le is opened. Returned by RMS. Set to buffer address after le is opened (Compaq Pascal RTL allocates buffer).
RAB$L_ROP RAB$V_NLK RAB$V_RAH RAB$V_TPT RAB$V_UIF RAB$V_WBH RAB$W_RSZ RAB$L_STS RAB$L_UBF

1
Record options No lock Read ahead Truncate le often PUT Update if record exists Write behind Record size Completion status code User record area address
1 After
the call to OPEN, this eld must contain the same value it had before the call.
Table 73 (Cont.) Setting of RMS Record Access Block Fields by a Call to the OPEN Procedure
Field RAB$W_USZ
1
Name User record area size
OPEN Parameters and Value Set to size of record area; for les other than TEXT, the size is equal to the size of the component type; for TEXT les, the size is equal to the value of RECORD_LENGTH; otherwise, 255.
1 After
the call to OPEN, this eld must contain the same value it had before the call.
Table 74 presents the status of RMS XAB elds when you call the OPEN procedure. If a eld is not included in the following table, it is initialized to zero. Table 74 Setting of Extended Attribute Block Fields by a Call to the OPEN Procedure
Field XAB$B_DTP XAB$B_FLG XAB$V_CHG XAB$V_DUP XAB$W_POS0 XAB$B_REF XAB$B_SIZ0 Name Data type of key Key option ags Changes allowed Duplicates allowed Key position Key of reference Key size 0 if key is 0, else 1 0 if key is 0, else 1 Position of key in indexed le Primary key is 0, rst alternate key is 1, second alternate key is 2, and so on Size of key PASCAL OPEN Keyword and Value Set to data type of key
Table 75 presents the status of RMS Name Block elds when you call the OPEN procedure. If a eld is not included in the following table, it is initialized to zero. Table 75 Setting of Name Block Fields by a Call to the OPEN Procedure
Field NAM$L_ESA
1 These 1
Name Expanded string area
OPEN Keyword and Value Address of RTL buffer
elds are only valid during execution of user-action functions; Compaq Pascal places 0 in these elds after the call to OPEN.
Table 75 (Cont.) Setting of Name Block Fields by a Call to the OPEN Procedure
Field NAM$B_ESS
1
Name Expanded string area Expanded string area Expanded string area
OPEN Keyword and Value NAM$C_MAXRSS Address of RTL buffer NAM$C_MAXRSS
NAM$L_RSA NAM$B_RSS
1 These
elds are only valid during execution of user-action functions; Compaq Pascal places 0 in these elds after the call to OPEN.
For More Information: On opening indexed les (Section 7.1.1)
7.1.7 Default Line Limits

Compaq Pascal determines a default line limit for TEXT les by translating the logical name PAS$LINELIMIT as a string of decimal digits. If this logical name has not been dened, there is no default line limit. You can override the default by calling the LINELIMIT procedure. For More Information: On LINELIMIT (Compaq Pascal Language Reference Manual)
7.2 User-Action Functions

The user-action parameter of the OPEN procedure allows you to access RMS facilities not explicitly available in the Compaq Pascal language by writing a function that controls the opening of the le. Inclusion of the user-action parameter causes the run-time library to call your function to open the le instead of calling RMS to open it according to its normal defaults. The user-action parameter of the CLOSE procedure is similar to that of the OPEN procedure. It allows you to access RMS facilities not directly available in Compaq Pascal by writing a function that controls the closing of the le. Including the user-action parameter causes the run-time library to call your function to close the le instead of calling RMS to close it according to its normal defaults. When an OPEN or CLOSE procedure is executed, the run-time library uses the procedures parameters to establish the RMS le access block (FAB) and the record access block (RAB), as well as to establish its own internal data structures. These blocks are used to transmit requests for le and record operations to RMS; they are also used to return the data contents of les, information about le characteristics, and status codes.
In order, the three parameters passed to a user-action function by the run-time library are as follows: FAB address RAB address File variable
A user-action function is usually written in Compaq Pascal and includes the following: Modications to the FAB or RAB, or both (optional) $OPEN and $CONNECT for existing les or $CREATE and $CONNECT for new les (required) Status check of the values returned by $OPEN or $CREATE and $CONNECT (required) Storage of FAB and RAB values in program variables (optional) Return of success or failure status value for the user-action function (required) Note Modication of any of the RMS le access blocks provided by the run-time library may interfere with the normal operation of the library.
Example 71 shows a Compaq Pascal program that copies one le into another. The program features two user-action functions, which allow the output le to be created with the same size as the input le and to be given contiguous allocation on the storage media.
Example 71 User-Action Function [INHERIT( SYS$LIBRARY:STARLET )] PROGRAM Contiguous_Copy( F_In, F_Out ); { The input file F_In is copied to the output file F_Out. F_Out has the same size as F_In and has contiguous allocation. } TYPE FType = FILE OF VARYING[133] OF CHAR; VAR F_In, F_Out : FType; Alloc_Quantity : UNSIGNED; FUNCTION User_Open( VAR FAB : FAB$TYPE; VAR RAB : RAB$TYPE; VAR F : FType) : INTEGER; VAR Status : INTEGER; BEGIN { Function User_Open } {Open file and remember allocation quantity } Status := $OPEN( FAB ); IF ODD( Status ) THEN Status := $CONNECT( RAB ); Alloc_Quantity := FAB.FAB$L_ALQ; User_Open := Status; END; { Function User_Open } FUNCTION User_Create( VAR FAB : FAB$TYPE; VAR RAB : RAB$TYPE; VAR F : FType ) : INTEGER; VAR Status : INTEGER; BEGIN { Function User_Create } { Set up contiguous allocation } FAB.FAB$L_ALQ := Alloc_Quantity; FAB.FAB$V_CBT := FALSE; FAB.FAB$V_CTG := TRUE; Status := $CREATE( FAB ); IF ODD( Status ) THEN Status := $CONNECT( RAB ); User_Create := Status; END; { Function User_Create }
Example 71 (Cont.) User-Action Function BEGIN { main program } { Open files } OPEN( F_In, HISTORY := READONLY, USER_ACTION := User_Open ); RESET(F_In ); OPEN( F_Out, HISTORY := NEW, USER_ACTION := User_Create ); REWRITE( F_Out ); { Copy F_In to F_Out } WHILE NOT EOF( F_In ) DO BEGIN WRITE( F_Out, F_In_^ ); GET( F_In ); END; { Close files } CLOSE( F_In ); CLOSE( F_Out ); END.
{ main program }
In this example, the record types FAB$TYPE and RAB$TYPE are dened in SYS$LIBRARY:STARLET, which the program inherits. The function User_ Open is called as a result of the OPEN procedure for the input le F_In. The function begins by opening the le with the RMS service $OPEN. If $OPEN succeeds, the value of Status is odd; in that case, $CONNECT is performed. The allocation quantity contained in the FAB.FAB$L_ALQ eld of the FAB is assigned to a variable so that this value can be used in the second user-action function. User_Open is then assigned the value of Status (in this case, TRUE), which is returned to the main program. The function User_Create is called as a result of the OPEN procedure for the output le F_Out. The function assigns the allocation quantity of the input le to the FAB.FAB$L_ALQ eld of the FAB, which contains the allocation size for the output le. The FAB eld FAB.FAB$V_CBT is set to FALSE to disable the request that le storage be allocated contiguously on a best try basis. Then, the FAB eld FAB.FAB$V_CTG is set to TRUE so that contiguous storage allocation is mandatory. Finally, the RMS service $CREATE is performed. If $CREATE is successful, $CONNECT will be done and the function return value will be that of $CREATE. Once the OPEN procedures have been performed successfully, the program can then accomplish its main task, copying the input le F_In to the output le F_Out, which is the same size as F_In and has contiguous allocation. The last step in the program is to close the les.
For More Information: On the OPEN and CLOSE procedures (Compaq Pascal Language Reference Manual) On RMS le access blocks (Section 7.1.6.2)
7.3 File Sharing

Through the RMS le sharing capability, a le can be accessed by more than one executingprogram at a time or by the same program through more than one le variable. There are two kinds of le sharing: read sharing and write sharing. Read sharing occurs when several programs are reading a le at the same time. Write sharing takes place when at least one program is writing a le and at least one other program is either reading or writing the same le. The extent to which le sharing can take place is determined by the following factors: Device type Sharing is possible only on disk les, since other les must be accessed sequentially. File organization All three le organizations permit read and write sharing on disk les. Explicit user-supplied information Whether or not le sharing actually takes place depends on two items of information that you provide for each program accessing the le. In Compaq Pascal programs, this information is supplied by the values of the SHARING and HISTORY parameters in the OPEN procedure. The HISTORY parameter determines how the program will access the le. HISTORY := NEW, HISTORY := OLD, and HISTORY := UNKNOWN determine that the program will read from and write to the le. HISTORY := READONLY determines that the program will only read from the le. If you try to open an existing le with HISTORY := OLD or HISTORY := UNKNOWN, the run-time library retries the OPEN procedure with HISTORY := READONLY if the initial OPEN fails with a privilege violation. The SHARING parameter determines what other programs are allowed to do with the le. Read sharing can occur when SHARING := READONLY is specied by all programs that access the le. Write sharing is accomplished when all programs specify SHARING := READWRITE. To prevent sharing, specify SHARING:= NONE with the rst program to access the le.
Programs that specify SHARING := READONLY or SHARING := READWRITE can access a le simultaneously; however, le sharing can fail under certain circumstances. For example, a program without either of these parameters will fail when it attempts to open a le currently being accessed by some other program. Or, a program that species SHARING := READONLY or SHARING := READWRITE can fail to open a le because a second program with a different specication is currently accessing that le. When two or more programs are write sharing a le, each program should include a condition handler. This error-processing mechanism prevents program failure due to a record-locking error. For More Information: On record-locking errors (Section 7.4) On condition handling (Chapter 8) On the OPEN procedure (Compaq Pascal Language Reference Manual)
7.4 Record Locking

The RMS record locking facility, along with the logic of the program, prevents two processes from accessing the same component simultaneously. It ensures that a program can add, delete, or update a component without having to do a synchronization check to determine whether that component is currently being accessed by another process. When a program opens a relative or indexed le and species SHARING := READWRITE, RMS locks each component as it is accessed. When a component is locked, any program that attempts to access it fails and a record-locked error results. A subsequent I/O operation on the le variable unlocks the previously accessed component. Thus, at most one component is locked for each le variable. If you use the READ procedure, Compaq Pascal will implicitly unlock the component by executing the UNLOCK procedure during the execution of the READ procedure. A Compaq Pascal program can explicitly unlock a component by executing the UNLOCK procedure. To minimize the time during which a component is locked against access by other programs, the UNLOCK procedure should be used in programs that retrieve components from a shared le but that do not attempt to update them. Compaq Pascal requires that a component be locked before a DELETE or an UPDATE procedure can be executed.
For More Information: On the OPEN, UNLOCK, DELETE, and UPDATE procedures (Compaq Pascal Language Reference Manual)
8
Error Processing and Condition Handling
An exception condition is an event, usually an error, that occurs during program execution and is detected by system hardware or software or the logic in a user application program. A condition handler is a routine that is used to resolve exception conditions. By default, the Condition Handling Facility (CHF) provides condition handling sufcient for most Compaq Pascal programs. The CHF also processes user-written condition handlers. This chapter discusses the following topics: Condition handling terms (Section 8.1) Overview of condition handling (Section 8.2) Writing condition handlers (Section 8.3) Fault and trap handling (Section 8.4) Examples of condition handlers (Section 8.5)
The use of condition handlers requires considerable programming experience. You should understand the discussions of condition handling in the following volumes before attempting to write your own condition handler: OpenVMS Programming Concepts Manual OpenVMS System Services Reference Manual OpenVMS Calling Standard
Error Processing and Condition Handling 81
8.1 Condition Handling Terms

The following terms are used in the discussion of condition handling: Condition valueAn integer value that identies a specic condition. Stack frameA standard data structure built on the stack during a routine call, starting from the location addressed by the frame pointer (FP) and proceeding to both higher and lower addresses; it is popped off the stack during the return from a routine. Routine activationThe environment in which a routine executes. This environment includes a unique stack frame on the run-time stack; the stack frame contains the address of a condition handler for the routine activation. A new routine activation is created every time a routine is called and is deleted when control passes from the routine. EstablishThe process of placing the address of a condition handler in the stack frame of the current routine activation. A condition handler established for a routine activation is automatically called when a condition occurs. In Compaq Pascal, condition handlers are established by means of the predeclared procedure ESTABLISH. A routine that establishes a condition handler is known as an establisher. Program exit statusThe status of the program at its completion. SignalThe means by which the occurrence of an exception condition is made known. Signals are generated by the operating system in response to I/O events and hardware errors, by the system-supplied library routines, and by user routines. All signals are initiated by a call to the signaling facility, for which there are two entry points: LIB$SIGNALUsed to signal a condition and, possibly, to continue program execution LIB$STOPUsed to signal a severe error and discontinue program execution, unless a condition handler performs an unwind operation ResignalThe means by which a condition handler indicates that the signaling facility is to continue searching for a condition handler to process a previously signaled error. To resignal, a condition handler returns the value SS$_RESIGNAL. UnwindThe return of control to a particular routine activation, bypassing any intermediate routine activations. For example, if X calls Y, and Y calls Z, and Z detects an error, then a condition handler associated with X or Y can unwind to X, bypassing Y. Control returns to X immediately following the point at which X called Y.
8.2 Overview of Condition Handling

When the OpenVMS system creates a user process, a system-dened condition handler is established in the absence of any user-written condition handler. The system-dened handler processes errors that occur during execution of the user image. Thus, by default, a run-time error causes the system-dened condition handler to print error messages and to terminate or continue execution of the image, depending on the severity of the error. When a condition is signaled, the system searches for condition handlers to process the condition. The system conducts the search for condition handlers by proceeding down the stack, frame by frame, until a condition handler is found that does not resignal. The default handler calls the systems message output routine to send the appropriate message to the user. Messages are sent to the SYS$OUTPUT and SYS$ERROR les. If the condition is not a severe error, program execution continues. If the condition is a severe error, the default handler forces program termination, and the condition value becomes the program exit status. You can create and establish your own condition handlers according to the needs of your application. For example, a condition handler could create and display messages that describe specic conditions encountered during the execution of your program, instead of relying on system error messages.
8.2.1 Condition Signals

A condition signal consists of a call to either LIB$SIGNAL or LIB$STOP, the two entry points to the signaling facility. These entry points can be inherited from SYS$LIBRARY:PASCAL$LIB_ROUTINES.PEN. If a condition occurs in a routine that is not prepared to handle it, a signal is issued to notify other active routines. If the current routine can continue after the signal is propagated, you can call LIB$SIGNAL. A higher-level routine can then determine whether program execution should continue. If the nature of the condition does not allow the current routine to continue, you can call LIB$STOP.
8.2.2 Handler Responses

A condition handler responds to an exception condition by taking action in three major areas: Condition correction Condition reporting Execution control
The handler rst determines whether the condition can be corrected. If so, it takes the appropriate action and execution continues. If the handler cannot correct the condition, the condition may be resignaled; that is, the handler requests that another condition handler be sought to process the condition. A handlers condition reporting can involve one or more of the following actions: Maintaining a count of exceptions encountered during program execution Resignaling the same condition to send the appropriate message to the output le Changing the severity eld of the condition value and resignaling the condition Signaling a different condition, for example, the production of a message designed for a specic application
A handler can control execution in several ways: By continuing from the signal. If the signal was issued through a call to LIB$STOP, the program exits. By doing a nonlocal GOTO operation (see Section 8.5, Example 5). By unwinding to the establisher at the point of the call that resulted in the exception. The handler can then determine the function value returned by the called routine. By unwinding to the establishers caller (the routine that called the routine which established the handler). The handler can then determine the function value returned by the called routine.
8.3 Writing Condition Handlers

The following sections describe how to write and establish condition handlers and provide some simple examples.
8.3.1 Establishing and Removing Handlers

To use a condition handler, you must rst declare the handler as a routine in the declaration section of your program; then, within the executable section, you must call the predeclared procedure ESTABLISH. The ESTABLISH procedure sets up a Compaq Pascal language-specic condition handler that in turn allows your handler to be called. User-written condition handlers set up by ESTABLISH must have the ASYNCHRONOUS attribute and two integer array formal parameters. Such routines can access only local, read-only, and volatile variables, and local, predeclared, and asynchronous routines.
Because condition handlers are asynchronous, any attempt to access a nonreadonly or nonvolatile variable declared in an enclosing block will result in a warning message. The predeclared le variables INPUT and OUTPUT are such nonvolatile variables; therefore, simultaneous access to these les from both an ordinary program and from an asynchronous condition handlers activation may have undened results. The following steps outline the recommended method for performing I/O operations from a condition handler: 1. Declare a le with the VOLATILE attribute at program level. 2. Open this le to refer to SYS$INPUT, SYS$OUTPUT, or another appropriate le. 3. Use this le in the condition handler. External routines (including system services) that are called by a condition handler require the ASYNCHRONOUS attribute in their declaration. You should set up a user-written condition handler with the predeclared procedure ESTABLISH rather than with the run-time library routine LIB$ESTABLISH. ESTABLISH follows the Compaq Pascal procedure-calling rules and is able to handle Compaq Pascal condition handlers more efciently than LIB$ESTABLISH. A condition handler set up by LIB$ESTABLISH might interfere with the default error handling of the Compaq Pascal run-time system, and cause unpredictable results. The following example shows how to establish a condition handler using the Compaq Pascal procedure ESTABLISH: [EXTERNAL,ASYNCHRONOUS] FUNCTION Handler (VAR Sigargs : Sigarr; VAR Mechargs : Mecharr) : INTEGER; EXTERN; . . . ESTABLISH (Handler); To establish the handler, call the ESTABLISH procedure. To remove an established handler, call the predeclared procedure REVERT, as follows: REVERT; As a result of this call, the condition handler established in the current stack frame is removed. When control passes from a routine, any condition handler established during the routines activation is automatically removed.
8.3.2 Declaring Parameters for Condition Handlers

A condition handler is an integer-valued function that is called when a condition is signaled. Two formal VAR parameters must be declared for a condition handler: An integer array to refer to the parameter list from the call to the signal routine (the signal array); that is, the list of parameters included in calls to LIB$SIGNAL or LIB$STOP (see Section 8.2.1) An integer array to refer to information concerning the routine activation that established the condition handler (the mechanism array). The size and contents of the mechanism array is different on OpenVMS Alpha and OpenVMS VAX systems.
For example, a condition handler can be dened as follows: TYPE Sigarr = ARRAY[0..9] OF INTEGER; Mecharr = ARRAY[0..(SIZE(CHF2$TYPE)-4) DIV 4] OF INTEGER; [EXTERNAL,ASYNCHRONOUS] FUNCTION Handler (VAR Sigargs : Sigarr; VAR Mechargs : Mecharr) : INTEGER; EXTERN; . . . ESTABLISH (Handler); . . . The signal procedure passes the following values to the array Sigargs:
Value Sigargs[0] Sigargs[1] Sigargs[2 to n] Description The number of parameters being passed in this array (parameter count). The primary condition being signaled (condition value). See Section 8.3.4 for a discussion of condition values. The optional parameters supplied in the call to LIB$SIGNAL or LIB$STOP; note that the index range of Sigargs should include as many entries as are needed to refer to the optional parameters.
The routine that established the condition handler passes the following values, which contain information about the establishers routine activation, to the array Mechargs:
OpenVMS VAX Value Mechargs[0] Mechargs[1] Mechargs[2]
OpenVMS Alpha Value Mechargs[0] Mechargs[2] Mechargs[4]
Description The number of parameters being passed in this array. The address of the stack frame that established the handler. The number of calls that have been made (that is, the stack frame depth) from the routine activation up to the point at which the condition was signaled. The value of register R0 at the time of the signal. The value of register R1 at the time of the signal.
Mechargs[3] Mechargs[4]
Mechargs[11] Mechargs[13]
For a complete description of the mechanism array, see OpenVMS Calling Standard.
8.3.3 Handler Function Return Values

Condition handlers are functions that return values to control subsequent execution. These values and their effects are listed as follows:
Value SS$_CONTINUE Effect Continues execution from the signal. If the signal was issued by a call to LIB$STOP, the program does not continue, but exits. Resignals to continue the search for a condition handler to process the condition.
SS$_RESIGNAL
In addition, a condition handler can request a stack unwind by calling the $UNWIND system service routine. You can inherit $UNWIND from SYS$LIBRARY:STARLET.PEN. When $UNWIND is called, the function return value of the condition handler is ignored. The handler modies the saved registers R0 and R1 in the mechanism parameters to specify the called functions return value. A stack unwind is usually made to one of two places: The point in the establisher at which the call was made that resulted in the exception. Specify the following: Status := $UNWIND (Mechargs[2],0); { OpenVMS VAX } Status := $UNWIND (Mechargs[4],0); { OpenVMS Alpha }
The routine that called the establisher. Specify the following: Status := $UNWIND (Mechargs[2]+1,0); { OpenVMS VAX } Status := $UNWIND (Mechargs[4]+1,0); { OpenVMS Alpha }
8.3.4 Condition Values and Symbols

The OpenVMS system uses condition values to indicate that a called routine has either executed successfully or failed, and to report exception conditions. Condition values are usually symbolic names that represent 32-bit packed records, consisting of elds (usually interpreted as integers) that indicate which system component generated the value, the reason the value was generated, and the severity of the condition. A warning severity code ( 0 ) indicates that although output was produced, the results may be unpredictable. An error severity code ( 2 ) indicates that output was produced even though an error was detected. Execution can continue, but the results may not be correct. A severe error code ( 4 ) indicates that the error was of such severity that no output was produced. A condition handler can alter the severity code of a condition value to allow execution to continue or to force an exit, depending on the circumstances. Occasionally a condition handler may require a particular condition to be identied by an exact match; that is, each bit of the condition value bits (0..31) must match the specied condition. For example, you may want to process a oating overow condition only if the severity code is still 4 (that is, if no previous condition handler has changed the severity code) and the control bits have not been modied. A typical condition handler response is to change the severity code and resignal. In most cases, however, you want some response to a condition, regardless of the value of the severity code or control bits. To ignore the severity and control elds of a condition value, declare and call the LIB$MATCH_COND function.
8.3.5 Using Condition Handlers that Return SS$_CONTINUE

Compaq Pascal condition handlers can do one of the following after appropriately responding to the error: Use a nonlocal GOTO to transfer control to a label in an enclosing block Return SS$_CONTINUE if the handler is conditioned to dismiss the error then signal to continue processing Return SS$_RESIGNAL if the handler is conditioned to continue searching for additional handlers to call
Call the $UNWIND system service to establish a new point to resume execution when the handler returns to the system
When an exception occurs, the system calls a handler in the Pascal RunTime Library that is established by the generated code. This handler in the RTL in turn calls the user condition handler that was established with the ESTABLISH built-in routine. The RTL handler contains a check to prevent a user handler from returning SS$_CONTINUE for a certain class of Pascal Run-Time Errors that could cause an innite loop if execution was to continue at the point of the error. There are two situations in which this check may cause unexpected behavior: 1. The user handler called $UNWIND and then returned with SS$_ CONTINUE. Because the $UNWIND service was called, execution will not resume at the point of the error even if SS$_CONTINUE is returned to the system. However, the RTL handler is not aware that $UNWIND has been called, and will report that program operation cannot continue for this type of error. The solution is to return SS$_RESIGNAL instead of SS$_CONTINUE after calling $UNWIND in the user handler. However, this solution is not possible if you establish the LIB$SIG_TO_ RET routine with the ESTABLISH built-in routine. LIB$SIG_TO_RET is a routine that can be used as a condition handler to convert a signal into a return to the caller of the routine that established LIB$SIG_TO_RET. Because LIB$SIG_TO_RET returns SS$_NORMAL, which in turn is the same value as SS$_CONTINUE, the handler in the Pascal RTL will report that program operation cannot continue for this type of error. The solution for this case is to establish your own handler with the ESTABLISH built-in routine that calls LIB$SIG_TO_RET and then returns SS$_RESIGNAL. You cannot establish LIB$SIG_TO_RET directly as a handler with the ESTABLISH built-in routine. 2. If you are moving code from OpenVMS VAX to OpenVMS Alpha. On OpenVMS VAX, only certain run-time errors are not allowed to return SS$_CONTINUE from a handler. These errors are those associated with the SUBSTR and PAD built-in routines, and those associated with checking code for set constructors. On OpenVMS Alpha, many more run-time errors are not allowed to return SS$_CONTINUE from a handler. It is recommended that you do not return SS$_CONTINUE for any Pascal run-time error that is not associated with a le operation.
For More Information: On the format of a condition value (OpenVMS Calling Standard) On calling the LIB$MATCH_COND function (Section 8.5)
8.4 Fault and Trap Handling
If a VAX processor detects an error while executing a machine instruction, it can take one of two actions. The rst action, called a fault, preserves the contents of registers and memory in a consistent state so that the instruction can be restarted. The second action, called a trap, completes the instruction, but with a predened result. For example, if an integer overow trap occurs, the result is the correct low-order part of the true value. The action taken when an exception occurs depends on the type of exception. For example, faults occur for access violations and for detection of a oating reserved operand. Traps occur for integer overow and for integer divideby-zero exceptions. However, when a oating overow, oating underow, or oating divide-by-zero exception occurs, the action taken depends on the type of VAX processor executing the instruction. The original VAX11/780 processor traps when these errors occur and stores a oating reserved operand in the destination. All other VAX processors fault on these exceptions, which allows the error to be corrected and the instruction restarted. If your program is written to handle oating traps but runs on a VAX processor that generates faults, execution may continue incorrectly. For example, if a condition handler causes execution to continue after a oating trap, a reserved operand is stored and the next instruction is executed. However, the same handler used on a processor that generates faults causes an innite loop of faults because it restarts the erroneous instruction. Therefore, you should write oating-point exception handlers that take the appropriate actions for both faults and traps. Separate sets of condition values are signaled by the processor for faults and traps. Exceptions and their condition code names are as follows:
Exception Floating overow Floating underow Floating divide-by-zero Fault SS$_FLTOVF_F SS$_FLTUND_F SS$_FLTDIV_F Trap SS$_FLTOVF SS$_FLTUND SS$_FLTDIV
To convert faults to traps, you can use the run-time library LIB$SIM_TRAP procedure either as a condition handler or as a called routine from a userwritten handler. When LIB$SIM_TRAP recognizes a oating fault, it simulates the instruction completion as if a oating trap had occurred.
8.5 Examples of Condition Handlers

The examples in this section inherit the $UNWIND system service routine from SYS$LIBRARY:STARLET.PEN. They also assume the following declaration has been made: [INHERIT( SYS$LIBRARY:STARLET, SYS$LIBRARY:PASCAL$LIB_ROUTINES )] PROGRAM Error_Handling( INPUT, OUTPUT); TYPE Sig_Args = ARRAY[0..100] OF INTEGER; { Signal parameters } Mech_Args = ARRAY[0..(SIZE(CHF2$TYPE)-4)DIV 4] OF [UNSAFE] INTEGER; { Mechanism parameters } Example 1 [ASYNCHRONOUS] FUNCTION Handler_0 (VAR SA : Sig_Args; VAR MA : Mech_Args) : [UNSAFE] INTEGER; BEGIN IF LIB$MATCH_COND (SA[1], condition-name ,...) <> 0 THEN BEGIN . . { do something appropriate } . Handler_0 := SS$_CONTINUE; { condition handled, propagate no further } END ELSE Handler_0 := SS$_RESIGNAL; { propagate condition status to other handlers } END; This example shows a simple condition handler. The handler identies the condition being signaled as one that it is prepared to handle and then takes appropriate action. Note that for all unidentied condition statuses, the handler resignals. A handler must always follow this behavior.
Example 2 [ASYNCHRONOUS] FUNCTION Handler_1 (VAR SA : Sig_Args; VAR MA : Mech_Args) : [UNSAFE] INTEGER; BEGIN IF SA[1] = SS$_UNWIND THEN BEGIN . . . END; Handler_1 := SS$_RESIGNAL; END;
{ cleanup }
When writing a handler, remember that it can be activated with a condition of SS$_UNWIND, signifying that the establishers stack frame is about to be unwound. If the establisher has special cleanup to perform, such as freeing dynamic memory, closing les, or releasing locks, the handler should check for the SS$_UNWIND condition status. If there is no cleanup, the required action of resignaling all unidentied conditions results in the correct behavior. On return from a handler activated with SS$_UNWIND, the stack frame of the routine that established the handler is deleted (unwound). Example 3 [ASYNCHRONOUS] FUNCTION Handler_2 (VAR SA : Sig_Args; VAR MA : Mech_Args) : [UNSAFE] INTEGER; BEGIN IF LIB$MATCH_COND (SA[1], THEN BEGIN . . . MA[3] := expression; $UNWIND; END; Handler_2 := SS$_RESIGNAL; END; A handler can perform a default unwind to force return to the caller of its establisher. If the establisher is a function whose result is expected in R0 or R0 and R1, the handler must establish the return value by modifying the appropriate positions of the mechanism array (the locations of the return R0 and R1 values). If the establisher is a function whose result is returned by
condition-name
,...) <> 0
{ cleanup } { establish function result seen by caller (MA[11] on OpenVMS Alpha)} { unwind to caller of establisher }
the extra-parameter method, the handler must establish the condition value by assignment to the function identier. In this case, you must observe two additional restrictions: The handler must be nested within the function The function result must be declared VOLATILE
Example 4 [ASYNCHRONOUS] FUNCTION Handler_3 (VAR SA : Sig_Args; VAR MA : Mech_Args) : [UNSAFE] INTEGER; BEGIN IF LIB$MATCH_COND (SA[1], condition-name ,...) <> 0 THEN BEGIN . . { cleanup } . MA[3] := expression; { establish function result seen by caller } $UNWIND (MA[2]); { unwind to establisher[4] for OpenVMS Alpha } END; Handler_3 := SS$_RESIGNAL; END; A handler can also force return to its establisher immediately following the point of call. In this case, you should make sure that the handler understands whether the currently uncompleted call was a function call (in which case a returned value is expected) or a procedure call. If the uncompleted call is a function call that will return a value in R0 or R0 and R1, then the handler can modify the mechanism array to supply a value. If, however, the uncompleted call is a function call that will return a value using the extra-parameter mechanism, then there is no way for the handler to supply a value. Example 5 [ASYNCHRONOUS] FUNCTION Handler_4 (VAR SA : Sig_Args; VAR MA : Mech_Args) : [UNSAFE] INTEGER; BEGIN IF LIB$MATCH_COND (SA[1], condition-name THEN GOTO 99; Handler_4 := SS$_RESIGNAL; END; ,...) <> 0
A handler can force control to resume at an arbitrary label in its scope. Note that this reference is to a label in an enclosing block, because a GOTO to a local label will remain within the handler. In accordance with the OpenVMS Calling Standard, Compaq Pascal implements references to labels in enclosing blocks by signaling SS$_UNWIND in all stack frames that must be deleted. Example 6 FUNCTION EXP_With_Status (X : REAL; VAR Status : INTEGER ) FUNCTION MTH$EXP (A : REAL) : REAL; EXTERNAL; [ASYNCHRONOUS] FUNCTION Math_Error (VAR SA : Sig_Args; VAR MA : Mech_Args) : [UNSAFE] INTEGER; BEGIN { Math_Error } IF LIB$MATCH_COND (SA[1], MTH$_FLOOVEMAT, MTH$_FLOUNDMAT) <> 0 THEN BEGIN IF ODD( Status ) { record condition status THEN if no previous error } Status := SA[1]::Cond_Status; Math_Error := SS$_CONTINUE; END ELSE Math_Error := SS$_RESIGNAL; END; BEGIN { EXP_With_Status } STATUS := SS$_SUCCESS; ESTABLISH (Math_Error); EXP_With_Status := MTH$EXP (X); END; This example shows a handler that records the condition status if a oating overow or underow error is detected during the execution of the mathematical function MTH$EXP. { condition handled, propagate no further } { propagate condition status to other handlers }
: REAL;
Example 7 [INHERIT(SYS$LIBRARY:STARLET)] PROGRAM Use_A_Handler(INPUT,OUTPUT); TYPE Sigarr = ARRAY [0..9] OF INTEGER; Mecharr = ARRAY [0..(Size(CHF2$TYPE)-4)DIV 4)] OF INTEGER; VAR F1,F2 : REAL; [ASYNCHRONOUS] FUNCTION My_Handler (VAR Sigargs : Sigarr; VAR Mechargs : Mecharr) : INTEGER; VAR Outfile : TEXT; [ASYNCHRONOUS] FUNCTION LIB$FIXUP_FLT (VAR Sigargs : Sigarr; VAR Mechargs : Mecharr; New_Opnd : REAL := %IMMED 0) : INTEGER; EXTERNAL; [ASYNCHRONOUS] FUNCTION LIB$SIM_TRAP (VAR Sigargs : Sigarr; VAR Mechargs : Mecharr) : INTEGER; EXTERNAL; BEGIN OPEN(Outfile,TT:); REWRITE(Outfile); { Handle various conditions } CASE Sigargs[1] OF { Convert floating faults to traps } SS$_FLTDIV_F, SS$_FLTOVF_F : LIB$SIM_TRAP(Sigargs,Mechargs); { Handle the floating divide by zero trap } SS$_FLTDIV : BEGIN WRITELN(Outfile,Floating divide by zero); My_Handler := SS$_CONTINUE; END; { Handle the floating overflow trap } SS$_FLTOVF : BEGIN WRITELN(Outfile,Floating overflow); My_Handler := SS$_CONTINUE; END;
{ Handle taking the square root } MTH$_SQUROONEG : BEGIN WRITELN(Outfile,Square root of a negative number); My_Handler := SS$_CONTINUE; END; { Handle the reserved operand left by SQRT } SS$_ROPRAND : BEGIN WRITELN(Outfile,Reserved floating operand); LIB$FIXUP_FLT(Sigargs,Mechargs); My_Handler := SS$_CONTINUE; END; OTHERWISE BEGIN WRITELN(Outfile,Condition occurred, ,HEX(Sigargs[1])); My_Handler := SS$_RESIGNAL; END; END; CLOSE(Outfile); END; BEGIN ESTABLISH(My_Handler); F1 := 0.0; F2 := 1E38; { Generate exception conditions } F1 := F2 / 0.0; F1 := F2 * f2; F1 := SQRT(-1.0); END. This example demonstrates the use of OpenVMS Alpha condition handlers with Compaq Pascal.
9
Migrating from OpenVMS VAX to OpenVMS Alpha
This chapter provides information on issues that affect programs being moved from OpenVMS VAX systems to OpenVMS Alpha systems.
9.1 Sharing Environment Files Across Platforms

Compaq Pascal inherits environment les created from a compiler for the same target platform. For example, you cannot inherit environment les generated by VAX Pascal with the Compaq Pascal for OpenVMS Alpha compiler.
9.2 Default Size for Enumerated Types and Booleans

The default size for enumerations and Booleans in unpacked structures is longword on OpenVMS Alpha systems. On OpenVMS VAX systems, the default size is byte for Booleans and small enumerations or words for larger enumerations. If you need the OpenVMS VAX behavior on OpenVMS Alpha systems, you can use one of the following: /ENUMERATION_SIZE=BYTE qualier [ENUMERATION_SIZE(BYTE)] attribute Individual [BYTE] or [WORD] attributes on the affected elds or components
The default for OpenVMS VAX compilers is /ENUMERATION_SIZE=BYTE, for compatibility.
Migrating from OpenVMS VAX to OpenVMS Alpha 91
9.3 Default Data Layout for Unpacked Arrays and Records

On OpenVMS Alpha systems, the default data layout is "natural" alignment, where record elds and array components are aligned on boundaries based on their size (for example, INTEGERs on longword boundaries, INTEGER64s on quadword boundaries) On OpenVMS VAX systems, the default alignment rule is to allocate such elds on the next byte boundary. If you need the OpenVMS VAX behavior on OpenVMS Alpha systems, you can use the /ALIGN=VAX qualier or the [ALIGN(VAX)] attribute.
9.4 Overow Checking

When overow checking is enabled on OpenVMS Alpha, the INT built-in signals a run-time error if its actual parameter cannot be represented as an INTEGER32 value. If you have a large unsigned value that you wish to convert to a negative integer, you must use a typecast to perform the operation.
9.5 Bound Procedure Values

On OpenVMS VAX systems, a Bound Procedure Value is a 2-longword data structure holding the address of the entry point and a frame-pointer to dene the nested environment. Compaq Pascal expects one of these 2-longword structures for PROCEDURE or FUNCTION parameters. A routine not written in Pascal needs different code depending on whether it will receive a Bound Procedure Value or a simple routine address. When passing routines to %IMMED formal routine parameters, Compaq Pascal passes the address of the entry point; otherwise, it passes the address of a Bound Procedure Value. On OpenVMS Alpha systems, a Bound Procedure Value is a special type of procedure descriptor that invokes a hidden jacket routine that in turn initializes the frame-pointer and calls the real routine. Compaq Pascal expects a procedure descriptor for PROCEDURE or FUNCTION parameters. A routine not written in Pascal does not require difference code for Bound Procedure Values. When passing routines to %IMMED formal routine parameters, (or asking for the IADDRESS of a routine) Compaq Pascal passes the address of a procedure descriptor as if the %IMMED was not present. There is no direct way in Compaq Pascal to obtain the actual code address of a routine because it is not generally useful without the associated procedure descriptor.
9.6 Different Descriptor Classes for Conformant Array Parameters

Compaq Pascal uses the "by descriptor" mechanism to pass conformant parameters from one routine to another. For conformant array parameters, Compaq Pascal uses a CLASS_A descriptor on VAX systems and a CLASS_ NCA descriptor on Alpha systems. The CLASS_NCA descriptors generate more efcient code when accessing array components and are able to describe arrays with alignment holes or padding (more common on Alpha systems). If you have a foreign routine that constructs CLASS_A descriptors for Pascal, you need to examine the code to see if changes are necessary: For certain actual parameters, the CLASS_A and CLASS_NCA descriptors are identical except for the DSC$B_CLASS eld (which Compaq Pascal does not examine). For other parameters, you will either have to generate a CLASS_NCA descriptor or you can add an explicit CLASS_A attribute to the formal conformant parameter in the Pascal routine.
9.7 Data Layout and Conversion

On Alpha systems (and to a lesser extent VAX systems), the layout of data can severely impact performance. The Alpha architecture and the current Alpha systems have strong preferences about data alignment and size. The Compaq Pascal compiler has several features to enable you to write Pascal code that will get the best performance on the target system. The remainder of this section describes the different types of record layouts, Compaq Pascal features that support them, how to get the best performance with your data structures, and how to convert existing code for better performance. This section focuses on records, but arrays also have similar properties. In almost all cases, where record elds are discussed, you can substitute array components.
9.7.1 Natural Alignment, VAX Alignment, and Enumeration Sizes

The compiler has the ability to lay out records in two ways: OpenVMS VAX alignment Fields and components less than or equal to 32 bits are allocated on the next available bit; otherwise they are allocated on the next available byte. Natural alignment where an object is aligned based on its size Essentially elds and components are allocated on the next naturally aligned address for their data type. For example: 8-bit character strings should start on byte boundaries 16-bit integers should start at addresses that are a multiple of 2 bytes (word alignment) 32-bit integers and single-precision real numbers should start at addresses that are a multiple of 4 bytes (longword alignment) 64-bit integers and double-precision real numbers should start at addresses that are a multiple of 8 bytes (quadword alignment)
For aggregates such as arrays and records, the data type to be considered for purposes of alignment is not the aggregate itself, but rather the elements of which the aggregate is composed. Varying 8-bit character strings must, for example, start at addresses that are a multiple of 2 bytes (word alignment) because of the 16-bit count at the beginning of the string. For records, the size is rounded up to a multiple of their natural alignment (a record with natural alignment of longword has a size that is a multiple of longwords, for example). The OpenVMS VAX and naturally aligned record formats are fully documented in the OpenVMS Calling Standard. The size as well as the alignment of record elds and array components can affect performance. For example, accessing a byte or word on an Alpha system requires more instructions than accessing a naturally aligned longword. On Alpha systems, Compaq Pascal uses larger allocation for unpacked Booleans and enumeration types to help performance, as shown in Table 91
Table 91 Unpacked Sizes of Fields and Components

Datatype Boolean Enumerated types Unpacked Size on VAX 1 byte 1 or 2 bytes Unpacked Size on Alpha 4 bytes 4 bytes
For compatibility reasons, the size of all data types in PACKED records and arrays are the same for both VAX and natural alignment formats.
9.7.2 Compaq Pascal Features Affecting Data Alignment and Size

The Compaq Pascal for OpenVMS compiler has the following DCL qualiers: /ALIGN=keyword, where keyword is either ALPHA_AXP or VAX /ENUMERATION_SIZE=keyword, where keyword is either BYTE or LONG
The /ALIGN qualier option controls the default record format used by the compiler. The /ENUMERATION_SIZE qualier option controls whether the compiler allocates Boolean and enumeration types as longwords or as 1 or 2 bytes. On VAX systems, the default alignment format is "VAX" and the default enumeration size is "BYTE". On Alpha systems, the default alignment format is "ALPHA_AXP" and the default enumeration size is "LONG". A corresponding pair of attributes can be used at the PROGRAM/MODULE level and on VAR and TYPE sections to specify the desired alignment format and enumeration size: ALIGN(keyword), where keyword is either ALPHA_AXP or VAX ENUMERATION_SIZE(keyword), where keyword is either BYTE or LONG
By using these attributes at the MODULE level, you can extract the records into a separate module and create an environment le with the desired alignment format. By using these attributes on VAR or TYPE sections, you can isolate the records in the same source le.
9.7.3 Optimal Record Layout

The optimal record layout is one where all the records elds are naturally sized on naturally aligned boundaries and the overall record is as small as possible (for example, the fewest number of padding bytes required for proper alignment).
On Alpha systems, the compiler automatically places all elds of unpacked records on naturally aligned boundaries. On VAX systems, you have to explicitly ask for natural alignment by using either a DCL qualier or the corresponding attribute. To allow the compiler to do this placement, you should refrain from using explicit positioning and alignment attributes on record elds unless required by your application. The keyword PACKED should be avoided in all cases except: PACKED ARRAY OF CHARs require the PACKED keyword to be manipulated as strings. Since chars are each 1 byte, using the PACKED keyword does not hurt their performance in any way. PACKED SETs may perform better than unpacked SETs. For PACKED SETs, the compiler can sometimes allocate fewer bits for the set eld or variable. These smaller sets can often be manipulated directly with longword or quadword instructions, instead of using a generic run-time library routine for larger sets. Inside unpacked records, PACKED SET elds are no slower than unpacked SET elds. The same holds true for variables of PACKED SETs. PACKED SETs of size 32 or 64 bits are the best performing set types; otherwise a multiple of 8 bits improves performance to a lesser degree. You may still need to use PACKED if you rely on the record for compatability with binary data les or when assuming that types like PACKED ARRAY OF BOOLEAN are implemented as bit strings. While the compiler can position record elds at natural boundaries, it cannot minimize the alignment bytes that are required between elds. The calling standard requires the compiler to allocate record elds in the same lexical order that they appear in the source le. For example: type t1 = record f1 : char; f2 : integer; f3 : char; f4 : integer; end; The size of this record is 16 bytes: F1 is a byte eld, followed by 3 padding bytes to position F2 at a longword boundary F2 is 4 bytes
F3 is a single byte, followed by 3 more padding bytes to position F4 at a longword boundary F4 is 4 bytes
The optimal layout would be: type t2 = record f1,f2 : integer; f3,f4 : char; end; The size of this record is only 12 bytes: F1 and F2 are placed on adjacent longword boundaries F3 and F4 can immediately follow, since they can appear on any byte boundary, they in turn are followed by 2 padding bytes to round the size of the record up to a multiple of its natural alignment of longword.
To achieve the fewest alignment bytes, you should place larger elds at the beginning of the record and smaller elds at the end. If you have record elds of schema types that have run-time size, you should place those at the very end of the record, since their offset requires run-time computation. You can get the optimal record layout by: Avoiding the PACKED keyword except for PACKED ARRAY OF CHARs and possibly PACKED SETs Avoiding explicit POS or ALIGNED attributes Placing larger elds before smaller elds Placing xed-size elds before run-time sized elds
9.7.4 Optimal Data Size

Data items that are smaller than 32-bits may impose a performance penalty, due to the additional instructions required to access them. The compiler will attempt to reorder loads and stores that manipulate adjacent items smaller than 32-bits to minimize the number of memory references required. For performance reasons, the compiler on Alpha systems will allocate Boolean and enumerated types as longwords in unpacked records or arrays. On VAX systems, you have to explicitly request this with a DCL qualier or the corresponding attribute.
You should avoid any explicit size attributes on subrange types. While it is true that [BYTE] 0..255 is smaller than 0..255 (which would allocate 4 bytes, since it is a subrange of INTEGER), the additional overhead of accessing the byte-sized subrange might be than the extra 3 bytes of storage. Using the BIT attribute on subranges is even less effective in terms of the extra instructions required to manipulate a 13-bit integer subrange inside a record. Use these attributes only where needed.
9.7.5 Converting Existing Records

When moving code from an OpenVMS VAX system to an Alpha system, you probably want to make sure that you are getting the best performance from your Alpha system. To do that, you must use natural alignment on your record types.
9.7.6 Applications with No External Data Dependencies

If your application has no external data dependencies (such as no stored binary data les, no binary data transmitted to some external device), then the conversion is as simple as: Using the default natural alignment. Using the default enumeration size. Removing any uses of PACKED that are not needed. Removing any explicit positioning or size attributes that are not needed. Optionally reordering elds to place larger elds before smaller elds. This does not make the record faster, but does make it smaller.
Depending on your datatypes, the removal of any PACKED keywords or attributes may make little improvement in performance. For example, a PACKED ARRAY OF REAL is identical in size and performance to an unpacked ARRAY OF REAL. The Compaq Pascal compiler has two features to help you identify poorly aligned records and how often they are used: The /USAGE=PERFORMANCE command line option This option causes the compiler to generate messages for declarations and uses of record elds that are poorly aligned or poorly sized. For example:
program a; type r = packed record f1 : boolean; f2 : integer; end; begin end. In this program the compiler can highlight the following: $ pascal/usage=performance test.pas f1 : boolean; .........^ %PASCAL-I-COMNOTSIZ, Component is not optimally sized at line number 4 in file DISK$:[DIR]TEST.PAS;32 f2 : integer; .........^ %PASCAL-I-COMNOTALN, Component is not optimally aligned at line number 5 in file DISK$:[DIR]TEST.PAS;32 %PASCAL-S-ENDDIAGS, PASCAL completed with 2 diagnostics In this example, the size of the Boolean eld in the PACKED ARRAY is only 1 bit. Single bit elds require additional instructions to process. The integer eld is not aligned on a well-aligned boundary for the target system. The /USAGE=PERFORMANCE qualier gives performance information customized to the target system. For example, on an OpenVMS VAX system, INTEGERs need only be aligned on a byte boundary for "good" performance; on an Alpha system, INTEGERs should be on a longword boundary. The /SHOW=STRUCTURE_LAYOUT command line option. This option causes the compiler to generate a structure layout summary in the listing le. This summary gives size and offset information about variables, types, and elds. It also ags the same information as the /USAGE=PERFORMANCE command line option. For example, compiling the above program with the following command produces the following in the listing le: $ pascal/list/show=structure_layout test.pas Comments Offset Size ----------- ----------- ----------5 Bytes Size 0 Bytes 1 Bit Align 1 Bit 4 Bytes
R {In PROGRAM A} = PACKED RECORD F1 : BOOLEAN F2 : INTEGER END
This output shows the size of the record "R" as well as the sizes and offsets of the records elds. It also highlights any components that were poorly sized or poorly aligned.
9.7.7 Applications with External Data Dependencies

If your application has external data dependencies, the process is more involved, since you have to isolate and understand the dependencies. Possible steps when porting the code include: Using the /ALIGN=VAX qualier Using the /ENUMERATION_SIZE=BYTE qualier Using the /FLOAT=D_FLOAT qualier (if you have any DOUBLE binary data) Leaving the code exactly as is
This should produce the same behavior on the OpenVMS Alpha system as you had on your OpenVMS VAX system (except that using D_Floating data on Alpha systems only give 53 bits of mantissa instead of 56 bits as on VAX systems). You then have to identify which records in your program have external data dependencies. These include binary les (for example, FILE OF xxx), shared memory sections with other programs, and binary information passed to a library routine (such as an OpenVMS item list). You can immediately begin to convert records without external data dependencies into optimal format (for example, remove any unneeded PACKED keywords and attributes as described earlier). You need to classify records with external dependencies further into: Records that cannot be naturally aligned due to a hard dependency that cannot be changed (like a record that maps onto an external piece of hardware, or a record that is passed to some software you cannot change). Records that can be changed after conversion of binary data or cooperating software.
Isolate records that you cannot change into their own environment le by using /ALIGN=VAX, /ENUM=BYTE, and /FLOAT=D_FLOAT. You can also attach the ALIGN and ENUMERATION_SIZE attributes to the TYPE or VAR sections that dene these records. In this case, you need to also change any uses of the DOUBLE datatype to the D_FLOAT datatype, to ensure that the proper oating format is used.
You do not need to isolate the record if it uses the PACKED keyword, since PACKED records are identical regardless of the /ALIGN or /ENUM qualiers. Nevertheless, isolating the records with dependencies might be useful in the future if you eventually intend to change the format. For records that you might change, you need to decide whether it is worthwhile to convert the record and any external binary data. If the record is of lowuse and you have a large quantity of external data, the cost of conversion is probably too high. If a record is of high-use but is mostly aligned, then the conversion also may not be worthwhile. However, a high-use record that is poorly aligned suggests conversion of external data regardless of the amount of effort required. There are two types of poorly aligned records: Records that use the PACKED keyword PACKED records lay out the same with either setting of the /ALIGN or /ENUMERATION_SIZE qualiers. To get natural alignment, you must remove the PACKED keyword. However, the keyword PACKED by itself does not guarantee poor alignment. For example: type t = packed record f1,f2 : integer; end; This record is well aligned with or without the PACKED keyword. It is also well aligned with /ALIGN=ALPHA_AXP and /ALIGN=VAX. You can remove the PACKED keyword for completeness, but nothing else needs to be done. Unpacked records that lay out differently with /ALIGN=ALPHA_AXP and /ALIGN=VAX These records automatically are well-aligned by the compiler when recompiled with /ALIGN=ALPHA_AXP. However, there are some unpacked records are already well-aligned with both alignment formats. For example: type t = record f1,f2 : integer; end; This unpacked record is well aligned with /ALIGN=ALPHA_AXP and /ALIGN=VAX. Nothing else needs to be done to this record. The /USAGE=PERFORMANCE and /SHOW=STRUCTURE_LAYOUT DCL qualiers can be helpful for identifying poorly aligned records.
For PACKED keywords, you can compile with and without the PACKED keyword to see if the elds are positioned at the same offsets or not. You have classied the records with external data dependencies into: Records that are well-aligned with both alignment/enumeration formats Records that are poorly aligned, where conversion is not worthwhile Records that are poorly aligned, where conversion is worthwhile
For the well-aligned records, no additional work is needed now, but be aware that you still have an external data dependency that might cause problems if you add elds to the record in the future. Isolate records that are not being converted into the same environment le or into the TYPE or VAR sections where you placed the records that you could not convert. For records that are worth converting, you need to plan how to convert the external binary data or cooperating software. For cooperating software, you need to ensure that it gets modied so it views the record with the "natural" layout. You can determine the layout by using the /SHOW=STRUCTURE_ LAYOUT command line option described above. For binary data, you need to write a conversion program. Converting existing binary data involves writing a program that reads the existing data into a poorly aligned record, copies the data into a well aligned record, and then writes out the new record. A simple conversion program would look like: program convert_it(oldfile,newfile); [align(vax),enumeration_size(byte)] type oldtype = packed record { Existing record fields } end; type newtype = record { Record fields reorganized for optimal alignment } end; var oldfile = file of oldtype; newfile = file of newtype; oldvar : oldtype; newvar : newtype;
begin reset(oldfile); rewrite(newfile); while not eof(oldfile) do begin read(oldfile,oldvar); { For each field, sub-field, etc. move the data } newvar.field1 := oldvar.field1; newvar.field2 := oldvar.field2; write(newfile,newtype); end; close(oldfile); close(newfile); end. Notice the "type" keyword before the denition of the "newtype" type. Without this keyword, "newtype" would be in the same type denition part as "oldtype" and would be processed with the same ALIGN and ENUMERATION_SIZE settings. If you have embedded DOUBLE data, you must use the D_FLOAT predened type in the "oldtype" denition since the default on OpenVMS Alpha is for G_Floating format. The compiler does not allow a simple assignment of a D_FLOAT value to a G_FLOAT or T_FLOAT variable. You need to use the CNV$CONVERT_FLOAT routine provided with OpenVMS to convert the oating data.
A
Errors Returned by STATUS and STATUSV Functions
This appendix lists the error conditions detected by the STATUS and STATUSV functions, their symbolic names, and the corresponding values. The symbolic names and their values are dened in the le SYS$LIBRARY:PASSTATUS.PAS, which you can include with a %INCLUDE directive in a CONST section of your program. To test for a specic condition, you compare the STATUS or STATUSV return values against the value of a symbolic name. The symbolic names correspond to some of the run-time errors listed in Appendix C; however, not all run-time errors can be detected by STATUS. There is a one-to-one correspondence between the symbolic constants returned by STATUS or STATUSV documented in PASSTATUS.PAS and the OpenVMS condition code values in SYS$LIBRARY:PASDEF.PAS. The following routine shows how to map the return value of STATUS to its corresponding condition code located in PASDEF.PAS: FUNCTION CONVERT_STATUS_TO_CONDITION(STAT:INTEGER):INTEGER; BEGIN CONVERT_STATUS_TO_CONDITION := 16#218644 + STAT * 8; END; Table A1 lists the symbolic names and the values returned by the STATUS and STATUSV functions and explains the error condition that corresponds to each value.
Errors Returned by STATUS and STATUSV Functions A1
Table A1 STATUS and STATUSV Return Values

Name PAS$K_ACCMETINC PAS$K_AMBVALENU PAS$K_CURCOMUND PAS$K_DELNOTALL PAS$K_EOF PAS$K_ERRDURCLO PAS$K_ERRDURDEL PAS$K_ERRDUREXT PAS$K_ERRDURFIN PAS$K_ERRDURGET PAS$K_ERRDUROPE PAS$K_ERRDURPRO PAS$K_ERRDURPUT PAS$K_ERRDURRES PAS$K_ERRDURREW PAS$K_ERRDURTRU PAS$K_ERRDURUNL PAS$K_ERRDURUPD PAS$K_ERRDURWRI PAS$K_EXTNOTALL PAS$K_FAIGETLOC PAS$K_FILALRCLO PAS$K_FILALROPE PAS$K_FILNAMREQ Value 5 30 73 100 1 16 101 127 102 103 2 36 104 105 106 107 108 109 50 128 74 15 1 14 Meaning Specied access method is not compatible with this le. String is an ambiguous value for the enumerated type type. DELETE or UPDATE was attempted while the current component was undened. DELETE is not allowed for a le with sequential organization. File is at end-of-le. Error occurred while the le was being closed. Error occurred during execution of DELETE. Error occurred during execution of EXTEND. Error occurred during execution of FIND or FINDK. Error occurred during execution of GET. Error occurred during execution of OPEN. Error occurred during prompting. Error occurred during execution of PUT. Error occurred during execution of RESET or RESETK. Error occurred during execution of REWRITE. Error occurred during execution of TRUNCATE. Error occurred during execution of UNLOCK. Error occurred during execution of UPDATE. Error occurred during execution of WRITELN. EXTEND is not allowed for a shared le. GET failed to retrieve a locked component. File is already closed. File is already open. File name must be specied in order to save, print, or submit an internal le. (continued on next page)
A2 Errors Returned by STATUS and STATUSV Functions
Table A1 (Cont.) STATUS and STATUSV Return Values

Name PAS$K_FILNOTDIR PAS$K_FILNOTFOU PAS$K_FILNOTGEN PAS$K_FILNOTINS PAS$K_FILNOTKEY PAS$K_FILNOTOPE PAS$K_FILNOTSEQ PAS$K_FILNOTTEX PAS$K_GENNOTALL PAS$K_GETAFTEOF PAS$K_INSNOTALL PAS$K_INSVIRMEM PAS$K_INVARGPAS PAS$K_INVFILSYN PAS$K_INVKEYDEF PAS$K_INVRECLEN PAS$K_INVSYNBIN PAS$K_INVSYNENU PAS$K_INVSYNHEX PAS$K_INVSYNINT PAS$K_INVSYNOCT PAS$K_INVSYNREA PAS$K_INVSYNUNS PAS$K_KEYCHANOT PAS$K_KEYDEFINC Value 110 3 111 112 113 114 115 116 117 118 119 120 121 4 9 12 37 31 38 32 39 33 34 72 10 Meaning File is not open for direct access. File was not found. File is not in generation mode. File is not in inspection mode. File is not open for keyed access. File is not open. File does not have sequential organization. File is not a text le. Generation mode is not allowed for a read-only le. GET attempted after end-of-le has been reached. Inspection mode is not allowed for a write-only le. Insufcient virtual memory. Invalid argument passed to a Compaq Pascal Run-Time Library procedure. Invalid syntax for le name. Key denition is invalid. Record length nnn is invalid. String is invalid syntax for a binary value. String is invalid syntax for a value of an enumerated type. String is invalid syntax for a hexadecimal value. String is invalid syntax for an integer. String is invalid syntax for an octal value. String is invalid syntax for a real number. String is invalid syntax for an unsigned integer. Changing the key eld is not allowed. KEY(nnn) denition is inconsistent with this le. (continued on next page)
Errors Returned by STATUS and STATUSV Functions A3
Table A1 (Cont.) STATUS and STATUSV Return Values

Name PAS$K_KEYDUPNOT PAS$K_KEYNOTDEF PAS$K_KEYVALINC PAS$K_LINTOOLON PAS$K_LINVALEXC PAS$K_NEGWIDDIG PAS$K_NOTVALTYP PAS$K_ORGSPEINC PAS$K_RECLENINC PAS$K_RECTYPINC PAS$K_RESNOTALL PAS$K_REWNOTALL PAS$K_SUCCESS PAS$K_TEXREQSEQ PAS$K_TRUNOTALL PAS$K_UPDNOTALL PAS$K_WRIINVENU Value 71 11 70 52 122 53 35 8 6 7 124 123 0 13 125 126 54 Meaning Duplication of key eld is not allowed. KEY(nnn) is not dened in this le. Key value is incompatible with les key nnn. Line is too long; exceeds record length by nnn characters. LINELIMIT value exceeded. Negative value in width or digits (of a eld width specication) is invalid. String is not a value of type type. Specied organization is inconsistent with this le. Specied record length is inconsistent with this le. Specied record type is inconsistent with this le. RESET is not allowed for an internal le that has not been opened. REWRITE is not allowed for a le opened for sharing. Last le operation completed successfully. Text les must have sequential organization and sequential access. TRUNCATE is not allowed for a le opened for sharing. UPDATE is not allowed for a le that has sequential organization. WRITE operation attempted on an invalid enumerated value
A4 Errors Returned by STATUS and STATUSV Functions
B
Entry Points to Compaq Pascal Utilities
This appendix describes the entry points to utilities in the OpenVMS Run-Time Library that can be called as external routines by a Compaq Pascal program. These utilities allow you to access Compaq Pascal extensions that are not directly provided by the language.
B.1 PAS$FAB( f )
The PAS$FAB function returns a pointer to the RMS le access block (FAB) of le f. After this function has been called, the FAB can be used to get information about the le and to access RMS facilities not explicitly available in the Compaq Pascal language. The component type of le f can be any type; the le must be open. PAS$FAB is an external function that must be explicitly declared by a declaration such as the following: TYPE Unsafe_File = [UNSAFE] FILE OF CHAR; Ptr_to_FAB = ^FAB$TYPE; FUNCTION PAS$FAB (VAR F : Unsafe_File) : Ptr_to_FAB; EXTERN; This declaration allows a le of any type to be used as an actual parameter to PAS$FAB. The type FAB$TYPE is dened in the Compaq Pascal environment le STARLET.PEN, which your program or module can inherit. You should take care that your use of the RMS FAB does not interfere with the normal operations of the OpenVMS Run-Time Library (RTL). Future changes to the RTL may change the way in which the FAB is used, which may in turn require you to change your program. For More Information: On the OpenVMS Run-Time Library (OpenVMS Programming Concepts Manual)
Entry Points to Compaq Pascal Utilities B1
B.2 PAS$RAB( f )
The PAS$RAB function returns a pointer to the RMS record access block (RAB) of le f. After this function has been called, the RAB can be used to get information about the le and to access RMS facilities not explicitly available in the Compaq Pascal language. The component type of le f can be any type; the le must be open. PAS$RAB is an external function that must be explicitly declared by a declaration such as the following: TYPE Unsafe_File = [UNSAFE] FILE OF CHAR; Ptr_to_RAB = ^RAB$TYPE; FUNCTION PAS$RAB (VAR F : Unsafe_File) : Ptr_to_RAB; EXTERN; This declaration allows a le of any type to be used as an actual parameter to PAS$RAB. The type RAB$TYPE is dened in the Compaq Pascal environment le STARLET.PEN, which your program or module can inherit. You should take care that your use of the RMS RAB does not interfere with the normal operations of the OpenVMS Run-Time Library. Future changes to the RTL may change the way in which the RAB is used, which may in turn require you to change your program. For More Information: On the OpenVMS Run-Time Library (OpenVMS Programming Concepts Manual)
B.3 PAS$MARK2( s )
The PAS$MARK2 function returns a pointer to a heap-allocated object of the size specied by s. If this pointer value is then passed to the PAS$RELEASE2 function, all objects allocated with NEW or PAS$MARK2 since the object was allocated are deallocated. PAS$MARK2 and PAS$RELEASE2 are provided only for compatibility with some other implementations of Compaq Pascal. Their use is not recommended in a modular programming environment. While a mark is in effect, any DISPOSE operation will not actually delete the storage, but merely mark the storage for deletion. To free the memory, you must use PAS$RELEASE2.
B2 Entry Points to Compaq Pascal Utilities
PAS$MARK2 is an external function that must be explicitly declared. Because the parameter to PAS$MARK2 is the size of the object (unlike the parameter to the predeclared procedure NEW), the best method for using this function is to declare a separate function name for each object you wish to mark. The following example shows how PAS$MARK2 could be declared and used as a function named Mark_Integer to allocate and mark an integer variable: TYPE Ptr_to_Integer = Înteger; VAR Marked_Integer: Ptr_to_Integer; [EXTERNAL(PAS$MARK2)] FUNCTION Mark_Integer (%IMMED S : Integer := SIZE(Integer)) : Ptr_to_Integer; EXTERN; . . . Marked_Integer := Mark_Integer; The parameter to PAS$MARK2 can be 0, in which case the function value is only a pointer to a marker, and cannot be used to store data.
B.4 PAS$RELEASE2( p )
The PAS$RELEASE2 function deallocates all storage allocated by NEW or PAS$MARK2 since the call to PAS$MARK2 allocates the parameter p. PAS$MARK2 and PAS$RELEASE2 are provided only for compatibility with some other implementations of Compaq Pascal. Their use is not recommended in a modular programming environment. PAS$RELEASE2 disables AST delivery during its execution, so it should not be used in a real-time environment. PAS$RELEASE2 is an external function that must be explicitly declared. An example of its declaration and use is as follows: TYPE Ptr_to_Integer = Înteger; VAR Marked_Integer : Ptr_to_Integer;
Entry Points to Compaq Pascal Utilities B3
[EXTERNAL(PAS$RELEASE2)] PROCEDURE Release (P :[UNSAFE] Ptr_to_Integer); EXTERN; . . . Release (Marked_Integer); In this example, Marked_Integer is assumed to contain the pointer value returned by a previous call to PAS$MARK2. For More Information: On PAS$MARK2 (Section B.3)
B4 Entry Points to Compaq Pascal Utilities
C
Diagnostic Messages
This appendix summarizes the error messages that can be generated by a Compaq Pascal program at compile time and at run time.
C.1 Compiler Diagnostics

The Compaq Pascal compiler reports compile-time diagnostics in the source listing (if one is being generated) and summarizes them on the terminal (in interactive mode) or in the batch log le (in batch mode). Compile-time diagnostics are preceded by the following: %PASCALIWEF-
Table C1 shows the severity level indicated by each letter. Table C1 Compiler Message Warning Levels
Letter I Meaning An informational message that ags extensions to the Pascal standard, identies unused or possibly uninitialized variables, or provides additional information about a more severe error. A warning that ags an error or construct that may cause unexpected results, but that does not prevent the program from linking and executing. An error that prevents generation of machine code; instead, the compiler produces an empty object module indicating that E-level messages were detected in the source program. A fatal error.
If the source program contains either E- or F-level messages, the errors must be corrected before the program can be linked and executed.
Diagnostic Messages C1
All diagnostic messages contain a brief explanation of the event that caused the error. This section lists compile-time diagnostic messages in alphabetical order, including their severity codes and explanatory message text. Where the message text is not self-explanatory, additional explanation follows. Portions of the message text enclosed in quotation marks are items that the compiler substitutes with the name of a data object when it generates the message. 64BITBASTYP, 64-bit pointer base types cannot contain le variables or schema types ERROR: File types and schema types may not be allocated in 64-bit S2 address space, because their implementation currently assumes 32-bit pointers in internal data structures. 64BITNOTALL, 64-bit pointers are not allowed in this context ERROR: File types and schema types may not be allocated in 64-bit S2 address space, because their implementation currently assumes 32-bit pointers in internal data structures. ABSALIGNCON, Absolute address / alignment conict Error: The address specied by the AT attribute does not have the number of low-order bits implied by the specied alignment attribute. ACCMETHCON, Specied ACCESS_METHOD conicts with les record organization Warning: You cannot specify ACCESS_METHOD:=DIRECT for a le that has indexed organization or sequential organization and variable-length records. You cannot specify ACCESS_METHOD:=KEYED for a le with sequential or relative organization. ACTHASNOFRML, Actual parameter has no corresponding formal parameter Error: The number of actual parameters specied in a routine call exceeds the number of formal parameters in the routines declaration, and the last formal parameter does not have the LIST attribute. ACTMULTPL, Actual parameter specied more than once Error: Each formal parameter (except one with the LIST attribute) can have only one corresponding actual parameter.
C2 Diagnostic Messages
ACTPASCNVTMP, Conversion: actual passed is resulting temporary ACTPASRDTMP, Formal requires read access: actual parameter is resulting temporary ACTPASSIZTMP, Size mismatch: actual passed is resulting temporary ACTPASWRTMP, Formal requires write access: actual parameter is resulting temporary Warning: A temporary variable is created if an actual parameter does not have the size, type, and accessibility properties required by the corresponding foreign formal parameter. ACTPRMORD, Actual parameter must be ordinal Error: The actual parameter that species the starting index of an array for the PACK or UNPACK procedure must have an ordinal type. ADDIWRDALIGN, ADD_INTERLOCKED requires variable with at least word alignment ADDIWRDSIZE, ADD_INTERLOCKED requires 16-bit variable Error: These restrictions are imposed by the instruction sequence that is used on the target architecture. ADDRESSVAR, parameter name is a VAR parameter, ADDRESS is illegal Warning: You should not use the ADDRESS function on a nonvolatile variable or component or on a formal VAR parameter. ADISCABSENT, Formal discriminant discriminant name has no corresponding actual discriminant Error: An actual discriminant must be specied for every formal discriminant in a schema type denition. ADISCHASNOFRML, Actual discriminant has no corresponding formal discriminant Error: The number of actual discriminants specied is greater than the number of formal discriminants dened in the schema type denition. AGGNOTALL, Aggregate variable access of this type not allowed, must be indexed Error. ALIATRTYPCON, Alignment attribute / type conict
ALIGNAUTO, Alignment greater than n conicts with automatic allocation Error: The value n has the value 3 on OpenVMS Alpha systems or 2 on OpenVMS VAX systems; OpenVMS Alpha hardware aligns the stack on a quadword boundary and OpenVMS VAX hardware aligns it on a longword boundary. You cannot specify a greater alignment for automatically allocated variables. ALIDOWN, Alignment down-graded from default of ALIGNED(n) Info: The value of n is based on the size of the object that is being downgraded. ALIGNFNCRES, Alignment greater than n not allowed on function result Error: The value n has the value 3 on OpenVMS Alpha systems or 2 on OpenVMS VAX systems. The use of an attribute on a routine conicts with the requirements of the objects type. ALIGNINT, ALIGNED expression must be integer value in range 0..n, defaulting to m Error: The value n has the value of the largest argument to the ALIGNED attribute allowed on the platform. ALIGNVALPRM, Alignment greater than n not allowed on value parameter Error: The value n has the value 3 on OpenVMS Alpha systems or 2 on OpenVMS VAX systems. The use of an attribute on a parameter conicts with the requirements of the objects type. ALLPRMSAM, All parameters to MIN or MAX must have the same type Error. APARMACTDEF, Anonymous parameter parameter number has neither actual nor default Error: If the declaration of a routine failed to specify a name for a formal parameter, a call to the routine will result in this error message. The routine declaration will also cause an error to be reported. ARITHOPNDREQ, Arithmetic operand(s) required Error. ARRCNTPCK, Array cannot be PACKED Error: At least one parameter to the PACK or UNPACK procedure must be unpacked.
ARRHAVSIZ, routine name requires that ARRAY component have compiletime known size Error: You cannot use the PACK and UNPACK procedures to pack or unpack one multidimensional conformant array into another. The component type of the dimension being copied must have a compile-time known size; that is, it must have some type other than a conformant schema. ARRMSTPCK, Array must be PACKED Error: At least one parameter to the PACK or UNPACK procedure must be of type PACKED. ARRNOTSTR, Array type is not a string type Error: You cannot write a value to a text le (using WRITE or WRITELN) or to a VARYING string (using WRITEV) if there is no textual representation for the type. Similarly, you cannot read a value from a text le (using READ or READLN) or from a VARYING string (using READV) if there is no textual representation for the type. The only legal array, therefore, is PACKED ARRAY [1..n] OF CHAR. ASYREQASY, ASYNCHRONOUS calling routine requires that called routine also be ASYNCHRONOUS Warning. ASYREQVOL, ASYNCHRONOUS routine name requires that variable name be VOLATILE Warning: A variable referred to in a nested asynchronous routine must have the VOLATILE attribute. ATINTUNS, AT address must be an integer value Error. ATREXTERN, attribute name attribute allowed only on external routines Error: The LIST and CLASS_S attributes can be specied only with the declarations of external routines. ATTRCONCMDLNE, Attribute contradicts command line qualier Error: The double-precision attribute specied contradicts the /FLOAT, /G_FLOATING, or /NOG_FLOATING qualier specied on the compile command line.
ATTRCONFLICT, Attribute conict: attribute name Information: This message can appear as additional information on other error messages. ATTRONTYP, Descriptor class attribute not allowed on this type Error: The use of the descriptor class attribute on the variable, parameter, or routine conicts with the requirements of the objects type. AUTOGTRMAXINT, Allocation of variable name causes automatic storage to exceed MAXINT bits Error: The Compaq Pascal implementation restricts automatic storage to a size of 2,147,483,647 bits. AUTOMAX, Unable to quadword align automatic variables, using long alignment Information. BADANAORG, Analysis data le le name is not on a random access device Fatal. BADENVORG, Environment le le name is not on a random access device Fatal. BADSETCMP, < and > not permitted in set comparisons Error. BINOCTHEX, Expecting BIN, OCT, or HEX Error: You must supply BIN, OCT, or HEX as a variable modier when reading the variable on a nondecimal basis. BLKNOTFND, routine block routine name declared FORWARD in block name is missing Error. BLKTOODEEP, Routine blocks nested too deeply Error: You cannot nest more than 31 routine blocks.
BNDACTDIFF, Actuals array bounds differ from those of other parameters in same section Error: All actual parameters passed to a formal parameter section whose type is a conformant schema must have identical bounds and be structurally compatible. BNDCNFRUN, Bounds of conformant ARRAY array name not known until run-time Error: You cannot use the UPPER and LOWER functions on a dynamic array parameter in a compile-time constant expression. BNDSUBORD, Bound expressions in a subrange type must be ordinal Error: The expressions that designate the upper and lower limits of a subrange must be of an ordinal type. BOOLOPREQ, BOOLEAN operand(s) required Error: The operation being performed requires operands of type BOOLEAN. Such operations include the AND, OR, and NOT operators and the SET_INTERLOCKED and CLEAR_INTERLOCKED functions. BOOSETREQ, BOOLEAN or SET operand(s) required Error. BYTEALIGN, Type larger than 32 bits can be positioned only on a byte boundary Error: See the Compaq Pascal Language Reference Manual for information on the types that are allocated more than 32 bits. CALLFUNC, Function function name called as procedure, function value discarded Warning. CARCONMNGLS, CARRIAGE_CONTROL parameter is meaningless given les type Warning: The carriage-control parameter is usually meaningful only for les of type TEXT and VARYING OF CHAR. CASLABEXPR, Case label and case selector expressions are not compatible Error: All case labels in a CASE statement must be compatible with the expression specied as the case selector.
CASORDRELPTR, Compile-time cast allowed only between ordinal, real, and pointer types CASSELORD, Case selector expression must be an ordinal type Error. CASSRCSIZ, Source type of a cast must have a size known at compile-time CASTARSIZ, Target type of a cast must have a size known at compile-time Error: A variable being cast by the type cast operator cannot be a conformant array or a conformant VARYING parameter. An expression being cast cannot be a conformant array parameter, a conformant VARYING parameter, or a VARYING OF CHAR expression. The target type of the cast cannot be VARYING OF CHAR. CDDABORT, %DICTIONARY processing of CDD record denition aborted Error: The Compaq Pascal compiler is unable to process the CDD record description. See the accompanying CDD messages for more information. CDDBADDIR, %DICTIONARY directive not allowed in deepest %INCLUDE, ignored Error: A program cannot use the %DICTIONARY directive in the fth nested %INCLUDE level. The compiler ignores all %DICTIONARY directives in the fth nested %INCLUDE level. CDDBADPTR, invalid pointer was specied in CDD record description Warning: The CDD pointer data type refers to a CDD path name that cannot be extracted, and is replaced by ÎNTEGER. CDDBIT, Ignoring bit eld in CDD record description Information: The Compaq Pascal compiler cannot translate a CDD bit data type that is not aligned on a byte boundary and whose size is greater than 32 bits. CDDBLNKZERO, Ignoring blank when zero attribute specied in CDD record description Information: The Compaq Pascal compiler does not support the CDD BLANK WHEN ZERO clause. CDDCOLMAJOR, CDD description species a column-major array Error: The Compaq Pascal compiler supports only row-major arrays. Change the CDD description to specify a row-major array.
CDDDEPITEM, Ignoring depends item attribute specied in CDD record description Information: The Compaq Pascal compiler does not support the CDD DEPENDING ON ITEM attribute. CDDDFLOAT, D_oating CDD datatype was specied when compiling with G_FLOATING Warning: The CDD record description contains a D_oating data type while compiling with G_oating enabled. It is replaced with [BYTE(8)] RECORD END. CDDFLDVAR, CDD record description contains eld(s) after CDD variant clause Error: The CDD record description contains elds after the CDD variant clause. Because Compaq Pascal translates a CDD variant clause into a Pascal variant clause, and a Pascal variant clause must be the last eld in a record type denition, the elds following the CDD variant clause are illegal. CDDGFLOAT, G_oating CDD datatype was specied when compiling with NOG_FLOATING Warning: The CDD record description contains a G_oating data type while compiling with D_oating enabled. It is replaced with [BYTE(8)] RECORD END. CDDILLARR, Aligned array elements can not be represented, replacing with [BIT(n)] RECORD END Information: The Compaq Pascal compiler does not support CDD record descriptions that specify an array whose array elements are aligned on a boundary greater than the size needed to represent the data type. It is replaced with [BIT(n)] RECORD END, where n is the appropriate length in bits. CDDINITVAL, Ignoring specied initial value specied in CDD record description Information: The Compaq Pascal compiler does not support the CDD INITIAL VALUE clause. CDDMINOCC, Ignoring minimum occurs attribute specied in CDD record description Information: The Compaq Pascal compiler does not support the CDD MINIMUM OCCURS attribute.
CDDONLYTYP, %DICTIONARY may only appear in a TYPE denition part Error: The %DICTIONARY directive is allowed only in the TYPE section of a program. CDDRGHTJUST, Ignoring right justied attribute specied in CDD record description Information: The Compaq Pascal compiler does not support the CDD JUSTIFIED RIGHT clause. CDDSCALE, Ignoring scaled attribute specied in CDD record description Information: The Compaq Pascal compiler does not support the CDD scaled data types. CDDSRCTYPE, Ignoring source type attribute specied in CDD record description Information: The Compaq Pascal compiler does not support the CDD source type attribute. CDDTAGDEEP, CDD description nested variants too deep Error: A CDD record description may not include more than 15 levels of CDD variants. The compiler ignores variants beyond the fteenth level. CDDTAGVAR, Ignoring tag variable and any tag values specied in CDD record description Information: The Compaq Pascal compiler does not fully support the CDD VARIANTS OF eld description statement. The specied tag variable and any tag values are ignored. CDDTOODEEP, CDD description nested too deep Error: Attributes for the CDD record description exceed the implementations limit for record complexity. Modify the CDD description to reduce the level of nesting in the record description. CDDTRUNCREF, Reference string which exceeds 255 characters has been truncated Information: The Compaq Pascal compiler does not support reference strings greater than 255 characters.
CDDUNSTYP, Unsupported CDD datatype standard data type name Information: The CDD record description for an item has attempted to use a data type that is not supported by Compaq Pascal. The Compaq Pascal compiler makes the data type accessible by declaring it as [BYTE(n)] RECORD END where n is the appropriate length in bytes. Change the data type to one that is supported by Compaq Pascal or manipulate the contents of the eld by passing it to external routines as variables or by using the Compaq Pascal type casting capabilities to perform an assignment. CLSCNFVAL, CLASS_S is only valid with conformant strings Error: When the CLASS_S attribute is used in the declaration of an internal routine, it can be used only on a conformant PACKED ARRAY OF CHAR. The conformant variable must also be passed by value semantics. CLSNOTALLW, descriptor class name not allowed on a parameter of this type Error: Descriptor class attributes are not allowed on formal parameters dened with either an immediate or a reference passing mechanism. CMTBEFEOF, Comment not terminated before end of input Error. CNFCANTCNF, Component of PACKED conformant parameter cannot be conformant Error. CNFREQNCA, Conformants of this parameter type require CLASS_NCA Error: The conformant parameter cannot be described with the default CLASS_A descriptor. Add the CLASS_NCA attribute to the parameter declaration. CNSTRNOTALL, Nonstandard constructors are not allowed on nonstatic types Error: You can write constructors for nonstatic types using the standard style of constructor. CNSTRONZERO, Record constructors only allow OTHERWISE ZERO Error.
CNTBEARRCMP, Not allowed on an array component CNTBEARRIDX, Not allowed on an array index CNTBECAST, Not allowed on a cast CNTBECNFCMP, Not allowed on a conformant array component CNTBECNFIDX, Not allowed on a conformant array index CNTBECNFVRY, Not allowed on a conformant varying component CNTBECOMP, Not allowed on a compilation unit CNTBECONST, Not allowed on a CONST denition part CNTBEDEFDECL, Not allowed on any denition or declaration part CNTBEDESPARM, Not allowed on a %DESCR foreign mechanism parameter CNTBEEXESEC, Not allowed on an executable section CNTBEFILCMP, Not allowed on a le component CNTBEFORMAL, Not allowed on a formal discriminant CNTBEFUNC, Not allowed on a function result CNTBEIMMPARM, Not allowed on a parameter passed by an immediate passing mechanism CNTBELABEL, Not allowed on a LABEL declaration part CNTBEPCKCNF, Not allowed on a PACKED conformant array component CNTBEPTRBAS, Not allowed on a pointer base CNTBERECFLD, Not allowed on a record eld CNTBEREFPARM, Not allowed on a parameter passed by a reference passing mechanism CNTBERTNDECL, Not allowed on a routine declaration CNTBERTNPARM, Not allowed on a routine parameter CNTBESCHEMA, Not allowed on a nonstatic type CNTBESETRNG, Not allowed on a set range CNTBESTDPARM, Not allowed on a %STDESCR foreign mechanism parameter CNTBETAGFLD, Not allowed on a variant tag eld CNTBETAGTYP, Not allowed on a variant tag type CNTBETO, Not allowed on TO BEGIN/END DO CNTBETYPDEF, Not allowed on a type denition CNTBETYPE, Not allowed on a TYPE denition part CNTBEVALPARM, Not allowed on a value parameter CNTBEVALUE, Not allowed on a VALUE initialization part CNTBEVALVAR, Not allowed on a VALUE variable CNTBEVAR, Not allowed on a VAR declaration part CNTBEVARBLE, Not allowed on a variable
CNTBEVARPARM, Not allowed on a VAR parameter CNTBEVRYCMP, Not allowed on a varying component Information: These messages can appear as additional information on other error messages. COMCONFLICT, COMMON block name conicts with another COMMON or PSECT of same name Error: You can allocate only one variable in a particular common block, and the name of the common block cannot be the same as the names of other common blocks or program sections used by your program. COMNOTALN, Component is not optimally aligned Information: The component indicated is not well aligned and accesses to it will be inefcient. COMNOTSIZ, Component is not optimally sized Information: The component indicated is not well sized and accesses to it will be inefcient. COMNOTALNSIZE, Component is not optimally aligned and sized Information: The component indicated is not naturally aligned and sized, accesses to it will be inefcient. COMNOTPOS, Fixed size eld positioned after a run-time sized eld is not optimal Information: Much better code can be generated for indicated component if it preceedes all of the run-time sized elds. CONTXTIGN, Text following constant denition ignored Warning: When dening constants with the /CONSTANT DCL qualier, any text appearing after a valid constant denition is ignored. CPPFILERR, Unable to open included le Error. CRETIMMOD, Creation time for module module name in environment environment le name differs from creation time in previous environments Warning: Two or more PEN les referred to a module, but the PEN les did not agree on the creation date/time for the module. This can occur if you recompile a module but do not recompile all the modules that inherited its PEN le.
CSTRBADTYP, Constructor: only ARRAY, RECORD, or SET type CSTRCOMISS, Constructor: component(s) missing CSTRNOVRNT, Constructor: no matching variant CSTRREFAARR, Repetition factor allowed only in ARRAY constructors CSTRREFAINT, Repetition factor must be integer CSTRREFALRG, Repetition factor too large CSTRREFANEG, Repetition factor cannot be negative CSTRTOOMANY, Constructor: too many components Error: You can write constructors only for data items of an ARRAY type. You must specify one and only one value in the constructor for each component of the type. In an array constructor, you cannot use a negative integer value as a repetition factor to specify values for consecutive components. CSTRREFAINT, Repetition factor must be an integer Error. CTESTRSIZ, Compile-time strings must be less than 8192 characters Error. CTGARRDESC, Contiguous array descriptor cannot describe size/alignment properties Information: Conformant array parameters, dynamic array parameters, and %DESCR array parameters all use the contiguous array descriptor mechanism in the OpenVMS Calling Standard. Size and alignment attributes are prohibited on such arrays, as these attributes can create noncontiguous allocation. This message can appear as additional information in other error messages. DEBUGOPT, /NOOPTIMIZE is recommended with /DEBUG Information: Unexpected results may be seen when debugging an optimized program. To prevent conicts between optimization and debugging, you should compile your program with /NOOPTIMIZE until it is thoroughly debugged. Then you can recompile the program with optimization enabled to produce more efcient code. DECLORDER, Declarations are out of order Error: The TO BEGIN DO and TO END DO declarations in a module must appear at the end of the module and may not be reordered.
DEFRTNPARM, Default parameter syntax not allowed on routine parameters DEFVARPARM, Default parameter syntax not allowed on VAR parameters Error. DESCOMABORT, Further processing of /DESIGN=COMMENTS has been aborted Error: An error has occurred that prohibits further comment processing. DESCOMERR, An error has occurred while processing design information Error. DESCOMSEVERR, An internal error has occurred while processing /DESIGN=COMMENTS - please submit an SPR Error: A fatal error has occurred during comment processing. Please submit an SPR including sufcient information to reproduce the program, including the version numbers of the Language-Sensitive Editor/Source Code Analyzer and the Compaq Pascal compiler. DESCTYPCON, Descriptor class / type conict Error: The descriptor class for parameter passing conicts with the parameters type. Refer to Section 5.3.3 of the Compaq Pascal User Manual for OpenVMS Systems for legal descriptor class/type combinations. DESIGNTOOOLD, The comment processing routines are too old for the compiler Error: The support routines for the /DESIGN=COMMENT qualier are obsolete. Contact your system manager. DIRCONVISIB, Directive contradicts visibility attribute Error: The EXTERN, EXTERNAL, and FORTRAN directives conict directly with the LOCAL and GLOBAL attributes. DIREXPECT, No matching directive for the %IF directive Error:: A %IF directive must contain a %THEN clause and be terminated by %ENDIF. DIRUNEXP Error:: Conditional compilation directives other than %IF are only valid after the parts of a %IF directive.
DISCLIMIT, Limit of 255 discriminants exceeded Error. DISNOTORD, Discriminant type must be an ordinal type Error: The formal discriminant in a schema type denition must be an ordinal type. DONTPACKVAR, routine name is illegal, variable can never appear in a packed context Error: You cannot call the BITSIZE and BITNEXT functions for conformant parameters. DUPLALIGN, Alignment already specied DUPLALLOC, Allocation already specied DUPLATTR, Attribute already specied DUPLCLASS, Descriptor class already specied DUPLDOUBLE, Double precision already specied Error: Only one member of a particular attribute class can appear in the same attribute list. DUPLFIN, TO END DO already specied DUPLINIT, TO BEGIN DO already specied Error: Only one TO BEGIN DO and one TO END DO section can appear in the same module. DUPLGBLNAM, Duplicated global name Warning: The GLOBAL attribute cannot appear on more than one variable or routine with the same name. DUPLMECH, Passing mechanism already specied DUPLOPT, Optimization already specied DUPLSIZE, Size already specied DUPLVISIB, Visibility already specied Error: Only one member of a particular attribute class can appear in the same attribute list.
DUPTYPALI, Alignment already specied by type identier type name DUPTYPALL, Allocation already specied by type identier type name DUPTYPATR, Attribute already specied by type identier type name DUPTYPDES, Descriptor class already specied by type identier type name DUPTYPSIZ, Size already specied by type identier type name DUPTYPVIS, Visibility already specied by the type identier type name Error: An attribute specied for an object was already specied in the denition of the objects type. ELEOUTRNG, Element out of range Error: A value specied in a set constructor used as a compile-time constant expression does not fall within the subrange dened as the sets base type. EMPTYCASE, Empty case body Error: You failed to specify any case labels and corresponding statements in the body of a CASE statement. ENVERROR, Environment resulted from a compilation with Errors Error: When a program inherits an environment le that compiled with errors, unexpected results may occur during the programs compilation. The environment le inherited by the program compiled with errors. Unexpected results may occur in the program now being compiled. ENVFATAL, Environment resulted from a compilation with Fatal Errors Error: The environment le inherited by the program compiled with fatal errors. Unexpected results may occur in the program now being compiled. ENVOLDVER, Environment was created by a VAX Pascal V2 compiler, please recompile Warning: The environment le inherited by the program was created by a Compaq Pascal Version 2.0 compiler. You should regenerate the environment le with a newer version of the compiler. ENVWARN, Environment resulted from a compilation with Warnings Warning: The environment le inherited by the program compiled with warnings. Unexpected results may occur in the program now being compiled. ENVWRGCMP, Environment identier was compiled by a Compaq Pascal for platform compiler Fatal.
ERREALCNST, Error in real constant: digit expected Error. ERRNONPOS, ERROR parameter can be specied only with nonpositional syntax Error. ERROR, %ERROR Error: This message is generated by the %ERROR directive. ERRORLIMIT, Error Limit = current error limit, source analysis terminated Fatal: The error limit specied for the programs compilation was exceeded; the compiler was unable to continue processing the program. By default, the error limit is set at 30, but you can use the error limit switch at compile time to change it. ESTBASYNCH, ESTABLISH requires that routine name be ASYNCHRONOUS Warning. EXPLCONVREQ, Explicit conversion to lower type required Error: An expression of a higher-ranked type cannot be assigned to a variable of a lower-ranked type; you must rst convert the higher-ranked expression by using DBLE, SNGL, TRUNC, ROUND, UTRUNC, or UROUND, as appropriate. EXPNOTRES, Expression does not contribute to result Information: The optimizer has determined that part of the expression does not affect the result of the expression and it will not evaluate that part of the expression. EXPR2ONVAL, Expression is allowed only on real, integer, or unsigned values Error: The second expression (and preceding colon) are allowed only if the value being written is of a real, integer, or unsigned type. EXPRARITH, Expression must be arithmetic Error: An expression whose type is not arithmetic cannot be assigned to a variable of a real type.
EXPRARRIDX, Expression is incompatible with unpacked arrays index type Error: The index type of the unpacked array is not compatible with the index type of either the PACK or UNPACK procedure it was passed to. EXPRCOMTAG, Expression is not compatible with tag type Error: A case label specied for a NEW, DISPOSE, or SIZE routine must be assignment compatible with the tag type of the variant. EXPRINFUNC, Expression allowed only in FUNCTION Error. EXPRNOTSET, Expression is not a SET type Error: The compiler encountered an expression of some type other than SET in a call to the CARD function. EXTRNALLOC, Allocation attribute conicts with EXTERNAL visibility Error: The storage for an external variable or routine is not allocated by the current compilation; therefore, the specication of an allocation attribute is meaningless. EXTRNAMDIFF, External names are different Information: This message can appear as additional information on other error messages. EXTRNCFLCT, PSECT or FORWARD conicts with EXTERNAL visibility Error: The storage for an external variable or routine is not allocated by the current compilation; therefore, the specication of an allocation attribute is meaningless. FILEVALASS, FILE evaluation / assignment is not allowed Error: You cannot attempt to evaluate a le variable or assign values to it. FILHASSCH, FILE component may not contain nonstatic types or discriminant identiers Error: Compaq Pascal restricts components of les to those with compiletime size. FILOPNDREQ, FILE operand required Error: The EOF, EOLN, and UFB functions require parameters of le types.
FILVARFIL, FILE_VARIABLE parameter must be of a FILE type Error: The le variable parameter to the OPEN and CLOSE procedures must denote a le variable. FLDIVPOS, Field eld name is illegally positioned Error: A POS attribute attempted to position a record eld before the end of the previous eld in the declaration. FLDNOTKNOWN, Unknown record eld Error. FLDONLYTXT, Field width allowed only when writing to a TEXT le Error. FLDRADINT, Field width or radix expression must be of type INTEGER Error: The eld-width or radix expression in a WRITE, WRITELN, or WRITEV routine must be of type INTEGER. FORACTORD, FOR loop control variable must be of an ordinal type FORACTVAR, FOR loop control must be a true variable Error: The control variable of a FOR statement must be a simple variable of an ordinal type and must be declared in a VAR section. For example, it cannot be a eld in a record that was specied by a WITH statement, or a function identier. FLDWDTHINT, Field-width expression must be of type integer Error. FORCTLVAR, variable name is a FOR control variable Warning: The control variable of a FOR statement cannot be assigned a value; used as a parameter to the ADDRESS function; passed as a writable VAR, %REF, %DESCR, or %STDESCR parameter; used as the control variable of a nested FOR statement; or written into by a READ, READLN, or READV procedure. FORINEXPR, Expression is incompatible with FOR loop control variable Error: The type of the initial or nal value specied in a FOR statement is variable.
FRMLPRMDESC, Formal parameters use different descriptor formats FRMLPRMINCMP, Formal routine parameters are not compatible FRMLPRMNAM, Formal parameters have different names FRMLPRMSIZ, Formal parameters have different size attributes FRMLPRMTYP, Formal parameters have different types Information: These messages can appear as additional information on other error messages. FRSTPRMSTR, READV requires rst parameter to be a string expression Error: You must specify at least two parameters for the READV procedurea character-string expression and a variable into which new values will be read. FRSTPRMVARY, WRITEV requires rst parameter to be a variable of type VARYING Error. FTRNOTHER, Feature not supported io this context Error. FTRNOTPOR, Feature not supported on platform(s) Information. FTRNOTSUP, Feature not supported on this platform Error. FUNCTRESTYP, Routine must be declared as FUNCTION to specify a result type Error: You cannot specify a result type on a PROCEDURE declaration. FUNRESTYP, Function result types are different Information: This message can appear as additional information on other error messages.
FWDREPATRLST, Declared FORWARD; repetition of attribute list not allowed FWDREPPRMLST, Declared FORWARD; repetition of formal parameter list not allowed FWDREPRESTYP, Declared FORWARD; repetition of result type not allowed Error: If the heading of a routine has the FORWARD directive, the declaration of the routine body cannot repeat the formal parameter list, the result type (applies only if the routine is a function), or any attribute lists that appeared in the heading. FWDWASFUNC, FORWARD declaration was FUNCTION FWDWASPROC, FORWARD declaration was PROCEDURE Error. GOTONOTALL, GOTO not allowed to jump into a structured statement Warning: Jumping into a structured statement may yield incorrect behavior and/or additional compile-time errors. GOTSZOVFL, GOT table overow for module "name" Error: The GOT (Global Offset Table) for the module is too large. Break up the module into multiple modules. GTR32BITS, routine name cannot accept parameters larger than 32 bits Error: DEC and UDEC cannot translate objects larger than 32 bits into their textual equivalent. HIDATOUTER, HIDDEN legal only on denitions and declarations at outermost level Error: When an environment le is being generated, it is possible to prevent information concerning a declaration from being included in the environment le by using the HIDDEN attribute. However, because an environment le consists only of declarations and denitions at the outermost level of a compilation unit, the HIDDEN attribute is legal only on these denitions and declarations. IDENTGTR31, Identier longer than 31 characters exceeds capacity of compiler Warning. IDNOTLAB, Identier symbol name not declared as a label Error.
IDXNOTCOMPAT, Index type is not compatible with declaration Error: The type of an index expression is not assignment compatible with the index type specied in the arrays type denition. IDXREQDKEY, Creating INDEXED organization requires dense keys Warning: When you specify ORGANIZATION:=INDEXED when opening a le with HISTORY := NEW or UNKNOWN, the les alternate keys must be dense; that is, you may not omit any key numbers in the range from 0 through the highest key number specied for the les component type. IDXREQKEY0, Creating INDEXED organization requires FILE OF RECORD with at least KEY(0) Warning: When you specify ORGANIZATION:=INDEXED when opening a le with HISTORY := NEW or UNKNOWN, the les component type must be a record for which a primary key, designated by the [KEY(0)] attribute, is dened. ILLINISCH, Illegal initialization of variable of nonstatic type Error: Nonstatic variables, such as those created from schema types, cannot be initialized in the VALUE declaration part. To initialize these variables, you must use the initial state feature. IMMEDBNDROU, Immediate passing mechanism may not be used on bound routine routine name Warning: You cannot prex a formal or an actual routine parameter with the immediate passing mechanism unless the routine was declared with the UNBOUND attribute. IMMEDUNBND, Routines passed by immediate passing mechanism must be UNBOUND Warning: A formal routine parameter that has the immediate passing mechanism must also have the UNBOUND attribute. IMMGTR32, Immediate passing mechanism not allowed on values larger than 32 or 64 bits Error: See the Compaq Pascal Language Reference Manual for more information on the types that are allocated more than 32 or 64 bits. IMMHAVSIZ, Type passed by immediate passing mechanism must have compile-time known size Error: You cannot specify an immediate passing mechanism for a conformant parameter or a formal parameter of type VARYING OF CHAR.
INCMPBASE, Incompatible with SET base type Error: If no type identier denotes the base type of a set constructor, the rst element of the constructor determines the base type. The type of all subsequent elements specied in the constructor must be compatible with the type of the rst. INCMPFLDS, Record elds are not the same type Error. INCMPOPND, Incompatible operand(s) Error: The types of one or more operands in an expression are not compatible with the operation being performed. INCMPPARM, Incompatible routine parameter Error: An actual routine parameter is incompatible with the corresponding formal parameter. INCMPTAGTYP, Incompatible variant tag types Error: This message can appear as additional information on other error messages. INCTOODEEP, %INCLUDE directives nested too deeply, ignored Error: A program cannot include more than ve levels of les with the %INCLUDE directive. The compiler ignores %INCLUDE les beyond the fth level. INDNOTORD, Index type must be an ordinal type Error: The index type of an array must be an ordinal type. INFO, %INFO Information: This message is generated by the %INFO directive. INITNOEXT, INITIALIZE routine may not be EXTERNAL INITNOFRML, INITIALIZE routine must have no formal parameter list Error. INITSYNVAR, Illegal initialization syntaxUse VALUE Error.
INPNOTDECL, INPUT not declared in heading Error: A call to EOF, EOLN, READLN, or WRITELN did not specify a le variable, and the default INPUT or OUTPUT was not listed in the program heading. INSTNEWLSE, Please install a new version of LSE Error: The version of the Language-Sensitive Editor/Source Code Analyzer on your system is too old for the compiler. Contact your system manager. INVCASERNG, Invalid range in case label list Error. INVEVAL, Array or Record evaluation not allowed Error. INVQUAVAL, Value for optimizer level is out of range. Default value will be used. Error. IVATTOPT, Unrecognized option for attribute Explanation: Explanation: You attempted to specify an invalid option for one of the following attributes: CHECK (Warning) FLOAT (Warning) KEY (Error) OPTIMIZE (Warning)
IVATTR, Unrecognized attribute Error. IVAUTOMOD, AUTOMATIC variable is illegal at the outermost level of a MODULE Error: You cannot specify the AUTOMATIC attribute for a variable declared at module level. IVCHKOPT, Unrecognized CHECK option Warning.
IVCOMBFLOAT, Illegal combination of D_oating and G_oating Error: You cannot combine D_oating and G_oating numbers in a binary operation. IVDIRECTIVE, Unrecognized directive Error: The directive following a procedure or function heading is not one of those recognized by the Compaq Pascal compiler. IVENVIRON, Environment environment name has illegal format, source analysis terminated Fatal: The environment le inherited by the program has an illegal format; compilation is immediately aborted. However, a listing will still be produced if one was being generated. IVFUNC, Invalid use of function function name IVFUNCALL, Invalid use of function call IVFUNCID, Invalid use of function identier Error: These messages result from illegal attempts to assign values or otherwise refer to the components of the function result (if its type is structured), use the type cast operator on a function identier or its result, or deallocate the storage reserved for the function result (if its type is a pointer). IVKEYOPT, Unrecognized KEY option Error. IVKEYVAL, FINDK KEY_VALUE cannot be an array (other than PACKED ARRAY [1..n] OF CHAR) Error. IVKEYWORD, Missing or unrecognized keyword Error: The compiler failed to nd an identier where it expected one in a call to the OPEN or CLOSE procedure, or it found an identier that was not legal in this position in the parameter list. IVMATCHTYP, Invalid MATCH_TYPE parameter to FINDK Error. IVOPTMOPT, Unrecognized OPTIMIZE option Warning.
IVOTHVRNT, Illegal use of OTHERWISE within CASE variant Error: The Compaq Pascal extension of using OTHERWISE in a record constructor is only dened at the outer level of a record. IVQUALFILE, Illegal switch switch name on le specication Warning: Only the /LIST and /NOLIST qualiers are allowed on the le specication of a %INCLUDE directive. IVQUOCHAR, Illegal nonprinting character (ASCII nnn) within quotes Warning: The only nonprinting characters allowed in a quoted string are the space and tab; the use of other nonprinting characters in a string causes this warning. To include nonprinting characters in a string, you should use the extended string syntax described in the Compaq Pascal Language Reference Manual. IVRADIX, Invalid radix was specied in the extended number Error. IVRADIXDGIT, Illegal digit in binary, octal, or hexadecimal constant Error. IVREDECL, Illegal redeclaration gives symbol name multiple meanings in scope name IVREDECLREC, Illegal redeclaration gives symbol name multiple meanings in this record IVREDEF, Illegal redenition gives symbol name multiple meanings in scope name Warning: When an identier is used in any given block, it must have the same meaning wherever it appears in the block. IVUSEALIGN, Invalid use of alignment attribute IVUSEALLOC, Invalid use of allocation attribute Error. IVUSEATTR, Invalid use of attribute name attribute Error: The use of an attribute on a variable, parameter, or routine conicts with the requirements of the objects type. IVUSEATTRLST, Invalid use of an attribute list Error.
IVUSEBNDID, Illegal use of bound identier identier name Error: An identier that represents one bound of a conformant schema was used where a variable was expected, such as in an assignment statement or in a formal VAR parameter section. The restrictions on the use of a bound identier are identical to those on a constant identier. IVUSEDES, Invalid use of descriptor class attribute Error: The use of an attribute on a variable, parameter, or routine conicts with the requirements of the objects type. IVUSEFNID, Illegal use of function identier identier name Error: Two examples of illegal uses are the assignment of values to the components of the function result (if its type is structured) and the passing of the function identier as a VAR parameter. IVUSEPOI, Illegal use of type POINTER or UNIV_PTR Error: Values of type POINTER and UNIV_PTR can not be dereferenced with the ^ operator or used with the built-in routines NEW and DISPOSE. IVUSESIZ, Invalid use of size attribute Error: The use of an attribute on a variable, parameter, or routine conicts with the requirements of the objects type. IVUSEVISIB, invalid use of visibility attribute Error: The use of a visibility attribute conicts with the requirements of the objects type. KEYINTRNG, KEY number must be an integer value in range 0..254 Error: The key number specied by a KEY attribute must fall in the integer subrange 0..254. KEYNOTALIGN, KEY key number eld eld name at bit position bit position is unaligned KEYORDSTR, KEY allowed only on ordinal and xed-length string elds KEYPCKREC, KEY eld in PACKED RECORD must have an alignment attribute KEYREDECL, Key number key number is multiply dened KEYSIZ1_2_4, Size of an ordinal key must be 1, 2 or 4 bytes KEYSIZ2_4, Size of a signed integer key must be 2 or 4 bytes KEYSIZSTR, Size of a string key cannot exceed 255 bytes KEYUNALIGN, KEY eld cannot be UNALIGNED Error.
LABDECIMAL, Label number must be expressed in decimal radix Error. LABINCTAG, Variant case labels type is incompatible with tag type Error: The type of a constant specied as a case label of a variant record is not assignment compatible with the type of the tag eld. LABNOTFND, No denition of label label name in statement part of block name Error: A label that you declared in a LABEL section does not prex a statement in the executable section. LABREDECL, Redenition of label label name in block name Error: A label cannot prex more than one statement in the same block. LABRNGTAG, Variant case label does not fall within range of tag type Error: A constant specied as a case label of a variant record is not within the range dened for the type of the tag eld. LABTOOBIG, Label label number is greater than MAXINT Error. LABUNDECL, Undeclared label label name Error: Compaq Pascal requires that you declare all labels in a LABEL declaration section before you use them in the executable section. LABUNSATDECL, Unsatised declaration of label label name is not local to block name Error: A label that prexes a statement in a nested block was declared in an enclosing block. LIBESTAB, LIB$ESTABLISH is incompatible with Compaq Pascal; use predeclared procedure ESTABLISH Warning: Compaq Pascal establishes its own condition handler for processing Pascal-specic run-time signals. Calling LIB$ESTABLISH directly replaces the handler supplied by the compiler with a user-written handler; the probable result is improper handling of run-time signals. You should use Pascals predeclared ESTABLISH procedure to establish user-written condition handlers.
LISTONEND, LIST attribute allowed only on nal formal parameter Error. LISTUSEARG, Formal parameter has LIST attribute, use predeclared function ARGUMENT Error: A formal parameter with the LIST attribute cannot be directly referenced. You should use the predeclared function ARGUMENT to reference the actual parameters corresponding to the formal parameter. LNETOOLNG, Line too long, is truncated to 255 characters Error: A source line cannot exceed 255 characters. If it does, the compiler disregards the remainder of the line. LOWGTRHIGH, Low-bound exceeds high-bound Error: The denition of the agged subrange type is illegal because the value specied for the lower limit exceeds that for the upper limit. MAXLENINT, Max-length must be a value of type integer Error: The maximum length specied for type VARYING OF CHAR must be an integer in the range 1..65535; that is, the type denition must denote a legal character string. MAXLENRNG, Max-length must be in range 1..65535 Error: The maximum length specied for type VARYING OF CHAR must be an integer in the range 1..65535; that is, the type denition must denote a legal character string. MAXNUMENV, Maximum number of environments exceeded Fatal: More than 512 environment les were used in the compilation. MECHEXTERN, Foreign mechanism specier allowed only on external routines Error. MISSINGEND, No matching END, expected near line line number Information: The compiler expected an END statement at a location where none was found. Compilation proceeds as though the END statement were correctly located.
MODINIT26, Module name limited to 26 characters when initialization required Error: When a module contains schema types, discriminated schema types, variables of discriminated schema types, or a TO BEGIN DO statement clause, the module name is limited to 26 characters. MODOFNEGNUM, MOD of a negative modulus has no mathematical denition Error: In the MOD operation A MOD B, the operand B must have a positive integer value. This message is issued only when the MOD operation occurs in a compile-time constant expression. MSTBEARRAY, Type must be ARRAY Error. MSTBEARRVRY, Type must be ARRAY or VARYING Error: You cannot use the syntax [index] to refer to an object that is not of type ARRAY or VARYING OF CHAR. MSTBEBOOL, Control expression must be of type BOOLEAN Error: The IF, REPEAT, and WHILE statements require a Boolean control expression. MSTBEDEREF, Must be dereferenced Information. MSTBEDISCR, Schema type must be discriminated Error: An undiscriminated schema type is not allowed everywhere that a regular type name is allowed. MSTBEORDSETARR, Type must be ordinal, SET, or ARRAY Error. MSTBEREC, Type must be RECORD Error. MSTBERECVRY, Type must be RECORD or VARYING Error: You cannot use the syntax Variable.Identier to refer to an object that is not of type RECORD or VARYING OF CHAR.
MSTBESTAT, Cannot initialize non-STATIC variables Error: You cannot initialize variables declared without the STATIC attribute in nested blocks, nor can you initialize program-level variables whose attributes give them some allocation other than static. MSTBETEXT, I/O routine requires FILE_VARIABLE of type TEXT Error: The READLN and WRITELN procedures operate only on text les. MULTDECL, symbol name has multiple conicting declarations, reason(s): Error. NCATOA, Cannot reformat content of actuals CLASS_NCA descriptor as CLASS_A Error: This message can appear as additional information on other error messages. NEWQUADAGN, type names base type is ALIGNED(nnn); NEW handles at most ALIGNED(3) Error: You cannot call the NEW procedure to allocate pointer variables whose base types specify alignment greater than a quadword. To allocate such variables, you must use external routines. NOACTCOM, No actuals are compatible with schema formal parameter Information: Undiscriminated schema formal parameters denoting subranges or sets cannot be used as value parameters. In these cases, no actual parameter can ever be compatible with the formal parameter. NOASSTOFNC, Block does not contain an assignment to function result function name Warning: The block of a function must include a statement that assigns a return value to the function identier. NOCONVAL, A constant value was not specied for eld eld name Error. NODECLVAR, symbol name is not declared in a VAR section of block name Error: You cannot initialize a variable using the VALUE section if the variable was not declared in the same block in which the VALUE section appears.
NODSCREC, No descriptor class for RECORD type Error: The OpenVMS Calling Standard does not dene a descriptor format for records; therefore, you cannot specify %DESCR for a parameter of type RECORD. NODSCRSCH, No descriptor class for schematic types Error. NOFLDREC, No eld eld name in RECORD type type name Error: The eld specied does not exist in the specied record. NOFRMINDECL, Declaration of routine parameter routine name supplied no formal parameter list Information: You specied actual parameters in a call on a formal routine parameter that was declared with no formal parameters. Although such a call was legal in VAX Pascal Version 1.0, it does not follow the rules of the Pascal standard. You should edit your program to reect this change. NOINITEXT, Initialization not allowed on EXTERNAL variables NOINITINH, Initialization not allowed on inherited variables Error: You can initialize only those variables whose storage is allocated in this compilation. NOINITVAR, Cannot initialize symbol nameit is not declared as a variable Error: Variables are the only data items that can be initialized, and they can be initialized only once. NOLISTATTR, Parameter to this predeclared function must have LIST attribute Error: ARGUMENT and ARGUMENT_LIST_LENGTH require their rst parameter to be a formal parameter with the LIST attribute. NONATOMIC, Unable to generate code for atomic access Warning: Due to poor alignment, the code generator is unable to generate an atomic code sequence to read or write the volatile object. NONGRNACC, Unable to generate code for requested granularity Warning: Due to poor alignment, the code generator is unable to generate a code sequence for the granularity requested.
NOREPRE, No textual representation for values of this type Error: You cannot write a value to a text le (using WRITE or WRITELN) or to a VARYING string (using WRITEV) if there is no textual representation for the type. Similarly, you cannot read a value from a text le (using READ or READLN) or from a VARYING string (using READV) if there is no textual representation for the type. Such types are RECORD, ARRAY (other than PACKED ARRAY [1..n] OF CHAR), SET, and pointer. NOTAFUNC, symbol name is not declared as a routine. Error: An identier followed by a left parenthesis, a semicolon, or one of the reserved words END, UNTIL, and ELSE is interpreted as a call to a routine with no parameters. This message is issued if the identier was not declared as a procedure or function identier. Note that in the current version, functions can be called with the procedure call statement. NOTASYNCH, routine name is not ASYNCHRONOUS Information: This message can appear as additional information on other error messages. NOTATAG, identier is not a tag-identier Error: The identier used with the CASE OF construct in a record constructor must be a tag identier. NOTATYPE, symbol name is not a type identier Error: An identier that does not represent a type was used in a context where the compiler expected a type identier. NOTAVAR, symbol name is not declared as a variable Error: You cannot assign a value to any object other than a variable. NOTAVARFNID, symbol name is not declared as a variable or a function identier Error: You cannot assign a value to any object other than a variable or a function identier. NOTAVARPARM, symbol name is not declared as a variable or parameter Error.
NOTBEADDR, May not be parameter to ADDRESS NOTBEARGV, May not be used as a parameter to ARGV NOTBEASSIGN, May not be assigned NOTBECALL, May not be called as a FUNCTION NOTBECAST, May not be type cast NOTBEDEREF, May not be dereferenced NOTBEDES, May not be passed by untyped %DESCR NOTBEEVAL, May not be evaluated NOTBEFILOP, May not be used in a le operation NOTBEFLD, May not be eld selected NOTBEFNCPRM, May not be passed as a FUNCTION parameter NOTBEFORCTL, May not be used as FOR loop variable NOTBEFORDES, May not be passed as a descriptor foreign parameter NOTBEFOREF, May not be passed as a reference foreign parameter NOTBEIADDR, May not be parameter to IADDRESS NOTBEIDX, May not be indexed NOTBEIMMED, May not be passed by untyped immediate passing mechanism NOTBENEW, May not be written into by NEW NOTBENSTCTL, May not be control variable for an inner FOR loop NOTBEREAD, May not be written into by READ NOTBEREF, May not be passed by untyped reference passing mechanism NOTBERODES, May not be passed as a READONLY descriptor foreign parameter NOTBEROFOR, May not be passed as a READONLY reference foreign parameter NOTBEROVAR, May not be passed as a READONLY VAR parameter NOTBETOUCH, May not be read/modied/written NOTBEVAR, May not be passed as a VAR parameter NOTBEWODES, May not be passed as a WRITEONLY descriptor foreign parameter NOTBEWOFOR, May not be passed as a WRITEONLY reference foreign parameter NOTBEWOVAR, May not be passed as a WRITEONLY VAR parameter NOTBEWRTV, May not be parameter to WRITEV Information: These messages can appear as additional information on other error messages. NOTBYTOFF, Field eld name is not aligned on a byte boundary Error.
NOTDECLROU, symbol name is not declared as a routine. NOTINITIAL, routine name is not INITIALIZE Information: These messages can appear as additional information on other error messages. NOTINRNG, Value does not fall within range of the tag type Error: The value specied as the case label of a variant record is not a legal value of the tag elds type. This message is also issued if a case label in a call to NEW, DISPOSE, or SIZE falls outside the range of the tag type. NOTNEWTYP, Schema must dene a new type Error: The type-denoter of a schema denition must dene a new type; for example, a subrange, an array, or a record. NOTXTLIB, No text library was specied at compile time Error: The specied %INCLUDE module could not be accessed because a text library was not specied on the command line or in the PASCAL$LIBRARY logical name. NOTSAMTYP, Not the same type NOTUNBOUND, routine name is not UNBOUND Information: These messages can appear as additional information on other error messages. NOTSCHEMA, symbol name is not a schema type Error. NOTVARNAM, Parameter to this predeclared function must be simple variable name Error: The parameter cannot be indexed, be dereferenced, have a eld selected, or be an expression. It must be the name of the entire variable. NOTVOLATILE, variable name is non-VOLATILE Warning: You should not use the ADDRESS function on a nonvolatile variable or component or on a formal VAR parameter. NOUNSATDECL, No unsatised declaration of label label name in block name Error.
NUMFRMLPARM, Different numbers of formal parameters Information: This message can appear as additional information on other error messages. NXTACTDIFF, NEXT of actuals component differs from that of other parameters in same section Error: All actual parameters passed to a formal parameter section whose type is a conformant schema must have identical bounds and be structurally compatible. This message refers to the allocation size and alignment of the arrays inner dimensions. OLDDECLSYN, Obsolete routine parameter declaration syntax Information: The declaration of a formal routine parameter uses the obsolete VAX Pascal Version 1.0 syntax. You should edit your program to incorporate the current version syntax, which is mandated by the Pascal standard. OPNDASSCOM, Operands are not assignment compatible OPNOTINT, Operand(s) must be of type integer Error. OPNDNAMCOM, Operands are not name compatible Error. ORDOPNDREQ, Ordinal operand(s) required Error or Warning: This message is at warning level if you try to use INT, ORD, or UINT on a pointer expression. It is at error level if you use PRED or SUCC on an expression whose type is not ordinal. OUTNOTDECL, OUTPUT not declared in heading Error: A call to EOF, EOLN, READLN, or WRITELN did not specify a le variable, and the default INPUT or OUTPUT was not listed in the program heading. OVRDIVZERO, Overow or division by zero in compile-time expression Error.
PACKSTRUCT, component name of a PACKED structured type Error or Warning: You cannot use the data items listed in a call to the ADDRESS function, nor can you pass them as writable VAR, %REF, %DESCR, or %STDESCR parameters. This message is at warning level if the variable or component has the UNALIGNED attribute, and at error level if the variable or component is actually unaligned. PARMACTDEF, Formal parameter parameter name has neither actual nor default Error: If a formal parameter is not declared with a default, you must pass an actual parameter to it when calling its routine. PARMCLAMAT, Parameter section classes do not match Information: This message can appear as additional information on other error messages. PARMLIMIT, OpenVMS architectural limit of 255 parameters exceeded Error: You cannot declare a procedure with more than 255 formal parameters. A function whose result type requires that the result be stored in more than 64 bits or whose result type is a character string cannot have more than 254 formal parameters. In a call to a routine declared with the LIST attribute, you also cannot pass more than 255 (or 254) actual parameters. PARMSECTMAT, Division into parameter sections does not match Information: This message can appear as additional information on other error messages. PARSEFAIL, error parsing command line; use PASCAL command Fatal: The Compaq Pascal compiler was invoked without using the PASCAL DCL command. PARSEFAIL, error parsing command line; using an invalid CLD table Fatal: The Compaq Pascal compiler was invoked with an incorrect or obsolete command-line denition in SYS$LIBRARY:DCLTABLES. Contact your system manager to reinstall SYS$LIBRARY: DCLTABLES.
PASPREILL, Passing predeclared routine name is illegal Error: You cannot use the IADDRESS function on a predeclared routine for which there is no corresponding routine in the run-time library (such as the interlocked functions). In addition, you cannot pass a predeclared routine as a parameter if there is no way to write the predeclared routines formal parameter list in Compaq Pascal. Examples of the latter case are the PRED and SUCC functions and many of the I/O routines. PASSEXTERN, Passing mechanism allowed only on external routines Error. PASSNOTLEG, Passing mechanism not legal for this type Error. PCKARRBOO, PACKED ARRAY OF BOOLEAN parameter expected Error. PCKUNPCKCON, Packed/unpacked conict Information: This message can appear as additional information on other error messages. PLACEBEFEOLN, Placeholder not terminated before end of line Error. PLACEIVCHAR, Illegal nonprinting character (ASCII decimal representation of character) within placeholder Warning. PLACENODOT, Repetition of pseudocode placeholders not allowed Error. PLACESEEN, Placeholder encountered Error. PLACEUNMAT, Unmatched placeholder delimiter Error. POSAFTNONPOS, Positional parameter cannot follow a nonpositional parameter Error.
POSALIGNCON, Position / alignment conict Error: The bit position specied by the POS attribute does not have the number of low-order bits implied by the specied alignment attribute. POSINT, POS expression must be a positive integer value Error. PRENAMRED, Predeclared name cannot be redened Error: A predeclared name may not be redened when dening constants with the /CONSTANT DCL qualier. PREREQPRMLST, Passing predeclared routine name requires formal to include parameter list Error: To pass one of the predeclared routines EXPO, ROUND, TRUNC, UNDEFINED, UTRUNC, UROUND, DBLE, SNGL, QUAD, INT, ORD, and UINT as an actual parameter to a routine, you must specify a formal parameter list in the corresponding formal routine parameter. PRMKWNSIZ, Parameter must have a size known at compile-time Error: The BIN, HEX, OCT, DEC, and UDEC functions cannot be used on conformant parameters. The SIZE and NEXT functions cannot be used on conformant parameters in compile-time constant expressions. PROCESSFILE, Compiling le le name Information. PROCESSRTN, Generating code for routine routine name Information. PROGSCHENV, PROGRAM with schema may not create environment Error: A program that declares a schema type cannot have the [ENVIRONMENT] attribute. Schema declarations should be placed in a separate module and inherited by the program.
PROPRMEXT, Declaration of program parameter name is EXTERNAL program parameter les must be locally allocated PROPRMFIL, A program parameter must be a variable of type FILE PROPRMINH, Declaration of program parameter name is inherited program parameter les must be locally allocated PROPRMLEV, Program parameter program parameter name is not declared as a variable at the outermost level Error: Any external le variable (other than INPUT and OUTPUT) that is listed in the program heading must also be declared as a le variable in a VAR section in the program block. PSECTMAXINT, Allocation of symbol name causes PSECT PSECT name to exceed MAXINT bits Error: The Compaq Pascal implementation restricts the size of a program section to 2,147,483,647 bits. PTRCMPEQL, Pointer values may only be compared for equality Error: The equality ( = ) and inequality ( <> ) operators are the only operators allowed for values of a pointer type; all other operators are illegal. PTREXPRCOM, Pointer expressions are not compatible Error: The base types of two pointer expressions being compared for equality ( = ) or inequality ( <> ) are not structurally compatible. QUOBEFEOL, Quoted string not terminated before end of line Error. QUOSTRING, Quoted string expected Error: The compiler expects the %DICTIONARY and %INCLUDE directives, and the radix notations for binary (%B), hexadecimal (%X), and octal constants (%O), to be followed by a quoted string of characters. RADIXTEXT, Radix input requires FILE_VARIABLE of type TEXT Error: The input radix speciers (BIN, OCT, and HEX) operate only on text les. READONLY, variable name is READONLY Warning: You cannot use a read-only variable in any context that would store a new value in the variable. For example, a read-only variable cannot be used in a le operation.
REALCNSTRNG, Real constant out of range Error: See the Compaq Pascal Language Reference Manual for details on the range of real numbers. REALOPNDREQ, Real (SINGLE, DOUBLE or QUADRUPLE) operand(s) required Error. RECHASFILE, Record contains one or more FILE components, POS is illegal Error. RECHASTMSTMP, Record contains one or more TIMESTAMP components, POS is illegal Error. RECLENINT, RECORD_LENGTH expression must be of type integer Error: The value of the record length parameter to the OPEN procedure must be an integer. RECLENMNGLS, RECORD_LENGTH parameter is meaningless given les type Warning: The record length parameter is usually relevant only for les of type TEXT and VARYING OF CHAR. RECMATCHTYP, MATCH_TYPE identier NXT or NXTEQL is recommended instead of GTR or GEQ Information. REDECL, A declaration of symbol name already exists in block name Error: You cannot redeclare an identier or a label in the same block in which it was declared. Inheriting an environment is equivalent to including all of its declarations at program or module level. REDECLATTR, attribute name already specied Error: Only one member of a particular attribute class can appear in the same attribute list. REDECLFLD, Record already contains a eld eld name Error: The names of the elds in a record must be unique; they cannot be duplicated between variants.
REINITVAR, variable name has already been initialized Error: Variables are the only data items that can be initialized, and they can be initialized only once. REPCASLAB, Value has already appeared as a label in this CASE statement Error: You cannot specify the same value more than once as a case label in a CASE statement. REPFACZERO, Repetition factor cannot be the function ZERO REQCLAORNCA, Arrays and conformants of this parameter type require either CLASS_A or CLASS_NCA REQCLS, Scalars and strings of this parameter type require CLASS_S Error. REGNATAGN, Operand must be naturally aligned Error. REQNOCH, Primary key requires NOCHANGES option Error. REQPKDARR, The combination of CLASS_S and %STDESCR requires a PACKED ARRAY OF CHAR structure Error. REQREADVAR, READ or READV requires at least one variable to read into Error: The READ and READV procedures require that you specify at least one variable to be read from a le. REQWRITELEM, WRITE requires at least one write-list-element Error: The WRITE procedure requires that you specify at least one item to be written to a le. RESPTRTYP, Result must be a pointer type Information. REVRNTLAB, Value has already appeared as a label in this variant part Error: You cannot specify the same value more than once as a case label in a variant part of a record.
RTNSTDESCR, Routines cannot be passed using %STDESCR Error. SCHCONST, Nonstatic constants are not allowed Error: Constants cannot be made for nonstatic types since that would yield constants without compile-time size and value. SCHFLDALN, Field in nonstatic type may not have greater than byte alignment Error. SCHOVERLAID, Use of schema types conicts with OVERLAID attribute Error: The OVERLAID attribute cannot be used on programs or modules that discriminate schema at the outermost level. SENDSPR, Internal Compiler Error Fatal: An error has occurred in the execution of the Compaq Pascal compiler. Along with this message, you will receive information that helps you nd the location in the source program and the name of the compilation phase at which the error occurred. You may be able to rewrite the section of your program that caused the error and thus successfully compile the program. However, even if you are able to remedy the problem, please submit a Software Performance Report (SPR) to Compaq and provide a machine-readable copy of the program. SEQ11FORT, PDP11 specic directive SEQ11 treated as equivalent to FORTRAN directive Information. SETBASCOM, SET base types are not compatible Error: The base type of two sets used in a set operation are not compatible. SETELEORD, SET element expression must be of an ordinal type Error: The expressions used to denote the elements of a set constructor or the bounds of a set type denition must have an ordinal type. SETNOTRNG, SET element is not in range 0..255 Error: In a set whose base type is a subrange of integers or unsigned integers, all set elements in the sets type denition or in a constructor for the set must be in the range 0..255.
SIZACTDIFF, SIZE of actual differs from that of other parameters in same section Error: All actual parameters passed to a formal parameter section whose type is a conformant schema must have identical bounds and be structurally compatible. This message refers to the allocation size of the arrays outermost dimension. SIZARRNCA, Explicit size on ARRAY dimension makes CLASS_NCA mandatory Error. SIZATRTYPCON, Size attribute / type conict Error: For an ordinal type, the size specied must be at least as large as the packed size but no larger than 32 bits on OpenVMS VAX or 64 bits on OpenVMS Alpha systems. Pointer types must be allocated, exactly 32 bits on OpenVMS systems and 64 bits on Tru64 UNIX systems. Type SINGLE exactly 32 bits, type DOUBLE exactly 64 bits, and type QUADRUPLE exactly 128 bits. For types ARRAY, RECORD, SET, and VARYING OF CHAR, the size specied must be at least as large as their packed sizes. For the details of allocation sizes in Compaq Pascal, see the Compaq Pascal Language Reference Manual. SIZCASTYP, Variables size conicts with casts target type Error: In a type cast operation, the size of the variable and the size of the type to which it is cast must be identical. SIZEDIFF, Sizes are different Information: This message can appear as additional information on other error messages. SIZEINT, Size expression must be a positive integer value Error. SIZGTRMAX, Size exceeds MAXINT bits Error: The size of a record or an array type or the size specied by a size attribute exceeds 2,147,483,647 bits. The Compaq Pascal implementation imposes this size restriction.
SIZMULTBYT, Size of component of array passed by descriptor is not a multiple of bytes Error: When an array or a conformant parameter is passed using the %DESCR mechanism specier, the descriptor built by the compiler must follow the OpenVMS Calling Standard. Such a descriptor can describe only an array whose components fall on byte boundaries. SPEOVRDECL, Foreign mechanism specier required to override parameter declaration Error: When you specify a default value for a formal VAR or routine parameter, you must also use a mechanism specier to override the characteristics of the parameter section. SPURIOUS, error message at line numbercolumn number Information: The compiler did not correctly note the location of this error in your program and later could not position and print the correct error message. You may be able to correct the section of your program that caused the error and thus avoid this error. Please submit a Software Performance Report (SPR) and provide a machine-readable copy of the program if you receive this error. SRCERRORS, Source errors inhibit continued compilationcorrect and recompile Fatal: A serious error previously detected in the source program has corrupted the compilers symbol tables and inhibits further compilation. Correct the serious error and recompile the program. SRCTXTIGNRD, Source text following end of compilation unit ignored Warning: The compiler ignores any text following the END statement that terminates a compilation unit. This error probably resulted from an unmatched END statement in your program. STDACTINCMP, Nonstandard: actual is not name compatible with other parameters in same section Information: According to the Pascal standard, all actual parameters passed to a parameter section must have the same type identier or the same type denition. This message is issued only if you have specied the standard switch on the compile command line.
STDATTRLST, Nonstandard: attribute list STDBIGLABEL, Nonstandard: label number greater than 9999 STDBLANKPAD, Nonstandard: blank-padding used during string operation STDBNDRMUSE, Nonstandard: usage of formal parameter for routine routine name STDCALLFUNC, Nonstandard: function function name called as a procedure STDCASLBLRNG, Nonstandard: label range in case selector STDCAST, Nonstandard: type cast operator STDCMPCOMPAT, Nonstandard: cannot PACK or UNPACK, array component types are incompatible STDCMPDIR, Nonstandard: compiler directive STDCOMFUNACC, Nonstandard: component function access STDCNFARR, Nonstandard: conformant array syntax Information: These messages refer to extensions to Pascal and are issued only if you have specied the standard switch on the compile command line. STDCNSTR, Nonstandard: array or record constructor Information: A VAX Pascal Version 1.0 style constructor was used. You should convert this constructor to the new constructor syntax provided in the current version of Compaq Pascal to be compatible with the Extended Pascal standard. STDCONCAT, Nonstandard: concatenation operator Information: This message refers to extensions to Pascal and is issued only if you have specied the standard switch on the compile command line. STDCONST, Nonstandard: type name constant Information: Binary, hexadecimal, and octal constants and constants of type DOUBLE, QUADRUPLE, UNSIGNED, INTEGER64, and UNSIGNED64 are extensions to Pascal. This message is issued only if you have specied the standard switch on the compile command line. STDCONSTACC, Nonstandard: structured constant access Information: This message is issued if you have specied a standard option other than extended on the compile command line.
STDCTLDECL, Nonstandard: control variable variable name not declared in VAR section of block name Information: The Pascal standard requires that the control variable of a FOR statement be declared in the same block in which the FOR statement appears. STDDECLSEC, Nonstandard: declaration sections either out of order or duplicated in block name Information: In the Pascal standard, the declaration sections must appear in the order LABEL, CONST, TYPE, VAR, PROCEDURE, and FUNCTION. The ability to specify the sections in any order is an extension. This message occurs only if you have specied the standard switch on the compile command line. STDDEFPARM, Nonstandard: default parameter declaration Information: This message refers to extensions to Pascal and is issued only if you have specied the standard switch on the compile command line. STDDIRECT, Nonstandard: directive name directive Information: The EXTERN, EXTERNAL, FORTRAN, and SEQ11 directives are extensions to Pascal. (FORWARD is the only directive specied by the Pascal standard.) This message is issued only if you have specied the standard switch on the compile command line. STDDISCREF, Nonstandard: schema discriminant reference STDDISCSCHEMA, Nonstandard: discriminated schema Information: These messages are issued if you have specied a standard argument other than extended on the compile command line. STDEMPCASLST, Nonstandard: empty case-list element Information: This message is issued if you do not specify any case labels and executable statements between two semicolons or between OF and a semicolon in the CASE statement. You must also have specied the standard switch on the compile command line. STDEMPPARM, Nonstandard: empty actual parameter position Information: This message refers to extensions to Pascal and is issued only if you have specied the standard switch on the compile command line.
STDEMPREC, Nonstandard: empty record section Information: The Pascal standard does not allow record type denitions of the form RECORD END. This message appears only if you have specied the standard switch on the compile command line. STDEMPSTR, Nonstandard: empty string Information: This message refers to extensions to Pascal and is issued only if you have specied the standard switch on the compile command line. STDEMPVRNT, Nonstandard: empty variant Information: This message occurs if you do not specify a variant between two semicolons or between OF and a semicolon. You must also have specied the standard switch on the compile command line. STDEOLCOM, Nonstandard: End of line comment Information: The message is issued if you use the exclamation point character to treat the remainder of the line as a comment. You must also have specied the standard switch on the compile command line. STDERRPARM, Nonstandard: error-recovery parameter STDEXPON, Nonstandard: exponentiation operator STDEXTSTR, Nonstandard: extended string syntax Information: These messages refer to extensions to Pascal and are issued only if you have specied the standard switch on the compile command line. STDFLDHIDPTR, Nonstandard: record eld identier eld identier name hides type identier eld identier name Information. STDFORIN, Nonstandard: SET-iteration in FOR statement Information: This message is issued if you have specied a standard argument other than extended on the compile command line. STDFORMECH, Nonstandard: foreign mechanism specier Information: This message refers to an extension to Pascal and is issued only if you have specied the standard switch on the compile command line.
STDFORWARD, Nonstandard: PROCEDURE/FUNCTION block routine name and its FORWARD heading are not in the same section Information: The Extended Pascal standard requires that FORWARD declared routines must specify their corresponding blocks without intervening LABEL, CONST, TYPE, or VAR sections. This message is issued only if you have specied the extended argument to the standard switch on the compile command line. STDFUNIDEVAR, Nonstandard: function identied variable Information: This message is issued if you have specied a standard argument other than extended on the compile command line. STDFUNCTRES, Nonstandard: FUNCTION returning a value of a type name type Information: The ability of functions to have structured result types is an extension to Pascal. This message is issued only if you have specied the standard switch on the compile command line. STDINCLUDE, Nonstandard: %INCLUDE directive STDINITVAR, Nonstandard: initialization syntax in VAR section Information: These messages refer to extensions to Pascal and are issued only if you have specied the standard switch on the compile command line. STDKEYWRD, Nonstandard: keyword name Information: This message is issued if you have specied a standard option other than extended on the compile command line. STDMATCHVRNT, Nonstandard: no matching variant label Information: This message is issued if you call the NEW or DISPOSE procedure, and one of the case labels specied in the call does not correspond to a case label in the record variable. You must also have specied the standard switch on the compile command line. STDMODCTL, Nonstandard: potential uplevel modication of variable name prohibits use as control variable Information: You cannot use as the control variable of a FOR statement any variable that might be modied in a nested block. This message is issued only if you have specied the standard switch on the compile command line.
STDMODULE, Nonstandard: MODULE declaration Information: The item listed in this message is an extension to Pascal. This message is issued only if you have specied the standard switch on the compile command line. STDNILCON, Nonstandard: use of reserved word NIL as a constant Information: Only simple constants and quoted strings are allowed by the Pascal standard to appear as constants. Simple constants are integers, character strings, real constants, symbolic constants, and constants of BOOLEAN and enumerated types. This message is issued only if you have specied the standard switch on the compile command line. STDNOFRML, Nonstandard: FUNCTION or PROCEDURE parameter declaration lacks formal parameter list Information: This message is issued if you try to pass actual parameters to a formal routine parameter for which you declared no formal parameter list. You must also have specied the standard switch on the compile command line. STDNONPOS, Nonstandard: nonpositional parameter syntax STDOTHER, Nonstandard: OTHERWISE clause STDPASSPRE, Nonstandard: passing predeclared routine name Information: These messages refer to extensions to Pascal and are issued only if you have specied the standard switch on the compile command line. STDNOTIN, Nonstandard: NOT IN operator Information: This message refers to an extension in Pascal and is issued only if you have specied the standard switch on the compile command line. STDPCKSET, Nonstandard: combination of packed and unpacked sets Information: The Pascal standard does not allow packed and unpacked sets to be combined in set operations. This message is issued only if you have specied the standard switch on the compile command line.
STDPRECONST, Nonstandard: predeclared constant constant name Information: The constants MAXCHAR, MAXINT64, MAXUNSIGNED, MAXUNSIGNED64, MAXREAL, MINREAL, EPSREAL, MAXDOUBLE, MINDOUBLE, EPSDOUBLE, MAXQUADRUPLE, MINQUADRUPLE, and EPSQUADRUPLE are extensions to Pascal. MAXCHAR, MAXREAL, MINREAL, and EPSREAL are contained in Extended Pascal. This message is issued only if you have specied the standard switch on the compile command line. STDPREDECL, Nonstandard: predeclared routine Information: Many predeclared procedures and functions are extensions to Pascal. The use of these routines causes this message to be issued if you have specied the standard switch on the compile command line. STDPRESCH, Nonstandard: predened schema type name Information: This message is issued if you have specied a standard switch other than extended on the compile command line. STDPRETYP, Nonstandard: predened type type name Information: The types SINGLE, DOUBLE, INTEGER64, QUADRUPLE, UNSIGNED, UNSIGNED64, and VARYING OF CHAR are extensions to Pascal. This message is issued only if you have specied the standard switch on the compile command line. STDQUOSTR, Nonstandard: quotes enclosing radix constant Information: This message is issued if you have specied the extended option on the compile command line. STDRADFORMAT, Nonstandard: use format radix#nnn for radix constant Information: This message refers to the use of an extension to Pascal. This message is issued only if you have specied the extended argument to the standard switch on the compile command line. STDRADIX, Nonstandard: radix constant Information: This message refers to the use of an extension to Pascal. This message is issued only if you have specied a standard switch other than extended on the compile command line.
STDRDBIN, Nonstandard: binary input from a TEXT le STDRDENUM, Nonstandard: enumerated type input from a TEXT le STDRDHEX, Nonstandard: hexadecimal input from a TEXT le STDRDOCT, Nonstandard: octal input from a TEXT le STDRDSTR, Nonstandard: string input from a TEXT le Information: The Pascal standard allows only INTEGER, CHAR, and REAL values to be read from a text le. The ability to read values of other types is an extension to Pascal. These messages are issued only if you have specied the standard switch on the compile command line. STDREDECLNIL, Nonstandard: redeclaration of reserved word NIL Information: The Pascal standard considers NIL a reserved word, while Compaq Pascal considers it to be a predeclared identier. Thus, if you have specied the standard switch on the compile command line, this message will be issued if you attempt to redene NIL. STDREM, Nonstandard: REM operator Information: The item listed in this message is an extension to Pascal. This message is issued only if you have specied the standard switch on the compile command line. STDSCHEMA, Nonstandard: schema type denition Information: This message is issued if you have specied a standard argument other than extended on the compile command line. STDSCHEMAUSE, Nonstandard: use of schema type Information: This message is issued if you have specied a standard argument other than extended on the compile command line. STDSIMCON, Nonstandard: only simple constant (optional sign) or quoted string Information: Only simple constants and quoted strings are allowed by the Pascal standard to appear as constants. Simple constants are integers, character strings, real constants, symbolic constants, constants of type BOOLEAN, and enumerated types. This message is issued only if you have specied the standard switch on the compile command line. STDSPECHAR, Nonstandard: $ or _ in identier STDSTRCOMPAT, Nonstandard: string compatibility Information: These messages refer to extensions to Pascal and are issued only if you have specied the standard switch on the compile command line.
STDSTRUCT, Nonstandard: types do not have same name Information: Because the Pascal standard does not recognize structural compatibility, two types must have the same type identier or type denition to be compatible. This message is issued only if you have specied the standard switch on the compile command line. STDSYMLABEL, Nonstandard: symbolic label Information: These messages refer to extensions to Pascal and are issued only if you have specied the standard switch on the compile command line. STDTAGFLD, Nonstandard: invalid use of tag eld Information: The tag eld of a variant record cannot be a parameter to the ADDRESS function, nor can you pass it as a writable VAR, %REF, %DESCR, or %STDESCR formal parameter. This message is issued only if you have specied the standard switch on the compile command line. STDTODECL, Nonstandard: TO BEGIN/END DO declaration Information: This message is issued if you have specied a standard argument other than extended on the compile command line. STDUNSAFE, Nonstandard: UNSAFE compatibility Information: If you have used the UNSAFE attribute on an object that is later tested for compatibility, you will receive this message. You must also have specied the standard switch on the compile command line. STDUSEDCNF, Nonstandard: conformant array used as a string STDUSEDPCK, Nonstandard: PACKED ARRAY [1..1] OF CHAR used as a string Information: These messages refer to extensions to Pascal and are issued only if you have specied the standard switch on the compile command line. STDVALCNFPRM, Nonstandard: conformant array may not be passed to value conformant parameter Information. STDVALUE, Nonstandard: VALUE initialization section STDVAXCDD, Nonstandard: %DICTIONARY directive Information: These messages refere to extensions to Pascal and are issued only if you have specied the standard switch on the compile command line.
STDVRNTCNSTR, Nonstandard: variant eld outside constructor variant part Information: This message refers to the use of an extension to Pascal. This message is issued only if you have specied the extended argument to the standard switch the compile command line. STDVRNTPART, Nonstandard: empty variant part Information: According to the Pascal standard, a variant part that declares no case labels and eld lists between the words OF and END is illegal. This message occurs only if you have specied the standard switch on the compile command line. STDVRNTRNG, Nonstandard: variant labels do not cover the range of the tag type Information: According to the Pascal standard, you must specify one case label for each value in the tag type of a variant record. This message is issued only if you have specied the standard switch on the compile command line. STDWRTBIN, Nonstandard: binary output to a TEXT le STDWRTENUM, Nonstandard: user dened enumerated type output to a TEXT le STDWRTHEX, Nonstandard: hexadecimal output to a TEXT le STDWRTOCT, Nonstandard: octal output to a TEXT le Information: The Pascal standard allows only INTEGER, BOOLEAN, CHAR, REAL, and PACKED ARRAY [1..n] OF CHAR values to be written to a text le. The ability to write values of other types is an extension to Pascal. These messages are issued only if you have specied the standard switch on the compile command line. STDSUBSTRING, Nonstandard: Substring notation Information. STDZERO, Nonstandard: ZERO function used in constructor Information: This message refers to an extension in Pascal and is issued only if you have specied the standard switch on the compile command line. STOREQEXC, Allocations to Psect "name" exceeded growth bounds Error: Too much data is allocated to the Psect. Either place variables into different Psects or break up the program into multiple modules
STREQLLEN, String values must be of equal length Error: You cannot perform string comparisons on character strings that have different lengths. STROPNDREQ, String (CHAR, PACKED ARRAY [1..n] OF CHAR, or VARYING) operand required STRPARMREQ, String (CHAR, PACKED ARRAY [1..n] OF CHAR, or VARYING) parameter required STRTYPREQ, String (CHAR, PACKED ARRAY [1..n] OF CHAR, or VARYING) type required Error: The le-name parameter to the OPEN procedure and the parameter to the LENGTH function must be character strings of the types listed. SYNASCII, Illegal ASCII character SYNASSERP, Syntax: :=, ; or ) expected SYNASSIGN, Syntax: := expected SYNASSIN, Syntax: := or IN expected SYNASSSEMI, Syntax: := or ; expected SYNATRCAST, Syntax: attribute list not allowed on a type cast SYNATTTYPE, Syntax: attribute-list or type specication SYNBEGDECL, Syntax: BEGIN or declaration expected SYNBEGEND, Syntax: BEGIN or END expected SYNBEGIN, Syntax: BEGIN expected SYNCOASSERP, Syntax: ,, :=, ; or ) expected SYNCOELRB, Syntax: ,, .. or ] expected SYNCOLCOMRP, Syntax: :, , or ) expected SYNCOLON, Syntax: : expected SYNCOMCOL, Syntax: , or : expected SYNCOMDO, Syntax: , or DO expected SYNCOMEQL, Syntax: , or = expected SYNCOMMA, Syntax: , expected SYNCOMRB, Syntax: , or ] expected SYNCOMRP, Syntax: , or ) expected SYNCOMSEM, Syntax: , or ; expected SYNCONTMESS, Syntax: CONTINUE or MESSAGE expected SYNCOSERP, Syntax: ,, ; or ) expected SYNDIRBLK, Syntax: directive or block expected Error: The compiler either failed to nd an important lexical or syntactical element where one was expected, or it detected an error in such an element that does exist in your program.
SYNDIRMIS, Syntax: directive missing, EXTERNAL assumed Error: In the absence of a directive where one is expected, the compiler assumes that EXTERNAL is the intended directive and proceeds with compilation based on that assumption. SYNDO, Syntax: DO expected SYNELIPSIS, Syntax: .. expected SYNELSESTMT, Syntax: ELSE or start of new statement expected SYNEND, Syntax: END expected SYNEQL, Syntax: = expected SYNEQLLP, Syntax: = or ( expected SYNERRCTE, Error in compile-time expression SYNEXPR, Syntax: expression expected SYNEXSEOTEN, Syntax: expression, ;, OTHERWISE or END expected SYNFUNPRO, Syntax: FUNCTION or PROCEDURE expected SYNHEADTYP, Syntax: routine heading or type identier expected SYNIDCAEND, Syntax: identier, CASE or END expected SYNIDCARP, Syntax: identier, CASE or ) expected SYNIDCASE, Syntax: identier or CASE expected SYNIDENT, Syntax: identier expected SYNILLEXPR, Syntax: ill-formed expression SYNINT, Syntax: integer expected SYNINTBOO, Syntax: integer, boolean, or string literal expected SYNINVSEP, Syntax: invalid token separator SYNIVATRLST, Syntax: illegal attribute list SYNIVPARM, Syntax: illegal actual parameter SYNIVPRMLST, Syntax: illegal actual parameter list SYNIVSYM, Syntax: illegal symbol SYNIVVAR, Syntax: illegal variable SYNLABEL, Syntax: label expected SYNLBRAC, Syntax: [ expected SYNLPAREN, Syntax: ( expected SYNLPASEM, Syntax: ( or ; expected SYNLPCORB, Syntax: (, , or ] expected SYNLPSECO, Syntax: (, ; or : expected SYNMECHEXPR, Syntax: mechanism specier or expression expected SYNNEWSTMT, Syntax: start of new statement expected SYNOF, Syntax: OF expected SYNPARMLST, Syntax: actual parameter list
SYNPARMSEC, Syntax: parameter section expected SYNPERIOD Syntax: . expected. SYNPROMOD, Syntax: PROGRAM or MODULE expected SYNQUOSTR, Syntax: quoted string expected SYNRBRAC, Syntax: ] expected SYNRESWRD, Syntax: reserved word cannot be redened SYNRPAREN, Syntax: ) expected SYNRPASEM, Syntax: ; or ) expected SYNRTNTYPCNF, Syntax: routine heading, type identier or conformant parameter expected SYNSEMI, Syntax: ; expected SYNSEMIEND, Syntax: ; or END expected SYNSEMMODI, Syntax: ;, ::, ^, or [ expected SYNSEMRB, Syntax: ; or ] expected SYNSEOTEN, Syntax: ;, OTHERWISE or END expected SYNTHEN, Syntax: THEN expected SYNTODOWN, Syntax: TO or DOWNTO expected SYNSEOTRP, Syntax: ;, OTHERWISE, or ) expected SYNTYPCNF, Syntax: type identier or conformant parameter expected SYNTYPID, Syntax: type identier expected Error: The compiler either failed to nd an important lexical or syntactical element where one was expected, or it detected an error in such an element that does exist in your program. SYNTYPPACK, Only ARRAY, FILE, RECORD or SET types can be PACKED Warning: You cannot pack any type other than the structured types listed in the message. SYNTYPSPEC, Syntax: type specication expected SYNUNEXDECL, Syntax: declaration encountered in executable section SYNUNTIL, Syntax: UNTIL expected SYNXTRASEMI, Syntax: ; ELSE is not valid Pascal, ELSE matched with IF on line line number Error: The compiler either detected an error in a lexical or syntactical element in your program, or it failed to nd such an element where one was expected. TAGNOTORD, Tag type must be an ordinal type Error: The type of a variant records tag eld must be one of the ordinal types.
TOOIDXEXPR, Too many index expressions; type has only number of dimensions dimensions Error: A call to the UPPER or LOWER function specied an index value that exceeds the number of dimensions in the dynamic array. TOOMANYIFS, Conditional compilation nesting level exceeds implementation limit Error: %IF directives may only be nested 32 deep. TOPROGRAM, TO BEGIN/END DO not allowed in PROGRAM Error: TO BEGIN DO and TO END DO declarations are only allowed in modules. TYPCNTDISCR, Type can not be discriminated in this context Error. TYPFILSIZ, Type contains one or more FILE components, size attribute is illegal Error: The allocation size of a FILE type cannot be controlled by a size attribute; therefore, you cannot use a size attribute on any type that has a le component. TYPHASFILE, Type contains one or more FILE components Error: Many operations are illegal on objects of type FILE and objects of structured types with le components; for example, you cannot initialize them, use them as value parameters, or read them with input procedures. TYPHASNOVRNT, Type contains no variant part Error: You can only use the formats of the NEW, DISPOSE, and SIZE routines that allow case labels to be specied when their parameters have variants. TYPPTRFIL, Type must be pointer or FILE Error: You cannot use the syntax Variable^ to refer to an object whose type is not pointer or FILE. TYPREF, %REF not allowed for this type Error: The %REF foreign mechanism specier cannot be used with schematic variables.
TYPSTDESCR, %STDESCR not allowed for this type Error: The %STDESCR mechanism specier is allowed only on objects of type CHAR, PACKED ARRAY [1..n] OF CHAR, VARYING OF CHAR, and arrays of these types. TYPVARYCHR, Component type of VARYING must be CHAR Error. UNALIGNED, variable name is UNALIGNED Error or Warning: You cannot use the data items listed in a call to the ADDRESS function, nor can you pass them as writable VAR, %REF, %DESCR, or %STDESCR parameters. This message is at warning level if the variable or component has the UNALIGNED attribute, and at error level if the variable or component is actually unaligned. UNAVOLACC, Volatile access appears unaligned, but must be aligned at run-time to ensure atomicity and byte granularity Warning: The code generator was unable to determine if a volatile access was aligned or not. It generated two sequences; one sequence will perform the atomic access if it was aligned properly; the second sequence accesses the object, but may contain a timing window where incorrect results may occur. UNBPNTRET, routine name is not UNBOUNDframe-pointer not returned Warning: The IADDRESS function returns only the address of the procedure value (on OpenVMS VAX systems, the entry mask of the routine is called). This address may be sufcient information to successfully invoke an unbound routine, but not a bound routine. (Bound routines are represented as a pair of addresses: one pointing to the procedure value and the other to the frame pointer to the routine in which the routine was declared.) UNCALLABLE, Routine "name" can never be called Information. UNCERTAIN, Variable name has not been initialized Information.
UNDECLFRML, Undeclared formal parameter symbol name Error: A formal parameter name listed in a nonpositional call to a routine does not match any of the formal parameters declared in the routine heading. UNDECLID, Undeclared identier symbol name Error: In Pascal, an identier must be declared before it is used. There are no default or implied declarations. UNDSCHILL, Undiscriminated schema type is illegal Error: An undiscriminated schema type does not have any actual discriminants. Without discriminants, the type size, any nested ARRAY bounds, and the offset of any nested RECORD elds are unknown. UNINIT, "Variable name" is fetched, not initialized Information. UNPREDRES, Calling FUNCTION function name declared FORWARD may yield unpredictable results Warning: By using FORWARD declared functions in actual discriminant expressions, you can cause innite loops at run time or access violations. UNREAD, Variable, variable name is assigned into, but never read Information. UNSCNFVRY, UNSAFE attribute not allowed on conformant VARYING parameter Error. UNSEXCRNG, constant exceeds range of "datatype" Error: The largest value allowed for an UNSIGNED value is 4,294,967,295. The largest value allowed for an UNSIGNED64 value is 18,446,744,073,709,551,615. UNUSED, Variable, variable name is never referenced Information. UNWRITTEN, Variable variable name is read, but never assigned into Warning.
UPLEVELACC, Unbound routine name precludes uplevel access to variable name Error: A routine that was declared with the UNBOUND attribute cannot refer to automatic variables, routines, or labels declared in outer blocks. UPLEVELGOTO, Unbound routine name precludes uplevel GOTO to label name Error: A routine that was declared with the UNBOUND attribute cannot refer to automatic variables, routines, or labels declared in outer blocks. USEDBFDECL, symbol name was used before being declared Warning. USEINISTA, Use initial-state (VALUE clause) on TYPE or VAR declaration Information: Nonstatic variables, such as those created from schema types, cannot be initialized in the VALUE declaration part. To initialize these variables, you must use the initial state feature. V1DYNARR, Decommitted Version 1 dynamic array type Error: The type syntax used to dene a dynamic array parameter has been decommitted for the current version of Compaq Pascal. You should edit your program to make the type denition conform to the current version conformant array syntax. V1DYNARRASN, Decommitted Version 1 dynamic array assignment Error: In VAX Pascal Version 1.0, dynamic arrays used in assignments could not be checked for compatibility until run time. This warning indicates that your program depends on an obsolete feature, which you should consider changing to reect the current version syntax for conformant array parameters. V1MISSPARM, Decommitted missing parameter syntax: correct by adding number of commas comma(s) Error: An OPEN procedure called with the decommitted VAX Pascal Version 1.0 syntax fails to mark omitted parameters with commas. Your program depends on this obsolete feature, and you should insert the correct number of commas as listed in the message.
V1PARMSYN, Use of unsupported V1 omitted parameter syntax with new V2 feature(s) Error: In a parameter list for the OPEN procedure, you cannot use both the Version 1.0 syntax for OPEN and the parameters that are new to subsequent versions of Compaq Pascal. V1RADIX, Decommitted Version 1 radix output specication Error: In VAX Pascal Version 1.0, octal and hexadecimal values could be written by placing the keywords OCT or HEX after a eld width expression. Your program uses this obsolete feature; you should consider changing it to use the current versions OCT or HEX predeclared functions. VALOUTBND, Value to be assigned is out of bounds Error: A value specied in an array or record constructor exceeds the subrange dened as the type of the corresponding component. VALUEINIT, VALUE variables must be initialized Error: Variables with both the VALUE and GLOBAL attributes must be given an initial value in either the VAR section or in the VALUE section. VALUETOOBIG, VALUE attribute not allowed on objects larger than 32 bits Error: Variables with the VALUE attribute cannot be larger than 32 bits because they are expressed to the linker as global symbol references. VALUETYP, VALUE allowed only on ordinal or real types Error. VALUEVISIB, GLOBAL or EXTERNAL visibility is required with the VALUE attribute Error: Variables with the VALUE attribute must be given either external or global visibility. (If the variable is given global visibility, then it must also be given an initial value.) VARCOMFRML, Variable is not compatible with formal parameter formal parameter name Error: A variable being passed as an actual parameter is not compatible with the corresponding formal parameter indicated. Variable parameters must be structurally compatible. The reason for the incompatibility is provided in an informational message that the compiler prints along with this error message.
VARNOTEXT, Variable must be of type TEXT Error: The EOLN function requires that its parameter be a le of type TEXT. VARPRMRTN, Formal VAR parameter may not be a routine Error: The reserved word VAR cannot precede the word PROCEDURE or FUNCTION in a formal parameter declaration. VARPTRTYP, Variable must be of a pointer type Error: The NEW and DISPOSE procedures operate only on pointer variables. VARYFLDS, LENGTH and BODY are the only elds in a VARYING type Error: You cannot use the syntax Variable.Identier to specify any elds of a VARYING OF CHAR variable other than LENGTH and BODY. VISAUTOCON, Visibility / AUTOMATIC allocation conict Error: The GLOBAL, EXTERNAL, WEAK_GLOBAL, and WEAK_ EXTERNAL attributes require static allocation and therefore conict with the AUTOMATIC attribute. VISGLOBEXT, Visibilities are not GLOBAL/EXTERNAL or EXTERNAL /EXTERNAL Information: In repeated declarations of a variable or routine, only one declaration at most can be global; all others must be external. This message can appear as additional information for other error messages. VRNTRNG, Variant labels do not cover the range of the tag type Error: According to the Pascal standard, you must specify one case label for each value in the tag type of a variant record or include an OTHERWISE clause. WDTHONREAL, Second eld width is allowed only when value is of a real type Error: The fraction value in a eld-width specication is allowed only for real-number values. WRITEONLY, variable name is WRITEONLY Warning: You cannot use a write-only variable in any context that requires the variable to be evaluated. For example, a write-only variable cannot be used as the control variable of a FOR statement.
XTRAERRORS, Additional diagnostics occurred on this line Information: The number of errors occurring on this line exceeds the implementations limit for outputting errors. You should correct the errors given and recompile your program. ZERNOTALL, ZERO is not allowed for type or types containing type name Error: ZERO may not be used to initialize objects of type FILE, TEXT, or TIMESTAMP or objects containing these types.
C.2 Run-Time Diagnostics

During execution, an image can generate a fatal error called an exception condition. When the Compaq Pascal run-time system detects such a condition, the system displays an error message and aborts program execution. Run-time errors can also be issued by other facilities, such as the OpenVMS Sort Utility or the OpenVMS operating system. Compaq Pascal run-time system diagnostics are preceded by the following: %PAS- FThe severity level of a run-time error is F, fatal error. Some conditions, particularly I/O errors, may cause several messages to be generated. The rst message is a diagnostic that species the le that was being accessed (if any) when the error occurred and the nature of the error. Next, an RMS error message may be generated. In most cases, you should be able to understand the error by looking up the rst message in the following list. If not, see the OpenVMS System Messages and Recovery Procedures Reference Manual (this manual is archived, and is available on the OpenVMS documentation CDROM) for an explanation of the RMS error message. All diagnostic messages contain a brief explanation of the event that caused the error. This section lists run-time diagnostic messages in alphabetical order, including explanatory message text. Where the message text is not self-explanatory, additional explanation follows. Portions of the message text enclosed in quotation marks are items that the compiler substitutes with the name of a data object when it generates the message.
ACCMETINC, ACCESS_METHOD specied is incompatible with this le Explanation: The value of the ACCESS_METHOD parameter for a call to the OPEN procedure is not compatible with the les organization or record type. You can use DIRECT access only with les that have relative organization or sequential organization and xed-length records. You can use KEYED access only with indexed les. User Action: Make sure that you are accessing the correct le. See Chapter 7 to determine which access method you should use. AMBVALENU, string is an ambiguous value for enumerated type type Explanation: While a value of an enumerated type was being read from a text le, not enough characters of the identier were found to specify an unambiguous value. User Action: Specify enough characters of the identier so that it is not ambiguous. ARRINDVAL, array index value is out of range Explanation: You enabled bounds checking for a compilation unit and attempted to specify an index that is outside the arrays index bounds. User Action: Correct the program or data so that all references to array indexes are within the declared bounds. ARRNOTCOM, conformant array is not compatible Explanation: You attempted to assign one dynamic array to another that did not have the same index bounds. This error occurs only when the arrays use the decommitted VAX Pascal Version 1.0 syntax for dynamic array parameters. User Action: Correct the program so that the two dynamic arrays have the same index bounds. You could also change the arrays to conform to the current syntax for conformant arrays; most incompatibilities could then be detected at compile time rather than at run time. See the Compaq Pascal Language Reference Manual for more information on current conformant arrays. ARRNOTSTR, conformant array is not a string Explanation: In a string operation, you used a conformant PACKED ARRAY OF CHAR value whose index had a lower bound not equal to 1 or an upper bound greater than 65535. User Action: Correct the arrays index so that the array is a character string.
ASSERTION, Pascal assertion failure Explanation: The expression used in the Pascal ASSERT built-in routine evaluated to false. User Action: Correct the problem that was being checked with the ASSERT built-in in the source program. BUGCHECK, internal consistency failure nnn in Pascal Run-Time Library Explanation: The run-time library has detected an internal error or inconsistency. This problem may be caused by an out-of-bounds array reference or a similar error in your program. User Action: Rerun your program with all CHECK options enabled. If you are unable to nd an error in your program, please submit a Software Performance Report (SPR) to Compaq, including a machine-readable copy of your program, data, and a sample execution illustrating the problem. CANCNTERR, handler cannot continue from a nonle error Explanation: A user condition handler attempted to return SS$_CONTINUE for an error not involving le input/output. To recover from such an error, you must use either an uplevel GOTO statement or the SYS$UNWIND system service. User Action: Modify the user handler to use one of the allowed recovery actions for nonle errors, or to resignal the error if no recovery action is possible. CASSELVAL, CASE selector value is out of range Explanation: The value of the case selector in a CASE statement does not equal any of the specied case labels, and the statement has no OTHERWISE clause. User Action: Either add an OTHERWISE clause to the CASE statement or change the value of the case selector so that it equals one of the case labels. See the Compaq Pascal Language Reference Manual for more information. CONCATLEN, string concatenation has more than 65535 characters Explanation: The result of a string concatenation operation would result in a string longer than 65,535 characters, which is the maximum length of a string. User Action: Correct the program so that all concatenations result in strings no longer than 65,535 characters.
CSTRCOMISS, invalid constructor: component(s) missing Explanation: The constructor did not specify sufcient component values to initialize a variable of the type. User Action: Specify more components in the constructor, use the OTHERWISE clause in the constructor, or modify the type denition to specify fewer components. CURCOMUND, current component is undened for DELETE or UPDATE Explanation: You attempted a DELETE or UPDATE procedure when no current component was dened. A current component is dened by a successful GET, FIND, FINDK, RESET, or RESETK that locks the component. Files opened with HISTORY:=READONLY never lock components. User Action: Correct the program so that a current component is dened before executing DELETE or UPDATE. DELNOTALL, DELETE is not allowed for a sequential organization le Explanation: You attempted a DELETE procedure for a le with sequential organization, which is not allowed. DELETE is valid only on les with relative or indexed organization. User Action: Make sure that the program is referencing the correct le. See Chapter 7 to determine what le characteristics are appropriate for your application. ERRDURCLO, error during CLOSE Explanation: RMS reported an unexpected error during execution of the CLOSE procedure. The RMS error message is also displayed. This message may also be issued with error severity when les are implicitly closed during a procedure or image exit. User Action: See the OpenVMS System Messages and Recovery Procedures Reference Manual for the description of the RMS error. ERRDURDEL, error during DELETE Explanation: RMS reported an unexpected error during execution of a DELETE procedure. The RMS error message is also displayed. User Action: See the OpenVMS System Messages and Recovery Procedures Reference Manual for the description of the RMS error.
ERRDURDIS, error during DISPOSE Explanation: An error occurred during execution of a DISPOSE procedure. An additional message that further describes the error may also be displayed. User Action: Make sure that the heap storage being freed was allocated by a successful call to the NEW procedure, and that it has not been already freed. If an additional message is shown, see the OpenVMS System Messages and Recovery Procedures Reference Manual for the description of that message. ERRDUREXT, error during EXTEND Explanation: RMS reported an unexpected error during execution of an EXTEND procedure. The RMS error message is also displayed. User Action: See the OpenVMS System Messages and Recovery Procedures Reference Manual for the description of the RMS error. ERRDURFIN, error during FIND or FINDK Explanation: RMS reported an unexpected error during execution of a FIND or FINDK procedure. The RMS error message is also displayed. User Action: See the OpenVMS System Messages and Recovery Procedures Reference Manual for the description of the RMS error. ERRDURGET, error during GET Explanation: RMS reported an unexpected error during execution of the GET procedure. The RMS error message is also displayed. User Action: See the OpenVMS System Messages and Recovery Procedures Reference Manual for the description of the RMS error. ERRDURMAR, error during MARK Explanation: An error occurred during execution of the PAS$MARK2 procedure. An additional message is displayed that further describes the error. User Action: See the OpenVMS System Messages and Recovery Procedures Reference Manual for a description of the additional message. ERRDURNEW, error during NEW Explanation: An error occurred during execution of the NEW procedure. An additional message is displayed that further describes the error. User Action: See the OpenVMS System Messages and Recovery Procedures Reference Manual for a description of the additional message.
ERRDUROPE, error during OPEN Explanation: An unexpected error occurred during execution of the OPEN procedure, or during an implicit open caused by a RESET or REWRITE procedure. An additional message is displayed that further describes the error. User Action: See the OpenVMS System Messages and Recovery Procedures Reference Manual for a description of the additional message. ERRDURPRO, error during prompting Explanation: RMS reported an unexpected error during output of partial lines to a terminal. The RMS error message is also displayed. User Action: See the OpenVMS System Messages and Recovery Procedures Reference Manual for a description of the RMS error. ERRDURPUT, error during PUT Explanation: RMS reported an unexpected error during execution of the PUT procedure. The RMS error message is also displayed. User Action: See the OpenVMS System Messages and Recovery Procedures Reference Manual for a description of the RMS message. ERRDURREL, error during RELEASE Explanation: An unexpected error occurred during execution of the PAS$RELEASE2 procedure. An additional message may be displayed that further describes the error. User Action: Make sure that the marker argument was returned from a successful call to PAS$MARK2 and that the storage has not been already freed. If an additional message is displayed, see the OpenVMS System Messages and Recovery Procedures Reference Manual for a description of that message. ERRDURRES, error during RESET or RESETK Explanation: RMS reported an unexpected error during execution of the RESET or RESETK procedure. The RMS error message is also displayed. User Action: See the OpenVMS System Messages and Recovery Procedures Reference Manual for a description of the RMS error. ERRDURREW, error during REWRITE Explanation: RMS reported an unexpected error during execution of the REWRITE procedure. The RMS error message is also displayed. User Action: See the OpenVMS System Messages and Recovery Procedures Reference Manual for a description of the RMS error.
ERRDURTRU, error during TRUNCATE Explanation: RMS reported an unexpected error during execution of the TRUNCATE procedure. The RMS error message is also displayed. User Action: See the OpenVMS System Messages and Recovery Procedures Reference Manual for a description of the RMS error. ERRDURUNL, error during UNLOCK Explanation: RMS reported an unexpected error during execution of the UNLOCK procedure. The RMS error message is also displayed. User Action: See the OpenVMS System Messages and Recovery Procedures Reference Manual for a description of the RMS error. ERRDURUPD, error during UPDATE Explanation: RMS reported an unexpected error during execution of the UPDATE procedure. The RMS error message is also displayed. User Action: See the OpenVMS System Messages and Recovery Procedures Reference Manual for a description of the RMS error. ERRDURWRI, error during WRITELN Explanation: RMS reported an unexpected error during execution of the WRITELN procedure. The RMS error message is also displayed. User Action: See the OpenVMS System Messages and Recovery Procedures Reference Manual for a description of the RMS error. EXTNOTALL, EXTEND is not allowed for a shared le Explanation: Your program attempted an EXTEND procedure for a le for which the program did not have exclusive access. EXTEND requires that no other users be allowed to access the le. Note that this message may also be issued if you do not have permission to extend to the le. User Action: Correct the program so that the le is opened with SHARING:=NONE, which is the default, before performing an EXTEND procedure. FAIGETLOC, failed to GET locked component Explanation: Your program attempted to access a component of a le that was locked by another user. You can usually expect this condition to occur when more than one user is accessing the same relative or indexed le. User Action: Determine whether this condition should be allowed to occur. If so, modify your program so that it detects the condition and retries the operation later. See Chapter 7 for more information.
FILALRACT, le le name is already active Explanation: Your program attempted a le operation on a le for which another operation was still in progress. This error can occur if a le is used in AST or condition-handling routines. User Action: Modify your program so that it does not try to use les that may currently be in use. FILALRCLO, le is already closed Explanation: Your program attempted to close a le that was already closed. User Action: Modify your program so that it does not try to close les that are not open. FILALROPE, le is already open Explanation: Your program attempted to open a le that was already open. User Action: Modify your program so that it does not try to open les that are already open. FILNAMREQ, FILE_NAME required for this HISTORY or DISPOSITION Explanation: Your program attempted to open a nonexternal le without specifying a le-name parameter to the OPEN procedure, but the HISTORY or DISPOSITION parameter specied requires a le name. User Action: Add a le-name parameter to the OPEN procedure call, specifying an appropriate le name. FILNOTDIR, le is not opened for direct access Explanation: Your program attempted to execute a DELETE, FIND, LOCATE, or UPDATE procedure on a le that was not opened for direct access. User Action: Modify the program to specify the ACCESS_ METHOD:=DIRECT parameter to the OPEN procedure when opening the le. See Chapter 7 to determine if direct access is appropriate for your application.
FILNOTFOU, le not found Explanation: Your program attempted to open a le that does not exist. An additional RMS message is displayed that further describes the problem. User Action: Make sure that you are specifying the correct le. See the OpenVMS System Messages and Recovery Procedures Reference Manual for a description of the additional RMS message. FILNOTGEN, le is not in Generation mode Explanation: Your program attempted a le operation that required the le to be in generation mode (ready for writing). User Action: Modify the program to use a REWRITE, TRUNCATE, or LOCATE procedure to place the le in generation mode as appropriate. See Chapter 7 for more information. FILNOTINS, le is not in Inspection mode Explanation: Your program attempted a le operation that required the le to be in inspection mode (ready for reading). User Action: Modify the program to use a RESET, RESETK, FIND, or FINDK procedure to place the le in inspection mode as appropriate. See Chapter 7 for more information. FILNOTKEY, le is not opened for keyed access Explanation: Your program attempted to execute a FINDK, RESETK, DELETE, or UPDATE procedure on a le that was not opened for keyed access. User Action: Modify the program to specify the ACCESS_ METHOD:=KEYED parameter to the OPEN procedure when opening the le. See Chapter 7 to make sure that keyed access is appropriate to your application. FILNOTOPE, le is not open Explanation: Your program attempted to execute a le manipulation procedure on a le that was not open. User Action: Correct the program to open the le using a RESET, REWRITE, or OPEN procedure as appropriate. See Chapter 7 for more information.
FILNOTSEQ, le is not sequential organization Explanation: Your program attempted to execute the TRUNCATE procedure on a le that does not have sequential organization. TRUNCATE is valid only on sequential les. User Action: Make sure that your program is accessing the correct le. Correct the program so that all TRUNCATE operations are performed on sequential les. FILNOTTEX, le is not a textle Explanation: Your program performed a le operation that required a le of type TEXT on a nontext le. Note that the type FILE OF CHAR is not equivalent to TEXT unless you have compiled the program with the /OLD_VERSION qualier. User Action: Make sure that your program is accessing the correct le. Correct the program so that a text le is always used when required. GENNOTALL, Generation mode is not allowed for a READONLY le Explanation: Your program attempted to place a le declared with the READONLY attribute into generation mode, which is not allowed. Note that the READONLY le attribute is not equivalent to the HISTORY:=READONLY parameter to the OPEN procedure. User Action: Correct the program so that the le either does not have the READONLY attribute or is not placed into generation mode. GETAFTEOF, GET attempted after end-of-le Explanation: Your program attempted a GET operation on a le while EOF(f) was TRUE. This situation occurs when a previous GET operation (possibly implicitly performed by a RESET, RESETK, or READ procedure) reads to the end of the le and causes the EOF(f) function to return TRUE. If another GET is then performed, this error is given. User Action: Correct the program so that it either tests whether EOF(f) is TRUE, before attempting a GET operation, or repositions the le before the end-of-le marker.
GOTOFAILED, non-local GOTO failed Explanation: An error occurred while a nonlocal GOTO statement was being executed. This error might occur because of an error in the user program, such as an out-of-bounds array reference. User Action: Rerun your program, enabling all CHECK options. If you cannot locate an error in your program and the problem persists, please submit a Software Performance Report (SPR) to Compaq, and include a machine-readable copy of your program, data, and results of a sample execution showing the problem. HALT, HALT procedure called Explanation: The program terminated its execution by executing the HALT procedure. This message is solely informational. User Action: None. ILLGOTO, illegal uplevel GOTO during routine activation Explanation: An uplevel GOTO was made into the body of a routine before the declaration part of the routine was completely processed. User Action: Correct the program to avoid the uplevel GOTO until the declaration part has been completely processed. INSNOTALL, Inspection mode is not allowed for a WRITEONLY le Explanation: Your program attempted to place a le declared with the WRITEONLY attribute into inspection mode, which is not allowed. User Action: Correct the program so that the le variable either does not have the WRITEONLY attribute or is not placed into inspection mode. INSVIRMEM, insufcient virtual memory Explanation: The run-time library was unable to allocate enough heap storage to open the le. User Action: Examine your program to see whether it is making excessive use of heap storage, which might be allocated using the NEW procedure or the run-time library procedure LIB$GET_VM. Modify your program to free any heap storage it does not need.
INVARGPAS, invalid argument to Pascal Run-Time Library Explanation: An invalid argument or inconsistent data structure was passed to the run-time library by the compiled code, or a system service returned an unrecognized value to the run-time library. User Action: Rerun your program with all CHECK options enabled. Make sure that the version of the current operating system is compatible with the version of the compiler. If you cannot locate an error in your program and the problem persists, please submit a Software Performance Report (SPR) to Compaq, and include a machine-readable copy of your program, data, and results of a sample execution showing the problem. INVFILSYN, invalid le name syntax Explanation: Your program attempted to open a le with an invalid le name. The le name used can be derived from the le variable name, the value of the le-name parameter to the OPEN procedure, or the logical name translations (if any) of the le variable name and portions of the le-name parameter and your default device and directory. The displayed text may include the erroneous le name. This error can also occur if the value of the le-name parameter is longer than 255 characters. Additional RMS messages may be displayed that further describe the error. User Action: Use the information provided in the displayed messages to determine which component of the le name is invalid. Verify that any logical names used are dened correctly. See the Compaq Pascal Language Reference Manual for information on le names. INVFILVAR, invalid le variable at location nnn Explanation: The le variable passed to a run-time library procedure was invalid or corrupted. This problem might be caused by an error in the user program, such as an out-of-bounds array access. It can also occur if a le variable is passed from a routine compiled with a version of VAX Pascal earlier than Version 2.0 to a routine compiled with a later version of the compiler, or if the new key options are used on OpenVMS systems earlier than Version 4.6. User Action: Rerun your program with all CHECK options enabled, and recompile all modules using the same compiler. If the problem persists, please submit a Software Performance Report (SPR) to Compaq and include a machine-readable copy of your program, data, and results of a sample execution showing the problem.
INVKEYDEF, invalid key denition Explanation: Your program attempted to open a le of type RECORD whose component type contained a eld with an invalid KEY attribute. One of the following errors occurred: A new le was being created and the key numbers were not dense. A key eld was dened at an offset of more than 65,535 bytes from the beginning of the record.
User Action: If a new le is being created, make sure that the key elds are numbered consecutively, starting with 0 for the required primary key. If you are opening an existing le, you must explicitly specify HISTORY:=OLD or HISTORY:=READONLY as a parameter to the OPEN procedure. Make sure that the length of the record is within the maximum permitted for the le organization being used. See Chapter 7 for more information. INVRADIX, specied radix must be in the range 2-36 Explanation: The specied radix for writing an ordinal value must be in the range of 2 through 36. User Action: Modify the program to specify a radix in the proper range INVRECLEN, invalid record length of nnn Explanation: A le was being opened, and one of the following errors occurred: The length of the le components was greater than that allowed for the le organization and record format (for most operations, the largest length allowed is 32,765 bytes). The value of the RECORD_LENGTH parameter to the OPEN procedure was greater than that allowed for the le organization and record format (for most operations, the largest value allowed is 32,765 bytes).
User Action: Correct the program so that the record length used is within the permitted limits for the type of le being used. See the OpenVMS Record Management Services Reference Manual for more information.
INVSYNBIN, string is invalid syntax for a binary value Explanation: While a READ or READV procedure was reading a binary value from a text le, the characters read did not conform to the syntax for a binary value. The displayed message includes the text actually read and the record number in which this text occurred. User Action: Correct the program or the input data so that the correct syntax is used. See the Compaq Pascal Language Reference Manual for more information. INVSYNHEX, string is invalid syntax for a hexadecimal value Explanation: While a READ or READV procedure was reading a hexadecimal value from a text le, the characters read did not conform to the syntax for an hexadecimal value. The displayed message includes the text actually read and the record number in which this text occurred. User Action: Correct the program or the input data so that the correct syntax is used. See the Compaq Pascal Language Reference Manual for more information. INVSYNENU, string is invalid syntax for an enumerated value Explanation: While a READ or READV procedure was reading an identier of an enumerated type from a text le, the characters read did not conform to the syntax for an enumerated value. The displayed message includes the text actually read and the record number in which this text occurred. User Action: Correct the program or the input data so that the correct syntax is used. See the Compaq Pascal Language Reference Manual for more information. INVSYNINT, string is invalid syntax for an integer value Explanation: While a READ or READV procedure was reading a value for an integer identier from a text le, the characters read did not conform to the syntax for an integer value. The displayed message includes the text actually read and the record number in which this text occurred. User Action: Correct the program or the input data so that the correct syntax is used. See the Compaq Pascal Language Reference Manual for more information.
INVSYNOCT, string is invalid syntax for an octal value Explanation: While a READ or READV procedure was reading an octal value from a text le, the characters read did not conform to the syntax for an octal value. The displayed message includes the text actually read and the record number in which this text occurred. User Action: Correct the program or the input data so that the correct syntax is used. See the Compaq Pascal Language Reference Manual for more information. INVSYNREA, string is invalid syntax for a real value Explanation: While a READ or READV procedure was reading a value for a real identier from a text le, the characters read did not conform to the syntax for a real value. The displayed message includes the text actually read and the record number in which this text occurred. User Action: Correct the program or the input data so that the correct syntax is used. See the Compaq Pascal Language Reference Manual for more information. INVSYNUNS, string is invalid syntax for an unsigned value Explanation: While a READ or READV procedure was reading a value for an unsigned identier from a text le, the characters read did not conform to the syntax for an unsigned value. The displayed message includes the text actually read and the record number in which this text occurred. User Action: Correct the program or the input data so that the correct syntax is used. See the Compaq Pascal Language Reference Manual for more information. KEYCHANOT, key eld change is not allowed Explanation: Your program attempted an UPDATE procedure for a record of an indexed le that would have changed the value of a key eld, and this situation was disallowed when the le was created. User Action: If the program needs to detect this situation when it occurs, specify the ERROR:=CONTINUE parameter for the UPDATE procedure, and use the STATUS function to determine which error, if any, occurred. If necessary, modify the program so that it does not improperly change a key eld, or recreate the le specifying that the key eld is permitted to change. See Chapter 7 for more information.
KEYDEFINC, KEY nnn denition is inconsistent with this le Explanation: An indexed le of type RECORD was opened, and the component type contained elds whose KEY attributes did not match those of the existing le. The number of the key in error is displayed in the message. User Action: Correct the RECORD denition so that it describes the correct KEY elds, or recreate the le so that it matches the declared keys. See Chapter 7 for more information. KEYDUPNOT, key eld duplication is not allowed Explanation: Your program attempted an UPDATE or PUT procedure for a record of an indexed le that would have duplicated a key eld value of an existing record, and this situation was disallowed when the le was created. User Action: If the program needs to detect this situation when it occurs, specify the ERROR:=CONTINUE parameter for the PUT or UPDATE procedure, and use the STATUS function to determine which error, if any, occurred. If necessary, modify the program so that it does not improperly duplicate a key eld, or recreate the le specifying that the key eld is permitted to be duplicated. See Chapter 7 for more information. KEYNOTDEF, KEY nnn is not dened for this le Explanation: Your program attempted a FINDK or RESETK procedure on an indexed le, and the key number specied does not exist in the le. User Action: Correct the program so that the correct key numbers are used when accessing the le. KEYVALINC, key value is incompatible with the les key nnn Explanation: The key value specied for the FINDK procedure was incompatible in type or size with the key eld of the le, or your program attempted an OPEN on an existing le and the key check failed. User Action: Make sure that the correct key value is being specied for FINDK and OPEN. Correct the program so that the type of the key value is compatible with the key of the le.
LINTOOLON, line is too long, exceeded record length by nnn character(s) Explanation: Your program attempted a WRITE, PUT, WRITEV, or other output procedure on a text le that would have placed more characters in the current line than the record length of the le would allow. The number of characters that did not t is displayed in the message. User Action: Correct the program so that it does not place too many characters in the current line. If appropriate, use the WRITELN procedure, or specify an increased record length parameter when opening the le with the OPEN procedure. LINVALEXC, LINELIMIT value exceeded Explanation: The number of lines written to the le exceeded the maximum specied as the line limit. The line limit value is determined by the translation of the logical name PAS$LINELIMIT, if any, or the value specied in a call to the LINELIMIT procedure for the le. User Action: As appropriate, correct the program so that it does not write as many lines, or increase the line limit for the le. Note that if a line limit is specied for a nontext le, each PUT procedure called for the le is considered to be one line. See the Compaq Pascal Language Reference Manual for more information. LOWGTRHIGH, low-bound exceeds high-bound Explanation: The lower bound of a subrange denition is larger than the higher bound. User Action: Modify the declaration so the lower bound is less than or equal to the higher bound. MAXLENRNG, maximum length must be in range 1..65535 Explanation: The maximum length for a string type is 65,535. User Action: Modify the declaration to specify a smaller amount. MODNEGNUM, MOD of a negative modulus has no mathematical denition Explanation: In the MOD operation A MOD B, the operand B must have a positive integer value. User Action: Correct the program so that the operand B has a positive integer value.
NEGDIGARG, negative Digits argument to BIN, HEX or OCT is not allowed Explanation: Your program attempted to specify a negative value for the Digits argument in a call to the BIN, HEX, or OCT procedure, which is not permitted. User Action: Correct the program so that only nonnegative Digits arguments are used for calls to BIN, HEX, and OCT. NEGWIDDIG, negative Width or Digits specication is not allowed Explanation: A WRITE or WRITEV procedure on a text le contained a eld width specication that included a negative Width or Digits value, which is not permitted. User Action: Correct the program so that only nonnegative Width and Digits parameters are used. NOTVALTYP, string is not a value of type type Explanation: Your program attempted a READ or READV procedure on a text le, but the value read could not be expressed in the specied type. For example, this error results if a real value read is outside the range of the identiers type, or if an enumerated value is read that does not match any of the valid constant identiers in its type. User Action: Correct the program or the input data so that the values read are compatible with the types of the identiers receiving the data. OPNDASSCOM, operands are not assignment compatible Explanation: The operands do not have the same type. User Action: Examine the declarations of the operands and make sure they have compatible types. ORDVALOUT, ordinal value is out of range Explanation: A value of an ordinal type is outside the range of values specied by the type. For example, this error results if you try to use the SUCC function on the last value in the type or the PRED function on the rst value. User Action: Correct the program so that all ordinal values are within the range of values specied by the ordinal type.
ORGSPEINC, ORGANIZATION specied is inconsistent with this le Explanation: The value of the ORGANIZATION parameter for the OPEN procedure that opened an existing le was inconsistent with the actual organization of the le. User Action: Correct the program so that the correct organization is specied. See Chapter 7 for more information. PADLENERR, PAD length error Explanation: The length of the character string to be padded by the PAD function is greater than the length specied as the nished size, or the nished size specied is greater than 65,535. User Action: Correct the call to PAD so that the nished size specied describes a character string of the correct length. See the Compaq Pascal Language Reference Manual for the rules governing the PAD function. PTRREFNIL, pointer reference to NIL Explanation: Your program attempted to evaluate a pointer value while its value was NIL. User Action: Make sure that the pointer has a value before you try to evaluate it. See the Compaq Pascal Language Reference Manual for more information on pointer values. RECLENINC, RECORD_LENGTH specied is inconsistent with this le Explanation: The record length obtained from the le components length or from the value of the record length parameter specied for the OPEN procedure was inconsistent with the actual record length of an existing le. User Action: Correct the program so that the record length specied, if any, is consistent with the le. See Chapter 7 for more information. RECTYPINC, RECORD_TYPE specied is inconsistent with this le Explanation: The value of the RECORD_LENGTH parameter specied for the OPEN procedure was inconsistent with the actual record type of an existing le. User Action: Correct the program so that the record type specied, if any, is consistent with the le. See Chapter 7 for more information. REFINAVAR, read or write of inactive variant Explanation: A eld of an inactive variant was read or written. User Action: Correct the program so the variant is active or remove the reference to the inactive eld.
RESNOTALL, RESET is not allowed on an unopened internal le Explanation: Your program attempted a RESET procedure for a nonexternal le that was not open. This operation is not permitted because RESET must operate on an existing le, and there is no information associated with a nonexternal le that allows RESET to open it. User Action: Correct the program so that nonexternal les are opened before using RESET. Either OPEN or REWRITE may be used to open a nonexternal le. See the Compaq Pascal Language Reference Manual for more information. REWNOTALL, REWRITE is not allowed for a shared le Explanation: Your program attempted a REWRITE procedure for a le for which the program did not have exclusive access. REWRITE requires that no other users be allowed to access the le while the les data is deleted. Note that this message may also be issued if you do not have permission to write to the le. User Action: Correct the program so that the le is opened with SHARING := NONE, which is the default, before performing a REWRITE procedure. SETASGVAL, set assignment value has element out of range Explanation: Your program attempted to assign to a set variable a value that is outside the range specied by the variables component type. User Action: Correct the assignment statement so that the value being assigned falls within the component type of the set variable. See the Compaq Pascal Language Reference Manual for more information on sets. SETCONVAL, set constructor value out of range Explanation: Your program attempted to include in a set constructor a value that is outside the range specied by the sets component type, or a value that is greater than 255 or less than 0. User Action: Correct the constructor so that it includes only those values within the range of the sets component type. See the Compaq Pascal Language Reference Manual for more information on sets. SETNOTRNG, set element is not in range 0..255 Explanation: Sets of INTEGER or UNSIGNED must be in the range of 0..255. User Action: Modify the declaration to specify a smaller range.
STRASGLEN, string assignment length error Explanation: Your program attempted to assign to a string variable a character string that is longer than the declared maximum length of the variable (if the variables type is VARYING) or that is not of the same length as the variable (if the variables type is PACKED ARRAY OF CHAR). User Action: Correct the program so that the string is of a correct length for the variable to which it is being assigned. STRCOMLEN, string comparison length error Explanation: Your program attempted to compare two character strings that do not have the same current length. User Action: Correct the program so that the two strings have the same length at the time of the comparison. SUBASGVAL, subrange assignment value out of range Explanation: Your program attempted to assign to a subrange variable a value that is not contained in the subrange type. User Action: Correct the program so that all values assigned to a subrange variable fall within the variables type. SUBSTRSEL, SUBSTR selection error Explanation: A SUBSTR function attempted to extract a substring that was not entirely contained in the original string. User Action: Correct the call to SUBSTR so that it species a substring that can be extracted from the original string. See the Compaq Pascal Language Reference Manual for complete information on the SUBSTR function. TEXREQSEQ, textles require sequential organization and access Explanation: Your program attempted to open a le of type TEXT that either did not have sequential organization, or had an ACCESS_METHOD other than SEQUENTIAL (the default) when opened by the OPEN procedure. User Action: Make sure that the program refers to the correct le. Correct the program so that only sequential organization and access are used for text les.
TRUNOTALL, TRUNCATE is not allowed for a shared le Explanation: Your program attempted to call the TRUNCATE procedure for a le that was opened for shared access. You cannot truncate les that might be shared by other users. This message may also be issued if you do not have permission to write to the le. User Action: Correct the program so that it does not try to truncate shared les. If the le is opened with the OPEN procedure, do not specify a value other than NONE (the default) for the SHARING parameter. UPDNOTALL, UPDATE not allowed for a sequential organization le Explanation: Your program attempted to call the UPDATE procedure for a sequential le. UPDATE is valid only on relative and indexed les. User Action: Correct the program so that it does not try to use UPDATE for sequential les, or recreate the le with relative or indexed organization. If you are using direct access on a sequential le, individual records can be updated with the LOCATE and PUT procedures. See Chapter 7 to determine whether a different le organization may be appropriate for your application. VARINDVAL, VARYING index value exceeds current length Explanation: The index value specied for a VARYING OF CHAR string is greater than the strings current length. User Action: Correct the index value so that it species a legal character in the string. WIDTOOLRG, totalwidth too large Explanation: The requested total-width for the oating point write operation overowed an internal buffer. User Action: Examine the source program to see if the specied totalwidth parameter is correct. If it is correct, please submit a problem report to Digital including a machine-readable copy of your program, data, and a samle execution illustrating the problem. WRIINVENU, WRITE of an invalid enumerated value Explanation: Your program attempted to write an enumerated value using a WRITE or WRITEV procedure, but the internal representation of that value was outside the possible range for the enumerated type. User Action: Verify that your program is not improperly using PRED, SUCC, or type casting to assign an invalid value to a variable of enumerated type.
Index
A
Access method type translations, 67 Action routine using the UNBOUND attribute with, 615 ADDRESS function effect on optimization, 317 /ALIGN qualier with PASCAL command, 13 Alpha emulation, 16 /ANALYSIS_DATA qualier, 14 /ARCHITECTURE qualier, 14 Argument pointer saved by routine call, 54 Assignment with unsupported CDD/Repository data type, 410 AST ASTADR parameter, 617 ASTPRM parameter, 617 using the UNBOUND attribute with, 615 AST routine attributes required for, 55 ASTADR parameter, 617 ASTPRM parameter, 617 ASYNCHRONOUS attribute, 615 in condition handler, 84 use with system routines, 66 Asynchronous system trap routine See AST routine Atomicity, 318 Attributes ASYNCHRONOUS, 615 BYTE use with CDD/Repository, 410 CHECK, C25 ENVIRONMENT, 22 FLOAT, C25 for by descriptor passing mechanism, 58 HIDDEN, 25 INHERIT, 22 KEY, C25 OPTIMIZE, C25 UNBOUND attribute, 615 used during OpenVMS programming, 614 VOLATILE, 614 Automatic variable in debugging, 42
B
Block RMS control, 71 Bound procedure values, 92 Branch to longword-aligned address, 311 /BRIEF qualier with /MAP on LINK command, 132 BYTE attribute use with CDD/Repository, 410
Index1
C
Call stack (OpenVMS Alpha systems only) contents of, 53 Call stack (OpenVMS VAX systems only) contents of, 54 Calling block function return value to, 52 Calling standard, 51 CASE statement effect on efciency, 314 run-time checking of, 16 CDD/Repository accessing from source program, 49 Compaq Pascal support of, 48 creating directory hierarchies, 48, 49 denition of, 48 entering denitions, 49 equivalent Compaq Pascal data types, 410 to 412 example of use, 412 CDDL, 49 Character string as function results, 52 run-time checking of, 16 CHECK attribute, C25 effect on efciency, 315 /CHECK qualier, 16 CHF, 81 CLASS_A attribute, 512 CLASS_NCA attribute, 512 CLASS_S attribute, 511 CLASS_VS descriptor used with LIB$FIND_ FILE, 618 CLOSE procedure with user-action parameters, 710 Code scheduling, 310 Comment processing (OpenVMS Alpha systems only), 43 Common Data Dictionary See CDD/Repository Common Data Dictionary Language See CDDL
Compilation statistics in listing le, 126 Compilation switch See Compile-time qualiers Compilation unit, 21 Compile-time qualiers, 13 to 123 Compiler command qualiers, 13 to 123 diagnostics, C1 generated labels, 311 invoking, 11 Compiling for optimal performance, 313 Condition handler controlling execution, 84 declaring parameters for, 86 denition of, 81 establishing, 84 examples of, 811 for faults, 810 for traps, 810 overview of, 83 performing I/O to and from, 85 removing, 85 reporting conditions, 84 return value of, 87 system-dened, 83 writing of, 84 to 816 Condition handlers returning SS$_ CONTINUE, 88 Condition Handling Facility See CHF Condition signal, 83 Condition value, 88 denition of, 82 for faults, 810 for traps, 810 matching, 88 severity code of, 88 Conformant array parameters different descriptor classes, 93 /CONSTANT qualier, 18
Index2
Constants compile-time evaluation of, 33 effect on efciency, 314 Constructor for schema variant record (example), 213 using OTHERWISE (example), 22 Conversion of constants, 33 Cross-reference section in listing le, 125 /CROSS_REFERENCE qualier, 19 with /MAP on LINK command, 132
D
Data conversion, 93 Data layout, 93 Data structure parameters, 610 Data type initial-state specier for, 213 Data types Compaq Pascal and CDDL equivalent, 410 D_oating restrictions, 410 G_oating restrictions, 410 /DEBUG qualier use with /NOOPTIMIZE qualier, 110 with LINK command, 131 with PASCAL command, 19 with RUN command, 135 Debugger Compaq Pascal support of assignment compatibility, 41 automatic variables, 42 examining LENGTH eld, 42 type cast operator, 42 unreferenced variables, 41 Debugging effects of optimization, 320 Declaration section multiply declared identiers, 24
Declaration sharing, 48 to 413 Default parameters in system services, 611 DEFINE command use with text library, 129 Denition le for system routines, 61 Denition sharing, 48 to 413 Dependency checking environment le, 25 %DESCR mechanism specier, 513 %DESCR mechanism used with LIB$FIND_ FILE, 618 Descriptor parameter, 58 /DESIGN qualier, 110 /DIAGNOSTICS qualier, 111 %DICTIONARY directive use with CDD/Repository, 49 example of, 412 Dictionary Management Utility See DMU Directives %DICTIONARY using to access CDD/Repository denitions, 48 DMU, 48
E
Emulation of Alpha instructions, 16 Enumerated type used in graphical data model (example), 212 /ENUMERATION_SIZE qualier, 111 Environment le cascading inheritance of (gure), 23 cascading interfaces (gure), 210 creating, 112, 22 linking, 130 Environment le dependency checking, 25 /ENVIRONMENT qualier, 112 Error conditions detected by STATUS and STATUSV, A1 to A4
Index3
Error messages compiler, C1 to C65 run-time, C65 to C86 syntax of, 136 /ERROR_LIMIT qualier, 112 ESTABLISH procedure, 84 Establisher routine denition of, 82 Evaluation order of operands effect of /NOOPTIMIZE, 117 EXAMINE command (debugger) automatic variables, 42 use with LENGTH eld, 42 Examples of an implementation module, 217 of an interface module, 216 of separate compilation, 216 Exception condition, 81 /EXECUTABLE qualier with LINK command, 132 EXPAND command (LSE), 43 Extended attribute block See XAB Extensions to standard Pascal detecting, 119 EXTERNAL attribute compared with UNBOUND, 55 effect on routine call, 55 External routine declaration example of, 64
Field initial-state specier for, 212 File locked component in, 716 sharing, 715 File access block See FAB File component locked, 716 File type of environment le, 112 of listing le, 114 of object le, 115 FLOAT attribute, C25 /FLOAT qualier, 112 FOR statement effect on efciency, 314 Foreign mechanism parameters, 514 Foreign mechanism specier, 617 Frame pointer of unbound routine, 54 saved by routine call, 54 /FULL qualier with /MAP on LINK command, 132 Function methods of returning result, 52 optimization of, 36 user-action, 710
F
FAB access to, B1 elds when calling OPEN, 75 passed as parameter by run-time library, 710 used to write user-action functions, 71 Fault condition handling for, 810 converting to trap, 811 Fetch with unsupported CDD/Repository data type, 410
G
GLOBAL attribute compared with UNBOUND, 55 effect on routine call, 55 GOTO statement effect on efciency, 314 Granularity, 318 /GRANULARITY qualier, 113
Index4
H
HELP LANGUAGE command (debugger), 41 HIDDEN attribute, 25 HISTORY parameter use when accessing les, 715
I
I/O processing, 71 to 717 IADDRESS function effect on optimization, 317 use with VOLATILE and item lists, 614 Identier multiply declared, 24 IF-THEN-ELSE statement effect on efciency, 314 %IMMED mechanism specier, 57 on default parameter, 611 %IMMED specier on actual parameters, 617 Implementation module, 28 See also Separate compilation example of, 217 Implementation-Dependent Behavior, 315 %INCLUDE directive use with text library, 127 /INCLUDE qualier, 114 with LINK command, 132 Indexed le, 72 locked component in, 716 Innite loop See Loop Information-level error message effect of /ERROR_LIMIT, 112 for extensions, 119 INHERIT attribute, 22 Initial-state specier for array variable (example), 22 on a data type, 213 on variant record elds (example), 212 variant record constructor (example), 213
INITIALIZE attribute compared with UNBOUND, 55 effect on routine call, 55 Inline summary in listing le, 126 Integer overow run-time checking of, 16, 315 Interface module, 28 See also Separate compilation example of, 216 Item list, 615 using the SIZE function, 616 using the VOLATILE attribute with, 614
K
KEY attribute, C25 Keywords use with LSE, 43
L
Language expression optimization of, 38, 316 order of evaluation, 38 reordering of, 32 Language-Sensitive Editor/Source Code Analyzer See LSE, LSE/SCA, 32 LIB$ESTABLISH routine establishing condition handler with, 85 LIB$FIND_FILE routine Resultant_Filespec parameter, 618 LIB$MATCH_COND function, 88 LIB$SIGNAL procedure signaling condition with, 83 LIB$SIM_TRAP procedure, 811 LIB$STOP procedure signaling condition with, 83 LIBDEF.PAS denition le, 63 Library object module, 134 of shareable images, 133 text, 127
Index5
/LIBRARY qualier, 113 with LINK command, 132 LINELIMIT procedure, 710 LINK command, 22 examples of, 130 qualiers for, 131 to 133 relinking implementation modules, 29 syntax of, 129 TO BEGIN DO execution order, 215 Linker including shareable image as input to, 133 LIST attribute with run-time library routine, 613 /LIST qualier, 114 use with %DICTIONARY directive, 49 Listing le compilation statistics, 126 cross-reference section, 125 inline summary, 126 machine code section, 125 OpenVMS VAX systems, 124 printing of, 124 source code, 125 table of contents, 125 Listing section structured layout, 126 Locked record, 716 in indexed le, 716 in relative le, 716 unlocking, 716 Logical expression optimization of, 38 LONG on pointer types, 621 Loop See Innite Loop Loop unrolling, 310 LSE comment processing example of, 46 Compaq Pascal support of for keywords or tokens, 43 placeholder processing, 43 example of use within program, 45 to 48
LSE (contd) placeholder processing example of, 46 LSE/SCA support of Compaq Pascal, 42 to 44
M
Machine code section in listing le, 125 /MACHINE_CODE qualier, 114 Map le, 132 /MAP qualier with LINK command, 132 /MATH_LIBRARY qualier, 114 MAXINT predeclared constant, 316 MAXINT64 predeclared constant, 316 Mechanism array, 86 Mechanism specier %DESCR, 513 %STDESCR, 512 Messages compiler, C1 to C65 run-time, C65 to C86 syntax of, 136 Methods to obtain access methods, 67 to obtain OpenVMS data types, 67 Migrate bound procedure values, 92 default data layout, 92 INT, 92 Migrating Compaq Pascal programs OpenVMS VAX to OpenVMS Alpha, 91 Module, 21 nalization section, 215 implementations and interfaces, 28 restrictions, 211 initialization and nalization sections restrictions, 211 initialization section, 215 interface inheritance path (gure), 28 multiply declared identiers, 24 relinking implementation modules, 29 requirements for linking modules, 23 TO BEGIN DO section, 212
Index6
MTHDEF.PAS denition le, 63 /MULTILANGUAGE qualier, 117
N
Name block elds when calling OPEN, 79 Nonstatic type restriction in modules, 211 /NOOPTIMIZE qualier, 117 effect on debugging, 110
/OPTIMIZE qualier, 115 /OPTIONS qualier with LINK command, 133 OTHERWISE reserved word initializing an array (example), 22 Overow run-time checking of, 16 Overow checking effect on efciency, 315 Overview of record layout, 93
O
Object le, 115 Object module library, 134 symbol table in, 134 /OBJECT qualier, 115 OPEN procedure default for VMS les, 74 related to RMS data structures, 75 to 710 with user-action parameters, 710 OpenVMS Calling Standard, 51 OpenVMS programming attributes used during, 614 Operation, optimization of, 37 Operator type cast use with CDD/Repository, 410 use with debugger, 42 Optimal performance, 313 compiling for, 313 Optimization, 115, 31 code scheduling, 310 denition of, 31 effect on debugging, 320 kinds of, 32 loop unrolling, 310 reducing errors through, 311 split lifetime analysis, 39 strength reduction, 39 OPTIMIZE attribute, C25
P
PACKED ARRAY OF CHAR type as function result type, 53 Parameter ASTADR parameter, 617 ASTPRM parameter, 617 data structure, 610 default value for, 611 descriptors, 58 for condition handler, 86 of LIB$FIND_FILE, 618 of SYS$QIO, SYS$QIOW, and SYS$FAO, 617 passing to run-time library routines, 613 passing to system services, 69 using foreign mechanisms on actual parameters, 617 Parameter list, 52 of arbitrary length, 613 Parameter-passing semantics, 55 Parentheses effect on efciency, 314 PAS$FAB function, B1 PAS$LINELIMIT logical name, 710 PAS$MARK2 function, B2 PAS$RAB function, B2 PAS$RELEASE2 function, B3 PASCAL command, 22 examples of, 13 qualiers with, 13 to 123 specifying text libraries in, 128 syntax of, 11
Index7
PASDEF.PAS denition le, 63 Passing mechanisms, 55 by descriptor, 57 by immediate value, 57 by reference, 57 foreign, 514 summary of, 514 type translations, 68 Passing semantics summary of, 514 PC saved by routine call, 54 Performance optimal, 313 Pipelining software, 312 Placeholder processing, 43 /PLATFORMS qualier, 117 Pointer run-time checking of, 16 Pointer types 64-bit, 618 POS attribute with data structure parameter, 610 Predeclared function optimization of, 36 Predeclared routines I/O processing, 71 to 717 Procedure calling standard, 51 Processor tuning, 313 Program See also Separate compilation compiling, 11 linking, 129 modularity of, 21 to 219 Program counter See PC Program exit status denition of, 82 Propagation value, 38
Q
QUAD on pointer types, 621 Qualiers /ALIGN, 13 /ANALYSIS_DATA, 14 /ARCHITECTURE, 14 /BRIEF, 132 /CHECK, 16 /CONSTANT, 18 /CROSS_REFERENCE, 19 /DEBUG, 19, 131 /DESIGN, 110 /DIAGNOSTICS, 111 /ENUMERATION_SIZE, 111 /ENVIRONMENT, 112 /ERROR_LIMIT, 112 /EXECUTABLE, 132 /FLOAT, 112 /FULL, 132 /GRANULARITY, 113 /INCLUDE, 114, 132 /LIBRARY, 113, 132 /LIST, 114 /MACHINE_CODE, 114 /MAP, 132 /MATH_LIBRARY, 114 /MULTILANGUAGE, 117 /OBJECT, 115 /OPTIMIZE, 115 /OPTIONS, 133 /PLATFORMS, 117 /SHAREABLE, 133 /SHOW, 118 /STANDARD, 119 /SYNCHRONOUS_EXCEPTIONS, 120 /TERMINAL, 120 /TIE, 120 /TRACEBACK, 133 /USAGE, 121 /WARNINGS, 123 with LINK command, 131 to 133 with PASCAL command, 13 to 123 with RUN command, 135
Index8
Qualiers (contd) /ZERO_HEAP, 123
R
RAB access to, B2 elds when calling OPEN, 77 passed as parameter by run-time library, 710 used to write user-action functions, 71 Read sharing, 715 READONLY attribute on pointer variables, 317 with system services, 69 Record efcient use of nonstatic elds, 315 locking, 716 variant (example), 212 Record access block See RAB Record le address See RFA Record layout overview, 93 Record Management Services See RMS %REF specier on actual parameters, 617 Register assignment of variables to, 32 contents saved by routine call, 54 effects of optimization, 320 Relative le locked component in, 716 REPEAT statement effect on efciency, 314 Resignal denition of, 82 Resultant_Filespec parameter of LIB$FIND_ FILE, 618 Return value function, 52 REVERT procedure, 85
RFA random access by, 73 RMS le sharing capability, 715 locking components with, 716 used to perform I/O tasks, 71 to 717 Routine I/O processing, 71 to 717 requirements for linking modules, 23 using foreign mechanisms on actual parameters, 617 Routine activation denition of, 82 RTL opening and closing a le, 710 symbol denitions for, 63 using LIST attribute with, 613 RUN command, 22 examples of, 135 qualier for, 135 syntax of, 135 Run-time error detected by STATUS or STATUSV, A1 messages, C65 to C86 Run-time library See RTL
S
SAVE FILE command (LSE), 43 SCA example of use within program, 45 to 48 support of Compaq Pascal, 42 to 48 Schema type efcient use of nonstatic elds, 315 of variant record (example), 212 Semantics of parameter passing, 55, 514 variable with system services, 69 Separate compilation, 21 to 219 cascading inheritance (gure), 23 cascading interfaces (gure), 210 examples, 216 interface inheritance path (gure), 28
Index9
Separate compilation (contd) multiply declared identiers, 24 relinking implementation modules, 29 TO BEGIN DO section, 212 SET type run-time checking of, 16 Shareable image, 133 library, 133 /SHAREABLE qualier with LINK command, 133 Sharing of les, 715 SHOW PLACEHOLDER command (LSE), 43 /SHOW qualier, 118 SHOW TOKEN command (LSE), 43 SIGDEF.PAS denition le, 63 Signal denition of, 82 Signal array, 86 in condition handlers, 86 Size attributes with data structure parameter, 610 SIZE function used with item lists, 616 Software pipelining, 312 Source code in listing le, 125 Source Code Analyzer See SCA Split lifetime analysis, 39 SS$_CONTINUE return value returned by condition handler, 87 SS$_RESIGNAL return value returned by condition handler, 87 Stack frame denition of, 82 /STANDARD qualier, 119 STARLET le contents of, 61 STATUS function conditions detected by, A1 to A4 STATUSV function conditions detected by, A1 to A4
%STDESCR mechanism specier, 512 Strength reduction, 39 STRING type as function result type, 53 Structure layout, 126 Structured statement effect on efciency, 314 optimization of, 35 Subexpression optimization of, 34 Subrange run-time checking of, 16 Symbol table in object module library, 134 /SYNCHRONOUS_EXCEPTIONS qualier, 120 SYS$FAO routine P1..P20 parameters, 617 SYS$LP_LINES logical name, 124 SYS$QIO routine P1..P6 parameters, 617 SYS$QIOW routine P1..P6 parameters, 617 System routines, 61 to 614 calling of, 613 data structure parameter, 610 declaring, 66 denition le for, 61 optional parameters for, 611 System services, 614
T
Table of contents in listing le, 125 Temporary variable effect on efciency, 315 /TERMINAL qualier, 120 Text library dening a default, 129 specifying with %INCLUDE directive, 127 specifying with PASCAL command, 128
Index10
/TIE qualier, 120 TO BEGIN DO section, 212, 215 Tokens use with LSE, 43 /TRACEBACK qualier with LINK command, 133 Trap condition handling for, 810 TRUNCATE attribute use with system routines, 612 Type cast operator use with CDD/Repository, 410 use with debugger, 42 Type conversion of constants, 33 Type translations access, 67 mechanism, 68
Variable (contd) types not checked for uninitialization, 122 Variable semantics with system services, 69 Variant record initial-state speciers for (example), 212 optimization considerations, 318 VARYING OF CHAR type as function result type, 53 used with LIB$FIND_FILE, 618 VMS programming using foreign mechanisms on actual parameters, 617 using item lists, 615 VOLATILE attribute, 614 on pointer variables, 317 Volatility, 318
U
UNBOUND attribute, 615 effect on routine call, 54 Uninitialized variables types not checked, 122 UNLOCK procedure, 716 Unwind denition of, 82 of stack by condition handler, 87 $UNWIND function called by condition handler, 87 /USAGE qualier, 121 User-action function example of, 712 parameters to, 710 to open a le, 710
W
Warning-level error messages, 123 effect of /ERROR_LIMIT, 112 /WARNINGS qualier, 123 WHILE statement effect on efciency, 314 WITH statement effect on efciency, 314 Write ordering, 318 Write sharing, 715
X
XAB elds when calling OPEN, 79 passed as parameter by run-time library, 710 used to write user-action functions, 71
V
Variable effect on efciency, 315 initialization of, 213 multiply declared identiers, 24 pointer, 317 requirements for linking modules, 23
Z
/ZERO_HEAP qualier, 123
Index11

Pascal Usermanual

Uploaded by

Pascal Usermanual

Uploaded by

Compaq Pascal User Manual for OpenVMS Systems

Revision/Update Information: Software Version:

This is an updated manual. Compaq Pascal Version 5.7

Compaq Computer Corporation Houston, Texas

ZK6140 This manual is available on CDROM.

This document was prepared using VAX DOCUMENT Version 2.1.

3 Program Correctness, Optimization, and Efciency

6 Programming on OpenVMS Systems

7 Input and Output Processing

8 Error Processing and Condition Handling

9 Migrating from OpenVMS VAX to OpenVMS Alpha

.... .... .... . . . . . . . . . . . . . . . . . . . . . . . .

A Errors Returned by STATUS and STATUSV Functions

B Entry Points to Compaq Pascal Utilities

Compaq Pascal Home Page

How to Order Additional Documentation

STATUS DISPOSE DISP

Square brackets enclose items that are optional. For example:

BLOCK DATA [nam]

Table 1 (Cont.) Conventions Used in This Manual

(OpenVMS Alpha systems only) (OpenVMS VAX systems only)

1.1 PASCAL Command

Compiling Compaq Pascal Programs 11

12 Compiling Compaq Pascal Programs

1.1.1 PASCAL Command Examples

1.1.2 PASCAL Qualiers

Compiling Compaq Pascal Programs 13

Default on OpenVMS VAX systems if /ALIGN is not specied.

14 Compiling Compaq Pascal Programs

Table 12 /ARCHITECTURE Qualier Options

Compiling Compaq Pascal Programs 15

Table 12 (Cont.) /ARCHITECTURE Qualier Options

16 Compiling Compaq Pascal Programs

Table 13 /CHECK Qualier Options

Compiling Compaq Pascal Programs 17

18 Compiling Compaq Pascal Programs

Compiling Compaq Pascal Programs 19

Table 14 /DEBUG Qualier Options

110 Compiling Compaq Pascal Programs

Table 15 /DESIGN Qualier Options

(continued on next page)

Compiling Compaq Pascal Programs 111

Table 16 (Cont.) /ENUMERATION_SIZE Qualier Options

112 Compiling Compaq Pascal Programs

Table 17 /FLOAT Qualier Options

Compiling Compaq Pascal Programs 113

114 Compiling Compaq Pascal Programs

Compiling Compaq Pascal Programs 115

Table 18 /OPTIMIZE Qualier Options

Negation NONE NOINLINE

Controls the optimization level (OpenVMS Values for num are:

116 Compiling Compaq Pascal Programs

Compiling Compaq Pascal Programs 117

Action Enables listing of all options.

SOURCE STATISTICS STRUCTURE_ LAYOUT

118 Compiling Compaq Pascal Programs

Compiling Compaq Pascal Programs 119

120 Compiling Compaq Pascal Programs

Compiling Compaq Pascal Programs 121

Table 113 (Cont.) /USAGE Qualier Options

122 Compiling Compaq Pascal Programs

Compiling Compaq Pascal Programs 123

1.1.3 Contents of the Compilation Listing File

Generated With /LIST /SHOW=TABLE_OF_CONTENTS /LIST /LIST /CROSS_REFERENCE /LIST /MACHINE_CODE

Source code Cross-reference Machine code Inline summary

/LIST /SHOW=INLINE /LIST /SHOW=STATISTICS /LIST /SHOW=STRUCTURE_LAYOUT

Compilation statistics Structure layout

124 Compiling Compaq Pascal Programs

Compiling Compaq Pascal Programs 125

126 Compiling Compaq Pascal Programs

1.1.4 Text Libraries

Compiling Compaq Pascal Programs 127

128 Compiling Compaq Pascal Programs

1.2 LINK Command

Compiling Compaq Pascal Programs 129

1.2.1 LINK Command Examples